VOLUME FOUR HUNDRED AND SIXTY-THREE

METHODS IN
ENZYMOLOGY
Guide to Protein Purification,
2nd Edition
METHODS IN ENZYMOLOGY
Editors-in-Chief

JOHN N. ABELSON AND MELVIN I. SIMON

Division of Biology
California Institute of Technology
Pasadena, California, USA

Founding Editors

SIDNEY P. COLOWICK AND NATHAN O. KAPLAN

 VOLUME FOUR HUNDRED AND SIXTY-THREE

METHODS IN
ENZYMOLOGY
Guide to Protein Purification,
2nd Edition

EDITED BY
RICHARD R. BURGESS
McArdle Laboratory for Cancer Research
University of Wisconsin-Madison
Madison, Wisconsin, USA

MURRAY P. DEUTSCHER
Department of Biochemistry and Molecular Biology
University of Miami School of Medicine
Miami, FL, USA

AMSTERDAM • BOSTON • HEIDELBERG • LONDON

NEW YORK • OXFORD • PARIS • SAN DIEGO
SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO
Academic Press is an imprint of Elsevier
Academic Press is an imprint of Elsevier
525 B Street, Suite 1900, San Diego, CA 92101-4495, USA
30 Corporate Drive, Suite 400, Burlington, MA 01803, USA
32 Jamestown Road, London NW1 7BY, UK

First edition 1990

Second edition 2009

Copyright # 2009, Elsevier Inc. All Rights Reserved.

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any
form or by any means electronic, mechanical, photocopying, recording or otherwise without the
prior written permission of the publisher

Permissions may be sought directly from Elsevier’s Science & Technology Rights Department
in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: permissions@
elsevier.com. Alternatively you can submit your request online by visiting the Elsevier web site at
http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material

Notice
No responsibility is assumed by the publisher for any injury and/or damage to persons or
property as a matter of products liability, negligence or otherwise, or from any use or operation
of any methods, products, instructions or ideas contained in the material herein. Because of rapid
advances in the medical sciences, in particular, independent verification of diagnoses and drug
dosages should be made

For information on all Academic Press publications

visit our website at elsevierdirect.com

ISBN: 978-0-12-374536-1 (hardback)

ISBN: 978-0-12-374978-9 (paperback)
ISSN: 0076-6879

Printed and bound in United States of America

09 10 11 12 10 9 8 7 6 5 4 3 2
CONTENTS

Contributors xix
Preface xxv
Volumes in Series xxvii

1. Why Purify Enzymes? 1

Arthur Kornberg (with preface by Murray Deutscher)

Section I. Developing Purification Procedures 7

2. Strategies and Considerations for Protein Purifications 9
Stuart Linn
1. General Considerations 10
2. Source of the Protein 15
3. Preparing Extracts 16
4. Bulk or Batch Procedures for Purification 17
5. Refined Procedures for Purification 18
6. Conclusions 19
References 19

3. Use of Bioinformatics in Planning a Protein Purification 21

Richard R. Burgess
1. What You Can Learn from an Amino Acid Sequence 22
2. What You Cannot yet Predict 25
3. Conclusion 26
References 27

4. Preparing a Purification Summary Table 29

Richard R. Burgess
1. Introduction 29
2. The Importance of Footnotes 32
3. The Value of an SDS–Polyacrylamide Gel Analysis on Main
Protein Fractions 32
4. Some Common Mistakes and Problems 32

v
vi Contents

Section II. General Methods for Handling

Proteins and Enzymes 35

5. Setting Up a Laboratory 37
Murray P. Deutscher
1. Supporting Materials 38
2. Detection and Assay Requirements 39
3. Fractionation Requirements 40

6. Buffers: Principles and Practice 43

Vincent S. Stoll and John S. Blanchard
1. Introduction 43
2. Theory 44
3. Buffer Selection 45
4. Buffer Preparation 48
5. Volatile Buffers 49
6. Broad-Range Buffers 50
7. Recipes for Buffer Stock Solutions 50
References 56

7. Measurement of Enzyme Activity 57

T. K. Harris and M. M. Keshwani
1. Introduction 58
2. Principles of Catalytic Activity 58
3. Measurement of Enzyme Activity 64
4. Formulation of Reaction Assay Mixtures 69
5. Discussion 71
Acknowledgments 71
References 71

8. Quantitation of Protein 73
James E. Noble and Marc J. A. Bailey
1. Introduction 74
2. General Instructions for Reagent Preparation 75
3. Ultraviolet Absorption Spectroscopy 80
4. Dye-Based Protein Assays 83
5. Coomassie Blue (Bradford) Protein Assay (Range: 1–50 mg) 85
6. Lowry (Alkaline Copper Reduction Assays) (Range: 5–100 mg) 86
7. Bicinchoninic Acid (BCA) (Range: 0.2–50 mg) 88
8. Amine Derivatization (Range: 0.05–25 mg) 89
Contents vii

9. Detergent-Based Fluorescent Detection (Range: 0.02–2 mg) 91

10. General Instructions 91
Acknowledgment 94
References 94

9. Concentration of Proteins and Removal of Solutes 97

David R. H. Evans, Jonathan K. Romero, and Matthew Westoby
1. Chromatography 98
2. Electrophoresis 103
3. Dialysis 104
4. Ultrafiltration 107
5. Lyophilization 113
6. Precipitation 116
7. Crystallization 118
References 118

10. Maintaining Protein Stability 121

Murray P. Deutscher
1. Causes of Protein Inactivation 121
2. General Handling Procedures 122
3. Concentration and Solvent Conditions 122
4. Stability Trials and Storage Conditions 123
5. Proteolysis and Protease Inhibitors 124
6. Loss of Activity 125

Section III. Recombinant Protein

Expression and Purification 129
11. Selecting an Appropriate Method for Expressing
a Recombinant Protein 131
William H. Brondyk
1. Introduction 132
2. Escherichia coli 133
3. Pichia pastoris 135
4. Baculovirus/Insect Cells 136
5. Mammalian Cells 138
6. Protein Characteristics 139
7. Recombinant Protein Applications 143
8. Conclusion 144
References 144
viii Contents

12. Bacterial Systems for Production of Heterologous Proteins 149

Sarah Zerbs, Ashley M. Frank, and Frank R. Collart
1. Introduction 150
2. Heterologous Protein Production Using Escherichia coli 150
3. Planning a Bacterial Expression Project 151
4. Evaluation of Project Requirements 152
5. Target Analysis 152
6. Cloning 153
7. Preparation of T4 DNA Polymerase-Treated DNA Fragments 155
8. Expression in the E. coli Cytoplasm 156
9. Expression of Cytoplasmic Targets in E. coli 157
10. Analysis of Heterologous Protein Expression in E. coli 157
11. Small-Scale Expression Cultures in Autoinduction
Media Protocol 160
12. Periplasmic Expression of Proteins 160
13. Expression of Periplasmic Targets in E. coli 161
14. Small-Scale Osmotic Shock Protocol 162
15. Alternative Bacterial Systems for Heterologous
Protein Production 164
16. Alternative Vector and Induction Conditions 165
17. Production Scale 166
Acknowledgment 166
References 166

13. Expression in the Yeast Pichia pastoris 169

James M. Cregg, Ilya Tolstorukov, Anasua Kusari, Jay Sunga,
Knut Madden, and Thomas Chappell
1. Introduction 170
2. Other Fungal Expression Systems 170
3. Culture Media and Microbial Manipulation Techniques 171
4. Genetic Strain Construction 172
5. Gene Preparation and Vector Selection 174
6. Transformation by Electroporation 176
7. DNA Preparation 176
8. Examination of Strains for Recombinant
Protein Production 178
9. Assay Development—The Yeastern Blot 182
10. Posttranslational Modification of the Recombinant Protein
(Proteinases and Glycosylation) 184
11. Selection for Multiple Copies of an Expression Cassette 185
References 187
Contents ix

14. Baculovirus–Insect Cell Expression Systems 191

Donald L. Jarvis
1. Introduction 192
2. A Brief Overview of Baculovirus Biology and
Molecular Biology 193
3. Baculovirus Expression Vectors 195
4. Baculovirus Expression Vector Technology—The Early Years 196
5. Baculovirus Expression Vector Technology—Improved 198
6. Baculovirus Transfer Plasmid Modifications 198
7. Parental Baculovirus Genome Modifications 200
8. The Other Half of the Baculovirus–Insect Cell System 210
9. A New Generation of Insect Cell Hosts for Baculovirus
Expression Vectors 212
10. Basic Baculovirus Protocols 214
References 218

15. Recombinant Protein Production by Transient Gene

Transfer into Mammalian Cells 223
Sabine Geisse and Cornelia Fux
1. Introduction 224
2. HEK293 and CHO Cell Lines Commonly Used
in TGE Approaches 224
3. Expression Vectors for HEK293 and CHO Cells 226
4. Cultivation of HEK293 Cells and CHO Cell Lines
in Suspension 228
5. Transfection Methods 228
6. Conclusions 234
Acknowledgments 234
References 235

16. Tagging for Protein Expression 239

Arun Malhotra
1. Introduction 240
2. Some Considerations When Designing a Tagged Protein 241
3. Protein Affinity Tags 245
4. Solubility Tags 249
5. Removal of Tags 251
6. Conclusions 253
Acknowledgment 254
References 254
x Contents

17. Refolding Solubilized Inclusion Body Proteins 259

Richard R. Burgess
1. Introduction 260
2. General Refolding Consideration 262
3. General Procedures 262
4. General Protocol 263
5. Comments on this General Procedure 264
6. Performing a Protein Refolding Test Screen 271
7. Other Refolding Procedures 275
8. Refolding Database: Refold 277
9. Strategies to Increase Proportion of Soluble Protein 277
10. Conclusion 279
References 279

Section IV. Preparation of Extracts

and Subcellular Fractionation 283

18. Advances in Preparation of Biological Extracts for

Protein Purification 285
Anthony C. Grabski
1. Introduction 286
2. Chemical and Enzymatic Cell Disruption 287
3. Mechanical Cell Disruption 290
4. Concluding Remarks 293
5. Procedures, Reagents, and Tips for Cell Disruption 293
References 301

19. Isolation of Subcellular Organelles and Structures 305

Uwe Michelsen and Jörg von Hagen
1. Introduction 306
2. Extraction and Prefractionation of Subproteomes 308
References 327

Section V. Purification Procedures: Bulk Methods 329

20. Protein Precipitation Techniques 331

Richard R. Burgess
1. Introduction 332
2. Ammonium Sulfate Precipitation 332
3. Polyethyleneimine Precipitation 337
Contents xi

4. Other Methods 341

5. General Procedures When Fractionating Proteins by Precipitation 341
References 342

21. Affi-Gel Blue for Nucleic Acid Removal and Early

Enrichment of Nucleotide Binding Proteins 343
Murray P. Deutscher
1. A Representative Protocol 344
Reference 345

Section VI. Purification Procedures:

Chromatographic Methods 347

22. Ion-Exchange Chromatography 349

Alois Jungbauer and Rainer Hahn
1. Introduction 349
2. Principle 351
3. Stationary Phases 353
4. Binding Conditions 355
5. Elution Conditions 361
6. Operation of Ion-Exchange Columns 363
7. Example: Separation of Complex Protein Mixture 366
8. Example: High-Resolution Separation with a Monolithic Column 367
References 370

23. Gel Filtration 373

Earle Stellwagen

1. Principle 373
2. Practice 374

24. Protein Chromatography on Hydroxyapatite Columns 387

Larry J. Cummings, Mark A. Snyder, and Kimberly Brisack
1. Introduction 388
2. Mechanisms 389
3. Chemical Characteristics 392
4. Purification Protocol Development 396
5. Packing Laboratory-Scale Columns 397
6. Process-Scale Column Packing 399
7. Applications 401
References 402
xii Contents

25. Theory and Use of Hydrophobic Interaction Chromatography

in Protein Purification Applications 405
Justin T. McCue

1. Theory 406
2. Latest Technology in HIC Adsorbents 408
3. Procedures for Use of HIC Adsorbents 409
References 413

Section VII. Purification Procedures: Affinity Methods 415

26. Affinity Chromatography: General Methods 417

Marjeta Urh, Dan Simpson, and Kate Zhao
1. Introduction 418
2. Selection of Affinity Matrix 419
3. Selection of Ligands 423
4. Attachment Chemistry 429
5. Purification Method 433
References 435

27. Immobilized-Metal Affinity Chromatography

(IMAC): A Review 439
Helena Block, Barbara Maertens, Anne Spriestersbach, Nicole Brinker,
Jan Kubicek, Roland Fabis, Jörg Labahn, and Frank Schäfer
1. Overview on IMAC Ligands and Immobilized Ions 440
2. IMAC Applications 444
3. Conclusions 467
Acknowledgments 468
References 468

28. Identification, Production, and Use of Polyol-Responsive

Monoclonal Antibodies for Immunoaffinity Chromatography 475
Nancy E. Thompson, Katherine M. Foley, Elizabeth S. Stalder,
and Richard R. Burgess
1. Introduction 476
2. Polyol-Responsive Monoclonal Antibodies 477
3. Conclusions 492
Disclosure 493
References 493
Contents xiii

Section VIII. Purification Procedures:

Electrophoretic Methods 495

29. One-Dimensional Gel Electrophoresis 497

David E. Garfin
1. Background 498
2. Polyacrylamide Gels 500
3. Principle of Method 501
4. Procedure 502
5. Detection of Proteins in Gels 508
6. Marker Proteins 510
7. Molecular Weight Determination 511
8. Preparative Electrophoresis 511
References 513

30. Isoelectric Focusing and Two-Dimensional Gel Electrophoresis 515

David B. Friedman, Sjouke Hoving, and Reiner Westermeier

1. Introduction 516
2. Materials 527
3. Methods 528
References 538

31. Protein Gel Staining Methods: An Introduction and Overview 541

Thomas H. Steinberg
1. Introduction 542
2. General Considerations 543
3. Instrumentation: Detection and Documentation 545
4. Total Protein Detection 545
5. Phosphoprotein Detection 556
6. Glycoprotein Detection 557
References 559

32. Elution of Proteins from Gels 565

Richard R. Burgess
1. Introduction 565
2. Elution of Proteins from Gels by Diffusion 566
3. Replacing the SDS Gel with a Reverse Phase HPLC 570
4. Electrophoretic Elution 570
5. Conclusion 571
References 571
xiv Contents

33. Performing and Optimizing Western Blots with an Emphasis

on Chemiluminescent Detection 573
Alice Alegria-Schaffer, Andrew Lodge, and Krishna Vattem

1. Western Blotting 574

2. Types of Western Blots 575
3. Detection Methods 579
4. The Chemiluminescence Signal 583
5. Common Problems and their Explanations 588
6. Blotting and Optimization Protocols using
Chemiluminescent Substrates 593
References 598

Section IX. Purification Procedures: Membrane

Proteins and Glycoproteins 601

34. Detergents: An Overview 603

Dirk Linke

1. Introduction 604
2. Detergent Structure 604
3. Properties of Detergents in Solution 605
4. Exploiting the Physicochemical Parameters of Detergents
for Membrane Protein Purification 612
5. Detergent Removal and Detergent Exchange 613
6. Choosing the Right Detergent 613
7. Conclusions 615
Acknowledgments 616
References 616

35. Purification of Membrane Proteins 619

Sue-Hwa Lin and Guido Guidotti
1. Introduction 619
2. Preparation of Membranes 620
3. Solubilization of Native Membrane Proteins 622
4. Purification of Membrane Proteins 625
5. Detergent Removal and Detergent Exchange 628
6. Expression and Purification of Recombinant Integral
Membrane Proteins 628
References 629
Contents xv

36. Purification of Recombinant G-Protein-Coupled Receptors 631

Reinhard Grisshammer
1. Introduction 632
2. Solubilization: General Considerations 633
3. Purification: General Considerations 634
4. Solubilization and Purification of a Recombinant
Neurotensin Receptor NTS1 637
5. Analysis of Purified NTS1 641
6. Conclusions 642
Acknowledgments 642
References 642

37. Cell-Free Translation of Integral Membrane Proteins

into Unilamelar Liposomes 647
Michael A. Goren, Akira Nozawa, Shin-ichi Makino,
Russell L. Wrobel, and Brian G. Fox
1. Introduction 648
2. Overview of Cell-Free Translation 650
3. Expression Vectors 651
4. Gene Cloning 652
5. PCR Product Cleanup 655
6. Flexi Vector and PCR Product Digestion Reaction 656
7. Ligation Reaction 657
8. Transformation Reaction 658
9. Purification of Plasmid DNA 659
10. Preparation of mRNA 660
11. Preparation of Liposomes 661
12. Wheat Germ Translation Reaction 662
13. Purification by Density Gradient
Ultracentrifugation 665
14. Characterization of Proteoliposomes 667
15. Considerations for Scale-Up 669
16. Isotopic Labeling for Structural Studies 669
17. Conclusions 670
Acknowledgments 670
References 670
xvi Contents

Section X. Characterization of Purified Proteins 675

38. Determination of Protein Purity 677

David G. Rhodes and Thomas M. Laue
1. Composition-Based and Activity-Based Analyses 680
2. Electrophoretic Methods 681
3. Chromatographic Methods 685
4. Sedimentation Velocity Methods 687
5. Mass Spectrometry Methods 687
6. Light Scattering Methods 688
References 689

39. Determination of Size, Molecular Weight,

and Presence of Subunits 691
David G. Rhodes, Robert E. Bossio, and Thomas M. Laue

1. Introduction 692
2. Chemical Methods 695
3. Transport Methods 698
4. Scattering Methods 716
5. Presence of Subunits 719
References 721

40. Identification and Quantification of Protein

Posttranslational Modifications 725
Adam R. Farley and Andrew J. Link

1. Introduction 726
2. Enrichment Techniques for Identifying PTMs 731
3. Nitrosative Protein Modifications 740
4. Methylation and Acetylation 741
5. Mass Spectrometry Analysis 744
6. CID versus ECD versus ETD 747
7. Quantifying PTMs 750
8. Future Directions 756
Acknowledgments 758
References 758
Contents xvii

Section XI. Additional Techniques 765

41. Parallel Methods for Expression and Purification 767

Scott A. Lesley
1. Introduction 767
2. Strategies Based on End-Use 768
3. Parallel Cloning Strategies for Creating Expression Constructs 771
4. Small-Scale Expression Screening to Identify Suitable Constructs 774
5. Analytical Testing of Proteins for Selection 778
6. Large-Scale Parallel Expression 780
7. Conclusion 783
Acknowledgments 783
References 784

42. Techniques to Isolate O2-Sensitive Proteins: [4FE–4S]-FNR

as an Example 787
Aixin Yan and Patricia J. Kiley
1. Introduction 788
2. Anaerobic Isolation of 4FE-FNR 790
3. Characterization of [4FE–4S]2þ Cluster Containing FNR 799
4. Summary 803
References 803

Section XII. Concluding Remarks 807

43. Rethinking Your Purification Procedure 809

Murray P. Deutscher

1. Introduction 809

44. Important but Little Known (or Forgotten) Artifacts in

Protein Biochemistry 813
Richard R. Burgess
1. Introduction 814
2. SDS Gel Electrophoresis Sample Preparation 814
3. Buffers 816
4. Chromatography 817
5. Protein Absorption During Filtration 818
6. Chemical Leaching from Plasticware 819
7. Cyanate in Urea 819
References 820

Author Index 821

Subject Index 835
CONTRIBUTORS

Alice Alegria-Schaffer
Thermo Fisher Scientific, Pierce Protein Research, Rockford, Illinois, USA
Marc J. A. Bailey
Nokia Research Centre - Eurolab, University of Cambridge, Cambridge, United
Kingdom
John S. Blanchard
Department of Biochemistry, Albert Einstein College of Medicine, Bronx,
New York
Helena Block
QIAGEN GmbH, Qiagen Strasse 1, Hilden, Germany
Robert E. Bossio
Department of Chemistry, University of Michigan–Dearborn, Dearborn,
Michigan, USA
Nicole Brinker
QIAGEN GmbH, Qiagen Strasse 1, Hilden, Germany
Kimberly Brisack
Bio-Rad Laboratories, Inc., Hercules, California, USA
William H. Brondyk
Genzyme Corporation, Framingham, Massachusetts, USA
Richard R. Burgess
McArdle Laboratory for Cancer Research, University of Wisconsin-Madison,
Madison, Wisconsin, USA
Thomas Chappell
Biogrammatics, Inc., Carlsbad, California, USA
Frank R. Collart
Biosciences Division, Argonne National Laboratory, Lemont, Illinois, USA
James M. Cregg
Keck Graduate Institute of Applied Life Sciences, Claremont, California, USA, and
Biogrammatics, Inc., Carlsbad, California, USA

xix
xx Contributors

Larry J. Cummings
Bio-Rad Laboratories, Inc., Hercules, California, USA
Murray P. Deutscher
Department of Biochemistry and Molecular Biology, University of Miami School
of Medicine, Miami, Florida, USA
David R. H. Evans
Process Biochemistry, Biopharmaceutical Development, Biogen Idec. Inc.,
Cambridge, Massachusetts, USA
Roland Fabis
QIAGEN GmbH, Qiagen Strasse 1, Hilden, Germany
Adam R. Farley
Department of Biochemistry, Vanderbilt University School of Medicine,
Nashville, Tennessee, USA
Katherine M. Foley
McArdle Laboratory for Cancer Research, University of Wisconsin-Madison,
Madison, Wisconsin, USA
Brian G. Fox
Department of Biochemistry, Dr. Brian Fox Lab, University of Wisconsin-
Madison, Madison, Wisconsin, USA
Ashley M. Frank
Biosciences Division, Argonne National Laboratory, Lemont, Illinois, USA
David B. Friedman
Proteomics Laboratory, Mass Spectrometry Research Center, Vanderbilt
University, Nashville, Tennessee, USA
Cornelia Fux
Novartis Pharma AG, Department of NBx/PSD: Scale up, Basel, Switzerland
David E. Garfin
Chemical Division, Research Products Group, Bio-Rad Laboratories,
Incorporated, Richmond, California
Sabine Geisse
Novartis Institutes for BioMedical Research, Department of NBC/PPA, Basel,
Switzerland
Michael A. Goren
Department of Biochemistry, Dr. Brian Fox Lab, University of Wisconsin-
Madison, Madison, Wisconsin, USA
Contributors xxi

Anthony C. Grabski
Department of Research and Development, Semba Biosciences, Inc., Madison,
Wisconsin, USA
Reinhard Grisshammer
Department of Health and Human Services, National Institutes of Health, National
Institute of Neurological Disorders and Stroke, Rockville, Maryland, USA
Guido Guidotti
Department of Molecular and Cellular Biology, Harvard University, Cambridge,
Massachusetts, USA
Rainer Hahn
Department of Biotechnology, University of Natural Resources and Applied Life
Sciences, Vienna, Austria
T. K. Harris
Department of Biochemistry and Molecular Biology, Miller School of Medicine,
University of Miami, Miami, Florida, USA
Sjouke Hoving
Novartis Institutes of BioMedical Research, Basel, Switzerland
Donald L. Jarvis
Department of Molecular Biology, University of Wyoming, Laramie, Wyoming,
USA
Alois Jungbauer
Department of Biotechnology, University of Natural Resources and Applied Life
Sciences, Vienna, Austria
M. M. Keshwani
Department of Biochemistry and Molecular Biology, Miller School of Medicine,
University of Miami, Miami, Florida, USA
Patricia J. Kiley
Department of Biomolecular Chemistry, University of Wisconsin, Madison,
Wisconsin, USA
Jan Kubicek
QIAGEN GmbH, Qiagen Strasse 1, Hilden, Germany
Anasua Kusari
Keck Graduate Institute of Applied Life Sciences, Claremont, California, USA
Jörg Labahn
Institute of Structural Biology (IBI-2), Research Center Jülich, Jülich, Germany
xxii Contributors

Thomas M. Laue
Department of Biochemistry and Molecular Biology, University of New
Hampshire, Durham, New Hampshire, USA
Scott A. Lesley
Genomics Institute of the Novartis Research Foundation, San Diego, California,
USA
Sue-Hwa Lin
Department of Molecular Pathology, University of Texas, Houston, Texas, USA
Andrew J. Link
Department of Microbiology and Immunology, Vanderbilt University School of
Medicine, Nashville, Tennessee, USA
Dirk Linke
Department I, Protein Evolution, Max Planck Institute for Developmental
Biology, Tübingen, Germany
Stuart Linn
Department of Molecular and Cellular Biology, University of California, Berkeley,
California, USA
Andrew Lodge
Thermo Fisher Scientific, Pierce Protein Research, Rockford, Illinois, USA
Knut Madden
Biogrammatics, Inc., Carlsbad, California, USA
Barbara Maertens
QIAGEN GmbH, Qiagen Strasse 1, Hilden, Germany
Shin-ichi Makino
Department of Biochemistry, Dr. Brian Fox Lab, University of Wisconsin-
Madison, Madison, Wisconsin, USA
Arun Malhotra
Department of Biochemistry and Molecular Biology, University of Miami Miller
School of Medicine, Miami, Florida, USA
Justin T. McCue
Biogen Idec Corporation, Cambridge, Massachusetts, USA
Uwe Michelsen
Merck KGaA, Darmstadt, Germany
James E. Noble
Analytical Science, National Physical Laboratory, Teddington, Middlesex, United
Kingdom
Contributors xxiii

Akira Nozawa
Department of Biochemistry, Dr. Brian Fox Lab, University of Wisconsin-
Madison, Madison, Wisconsin, USA
David G. Rhodes
Lipophilia Consulting, Storrs, Connecticut, USA
Jonathan K. Romero
Process Biochemistry, Biopharmaceutical Development, Biogen Idec. Inc.,
Cambridge, Massachusetts, USA
Frank Schäfer
QIAGEN GmbH, Qiagen Strasse 1, Hilden, Germany
Dan Simpson
Promega Corporation, Madison, Wisconsin, USA
Mark A. Snyder
Bio-Rad Laboratories, Inc., Hercules, California, USA
Anne Spriestersbach
QIAGEN GmbH, Qiagen Strasse 1, Hilden, Germany
Elizabeth S. Stalder
McArdle Laboratory for Cancer Research, University of Wisconsin-Madison,
Madison, Wisconsin, USA
Thomas H. Steinberg
McArdle Laboratory for Cancer Research, University of Wisconsin–Madison,
Madison, Wisconsin, USA
Earle Stellwagen
Department of Biochemistry, University of Iowa, Iowa City, Iowa
Vincent S. Stoll
Department of Biochemistry, Albert Einstein College of Medicine, Bronx,
New York
Jay Sunga
Keck Graduate Institute of Applied Life Sciences, Claremont, California, USA
Nancy E. Thompson
McArdle Laboratory for Cancer Research, University of Wisconsin-Madison,
Madison, Wisconsin, USA
Ilya Tolstorukov
Keck Graduate Institute of Applied Life Sciences, Claremont, California, USA, and
Biogrammatics, Inc., Carlsbad, California, USA
Marjeta Urh
Promega Corporation, Madison, Wisconsin, USA
xxiv Contributors

Krishna Vattem
Thermo Fisher Scientific, Pierce Protein Research, Rockford, Illinois, USA
Jörg von Hagen
Merck KGaA, Darmstadt, Germany
Reiner Westermeier
Gelcompany GmbH, Paul-Ehrlich-Strasse, Tübingen, Germany
Matthew Westoby
Process Biochemistry, Biopharmaceutical Development, Biogen Idec. Inc., San
Diego, California, USA
Russell L. Wrobel
Department of Biochemistry, Dr. Brian Fox Lab, University of Wisconsin-
Madison, Madison, Wisconsin, USA
Aixin Yan
School of Biological Sciences, The University of Hong Kong, Hong Kong,
Hong Kong SAR
Sarah Zerbs
Biosciences Division, Argonne National Laboratory, Lemont, Illinois, USA
Kate Zhao
Promega Corporation, Madison, Wisconsin, USA
PREFACE

Protein biochemistry continues to be an essential part of modern biological

research. With the huge increase in genome sequencing, the sequences of
most studied organisms are available, facilitating gene cloning, protein
production, and the study of protein properties, structure, and function.
Tremendous advances in heterologous expression of recombinant proteins
have greatly increased our ability to produce proteins of interest. However,
a protein must still be purified and characterized. This is particularly impor-
tant if the gene is unknown.
Much has happened since the First Edition of the Guide to Protein
Purification (Methods in Enzymology, Volume 182) was published in 1990.
Major changes have occurred, including the explosion of genomics and
proteomics, high-throughput expression/purification for the Protein Struc-
ture Initiative, protein target-based high throughput screening of chemicals
to find new pharmaceuticals, a growing need to express and purify mem-
brane proteins, the use of purification and solubilization tags, the systematic
testing of refolding conditions to facilitate and optimize protein refolding,
mass spectrometry analysis of post-translational modifications, and expres-
sion of recombinant proteins in non-bacterial hosts. These changes have
necessitated significant advances in techniques, materials, reagents and
equipment.
In the Second Edition of the Guide to Protein Purification, we have
tried to identify the areas of greatest change in the last 20 years and to
present authoritative reviews and methods that reflect the current best
practice in each area of protein purification. First and foremost, we wanted
these chapters to be educational and highly useful to the reader, and
included typical best use protocols, cautions, technical tips, and limitations.
While many of the chapters are by experts in academia, a significant number
are by experts in certain areas in the biotechnology and pharmaceutical
industry. We feel that high-level scientists in companies that use, develop,
sell, or provide technical support for products in technical areas are particu-
larly valuable in writing such chapters. We have encouraged such chapters
and tried to ensure that they are even-handed and not focused entirely on
the products of a particular company.
This Guide is a self-contained volume covering all the important
procedures for purifying proteins as well as the more specialized techniques.

xxv
xxvi Preface

We hope that the Guide will be an invaluable resource and reference text
for researchers new to protein purification as well as for the more experi-
enced researchers, and that it will find an important place in every protein
biochemistry laboratory.

RICHARD R. BURGESS
MURRAY P. DEUTSCHER
METHODS IN ENZYMOLOGY

VOLUME I. Preparation and Assay of Enzymes

Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN
VOLUME II. Preparation and Assay of Enzymes
Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN
VOLUME III. Preparation and Assay of Substrates
Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN
VOLUME IV. Special Techniques for the Enzymologist
Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN
VOLUME V. Preparation and Assay of Enzymes
Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN
VOLUME VI. Preparation and Assay of Enzymes (Continued)
Preparation and Assay of Substrates
Special Techniques
Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN
VOLUME VII. Cumulative Subject Index
Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN
VOLUME VIII. Complex Carbohydrates
Edited by ELIZABETH F. NEUFELD AND VICTOR GINSBURG
VOLUME IX. Carbohydrate Metabolism
Edited by WILLIS A. WOOD
VOLUME X. Oxidation and Phosphorylation
Edited by RONALD W. ESTABROOK AND MAYNARD E. PULLMAN
VOLUME XI. Enzyme Structure
Edited by C. H. W. HIRS
VOLUME XII. Nucleic Acids (Parts A and B)
Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE
VOLUME XIII. Citric Acid Cycle
Edited by J. M. LOWENSTEIN
VOLUME XIV. Lipids
Edited by J. M. LOWENSTEIN
VOLUME XV. Steroids and Terpenoids
Edited by RAYMOND B. CLAYTON

xxvii
xxviii Methods in Enzymology

VOLUME XVI. Fast Reactions

Edited by KENNETH KUSTIN
VOLUME XVII. Metabolism of Amino Acids and Amines (Parts A and B)
Edited by HERBERT TABOR AND CELIA WHITE TABOR
VOLUME XVIII. Vitamins and Coenzymes (Parts A, B, and C)
Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT
VOLUME XIX. Proteolytic Enzymes
Edited by GERTRUDE E. PERLMANN AND LASZLO LORAND
VOLUME XX. Nucleic Acids and Protein Synthesis (Part C)
Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN
VOLUME XXI. Nucleic Acids (Part D)
Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE
VOLUME XXII. Enzyme Purification and Related Techniques
Edited by WILLIAM B. JAKOBY
VOLUME XXIII. Photosynthesis (Part A)
Edited by ANTHONY SAN PIETRO
VOLUME XXIV. Photosynthesis and Nitrogen Fixation (Part B)
Edited by ANTHONY SAN PIETRO
VOLUME XXV. Enzyme Structure (Part B)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF
VOLUME XXVI. Enzyme Structure (Part C)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF
VOLUME XXVII. Enzyme Structure (Part D)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF
VOLUME XXVIII. Complex Carbohydrates (Part B)
Edited by VICTOR GINSBURG
VOLUME XXIX. Nucleic Acids and Protein Synthesis (Part E)
Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE
VOLUME XXX. Nucleic Acids and Protein Synthesis (Part F)
Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN
VOLUME XXXI. Biomembranes (Part A)
Edited by SIDNEY FLEISCHER AND LESTER PACKER
VOLUME XXXII. Biomembranes (Part B)
Edited by SIDNEY FLEISCHER AND LESTER PACKER
VOLUME XXXIII. Cumulative Subject Index Volumes I-XXX
Edited by MARTHA G. DENNIS AND EDWARD A. DENNIS
VOLUME XXXIV. Affinity Techniques (Enzyme Purification: Part B)
Edited by WILLIAM B. JAKOBY AND MEIR WILCHEK
Methods in Enzymology xxix

VOLUME XXXV. Lipids (Part B)

Edited by JOHN M. LOWENSTEIN
VOLUME XXXVI. Hormone Action (Part A: Steroid Hormones)
Edited by BERT W. O’MALLEY AND JOEL G. HARDMAN
VOLUME XXXVII. Hormone Action (Part B: Peptide Hormones)
Edited by BERT W. O’MALLEY AND JOEL G. HARDMAN
VOLUME XXXVIII. Hormone Action (Part C: Cyclic Nucleotides)
Edited by JOEL G. HARDMAN AND BERT W. O’MALLEY
VOLUME XXXIX. Hormone Action (Part D: Isolated Cells, Tissues,
and Organ Systems)
Edited by JOEL G. HARDMAN AND BERT W. O’MALLEY
VOLUME XL. Hormone Action (Part E: Nuclear Structure and Function)
Edited by BERT W. O’MALLEY AND JOEL G. HARDMAN
VOLUME XLI. Carbohydrate Metabolism (Part B)
Edited by W. A. WOOD
VOLUME XLII. Carbohydrate Metabolism (Part C)
Edited by W. A. WOOD
VOLUME XLIII. Antibiotics
Edited by JOHN H. HASH
VOLUME XLIV. Immobilized Enzymes
Edited by KLAUS MOSBACH
VOLUME XLV. Proteolytic Enzymes (Part B)
Edited by LASZLO LORAND
VOLUME XLVI. Affinity Labeling
Edited by WILLIAM B. JAKOBY AND MEIR WILCHEK
VOLUME XLVII. Enzyme Structure (Part E)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF
VOLUME XLVIII. Enzyme Structure (Part F)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF
VOLUME XLIX. Enzyme Structure (Part G)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF
VOLUME L. Complex Carbohydrates (Part C)
Edited by VICTOR GINSBURG
VOLUME LI. Purine and Pyrimidine Nucleotide Metabolism
Edited by PATRICIA A. HOFFEE AND MARY ELLEN JONES
VOLUME LII. Biomembranes (Part C: Biological Oxidations)
Edited by SIDNEY FLEISCHER AND LESTER PACKER
xxx Methods in Enzymology

VOLUME LIII. Biomembranes (Part D: Biological Oxidations)

Edited by SIDNEY FLEISCHER AND LESTER PACKER
VOLUME LIV. Biomembranes (Part E: Biological Oxidations)
Edited by SIDNEY FLEISCHER AND LESTER PACKER
VOLUME LV. Biomembranes (Part F: Bioenergetics)
Edited by SIDNEY FLEISCHER AND LESTER PACKER
VOLUME LVI. Biomembranes (Part G: Bioenergetics)
Edited by SIDNEY FLEISCHER AND LESTER PACKER
VOLUME LVII. Bioluminescence and Chemiluminescence
Edited by MARLENE A. DELUCA
VOLUME LVIII. Cell Culture
Edited by WILLIAM B. JAKOBY AND IRA PASTAN
VOLUME LIX. Nucleic Acids and Protein Synthesis (Part G)
Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN
VOLUME LX. Nucleic Acids and Protein Synthesis (Part H)
Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN
VOLUME 61. Enzyme Structure (Part H)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF
VOLUME 62. Vitamins and Coenzymes (Part D)
Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT
VOLUME 63. Enzyme Kinetics and Mechanism (Part A: Initial Rate and
Inhibitor Methods)
Edited by DANIEL L. PURICH
VOLUME 64. Enzyme Kinetics and Mechanism
(Part B: Isotopic Probes and Complex Enzyme Systems)
Edited by DANIEL L. PURICH
VOLUME 65. Nucleic Acids (Part I)
Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE
VOLUME 66. Vitamins and Coenzymes (Part E)
Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT
VOLUME 67. Vitamins and Coenzymes (Part F)
Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT
VOLUME 68. Recombinant DNA
Edited by RAY WU
VOLUME 69. Photosynthesis and Nitrogen Fixation (Part C)
Edited by ANTHONY SAN PIETRO
VOLUME 70. Immunochemical Techniques (Part A)
Edited by HELEN VAN VUNAKIS AND JOHN J. LANGONE
Methods in Enzymology xxxi

VOLUME 71. Lipids (Part C)

Edited by JOHN M. LOWENSTEIN
VOLUME 72. Lipids (Part D)
Edited by JOHN M. LOWENSTEIN
VOLUME 73. Immunochemical Techniques (Part B)
Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS
VOLUME 74. Immunochemical Techniques (Part C)
Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS
VOLUME 75. Cumulative Subject Index Volumes XXXI, XXXII, XXXIV–LX
Edited by EDWARD A. DENNIS AND MARTHA G. DENNIS
VOLUME 76. Hemoglobins
Edited by ERALDO ANTONINI, LUIGI ROSSI-BERNARDI, AND EMILIA CHIANCONE
VOLUME 77. Detoxication and Drug Metabolism
Edited by WILLIAM B. JAKOBY
VOLUME 78. Interferons (Part A)
Edited by SIDNEY PESTKA
VOLUME 79. Interferons (Part B)
Edited by SIDNEY PESTKA
VOLUME 80. Proteolytic Enzymes (Part C)
Edited by LASZLO LORAND
VOLUME 81. Biomembranes (Part H: Visual Pigments and Purple Membranes, I)
Edited by LESTER PACKER
VOLUME 82. Structural and Contractile Proteins (Part A: Extracellular Matrix)
Edited by LEON W. CUNNINGHAM AND DIXIE W. FREDERIKSEN
VOLUME 83. Complex Carbohydrates (Part D)
Edited by VICTOR GINSBURG
VOLUME 84. Immunochemical Techniques (Part D: Selected Immunoassays)
Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS
VOLUME 85. Structural and Contractile Proteins (Part B: The Contractile Apparatus
and the Cytoskeleton)
Edited by DIXIE W. FREDERIKSEN AND LEON W. CUNNINGHAM
VOLUME 86. Prostaglandins and Arachidonate Metabolites
Edited by WILLIAM E. M. LANDS AND WILLIAM L. SMITH
VOLUME 87. Enzyme Kinetics and Mechanism (Part C: Intermediates,
Stereo-chemistry, and Rate Studies)
Edited by DANIEL L. PURICH
VOLUME 88. Biomembranes (Part I: Visual Pigments and Purple Membranes, II)
Edited by LESTER PACKER
xxxii Methods in Enzymology

VOLUME 89. Carbohydrate Metabolism (Part D)

Edited by WILLIS A. WOOD
VOLUME 90. Carbohydrate Metabolism (Part E)
Edited by WILLIS A. WOOD
VOLUME 91. Enzyme Structure (Part I)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF
VOLUME 92. Immunochemical Techniques (Part E: Monoclonal Antibodies and
General Immunoassay Methods)
Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS
VOLUME 93. Immunochemical Techniques (Part F: Conventional Antibodies, Fc
Receptors, and Cytotoxicity)
Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS
VOLUME 94. Polyamines
Edited by HERBERT TABOR AND CELIA WHITE TABOR
VOLUME 95. Cumulative Subject Index Volumes 61–74, 76–80
Edited by EDWARD A. DENNIS AND MARTHA G. DENNIS
VOLUME 96. Biomembranes [Part J: Membrane Biogenesis: Assembly and
Targeting (General Methods; Eukaryotes)]
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
VOLUME 97. Biomembranes [Part K: Membrane Biogenesis: Assembly and
Targeting (Prokaryotes, Mitochondria, and Chloroplasts)]
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
VOLUME 98. Biomembranes (Part L: Membrane Biogenesis: Processing
and Recycling)
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
VOLUME 99. Hormone Action (Part F: Protein Kinases)
Edited by JACKIE D. CORBIN AND JOEL G. HARDMAN
VOLUME 100. Recombinant DNA (Part B)
Edited by RAY WU, LAWRENCE GROSSMAN, AND KIVIE MOLDAVE
VOLUME 101. Recombinant DNA (Part C)
Edited by RAY WU, LAWRENCE GROSSMAN, AND KIVIE MOLDAVE
VOLUME 102. Hormone Action (Part G: Calmodulin and
Calcium-Binding Proteins)
Edited by ANTHONY R. MEANS AND BERT W. O’MALLEY
VOLUME 103. Hormone Action (Part H: Neuroendocrine Peptides)
Edited by P. MICHAEL CONN
VOLUME 104. Enzyme Purification and Related Techniques (Part C)
Edited by WILLIAM B. JAKOBY
Methods in Enzymology xxxiii

VOLUME 105. Oxygen Radicals in Biological Systems

Edited by LESTER PACKER
VOLUME 106. Posttranslational Modifications (Part A)
Edited by FINN WOLD AND KIVIE MOLDAVE
VOLUME 107. Posttranslational Modifications (Part B)
Edited by FINN WOLD AND KIVIE MOLDAVE
VOLUME 108. Immunochemical Techniques (Part G: Separation and
Characterization of Lymphoid Cells)
Edited by GIOVANNI DI SABATO, JOHN J. LANGONE, AND HELEN VAN VUNAKIS
VOLUME 109. Hormone Action (Part I: Peptide Hormones)
Edited by LUTZ BIRNBAUMER AND BERT W. O’MALLEY
VOLUME 110. Steroids and Isoprenoids (Part A)
Edited by JOHN H. LAW AND HANS C. RILLING
VOLUME 111. Steroids and Isoprenoids (Part B)
Edited by JOHN H. LAW AND HANS C. RILLING
VOLUME 112. Drug and Enzyme Targeting (Part A)
Edited by KENNETH J. WIDDER AND RALPH GREEN
VOLUME 113. Glutamate, Glutamine, Glutathione, and Related Compounds
Edited by ALTON MEISTER
VOLUME 114. Diffraction Methods for Biological Macromolecules (Part A)
Edited by HAROLD W. WYCKOFF, C. H. W. HIRS, AND SERGE N. TIMASHEFF
VOLUME 115. Diffraction Methods for Biological Macromolecules (Part B)
Edited by HAROLD W. WYCKOFF, C. H. W. HIRS, AND SERGE N. TIMASHEFF
VOLUME 116. Immunochemical Techniques
(Part H: Effectors and Mediators of Lymphoid Cell Functions)
Edited by GIOVANNI DI SABATO, JOHN J. LANGONE, AND HELEN VAN VUNAKIS
VOLUME 117. Enzyme Structure (Part J)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF
VOLUME 118. Plant Molecular Biology
Edited by ARTHUR WEISSBACH AND HERBERT WEISSBACH
VOLUME 119. Interferons (Part C)
Edited by SIDNEY PESTKA
VOLUME 120. Cumulative Subject Index Volumes 81–94, 96–101
VOLUME 121. Immunochemical Techniques (Part I: Hybridoma Technology and
Monoclonal Antibodies)
Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS
VOLUME 122. Vitamins and Coenzymes (Part G)
Edited by FRANK CHYTIL AND DONALD B. MCCORMICK
xxxiv Methods in Enzymology

VOLUME 123. Vitamins and Coenzymes (Part H)

Edited by FRANK CHYTIL AND DONALD B. MCCORMICK
VOLUME 124. Hormone Action (Part J: Neuroendocrine Peptides)
Edited by P. MICHAEL CONN
VOLUME 125. Biomembranes (Part M: Transport in Bacteria, Mitochondria, and
Chloroplasts: General Approaches and Transport Systems)
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
VOLUME 126. Biomembranes (Part N: Transport in Bacteria, Mitochondria, and
Chloroplasts: Protonmotive Force)
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
VOLUME 127. Biomembranes (Part O: Protons and Water: Structure
and Translocation)
Edited by LESTER PACKER
VOLUME 128. Plasma Lipoproteins (Part A: Preparation, Structure, and
Molecular Biology)
Edited by JERE P. SEGREST AND JOHN J. ALBERS
VOLUME 129. Plasma Lipoproteins (Part B: Characterization, Cell Biology,
and Metabolism)
Edited by JOHN J. ALBERS AND JERE P. SEGREST
VOLUME 130. Enzyme Structure (Part K)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF
VOLUME 131. Enzyme Structure (Part L)
Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF
VOLUME 132. Immunochemical Techniques (Part J: Phagocytosis and
Cell-Mediated Cytotoxicity)
Edited by GIOVANNI DI SABATO AND JOHANNES EVERSE
VOLUME 133. Bioluminescence and Chemiluminescence (Part B)
Edited by MARLENE DELUCA AND WILLIAM D. MCELROY
VOLUME 134. Structural and Contractile Proteins (Part C: The Contractile
Apparatus and the Cytoskeleton)
Edited by RICHARD B. VALLEE
VOLUME 135. Immobilized Enzymes and Cells (Part B)
Edited by KLAUS MOSBACH
VOLUME 136. Immobilized Enzymes and Cells (Part C)
Edited by KLAUS MOSBACH
VOLUME 137. Immobilized Enzymes and Cells (Part D)
Edited by KLAUS MOSBACH
VOLUME 138. Complex Carbohydrates (Part E)
Edited by VICTOR GINSBURG
Methods in Enzymology xxxv

VOLUME 139. Cellular Regulators (Part A: Calcium- and

Calmodulin-Binding Proteins)
Edited by ANTHONY R. MEANS AND P. MICHAEL CONN
VOLUME 140. Cumulative Subject Index Volumes 102–119, 121–134
VOLUME 141. Cellular Regulators (Part B: Calcium and Lipids)
Edited by P. MICHAEL CONN AND ANTHONY R. MEANS
VOLUME 142. Metabolism of Aromatic Amino Acids and Amines
Edited by SEYMOUR KAUFMAN
VOLUME 143. Sulfur and Sulfur Amino Acids
Edited by WILLIAM B. JAKOBY AND OWEN GRIFFITH
VOLUME 144. Structural and Contractile Proteins (Part D: Extracellular Matrix)
Edited by LEON W. CUNNINGHAM
VOLUME 145. Structural and Contractile Proteins (Part E: Extracellular Matrix)
Edited by LEON W. CUNNINGHAM
VOLUME 146. Peptide Growth Factors (Part A)
Edited by DAVID BARNES AND DAVID A. SIRBASKU
VOLUME 147. Peptide Growth Factors (Part B)
Edited by DAVID BARNES AND DAVID A. SIRBASKU
VOLUME 148. Plant Cell Membranes
Edited by LESTER PACKER AND ROLAND DOUCE
VOLUME 149. Drug and Enzyme Targeting (Part B)
Edited by RALPH GREEN AND KENNETH J. WIDDER
VOLUME 150. Immunochemical Techniques (Part K: In Vitro Models of B and
T Cell Functions and Lymphoid Cell Receptors)
Edited by GIOVANNI DI SABATO
VOLUME 151. Molecular Genetics of Mammalian Cells
Edited by MICHAEL M. GOTTESMAN
VOLUME 152. Guide to Molecular Cloning Techniques
Edited by SHELBY L. BERGER AND ALAN R. KIMMEL
VOLUME 153. Recombinant DNA (Part D)
Edited by RAY WU AND LAWRENCE GROSSMAN
VOLUME 154. Recombinant DNA (Part E)
Edited by RAY WU AND LAWRENCE GROSSMAN
VOLUME 155. Recombinant DNA (Part F)
Edited by RAY WU
VOLUME 156. Biomembranes (Part P: ATP-Driven Pumps and Related Transport:
The Na, K-Pump)
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
xxxvi Methods in Enzymology

VOLUME 157. Biomembranes (Part Q: ATP-Driven Pumps and Related Transport:

Calcium, Proton, and Potassium Pumps)
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
VOLUME 158. Metalloproteins (Part A)
Edited by JAMES F. RIORDAN AND BERT L. VALLEE
VOLUME 159. Initiation and Termination of Cyclic Nucleotide Action
Edited by JACKIE D. CORBIN AND ROGER A. JOHNSON
VOLUME 160. Biomass (Part A: Cellulose and Hemicellulose)
Edited by WILLIS A. WOOD AND SCOTT T. KELLOGG
VOLUME 161. Biomass (Part B: Lignin, Pectin, and Chitin)
Edited by WILLIS A. WOOD AND SCOTT T. KELLOGG
VOLUME 162. Immunochemical Techniques (Part L: Chemotaxis
and Inflammation)
Edited by GIOVANNI DI SABATO
VOLUME 163. Immunochemical Techniques (Part M: Chemotaxis
and Inflammation)
Edited by GIOVANNI DI SABATO
VOLUME 164. Ribosomes
Edited by HARRY F. NOLLER, JR., AND KIVIE MOLDAVE
VOLUME 165. Microbial Toxins: Tools for Enzymology
Edited by SIDNEY HARSHMAN
VOLUME 166. Branched-Chain Amino Acids
Edited by ROBERT HARRIS AND JOHN R. SOKATCH
VOLUME 167. Cyanobacteria
Edited by LESTER PACKER AND ALEXANDER N. GLAZER
VOLUME 168. Hormone Action (Part K: Neuroendocrine Peptides)
Edited by P. MICHAEL CONN
VOLUME 169. Platelets: Receptors, Adhesion, Secretion (Part A)
Edited by JACEK HAWIGER
VOLUME 170. Nucleosomes
Edited by PAUL M. WASSARMAN AND ROGER D. KORNBERG
VOLUME 171. Biomembranes (Part R: Transport Theory: Cells and Model
Membranes)
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
VOLUME 172. Biomembranes (Part S: Transport: Membrane Isolation
and Characterization)
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
Methods in Enzymology xxxvii

VOLUME 173. Biomembranes [Part T: Cellular and Subcellular Transport:

Eukaryotic (Nonepithelial) Cells]
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
VOLUME 174. Biomembranes [Part U: Cellular and Subcellular Transport:
Eukaryotic (Nonepithelial) Cells]
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
VOLUME 175. Cumulative Subject Index Volumes 135–139, 141–167
VOLUME 176. Nuclear Magnetic Resonance (Part A: Spectral Techniques and
Dynamics)
Edited by NORMAN J. OPPENHEIMER AND THOMAS L. JAMES
VOLUME 177. Nuclear Magnetic Resonance (Part B: Structure and Mechanism)
Edited by NORMAN J. OPPENHEIMER AND THOMAS L. JAMES
VOLUME 178. Antibodies, Antigens, and Molecular Mimicry
Edited by JOHN J. LANGONE
VOLUME 179. Complex Carbohydrates (Part F)
Edited by VICTOR GINSBURG
VOLUME 180. RNA Processing (Part A: General Methods)
Edited by JAMES E. DAHLBERG AND JOHN N. ABELSON
VOLUME 181. RNA Processing (Part B: Specific Methods)
Edited by JAMES E. DAHLBERG AND JOHN N. ABELSON
VOLUME 182. Guide to Protein Purification
Edited by MURRAY P. DEUTSCHER
VOLUME 183. Molecular Evolution: Computer Analysis of Protein and
Nucleic Acid Sequences
Edited by RUSSELL F. DOOLITTLE
VOLUME 184. Avidin-Biotin Technology
Edited by MEIR WILCHEK AND EDWARD A. BAYER
VOLUME 185. Gene Expression Technology
Edited by DAVID V. GOEDDEL
VOLUME 186. Oxygen Radicals in Biological Systems (Part B: Oxygen Radicals and
Antioxidants)
Edited by LESTER PACKER AND ALEXANDER N. GLAZER
VOLUME 187. Arachidonate Related Lipid Mediators
Edited by ROBERT C. MURPHY AND FRANK A. FITZPATRICK
VOLUME 188. Hydrocarbons and Methylotrophy
Edited by MARY E. LIDSTROM
VOLUME 189. Retinoids (Part A: Molecular and Metabolic Aspects)
Edited by LESTER PACKER
xxxviii Methods in Enzymology

VOLUME 190. Retinoids (Part B: Cell Differentiation and Clinical Applications)

Edited by LESTER PACKER
VOLUME 191. Biomembranes (Part V: Cellular and Subcellular Transport:
Epithelial Cells)
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
VOLUME 192. Biomembranes (Part W: Cellular and Subcellular Transport:
Epithelial Cells)
Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER
VOLUME 193. Mass Spectrometry
Edited by JAMES A. MCCLOSKEY
VOLUME 194. Guide to Yeast Genetics and Molecular Biology
Edited by CHRISTINE GUTHRIE AND GERALD R. FINK
VOLUME 195. Adenylyl Cyclase, G Proteins, and Guanylyl Cyclase
Edited by ROGER A. JOHNSON AND JACKIE D. CORBIN
VOLUME 196. Molecular Motors and the Cytoskeleton
Edited by RICHARD B. VALLEE
VOLUME 197. Phospholipases
Edited by EDWARD A. DENNIS
VOLUME 198. Peptide Growth Factors (Part C)
Edited by DAVID BARNES, J. P. MATHER, AND GORDON H. SATO
VOLUME 199. Cumulative Subject Index Volumes 168–174, 176–194
VOLUME 200. Protein Phosphorylation (Part A: Protein Kinases: Assays,
Purification, Antibodies, Functional Analysis, Cloning, and Expression)
Edited by TONY HUNTER AND BARTHOLOMEW M. SEFTON
VOLUME 201. Protein Phosphorylation (Part B: Analysis of Protein
Phosphorylation, Protein Kinase Inhibitors, and Protein Phosphatases)
Edited by TONY HUNTER AND BARTHOLOMEW M. SEFTON
VOLUME 202. Molecular Design and Modeling: Concepts and Applications
(Part A: Proteins, Peptides, and Enzymes)
Edited by JOHN J. LANGONE
VOLUME 203. Molecular Design and Modeling: Concepts and Applications
(Part B: Antibodies and Antigens, Nucleic Acids, Polysaccharides, and Drugs)
Edited by JOHN J. LANGONE
VOLUME 204. Bacterial Genetic Systems
Edited by JEFFREY H. MILLER
VOLUME 205. Metallobiochemistry (Part B: Metallothionein and
Related Molecules)
Edited by JAMES F. RIORDAN AND BERT L. VALLEE
Methods in Enzymology xxxix

VOLUME 206. Cytochrome P450

Edited by MICHAEL R. WATERMAN AND ERIC F. JOHNSON
VOLUME 207. Ion Channels
Edited by BERNARDO RUDY AND LINDA E. IVERSON
VOLUME 208. Protein–DNA Interactions
Edited by ROBERT T. SAUER
VOLUME 209. Phospholipid Biosynthesis
Edited by EDWARD A. DENNIS AND DENNIS E. VANCE
VOLUME 210. Numerical Computer Methods
Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON
VOLUME 211. DNA Structures (Part A: Synthesis and Physical Analysis of DNA)
Edited by DAVID M. J. LILLEY AND JAMES E. DAHLBERG
VOLUME 212. DNA Structures (Part B: Chemical and Electrophoretic
Analysis of DNA)
Edited by DAVID M. J. LILLEY AND JAMES E. DAHLBERG
VOLUME 213. Carotenoids (Part A: Chemistry, Separation, Quantitation,
and Antioxidation)
Edited by LESTER PACKER
VOLUME 214. Carotenoids (Part B: Metabolism, Genetics, and Biosynthesis)
Edited by LESTER PACKER
VOLUME 215. Platelets: Receptors, Adhesion, Secretion (Part B)
Edited by JACEK J. HAWIGER
VOLUME 216. Recombinant DNA (Part G)
Edited by RAY WU
VOLUME 217. Recombinant DNA (Part H)
Edited by RAY WU
VOLUME 218. Recombinant DNA (Part I)
Edited by RAY WU
VOLUME 219. Reconstitution of Intracellular Transport
Edited by JAMES E. ROTHMAN
VOLUME 220. Membrane Fusion Techniques (Part A)
Edited by NEJAT DÜZGÜNES,
VOLUME 221. Membrane Fusion Techniques (Part B)
Edited by NEJAT DÜZGÜNES,
VOLUME 222. Proteolytic Enzymes in Coagulation, Fibrinolysis, and Complement
Activation (Part A: Mammalian Blood Coagulation Factors and Inhibitors)
Edited by LASZLO LORAND AND KENNETH G. MANN
xl Methods in Enzymology

VOLUME 223. Proteolytic Enzymes in Coagulation, Fibrinolysis, and Complement

Activation (Part B: Complement Activation, Fibrinolysis, and Nonmammalian
Blood Coagulation Factors)
Edited by LASZLO LORAND AND KENNETH G. MANN
VOLUME 224. Molecular Evolution: Producing the Biochemical Data
Edited by ELIZABETH ANNE ZIMMER, THOMAS J. WHITE, REBECCA L. CANN,
AND ALLAN C. WILSON

VOLUME 225. Guide to Techniques in Mouse Development

Edited by PAUL M. WASSARMAN AND MELVIN L. DEPAMPHILIS
VOLUME 226. Metallobiochemistry (Part C: Spectroscopic and Physical Methods
for Probing Metal Ion Environments in Metalloenzymes and Metalloproteins)
Edited by JAMES F. RIORDAN AND BERT L. VALLEE
VOLUME 227. Metallobiochemistry (Part D: Physical and Spectroscopic Methods
for Probing Metal Ion Environments in Metalloproteins)
Edited by JAMES F. RIORDAN AND BERT L. VALLEE
VOLUME 228. Aqueous Two-Phase Systems
Edited by HARRY WALTER AND GÖTE JOHANSSON
VOLUME 229. Cumulative Subject Index Volumes 195–198, 200–227
VOLUME 230. Guide to Techniques in Glycobiology
Edited by WILLIAM J. LENNARZ AND GERALD W. HART
VOLUME 231. Hemoglobins (Part B: Biochemical and Analytical Methods)
Edited by JOHANNES EVERSE, KIM D. VANDEGRIFF, AND ROBERT M. WINSLOW
VOLUME 232. Hemoglobins (Part C: Biophysical Methods)
Edited by JOHANNES EVERSE, KIM D. VANDEGRIFF, AND ROBERT M. WINSLOW
VOLUME 233. Oxygen Radicals in Biological Systems (Part C)
Edited by LESTER PACKER
VOLUME 234. Oxygen Radicals in Biological Systems (Part D)
Edited by LESTER PACKER
VOLUME 235. Bacterial Pathogenesis (Part A: Identification and Regulation of
Virulence Factors)
Edited by VIRGINIA L. CLARK AND PATRIK M. BAVOIL
VOLUME 236. Bacterial Pathogenesis (Part B: Integration of Pathogenic Bacteria
with Host Cells)
Edited by VIRGINIA L. CLARK AND PATRIK M. BAVOIL
VOLUME 237. Heterotrimeric G Proteins
Edited by RAVI IYENGAR
VOLUME 238. Heterotrimeric G-Protein Effectors
Edited by RAVI IYENGAR
Methods in Enzymology xli

VOLUME 239. Nuclear Magnetic Resonance (Part C)

Edited by THOMAS L. JAMES AND NORMAN J. OPPENHEIMER
VOLUME 240. Numerical Computer Methods (Part B)
Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND
VOLUME 241. Retroviral Proteases
Edited by LAWRENCE C. KUO AND JULES A. SHAFER
VOLUME 242. Neoglycoconjugates (Part A)
Edited by Y. C. LEE AND REIKO T. LEE
VOLUME 243. Inorganic Microbial Sulfur Metabolism
Edited by HARRY D. PECK, JR., AND JEAN LEGALL
VOLUME 244. Proteolytic Enzymes: Serine and Cysteine Peptidases
Edited by ALAN J. BARRETT
VOLUME 245. Extracellular Matrix Components
Edited by E. RUOSLAHTI AND E. ENGVALL
VOLUME 246. Biochemical Spectroscopy
Edited by KENNETH SAUER
VOLUME 247. Neoglycoconjugates (Part B: Biomedical Applications)
Edited by Y. C. LEE AND REIKO T. LEE
VOLUME 248. Proteolytic Enzymes: Aspartic and Metallo Peptidases
Edited by ALAN J. BARRETT
VOLUME 249. Enzyme Kinetics and Mechanism (Part D: Developments in
Enzyme Dynamics)
Edited by DANIEL L. PURICH
VOLUME 250. Lipid Modifications of Proteins
Edited by PATRICK J. CASEY AND JANICE E. BUSS
VOLUME 251. Biothiols (Part A: Monothiols and Dithiols, Protein Thiols, and
Thiyl Radicals)
Edited by LESTER PACKER
VOLUME 252. Biothiols (Part B: Glutathione and Thioredoxin; Thiols in Signal
Transduction and Gene Regulation)
Edited by LESTER PACKER
VOLUME 253. Adhesion of Microbial Pathogens
Edited by RON J. DOYLE AND ITZHAK OFEK
VOLUME 254. Oncogene Techniques
Edited by PETER K. VOGT AND INDER M. VERMA
VOLUME 255. Small GTPases and Their Regulators (Part A: Ras Family)
Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL
VOLUME 256. Small GTPases and Their Regulators (Part B: Rho Family)
Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL
xlii Methods in Enzymology

VOLUME 257. Small GTPases and Their Regulators (Part C: Proteins Involved
in Transport)
Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL
VOLUME 258. Redox-Active Amino Acids in Biology
Edited by JUDITH P. KLINMAN
VOLUME 259. Energetics of Biological Macromolecules
Edited by MICHAEL L. JOHNSON AND GARY K. ACKERS
VOLUME 260. Mitochondrial Biogenesis and Genetics (Part A)
Edited by GIUSEPPE M. ATTARDI AND ANNE CHOMYN
VOLUME 261. Nuclear Magnetic Resonance and Nucleic Acids
Edited by THOMAS L. JAMES
VOLUME 262. DNA Replication
Edited by JUDITH L. CAMPBELL
VOLUME 263. Plasma Lipoproteins (Part C: Quantitation)
Edited by WILLIAM A. BRADLEY, SANDRA H. GIANTURCO, AND JERE P. SEGREST
VOLUME 264. Mitochondrial Biogenesis and Genetics (Part B)
Edited by GIUSEPPE M. ATTARDI AND ANNE CHOMYN
VOLUME 265. Cumulative Subject Index Volumes 228, 230–262
VOLUME 266. Computer Methods for Macromolecular Sequence Analysis
Edited by RUSSELL F. DOOLITTLE
VOLUME 267. Combinatorial Chemistry
Edited by JOHN N. ABELSON
VOLUME 268. Nitric Oxide (Part A: Sources and Detection of NO; NO Synthase)
Edited by LESTER PACKER
VOLUME 269. Nitric Oxide (Part B: Physiological and Pathological Processes)
Edited by LESTER PACKER
VOLUME 270. High Resolution Separation and Analysis of Biological
Macromolecules (Part A: Fundamentals)
Edited by BARRY L. KARGER AND WILLIAM S. HANCOCK
VOLUME 271. High Resolution Separation and Analysis of Biological
Macromolecules (Part B: Applications)
Edited by BARRY L. KARGER AND WILLIAM S. HANCOCK
VOLUME 272. Cytochrome P450 (Part B)
Edited by ERIC F. JOHNSON AND MICHAEL R. WATERMAN
VOLUME 273. RNA Polymerase and Associated Factors (Part A)
Edited by SANKAR ADHYA
VOLUME 274. RNA Polymerase and Associated Factors (Part B)
Edited by SANKAR ADHYA
Methods in Enzymology xliii

VOLUME 275. Viral Polymerases and Related Proteins

Edited by LAWRENCE C. KUO, DAVID B. OLSEN, AND STEVEN S. CARROLL
VOLUME 276. Macromolecular Crystallography (Part A)
Edited by CHARLES W. CARTER, JR., AND ROBERT M. SWEET
VOLUME 277. Macromolecular Crystallography (Part B)
Edited by CHARLES W. CARTER, JR., AND ROBERT M. SWEET
VOLUME 278. Fluorescence Spectroscopy
Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON
VOLUME 279. Vitamins and Coenzymes (Part I)
Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER
VOLUME 280. Vitamins and Coenzymes (Part J)
Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER
VOLUME 281. Vitamins and Coenzymes (Part K)
Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER
VOLUME 282. Vitamins and Coenzymes (Part L)
Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER
VOLUME 283. Cell Cycle Control
Edited by WILLIAM G. DUNPHY
VOLUME 284. Lipases (Part A: Biotechnology)
Edited by BYRON RUBIN AND EDWARD A. DENNIS
VOLUME 285. Cumulative Subject Index Volumes 263, 264, 266–284, 286–289
VOLUME 286. Lipases (Part B: Enzyme Characterization and Utilization)
Edited by BYRON RUBIN AND EDWARD A. DENNIS
VOLUME 287. Chemokines
Edited by RICHARD HORUK
VOLUME 288. Chemokine Receptors
Edited by RICHARD HORUK
VOLUME 289. Solid Phase Peptide Synthesis
Edited by GREGG B. FIELDS
VOLUME 290. Molecular Chaperones
Edited by GEORGE H. LORIMER AND THOMAS BALDWIN
VOLUME 291. Caged Compounds
Edited by GERARD MARRIOTT
VOLUME 292. ABC Transporters: Biochemical, Cellular, and Molecular Aspects
Edited by SURESH V. AMBUDKAR AND MICHAEL M. GOTTESMAN
VOLUME 293. Ion Channels (Part B)
Edited by P. MICHAEL CONN
xliv Methods in Enzymology

VOLUME 294. Ion Channels (Part C)

Edited by P. MICHAEL CONN
VOLUME 295. Energetics of Biological Macromolecules (Part B)
Edited by GARY K. ACKERS AND MICHAEL L. JOHNSON
VOLUME 296. Neurotransmitter Transporters
Edited by SUSAN G. AMARA
VOLUME 297. Photosynthesis: Molecular Biology of Energy Capture
Edited by LEE MCINTOSH
VOLUME 298. Molecular Motors and the Cytoskeleton (Part B)
Edited by RICHARD B. VALLEE
VOLUME 299. Oxidants and Antioxidants (Part A)
Edited by LESTER PACKER
VOLUME 300. Oxidants and Antioxidants (Part B)
Edited by LESTER PACKER
VOLUME 301. Nitric Oxide: Biological and Antioxidant Activities (Part C)
Edited by LESTER PACKER
VOLUME 302. Green Fluorescent Protein
Edited by P. MICHAEL CONN
VOLUME 303. cDNA Preparation and Display
Edited by SHERMAN M. WEISSMAN
VOLUME 304. Chromatin
Edited by PAUL M. WASSARMAN AND ALAN P. WOLFFE
VOLUME 305. Bioluminescence and Chemiluminescence (Part C)
Edited by THOMAS O. BALDWIN AND MIRIAM M. ZIEGLER
VOLUME 306. Expression of Recombinant Genes in Eukaryotic Systems
Edited by JOSEPH C. GLORIOSO AND MARTIN C. SCHMIDT
VOLUME 307. Confocal Microscopy
Edited by P. MICHAEL CONN
VOLUME 308. Enzyme Kinetics and Mechanism (Part E: Energetics of
Enzyme Catalysis)
Edited by DANIEL L. PURICH AND VERN L. SCHRAMM
VOLUME 309. Amyloid, Prions, and Other Protein Aggregates
Edited by RONALD WETZEL
VOLUME 310. Biofilms
Edited by RON J. DOYLE
VOLUME 311. Sphingolipid Metabolism and Cell Signaling (Part A)
Edited by ALFRED H. MERRILL, JR., AND YUSUF A. HANNUN
Methods in Enzymology xlv

VOLUME 312. Sphingolipid Metabolism and Cell Signaling (Part B)

Edited by ALFRED H. MERRILL, JR., AND YUSUF A. HANNUN
VOLUME 313. Antisense Technology (Part A: General Methods, Methods of
Delivery, and RNA Studies)
Edited by M. IAN PHILLIPS
VOLUME 314. Antisense Technology (Part B: Applications)
Edited by M. IAN PHILLIPS
VOLUME 315. Vertebrate Phototransduction and the Visual Cycle (Part A)
Edited by KRZYSZTOF PALCZEWSKI
VOLUME 316. Vertebrate Phototransduction and the Visual Cycle (Part B)
Edited by KRZYSZTOF PALCZEWSKI
VOLUME 317. RNA–Ligand Interactions (Part A: Structural Biology Methods)
Edited by DANIEL W. CELANDER AND JOHN N. ABELSON
VOLUME 318. RNA–Ligand Interactions (Part B: Molecular Biology Methods)
Edited by DANIEL W. CELANDER AND JOHN N. ABELSON
VOLUME 319. Singlet Oxygen, UV-A, and Ozone
Edited by LESTER PACKER AND HELMUT SIES
VOLUME 320. Cumulative Subject Index Volumes 290–319
VOLUME 321. Numerical Computer Methods (Part C)
Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND
VOLUME 322. Apoptosis
Edited by JOHN C. REED
VOLUME 323. Energetics of Biological Macromolecules (Part C)
Edited by MICHAEL L. JOHNSON AND GARY K. ACKERS
VOLUME 324. Branched-Chain Amino Acids (Part B)
Edited by ROBERT A. HARRIS AND JOHN R. SOKATCH
VOLUME 325. Regulators and Effectors of Small GTPases (Part D: Rho Family)
Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL
VOLUME 326. Applications of Chimeric Genes and Hybrid Proteins (Part A: Gene
Expression and Protein Purification)
Edited by JEREMY THORNER, SCOTT D. EMR, AND JOHN N. ABELSON
VOLUME 327. Applications of Chimeric Genes and Hybrid Proteins (Part B: Cell
Biology and Physiology)
Edited by JEREMY THORNER, SCOTT D. EMR, AND JOHN N. ABELSON
VOLUME 328. Applications of Chimeric Genes and Hybrid Proteins (Part C:
Protein–Protein Interactions and Genomics)
Edited by JEREMY THORNER, SCOTT D. EMR, AND JOHN N. ABELSON
xlvi Methods in Enzymology

VOLUME 329. Regulators and Effectors of Small GTPases (Part E: GTPases

Involved in Vesicular Traffic)
Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL
VOLUME 330. Hyperthermophilic Enzymes (Part A)
Edited by MICHAEL W. W. ADAMS AND ROBERT M. KELLY
VOLUME 331. Hyperthermophilic Enzymes (Part B)
Edited by MICHAEL W. W. ADAMS AND ROBERT M. KELLY
VOLUME 332. Regulators and Effectors of Small GTPases (Part F: Ras Family I)
Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL
VOLUME 333. Regulators and Effectors of Small GTPases (Part G: Ras Family II)
Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL
VOLUME 334. Hyperthermophilic Enzymes (Part C)
Edited by MICHAEL W. W. ADAMS AND ROBERT M. KELLY
VOLUME 335. Flavonoids and Other Polyphenols
Edited by LESTER PACKER
VOLUME 336. Microbial Growth in Biofilms (Part A: Developmental and
Molecular Biological Aspects)
Edited by RON J. DOYLE
VOLUME 337. Microbial Growth in Biofilms (Part B: Special Environments and
Physicochemical Aspects)
Edited by RON J. DOYLE
VOLUME 338. Nuclear Magnetic Resonance of Biological Macromolecules (Part A)
Edited by THOMAS L. JAMES, VOLKER DÖTSCH, AND ULI SCHMITZ
VOLUME 339. Nuclear Magnetic Resonance of Biological Macromolecules (Part B)
Edited by THOMAS L. JAMES, VOLKER DÖTSCH, AND ULI SCHMITZ
VOLUME 340. Drug–Nucleic Acid Interactions
Edited by JONATHAN B. CHAIRES AND MICHAEL J. WARING
VOLUME 341. Ribonucleases (Part A)
Edited by ALLEN W. NICHOLSON
VOLUME 342. Ribonucleases (Part B)
Edited by ALLEN W. NICHOLSON
VOLUME 343. G Protein Pathways (Part A: Receptors)
Edited by RAVI IYENGAR AND JOHN D. HILDEBRANDT
VOLUME 344. G Protein Pathways (Part B: G Proteins and Their Regulators)
Edited by RAVI IYENGAR AND JOHN D. HILDEBRANDT
VOLUME 345. G Protein Pathways (Part C: Effector Mechanisms)
Edited by RAVI IYENGAR AND JOHN D. HILDEBRANDT
Methods in Enzymology xlvii

VOLUME 346. Gene Therapy Methods

Edited by M. IAN PHILLIPS
VOLUME 347. Protein Sensors and Reactive Oxygen Species (Part A:
Selenoproteins and Thioredoxin)
Edited by HELMUT SIES AND LESTER PACKER
VOLUME 348. Protein Sensors and Reactive Oxygen Species (Part B: Thiol
Enzymes and Proteins)
Edited by HELMUT SIES AND LESTER PACKER
VOLUME 349. Superoxide Dismutase
Edited by LESTER PACKER
VOLUME 350. Guide to Yeast Genetics and Molecular and Cell Biology (Part B)
Edited by CHRISTINE GUTHRIE AND GERALD R. FINK
VOLUME 351. Guide to Yeast Genetics and Molecular and Cell Biology (Part C)
Edited by CHRISTINE GUTHRIE AND GERALD R. FINK
VOLUME 352. Redox Cell Biology and Genetics (Part A)
Edited by CHANDAN K. SEN AND LESTER PACKER
VOLUME 353. Redox Cell Biology and Genetics (Part B)
Edited by CHANDAN K. SEN AND LESTER PACKER
VOLUME 354. Enzyme Kinetics and Mechanisms (Part F: Detection and
Characterization of Enzyme Reaction Intermediates)
Edited by DANIEL L. PURICH
VOLUME 355. Cumulative Subject Index Volumes 321–354
VOLUME 356. Laser Capture Microscopy and Microdissection
Edited by P. MICHAEL CONN
VOLUME 357. Cytochrome P450, Part C
Edited by ERIC F. JOHNSON AND MICHAEL R. WATERMAN
VOLUME 358. Bacterial Pathogenesis (Part C: Identification, Regulation, and
Function of Virulence Factors)
Edited by VIRGINIA L. CLARK AND PATRIK M. BAVOIL
VOLUME 359. Nitric Oxide (Part D)
Edited by ENRIQUE CADENAS AND LESTER PACKER
VOLUME 360. Biophotonics (Part A)
Edited by GERARD MARRIOTT AND IAN PARKER
VOLUME 361. Biophotonics (Part B)
Edited by GERARD MARRIOTT AND IAN PARKER
VOLUME 362. Recognition of Carbohydrates in Biological Systems (Part A)
Edited by YUAN C. LEE AND REIKO T. LEE
xlviii Methods in Enzymology

VOLUME 363. Recognition of Carbohydrates in Biological Systems (Part B)

Edited by YUAN C. LEE AND REIKO T. LEE
VOLUME 364. Nuclear Receptors
Edited by DAVID W. RUSSELL AND DAVID J. MANGELSDORF
VOLUME 365. Differentiation of Embryonic Stem Cells
Edited by PAUL M. WASSAUMAN AND GORDON M. KELLER
VOLUME 366. Protein Phosphatases
Edited by SUSANNE KLUMPP AND JOSEF KRIEGLSTEIN
VOLUME 367. Liposomes (Part A)
Edited by NEJAT DÜZGÜNES,
VOLUME 368. Macromolecular Crystallography (Part C)
Edited by CHARLES W. CARTER, JR., AND ROBERT M. SWEET
VOLUME 369. Combinational Chemistry (Part B)
Edited by GUILLERMO A. MORALES AND BARRY A. BUNIN
VOLUME 370. RNA Polymerases and Associated Factors (Part C)
Edited by SANKAR L. ADHYA AND SUSAN GARGES
VOLUME 371. RNA Polymerases and Associated Factors (Part D)
Edited by SANKAR L. ADHYA AND SUSAN GARGES
VOLUME 372. Liposomes (Part B)
Edited by NEJAT DÜZGÜNES,
VOLUME 373. Liposomes (Part C)
Edited by NEJAT DÜZGÜNES,
VOLUME 374. Macromolecular Crystallography (Part D)
Edited by CHARLES W. CARTER, JR., AND ROBERT W. SWEET
VOLUME 375. Chromatin and Chromatin Remodeling Enzymes (Part A)
Edited by C. DAVID ALLIS AND CARL WU
VOLUME 376. Chromatin and Chromatin Remodeling Enzymes (Part B)
Edited by C. DAVID ALLIS AND CARL WU
VOLUME 377. Chromatin and Chromatin Remodeling Enzymes (Part C)
Edited by C. DAVID ALLIS AND CARL WU
VOLUME 378. Quinones and Quinone Enzymes (Part A)
Edited by HELMUT SIES AND LESTER PACKER
VOLUME 379. Energetics of Biological Macromolecules (Part D)
Edited by JO M. HOLT, MICHAEL L. JOHNSON, AND GARY K. ACKERS
VOLUME 380. Energetics of Biological Macromolecules (Part E)
Edited by JO M. HOLT, MICHAEL L. JOHNSON, AND GARY K. ACKERS
VOLUME 381. Oxygen Sensing
Edited by CHANDAN K. SEN AND GREGG L. SEMENZA
Methods in Enzymology xlix

VOLUME 382. Quinones and Quinone Enzymes (Part B)

Edited by HELMUT SIES AND LESTER PACKER
VOLUME 383. Numerical Computer Methods (Part D)
Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON
VOLUME 384. Numerical Computer Methods (Part E)
Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON
VOLUME 385. Imaging in Biological Research (Part A)
Edited by P. MICHAEL CONN
VOLUME 386. Imaging in Biological Research (Part B)
Edited by P. MICHAEL CONN
VOLUME 387. Liposomes (Part D)
Edited by NEJAT DÜZGÜNES,
VOLUME 388. Protein Engineering
Edited by DAN E. ROBERTSON AND JOSEPH P. NOEL
VOLUME 389. Regulators of G-Protein Signaling (Part A)
Edited by DAVID P. SIDEROVSKI
VOLUME 390. Regulators of G-Protein Signaling (Part B)
Edited by DAVID P. SIDEROVSKI
VOLUME 391. Liposomes (Part E)
Edited by NEJAT DÜZGÜNES,
VOLUME 392. RNA Interference
Edited by ENGELKE ROSSI
VOLUME 393. Circadian Rhythms
Edited by MICHAEL W. YOUNG
VOLUME 394. Nuclear Magnetic Resonance of Biological Macromolecules (Part C)
Edited by THOMAS L. JAMES
VOLUME 395. Producing the Biochemical Data (Part B)
Edited by ELIZABETH A. ZIMMER AND ERIC H. ROALSON
VOLUME 396. Nitric Oxide (Part E)
Edited by LESTER PACKER AND ENRIQUE CADENAS
VOLUME 397. Environmental Microbiology
Edited by JARED R. LEADBETTER
VOLUME 398. Ubiquitin and Protein Degradation (Part A)
Edited by RAYMOND J. DESHAIES
VOLUME 399. Ubiquitin and Protein Degradation (Part B)
Edited by RAYMOND J. DESHAIES
VOLUME 400. Phase II Conjugation Enzymes and Transport Systems
Edited by HELMUT SIES AND LESTER PACKER
l Methods in Enzymology

VOLUME 401. Glutathione Transferases and Gamma Glutamyl Transpeptidases

Edited by HELMUT SIES AND LESTER PACKER
VOLUME 402. Biological Mass Spectrometry
Edited by A. L. BURLINGAME
VOLUME 403. GTPases Regulating Membrane Targeting and Fusion
Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL
VOLUME 404. GTPases Regulating Membrane Dynamics
Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL
VOLUME 405. Mass Spectrometry: Modified Proteins and Glycoconjugates
Edited by A. L. BURLINGAME
VOLUME 406. Regulators and Effectors of Small GTPases: Rho Family
Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL
VOLUME 407. Regulators and Effectors of Small GTPases: Ras Family
Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL
VOLUME 408. DNA Repair (Part A)
Edited by JUDITH L. CAMPBELL AND PAUL MODRICH
VOLUME 409. DNA Repair (Part B)
Edited by JUDITH L. CAMPBELL AND PAUL MODRICH
VOLUME 410. DNA Microarrays (Part A: Array Platforms and Web-Bench
Protocols)
Edited by ALAN KIMMEL AND BRIAN OLIVER
VOLUME 411. DNA Microarrays (Part B: Databases and Statistics)
Edited by ALAN KIMMEL AND BRIAN OLIVER
VOLUME 412. Amyloid, Prions, and Other Protein Aggregates (Part B)
Edited by INDU KHETERPAL AND RONALD WETZEL
VOLUME 413. Amyloid, Prions, and Other Protein Aggregates (Part C)
Edited by INDU KHETERPAL AND RONALD WETZEL
VOLUME 414. Measuring Biological Responses with Automated Microscopy
Edited by JAMES INGLESE
VOLUME 415. Glycobiology
Edited by MINORU FUKUDA
VOLUME 416. Glycomics
Edited by MINORU FUKUDA
VOLUME 417. Functional Glycomics
Edited by MINORU FUKUDA
VOLUME 418. Embryonic Stem Cells
Edited by IRINA KLIMANSKAYA AND ROBERT LANZA
Methods in Enzymology li

VOLUME 419. Adult Stem Cells

Edited by IRINA KLIMANSKAYA AND ROBERT LANZA
VOLUME 420. Stem Cell Tools and Other Experimental Protocols
Edited by IRINA KLIMANSKAYA AND ROBERT LANZA
VOLUME 421. Advanced Bacterial Genetics: Use of Transposons and Phage for
Genomic Engineering
Edited by KELLY T. HUGHES
VOLUME 422. Two-Component Signaling Systems, Part A
Edited by MELVIN I. SIMON, BRIAN R. CRANE, AND ALEXANDRINE CRANE
VOLUME 423. Two-Component Signaling Systems, Part B
Edited by MELVIN I. SIMON, BRIAN R. CRANE, AND ALEXANDRINE CRANE
VOLUME 424. RNA Editing
Edited by JONATHA M. GOTT
VOLUME 425. RNA Modification
Edited by JONATHA M. GOTT
VOLUME 426. Integrins
Edited by DAVID CHERESH
VOLUME 427. MicroRNA Methods
Edited by JOHN J. ROSSI
VOLUME 428. Osmosensing and Osmosignaling
Edited by HELMUT SIES AND DIETER HAUSSINGER
VOLUME 429. Translation Initiation: Extract Systems and Molecular Genetics
Edited by JON LORSCH
VOLUME 430. Translation Initiation: Reconstituted Systems and Biophysical
Methods
Edited by JON LORSCH
VOLUME 431. Translation Initiation: Cell Biology, High-Throughput and
Chemical-Based Approaches
Edited by JON LORSCH
VOLUME 432. Lipidomics and Bioactive Lipids: Mass-Spectrometry–Based Lipid
Analysis
Edited by H. ALEX BROWN
VOLUME 433. Lipidomics and Bioactive Lipids: Specialized Analytical Methods and
Lipids in Disease
Edited by H. ALEX BROWN
VOLUME 434. Lipidomics and Bioactive Lipids: Lipids and Cell Signaling
Edited by H. ALEX BROWN
VOLUME 435. Oxygen Biology and Hypoxia
Edited by HELMUT SIES AND BERNHARD BRÜNE
lii Methods in Enzymology

VOLUME 436. Globins and Other Nitric Oxide-Reactive Protiens (Part A)

Edited by ROBERT K. POOLE
VOLUME 437. Globins and Other Nitric Oxide-Reactive Protiens (Part B)
Edited by ROBERT K. POOLE
VOLUME 438. Small GTPases in Disease (Part A)
Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL
VOLUME 439. Small GTPases in Disease (Part B)
Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL
VOLUME 440. Nitric Oxide, Part F Oxidative and Nitrosative Stress in Redox
Regulation of Cell Signaling
Edited by ENRIQUE CADENAS AND LESTER PACKER
VOLUME 441. Nitric Oxide, Part G Oxidative and Nitrosative Stress in Redox
Regulation of Cell Signaling
Edited by ENRIQUE CADENAS AND LESTER PACKER
VOLUME 442. Programmed Cell Death, General Principles for Studying Cell
Death (Part A)
Edited by ROYA KHOSRAVI-FAR, ZAHRA ZAKERI, RICHARD A. LOCKSHIN,
AND MAURO PIACENTINI

VOLUME 443. Angiogenesis: In Vitro Systems

Edited by DAVID A. CHERESH
VOLUME 444. Angiogenesis: In Vivo Systems (Part A)
Edited by DAVID A. CHERESH
VOLUME 445. Angiogenesis: In Vivo Systems (Part B)
Edited by DAVID A. CHERESH
VOLUME 446. Programmed Cell Death, The Biology and Therapeutic
Implications of Cell Death (Part B)
Edited by ROYA KHOSRAVI-FAR, ZAHRA ZAKERI, RICHARD A. LOCKSHIN,
AND MAURO PIACENTINI

VOLUME 447. RNA Turnover in Bacteria, Archaea and Organelles

Edited by LYNNE E. MAQUAT AND CECILIA M. ARRAIANO
VOLUME 448. RNA Turnover in Eukaryotes: Nucleases, Pathways
and Analysis of mRNA Decay
Edited by LYNNE E. MAQUAT AND MEGERDITCH KILEDJIAN
VOLUME 449. RNA Turnover in Eukaryotes: Analysis of Specialized and Quality
Control RNA Decay Pathways
Edited by LYNNE E. MAQUAT AND MEGERDITCH KILEDJIAN
VOLUME 450. Fluorescence Spectroscopy
Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON
Methods in Enzymology liii

VOLUME 451. Autophagy: Lower Eukaryotes and Non-Mammalian Systems (Part A)

Edited by DANIEL J. KLIONSKY
VOLUME 452. Autophagy in Mammalian Systems (Part B)
Edited by DANIEL J. KLIONSKY
VOLUME 453. Autophagy in Disease and Clinical Applications (Part C)
Edited by DANIEL J. KLIONSKY
VOLUME 454. Computer Methods (Part A)
Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND
VOLUME 455. Biothermodynamics (Part A)
Edited by MICHAEL L. JOHNSON, JO M. HOLT, AND GARY K. ACKERS (RETIRED)
VOLUME 456. Mitochondrial Function, Part A: Mitochondrial Electron Transport
Complexes and Reactive Oxygen Species
Edited by WILLIAM S. ALLISON AND IMMO E. SCHEFFLER
VOLUME 457. Mitochondrial Function, Part B: Mitochondrial Protein Kinases,
Protein Phosphatases and Mitochondrial Diseases
Edited by WILLIAM S. ALLISON AND ANNE N. MURPHY
VOLUME 458. Complex Enzymes in Microbial Natural Product Biosynthesis,
Part A: Overview Articles and Peptides
Edited by DAVID A. HOPWOOD
VOLUME 459. Complex Enzymes in Microbial Natural Product Biosynthesis,
Part B: Polyketides, Aminocoumarins and Carbohydrates
Edited by DAVID A. HOPWOOD
VOLUME 460. Chemokines, Part A
Edited by TRACY M. HANDEL AND DAMON J. HAMEL
VOLUME 461. Chemokines, Part B
Edited by TRACY M. HANDEL AND DAMON J. HAMEL
VOLUME 462. Non-Natural Amino Acids
Edited by TOM W. MUIR AND JOHN N. ABELSON
VOLUME 463. Guide to Protein Purification, 2nd Edition
Edited by RICHARD R. BURGESS AND MURRAY P. DEUTSCHER
PREFACE TO CHAPTER 1

It is with great admiration that we reprint in its entirety Chapter 1 of the first
edition of this volume, ‘‘Why Purify Enzymes’’ by the late Arthur Kornberg.
Arthur was one of the giants of 20th century biochemistry and his forte was
enzymology. Upon reading the chapter, one can appreciate why this was so.
His passion for identifying an enzyme activity and purifying it to homo-
geneity so that many of its properties could be directly studied become clearly
evident from Arthur’s words. Yet, he was clearly mindful of what he called
the enzyme’s ‘‘social face,’’ its interactions with other cellular components,
an attribute that has gained more prominence as we have come to understand
the importance of cell organization.
There have been many advances in the 20 years since the chapter was
written, particularly in the ease with which a protein can be purified if its
gene is known. Nevertheless, the admonition to ‘‘purify, purify, purify’’
remains relevant when one is studying the catalytic properties of an enzyme,
elucidating the structure of a protein, or identifying possible regulatory
factors. Protein purification remains of prime importance if one endeavors
to understand the enzyme responsible for a newly discovered cellular
reaction or the proteins involved in a cellular process. With this in mind,
the lessons to be learned from reading, or re-reading, Arthur’s wonderfully
lucid and pertinent chapter will be extremely rewarding.

MURRAY P. DEUTSCHER

1
 C H A P T E R O N E

Why Purify Enzymes?

Arthur Kornbergw

‘‘Don’t waste clean thinking on dirty enzymes’’ is an admonition of

Efraim Racker’s which is at the core of enzymology and good chemical
practice. It simply says that detailed studies of how an enzyme catalyzes the
conversion of one substance to another is generally a waste of time until the
enzyme has been purified away from the other enzymes and substances that
make up a crude cell extract. The mixture of thousands of different enzymes
released from a disrupted liver, yeast, or bacterial cell likely contains several
that direct other rearrangement of the starting material and the product of
the particular enzyme’s action. Only when we have purified the enzyme to
the point that no other enzymes can be detected can we fell assured that
a single type of enzyme molecule directs the conversion of substance A
to substance B, and does nothing more. Only then can we learn how the
enzyme does its work.
The rewards for the labor of purifying an enzyme were laid out in a series
of inspirational papers by Otto Warburg in the 1930s. From his laboratory in
Berlin-Dahlem came the discipline and many of the methods of purifying
enzymes and with those the clarification of key reactions and vitamin func-
tions in respiration and the fermentation of glucose. Warburg’s contributions
strengthened the classic approach to enzymology inaugurated with Eduard
Büchner’s accidental discovery, at the turn of this century, of cell-free
conversion of sucrose to ethanol. One tracks the molecular basis of cellular
function—alcoholic fermentation in yeast, glycolysis in muscle, lumines-
cence in a fly, or the replication of DNA—by first observing the phenome-
non in a cell-free system. Then one isolates the responsible enzyme
(or enzymes) by fractionation of the cell extract and purifies it to homogene-
ity. Then one hopes to learn enough about the structure of the enzyme to
explain how it performs its catalytic functions, responds to regulatory signals,
and is associated with other enzymes and structures in the cell.
By a reverse approach, call it neoclassical, especially popular in recent
decades, one first obtains a structure and then looks for its function. The
protein is preferably small and stable, and has been amplified by cloning or is
commercially available. By intensive study of the protein and homologous
proteins, one hopes to get some clues to how it functions. As the popularity
w
Deceased

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63001-9 All rights reserved.

3
4 Arthur Kornberg

of the neoclassical approach has increased, so has there been a corresponding

decrease in interest in the classical route: pursuit of a function to isolate the
responsible structure.
Implicit in the devotion to purifying enzymes is the faith of a dedicated
biochemist of being able to reconstitute in a test tube anything a cell can do.
In fact, the biochemist with the advantage of manipulating the medium:
pH, ionic strength, etc., by creating high concentrations of reactants, by
trapping products and so on, should have an easier time of it. Another article
of faith is that everything that goes on in a cell is catalyzed by an enzyme.
Chemists sometimes find this conviction difficult to swallow.
On a recent occasion I was told by a mature and well-known physical
chemist that what fascinated him most in my work was that DNA replica-
tion was catalyzed by enzymes! This reminded me of a seminar I gave to the
Washington University chemistry department when I arrived in St. Louis in
1953. I was describing the enzymes that make and degrade orotic acid,
and began to realize that my audience was rapidly slipping away. Perhaps,
they had been expecting to hear about an organic synthesis of erotic acid.
In a last-ditch attempt to retrieve their attention, I said loudly that every
chemical event in the cell depends on the action of an enzyme. At that
point, the late Joseph Kennedy, the brilliant young chairman, awoke:
‘‘Do you mean to tell us that something as simple as the hydration of carbon
dioxide (to form bicarbonate) needs an enzyme?’’ The Lord had delivered
him into my hands. ‘‘Yes, Joe, cells have an enzyme, called carbonic
anhydrase. It enhances the rate of that reaction more than a million fold.’’
Enzymes are awesome machines with a suitable level of complexity.
One may feel ill at ease grappling with the operations of a cell, let alone
those of a multicellular creature, or feel inadequate in probing the fine
chemistry of small molecules. Becoming familiar with the personality of an
enzyme performing in a major synthetic pathway can be just right. To gain
his intimacy, the enzyme must first be purified to near homogeneity. For
the separation of a protein species present as one-tenth or one-hundredth of
1% of the many thousands of other kinds in the cellular community, we
need to devise and be guided by a quick and reliable assay of its catalytic
activity.
No enzyme is purified to the point of absolute homogeneity. Even when
other proteins constitute less than 1% of the purified protein and escape
detection by our best methods, there are likely to be many millions of
foreign molecules in a reaction mixture. Generally, such contaminants do
not matter unless they are preponderantly of one kind and are highly active
on one of the components being studied.
Only after the properties of the pure enzyme are known is it profitable
to examine its behavior in a crude state. ‘‘Don’t waste clean thinking on
dirty enzymes’’ is sound dogma. I cannot recall a single instance in which I
begrudged the time spent on the purification of an enzyme, whether it led
Why Purify Enzymes? 5

to the clarification of a reaction pathway, to discovering new enzymes, to

acquiring a unique analytical reagent, or led merely to greater expertise with
purification procedures. So, purify, purify, purify.
Purifying an enzyme is rewarding all the way, from first starting to free it
from the mob of proteins in a broken cell to having it finally in splendid
isolation. It matters that, upon removing the enzyme from its snug cellular
niche, one cares about many inclemencies: high dilution in unfriendly
solvents, contact with glass surfaces and harsh temperatures, and exposure
to metals, oxygen, and untold other perils. Failures are often attributed to
the fragility of the enzyme and its ready denaturability, whereas the blame
should rest on the scientist for being more easily denatured. Like a parent
concerned for a child’s whereabouts and safety, one cannot leave the
laboratory at night without knowing how much of the enzyme has been
recovered in that day’s procedure and how much of the contaminating
proteins still remain.
To attain the goal of a pure protein, the cardinal rule is that the ratio of
enzyme activity to the total protein is increased to the limit. Units of activity
and amounts of protein must be strictly accounted for in each manipulation
and at every stage. In this vein, the notebook record of an enzyme purifica-
tion should withstand the scrutiny of an auditor or bank examiner. Not that
one should ever regard the enterprise as a business or banking operation.
Rather, it often many seem like the ascent of an uncharted mountain: the
logistics like those of supplying successively higher base camps. Protein
fatalities and confusing contaminants may resemble the adventure of unex-
pected storms and hardships. Gratifying views along the way feed the
anticipation of what will be seen from the top. The ultimate reward of a
pure enzyme is tantamount to the unobstructed and commanding view
from the summit. Beyond the grand vista and thrill of being there first, there
is no need for descent, but rather the prospect of even more inviting
mountains, each with the promise of even grander views.
With the purified enzyme, we learn about its catalytic activities and its
responsiveness to regulatory molecules that raise or lower activity. Beyond
the catalytic and regulatory aspects, enzymes have a social face that dictates
crucial interactions with other enzymes, nucleic acids, and membrane
surfaces. To gain a perspective on the enzyme’s contributions to the cellular
economy, we must also identify the factors that induce or repress the genes
responsible for producing the enzyme. Tracking a metabolic or biosynthetic
enzyme uncovers marvelous intricacies by which a bacterial cell gears
enzyme production precisely to its fluctuating needs.
Popular interest now centers on understanding the growth and develop-
ment of flies and worms, their cells and tissues. Many laboratories focus on
the aberrations of cancer and hope that their studies will furnish insights into
the normal patterns. Enormous efforts are also devoted to AIDS, both to the
virus and its destructive action on the immune system. In these various
6 Arthur Kornberg

Figure 1.1 Personalized license plate expressing a commitment to enzymology.

studies, the effects of manipulating the cell’s genome and the actions of
viruses and agents are almost always monitored with intact cells and organ-
isms. Rarely are attempts made to examine a stage in an overall process in a
cell-free system. This reliance in current biological research on intact cells
and organisms to fathom their chemistry is a modern version of the vitalism
that befell Pasteur and that has permeated the attitudes of generations of
biologists before and since.
It baffles me that the utterly simple and proven enzymologic approach to
solving basic problems in metabolism is so commonly ignored. The precept
that discrete substances and their interactions must be understood before
more complex phenomena can be explained is rooted in the history of
biochemistry and should by now be utterly commonsensical. Robert Koch,
in identifying the causative agent of an infectious disease, taught us a century
ago that we must first isolate the responsible microbe from all others.
Organic chemists have known even longer that we must purify and crystal-
lize a substance to prove its identity. More recently in history, the vitamin
hunters found it futile to try to discover the metabolic and nutritional roles
of vitamins without having isolated each in pure form. And so with enzymes
it is only by purifying enzymes that we can clearly identify each of the
molecular machines responsible for a discrete metabolic operation. Con-
vinced of this, one of my graduate students expressed it in a personalized
license plate (Fig. 1.1).

ACKNOWLEDGMENT
This article borrows extensively from ‘‘For the Love of Enzymes: The Odyssey of a
Biochemist,’’ Harvard University Press, 1989.
 C H A P T E R T W O

Strategies and Considerations

for Protein Purifications
Stuart Linn

Contents
1. General Considerations 10
1.1. Properties and sensitivities of the protein 10
1.2. For what is the protein to be used 10
1.3. Assays 11
1.4. What should be added to the suspension, storage,
and assay buffers 13
1.5. Storage of protein solutions 14
1.6. Contaminating activities 15
2. Source of the Protein 15
2.1. Preliminary studies to obtain sequence information 15
2.2. Overexpressed protein as the source material 16
3. Preparing Extracts 16
4. Bulk or Batch Procedures for Purification 17
5. Refined Procedures for Purification 18
5.1. High-capacity steps 18
5.2. Intermediate-capacity steps 19
5.3. Low-capacity steps 19
6. Conclusions 19
References 19

Abstract
Prior to embarking upon the purification of a protein, one should begin by
considering what the protein is to be used for. In particular, how much of the
protein is needed, what should be its state of purity, and must it be folded
correctly and associated with various other peptides or cofactors. Using such
criteria, an appropriate assay should be chosen and a procedure be planned
taking into account the source of the protein, how it is to be extracted from the
source, and what agents the protein ought to be exposed to or ultimately be
stored in.

Department of Molecular and Cellular Biology, University of California, Berkeley, California, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63002-0 All rights reserved.

9
10 Stuart Linn

One is often surprised at the time necessary to develop an appropriate

protein purification procedure relative to the time required to clone a gene or
to accumulate information with the purified protein. There are an overwhelming
number of options for protein purification steps, so forethought is necessary to
expedite the tedious job of developing the purification scheme, or to avoid
having to redesign it upon attempting to use the protein. This chapter points out
general considerations to be undertaken in designing, organizing, and execut-
ing the purification, while subsequent chapters of this volume supply more
specific options and technical details.

1. General Considerations
1.1. Properties and sensitivities of the protein
If known, one must consider whether the protein is soluble, and, if not,
what agents might help to solubilize it. Is the protein labile at high or low
concentrations? Is the protein sensitive to high or low salt concentrations,
high or low temperatures, high or low pH, or oxidation, particularly by
oxygen. A few preliminary experiments in this regard might be well worth
the effort so as to learn what reagents ought to be present during the
purification, what agents must be avoided (e.g., oxygen), and under what
conditions the protein ought to be stored.

1.2. For what is the protein to be used

The required amount of purified protein may vary from a few micrograms
needed for a cloning or identification endeavor to several kilograms required
for industrial or pharmaceutical applications. Therefore, a very major con-
sideration is the amount of material required. One should be aware of the
scale-up that will be necessary and the final scheme should be appropriate for
expansion to those levels. There are very real limitations to scaling up a
procedure which are brought about not only by considerations of cost and
availability of facilities but also by physical constraints by such factors as
chromatographic column size limits or electrophoresis (ohmic) heating.
As outlined below, individual steps of the procedure should flow from
high-capacity/low-cost techniques toward low-capacity/high-cost ones.
In some cases, alternative procedures may be required: for example, one to
obtain microgram quantities for sequencing and cloning and a second to
produce kilogram amounts of the cloned and overproduced material in its
native state. The protein chemist should anticipate such variable needs and
remain flexible for adopting new procedures as the need arises.
A corollary to the amount is the concentration of protein necessary.
Concentration comes into play not only for applications of the protein but
General Strategies and Considerations 11

also for detection or assay of the amount and activity of the protein.
Concentration techniques such as those covered in Chapter 9 often must
be included after an intermediate or the final step of purification.
Other considerations include whether the protein must be active (as an
enzyme, a regulatory protein, or an antibody, for example). If so, must it be
folded correctly into its normal native configuration? (Refolding of dena-
tured- or overexpressed proteins may alter precise kinetic properties, for
example, if subtle, but stable alternative configurations are assumed.) Or,
folding may not be an issue, for example, if the protein is to be utilized for
sequence information or identification. The techniques employed during a
procedure should be as gentle as are necessary. But, whenever possible,
some of the harsher, but often spectacularly successful procedures such as
those that involve extremes of pH, organic solvents, detergents, or hydro-
phobic or strong affinity chromatographic media should also be tried.
How free from contaminants should a protein be and, in particular,
which contaminants? For example, proteins used as pharmaceuticals must
be ultra-pure in all respects, whereas polymerases must be free of enzymes
that degrade the polymer product of the polymerase or proteins that add or
remove protein modifications must be free of their counterparts.
A consideration that is currently receiving a great deal of attention is that
of the appropriate subunit content of a multimeric protein. What are the
subunits of the desired protein? Is an isolated complex identical in content
to that within the starting material? And, how might this content change
with the growth state, growth conditions, or storage of the source material?
These considerations are further exacerbated by the possibility that a protein
in question might exist simultaneously in several or many different com-
plexes or states of aggregation. Likewise, it might exist simultaneously with
alternative posttranslational modifications. In either or both of these
instances, the distribution of the protein among these various alternatives
might be extremely sensitive to stresses imposed upon the cells used as the
source material prior to purification. One should take these possibilities
into account, particularly when the protein seems to partition into several
peaks during purification. These variables may preclude an easy or successful
purification scheme. A case in point is p53 (the product of the human TP53
gene) for which a satisfactory purification of active protein has yet to be
achieved, in spite of its importance as the ‘‘guardian of the human genome,’’
because of its numerous patterns of associated peptides, alternative
modifications, and states of aggregation.

1.3. Assays
Classically, a most important consideration is the development of appropri-
ate assays for the protein. The success of the purification is often dependent
on this. Six considerations are relevant: sensitivity, accuracy, precision,
linearity, substrate availability, and cost as measured in time and money.
12 Stuart Linn

Sensitivity can be the limiting factor as the protein becomes diluted into
column effluents, etc. Before formulating a step, dilution and losses ought to
be estimated so that the ability to detect and utilize the protein will be
possible.
Accuracy and precision are often compromised, but clearly these factors
must be controlled to the extent that the assay is reliable for assuring
reproducibility and the recovery of sufficient material.
The linear range of the assay must be measured both with respect to time
and to protein amount. As a procedure progresses, these ranges ought to be
reevaluated, particularly for activity assays. For example, removal of protein
contaminants might affect the limits if an inhibitor or stimulator were
removed. Or, the limits could be altered if the highly purified protein
becomes less stable during assay incubations.
Substrate availability and cost refer to the practicality of the assay:
Is enough substrate available to perform the entire purification without
interruption? Can the assay be performed simultaneously on a reasonable
number of fractions in a reasonable amount of time? Stopping to prepare
more substrate or skimping on material and/or the number of fractions
assayed usually compromises the purification’s success. On the other hand,
assays at some steps might be compromised, for example, by omitting
specificity controls at later stages of purification or assaying alternate or
combined chromatography fractions.
It is often tempting not to use assays for activity, but to purify a gel band
or an antigen instead. Although this approach is appropriate in some
instances, particularly when an activity is not being sought, it is to be
strongly discouraged when activity is in fact what is desired. It cannot be
emphasized strongly enough that an activity assay is necessary to obtain
optimal yields of unaltered activity, be it one associated with an enzyme, an
antibody, a hormone, a regulatory protein, or a protein that modifies the
secondary or tertiary structure of a polymer.
An inherent part of an assay is the assay conditions. Optimal conditions
must be determined empirically, but, beware: optimal conditions often
change with purification due to removal of interfering inhibitory or stimu-
latory factors, lower overall protein concentrations, etc. Moreover, always
take into account materials added with the substrate or with the protein
storage, suspension or dilution buffers as these may change the pH and ionic
strength or otherwise affect the reliability of the assay.
A final comment pertains to the protein assay. While the goals also
include simplicity, linearity, reproducibility, specificity, and reliability,
accuracy is generally compromised, as no commonly used assay is absolute
with regard to all proteins. With crude fractions, color reactions are proba-
bly best. While the Bradford Method (Bradford, 1976) is the simplest of
these, in our laboratory we find it to be unreliable with crude fractions from
animal cells or when detergents are present. Protein assays can and often
General Strategies and Considerations 13

must by changed as the purification progresses. For example, for column

effluents, ultraviolet absorption may be utilized: it is simple, sensitive, does
not consume the material, and can even be continually monitored with a
flow detector. For extremes of sensitivity, wavelengths between 210 and
230 nm can be utilized (Tombs et al., 1959; Waddell, 1956). Beware,
however, that blanks must be adjusted to take into account the UV-
absorption of materials in the solutes and that nonprotein contaminants,
particularly nucleic acids, may have significant UV-absorption. Remember,
protein assays are done to follow the extent of purification and to
monitor reproducibility; they are not absolute. In the rare cases where
absolute values must be obtained, such assays as nitrogen determinations
are necessary (see Chapter 8).

1.4. What should be added to the suspension, storage,

and assay buffers
A common query is, ‘‘why is the protein suspended in x?’’ And its answer is
often, ‘‘if I change the components, I don’t know what will happen.’’
The obvious lesson is to add something only with good reason in the first
place. Even more of a problem in this regard pertains to proteins obtained
commercially in complex buffers, the origin of which may be obscure and,
even worse, the content of which may be proprietary and not disclosed.
In this instance, replacement with a known buffer cocktail may be advisable
prior to embarking upon a study of the protein’s properties or its application
in complex experiments (see Chapter 9).
Solutes are added usually to improve stability, prevent the growth of
microorganisms, reduce the freezing point, or keep the protein in solution.
Table 2.1 lists several classes and examples of such materials, but the reader is
referred to chapters later in this volume for comprehensive discussions of
them.
It is well worth the effort to carry out stability studies (e.g., heat inacti-
vation or storage trials) in order to learn how to maintain a stable, soluble
protein. Three notes of caution: (1) optimal storage conditions may change
with purification; (2) storage conditions may alter peptide–peptide interac-
tions; and (3) optimal storage conditions need not relate to optimal condi-
tions for activity. Indeed, materials that stabilize a protein might inhibit it
when added to activity mixes. Of course, the latter situation must be
considered when utilizing the protein—interfering substances will have to
be removed or ‘‘diluted out’’ for utilization of the protein.
In our experience, reducing agents are particularly effective for storage
of bacterial enzymes that often derive from a reducing environment,
whereas enzymes from animals often take kindly to surfactants or protease
inhibitors. Fungal proteins also respond to protease inhibitors. Optimal pH
and salt concentrations vary widely.
14 Stuart Linn

Table 2.1 Additions to protein solutions

Class Examples Purpose

Buffer Potassium phosphate, Tris–HCl, Stability
sodium acetate, Hepes–NaOH
Salt KCl, NaCl, (NH4)2SO4 Stability
Ionic Sodium deoxycholate, sodium Stability, solubility
detergents cholate
Nonionic Triton X-100, Nonidet P-40, Stability
detergents Tween 20, Brij 35
Glycerol, Stability; unfrozen
sucrose storage below
0
C
Sodium azide Bacteriostatic
Metal chelators EDTA (ethylenediaminetetraacetic Stability
acid), EGTA (ethylene glycol bis
(b-aminoethylether) N,N 0 -
tetraacetic acid)
Reducing b-Mercaptoethanol (BME), Stability; reduce
agents dithiothreitol (DTT), TCEP disulfide bonds
Ligands Mg2þ, ATP, phosphate Stability
Protease PMSF (phenylmethylsolfonyl Stability
inhibitors fluoride), TPCK (N-tosyl-L-
phenylalanine chloromethyl
ketone), TLCK (N a-p-tosyl-L-
lysine chloromethyl ketone)

In a converse context, what might be absolutely necessary to eliminate

from buffers and resins? Special precautions could be necessary to deal with
heavy metals (e.g., the use of ultra-pure reagents and double-distilled water)
or with oxygen, for which the purging of all buffers with N2 and/or
working in an anoxic glove box might be required. Keep in mind that a
seemingly minor amount of contaminating heavy metal or O2 in a liter of
dialysis buffer can be a relatively large amount in molar terms versus a few
micrograms of purified protein being dialyzed.
Special precautions, which must be taken for large protein complexes,
are noted in Chapter 10.

1.5. Storage of protein solutions

Most proteins prefer to be stored at the lowest temperature possible. If not
frozen, 0
C (on ice, rather than in a refrigerator in order to minimize
degradation or denaturation and bacterial or fungal growth) or  20
C
General Strategies and Considerations 15

with sufficient glycerol or sucrose present to avoid freezing. If frozen,

storage above liquid nitrogen or at  70
C is usually best and ‘‘frost-
free’’ freezers ought to be avoided if possible due to the possibility of
repeated thawing of small volumes during the thaw cycles. Allocation into
small aliquots avoids losses brought about by multiple freezing/thawing and
thawing should be as rapid as possible to minimize exposure of the protein
to concentrated solute in partially thawed solutions.
Concerning the containers utilized for purified proteins, particularly at
low concentrations, plastic rather than glass should be used. In our experi-
ence, polypropylene is superior to polyethylene or polystyrene and other
clear plastics. In rare instances, we have had to precoat storage tubes with an
inert protein prior to their use. Be sure to have tight-fitting caps if storage is
in ‘‘frost-free’’ freezers. These and other precautions are discussed in detail
in Chapter 10.

1.6. Contaminating activities

Often proteins need not or cannot be obtained in a pure state, but particular
interfering activities (e.g., nucleases, proteases, phosphatases) must not be
present. In our experience, attempting to purify one activity against one or
more others by utilizing multiple types of assays, as a criterion of purity
is extremely frustrating. Instead, purifying so as to optimize yield of
the specific protein and minimize total protein content (optimize specific
activity) with selective choice of fractions only at the last or penultimate step
is more likely to be successful.
Finally, to avoid an extremely embarrassing—though surprisingly
common—error, be certain that the activity that you have purified is
associated physically with the protein species being characterized. Genetic
experiments or physical criteria that confirm that these two entities comi-
grate under independent procedures (e.g., sedimentation, nondenaturating
gel electrophoresis) are most strongly encouraged.

2. Source of the Protein

2.1. Preliminary studies to obtain sequence information
If gene or peptide sequence is not available and a low-abundance protein is
to be purified in order to obtain such information, careful consideration of
the source material (e.g., organism, tissue, culture cell type) is worth the
time and effort, and trial extracts from a number of sources should be
explored. These should be tested for content of the protein (per gram of
starting material or per unit cost), the starting specific activity if relevant, and
the stability of the protein. If relevant, the classical microbiological approach
16 Stuart Linn

of isolating microorganisms with unique growth requirements can lead to

unexpected success.
The cost and availability of the source material, particularly if a largely
scaled-up preparation might be desirable in the future, should be consid-
ered. Genetic knowledge and technology for the organism or cell type
ought to be available should regulatory characterization and/or gene
sequence manipulations be envisioned for the future.

2.2. Overexpressed protein as the source material

With the recent dramatic technological improvements in gene manipula-
tions, one usually expects ultimately to overexpress a protein, whether for
in-depth characterization or for reagent use. For overexpression, is a bacte-
rial, fungal, insect cell, or mammalian system best? (see Chapters 11–15).
What special precautions are necessary for each source? Will the protein be
appropriately posttranslationally modified and will such modifications be
easily identifiable? Will the protein be obtained in a biologically relevant
state of aggregation? If a heterologous source is to be utilized (e.g., expres-
sion of a mammalian protein in insect cells), might an endogenous homo-
logue cause any problems? Should a mutant host cell be utilized (e.g., one
lacking a similar protein or an interfering activity such as a nuclease or
protease)?
All of the subunits of a heteromultimeric protein often must be over-
expressed together in order to obtain the protein in its ‘‘native’’ state. To
accomplish this, it is necessary to learn the efficiency of expression of each of
the peptide constructs in order that the peptides are produced in roughly the
desired stoichiometric amounts. This is often ultimately not difficult to
accomplish, but a good deal of effort must be expended in learning the
conditions to differentially control such joint expression systems by varia-
tion of transfection levels or promoter efficiencies as the case may be.
If tags are to be attached to a protein to aid in purification and/or
detection, will they interfere with subsequent studies? Where in the protein
should the tags be placed? If on the N-terminus, prematurely terminated
peptides will contaminate affinity-purified material. If on the C-terminus,
functional or binding properties of that part of the protein might be
compromised (see Chapter 16 for discussion of this important topic).

3. Preparing Extracts
Preparation of extracts is discussed in details in Section IV, so only
general considerations are noted here. In our experience, the manner in
which cells are disrupted has a profound, but unpredictable effect on the
General Strategies and Considerations 17

yield and quality of the protein preparation. Trials of alternative procedures

are clearly necessary.
Thought should always be given to scaling up of the preparation. Will
the volumes or time required become unreasonable? Can a subsequent
clarification of the extract be conveniently scaled up? In general, volumes
should be kept to a minimum—for example, extracts should be as concen-
trated as possible. If relevant and convenient, tissue, cell type, or organelle
fractionation is almost always worthwhile prior to general disruption.
The buffer that favors the most efficient disruption may not prove to be
one that favors stability or a subsequent step, and, in extreme cases, the
buffer might need immediate alteration or a subsequent step might imme-
diately be necessary upon completion of the disruption protocol. For
example, hypotonicity, chelators, or pH extremes might aid in total release
of the desired protein, but labialize the protein in the resulting extract. Or,
an overproduced protein might not remain soluble in the disruption
medium. Parenthetically, such buffer modifications are also often necessary
during subsequent purification protocols in which, for example, it is not
uncommon to collect chromatography fractions into tubes containing
additional buffer constituents, or to dialyze individual fractions immediately
as they are eluted.

4. Bulk or Batch Procedures for Purification

These procedures are almost always utilized early in the purification as
they are usually most effective in removing nonprotein material and are
amenable to the relatively large volume and amounts of material that exist in
earlier stages of the preparation. A great deal of effort went into designing
these steps in the early days of protein chemistry, and much frustration,
time, and money can often be avoided by including some of these ‘‘old-
fashioned’’ procedures.
Section V details many of these procedures. Gentle procedures including
‘‘salting out’’ or phase partition with organic polymers are most common.
More drastic methods such as heat, extremes of pH, or phase partition with
organic solvents might be particularly effective with stable proteins, though
subtle forms of damage may be difficult to foresee or to detect. In addition,
one might consider the use of high-capacity ion-exchange resins either
added as a slurry to crude material or used in columns for batch elution.
The time expended in developing and optimizing these early steps is always
worthwhile—even removal of half of the contaminating protein may dic-
tate the feasibility of a subsequent step from both cost and technical
considerations.
18 Stuart Linn

5. Refined Procedures for Purification

Once the bulk methods have yielded a protein preparation that is
reasonably free of nucleic acids, polysaccharides, and lipids, the preparation
becomes amenable to the more interesting and often spectacular proce-
dures, which have been developed in recent years. The general strategy is to
proceed from high- to low-capacity procedures and to attempt to exploit
specific affinity materials whenever possible.
Applications, specific examples, and technical details for these proce-
dures are discussed in Sections VI and VII and will not be described here. As
a general consideration, in proceeding from one procedure to the next, one
ought to reduce as much as possible the necessity for dialysis and concen-
tration. Hence, procedures that separate by size and shape can simulta-
neously remove salt or otherwise modify the buffer. Procedures utilizing
high-capacity resins can concentrate proteins as well as purify them, or
resins from which the desired protein elutes at relatively low salt concentra-
tions can be directly followed with a resin to which the protein binds at this
concentration. Also, some steps (e.g., sedimentation through gradients of
sucrose or glycerol) may leave the protein in a medium that might be ideal
for long-term storage, and appropriate (or inappropriate) for use directly for
studies or a subsequent purification step. Finally, interchanging the order of
the steps of a procedure can, and often does, have a profound effect on the
success of a purification scheme.
Two notes of caution: it has been our experience that rapid dilution of a
fraction that has a high concentration of salt, glycerol, or sucrose often
results in a loss of activity. If this phenomenon is experienced, the use of
dialysis or gel exclusion chromatography may be a useful alternative. Also,
during a purification scheme, the protein solution sometimes becomes
cloudy. This solution must be clarified prior to chromatography to avoid
unacceptably low flow rates. In our experience, the insoluble material rarely
contains native protein.
Some procedures which cannot be effectively scaled up (e.g., sedimen-
tation or HPLC/FPLC) can be carried out with small aliquots of the
preparation if left to the final stages. In fact, in some cases the utilization
of such aliquots is desirable as the less purified fractions might be more stable
to long-term storage.

5.1. High-capacity steps

Generally, these include ion-exchange resins or very general affinity agents
such as dyes or glass. When used in the initial stages of a purification scheme
with large amounts of material, a particular type of ion-exchange resin can
often be successfully reutilized for purification at a later stage (especially if
the pH is changed) or for concentration.
General Strategies and Considerations 19

5.2. Intermediate-capacity steps

These might include the hydrophobic resins for which long chro-
matographic times reduce activity yields. Many affinity agents bound to
inert chromatography scaffolds (e.g., bulk DNA, simple oligonucleotides,
antibodies, or ligands of a protein) fall into this class. In these instances,
thought and effort must be given to finding conditions and/or materials that
can successfully elute the protein without affecting its subunit structure or
destabilizing or inactivating it. Free, unbound ligand often is successful,
although it may be expensive and/or difficult to unbind from the protein.
Finally, gel filtration is considered as a step with intermediate capacity.

5.3. Low-capacity steps

Affinity steps utilizing valuable ligands such as substrate analogs, complex
polynucleotide sequences, monoclonal antibodies, or lectins might be
included here. Also included are electrophoresis methods, including
isoelectric focusing, as precipitation of the protein may be a problem with
moderate amounts of protein. Ultracentrifugation steps are usually limited
to small volumes of material and HPLC/FPLC steps are also often limited
by cost and/or capacity of the column. Small hydrophobic columns
might be successfully utilized when larger ones of the same resin result in
activity loss.

6. Conclusions
Although the development and execution of protein purification pro-
cedures are often difficult and frustrating processes, the rewards are great.
Moreover, given the continual addition of new technology, high-quality
commercial materials utilized for purification procedures, and genetically
altered sources of proteins, the future bodes well for ever-simpler procedures
accompanied by ever-greater rewards.

REFERENCES
Bradford, M. M. (1976). A rapid and sensitive method for quantitation of microgram
quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72,
248–254.
Tombs, M. P., Souter, F., and MacLagan, N. F. (1959). The spectrophotometric determi-
nation of protein at 210 mm. Biochem. J. 73, 167–171.
Waddell, W. J. (1956). A simple ultraviolet spectrophotometric method for the determina-
tion of protein. J. Lab. Clin. Med. 48, 311–314.
 C H A P T E R T H R E E

Use of Bioinformatics in Planning

a Protein Purification
Richard R. Burgess

Contents
1. What You Can Learn from an Amino Acid Sequence 22
1.1. Molecular weight of the polypeptide chain 22
1.2. Charge versus pH/titration curve, isoelectric point 22
1.3. Molar extinction coefficient 23
1.4. Cysteine content 23
1.5. Secondary structure 24
1.6. Stability 24
1.7. Hydrophobicity and membrane-spanning regions 24
1.8. Sequence similarity suggests homology
and possible cofactor affinity 24
1.9. Potential modification sites 25
1.10. Solubility on overexpression in E. coli 25
2. What You Cannot yet Predict 25
2.1. Three-dimension structure; shape, surface features 26
2.2. Multisubunit features; homomultimers, heteromultimers? 26
2.3. Precipitation properties 26
3. Conclusion 26
3.1. Protein bioinformatic resources 27
3.2. Purification in the denatured state 27
References 27

Abstract
Now that many hundreds and even thousands of whole genomes have been
sequenced, it is rare to be studying a target protein whose amino acid sequence
is not known. However, it is still often necessary to obtain large amounts of the
target protein for a variety of purposes including structural studies, drug
discovery, enzymology, protein biochemistry, and industrial application. It
would seem that knowing the amino acid sequences would make it much easier

McArdle Laboratory for Cancer Research, University of Wisconsin-Madison, Madison, Wisconsin, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63003-2 All rights reserved.

21
22 Richard R. Burgess

to design an effective purification of that protein. We examine in this chapter

what you can and cannot predict from an amino acid sequence and conclude
that protein purification is still largely an empirical science.

1. What You Can Learn from an

Amino Acid Sequence
Often there are two common scenarios that a protein biochemist faces
these days: (1) you know the protein and at least some of its functions; and
(2) you have identified a gene and its protein through transcriptional
profiling or proteomic studies by two-dimensional gel analysis or mass
spectroscopy, but it is of unknown function. In either case, you know the
sequence and want to purify it. Here are some of the types of information
you can learn from its sequence that might help you in designing a protein
purification procedure. Many of these properties can readily be obtained
using sequence analysis software as described below.

1.1. Molecular weight of the polypeptide chain

It is very straightforward to calculate the molecular weight (MW) of a
polypeptide chain protein by multiplying the MW of an amino acid in a
polypeptide (the MW of an amino acid –18 Da) by the number of times that
amino acid occurs in the polypeptide. Remember to add an extra 18 Da to
the total to account for the free amino and carboxy termini. Of course, one
cannot know, except by N-terminal sequencing or mass spectroscopy,
whether the N-terminal initiating methionine has been removed or not.

1.2. Charge versus pH/titration curve, isoelectric point

Assuming a pKa for each ionizable group in a protein (the pKa is the pH at
which that residue is half ionized), one can calculate the total charge as a
function of pH of a given protein from its amino acid sequence. The pH at
which the net charge is zero is defined as its isoelectric point or pI. This
information is helpful in deciding which ion exchange chromatography
resin to use. For example, if you assume that the protein is a monomer and
not modified to change the charge, then an acidic protein (one that is
negatively charged at pH 7) is likely to bind to a positively charged anion
exchange resin such as MonoQ while a basic protein (one that is positively
charged at pH 7) is likely to bind to a negatively charged cation exchange
resin such as MonoS. There are exceptions to this generalization if the
charge on the surface is not uniformly distributed. A protein could have a
positive patch and a separate negative patch and be able to bind to both
Bioinformatics in Protein Purification 23

anion and cation exchange columns under the same conditions. To the
first approximation, the greater the charge on a protein at the pH of
the column buffer, the tighter it will bind and the higher the salt needed
to elute it from the column resin. Of course, if the protein is part of a
multiprotein complex, then you have no idea what its ion exchange column
binding properties will be.
Finally, since a protein is generally least soluble at its pI, one can consider
an isoelectric precipitation step (assuming that the protein is not in a stable
complex with another protein or proteins) (see Chapter 20).

1.3. Molar extinction coefficient

At a wavelength of 280 nm, all of the absorbance of an unmodified protein
is due to the absorption of its amino acids tryptophan, tyrosine, and cysteine.
Gill and von Hippel (1989) and Pace et al. (1995) have described similar
methods to estimate the molar absorption/extinction coefficient of a pro-
tein given its amino acid composition. This involves some assumptions,
based on experimental data, on the average proportion of the tryptophan
and tyrosine that are exposed on the surface versus buried in the interior.
This method is perhaps the most useful practical method of determining the
concentration of a purified protein. For example, if the protein has six
tryptophans, seven tyrosines, and no cysteines, the molar extinction coeffi-
cient (e280 nm) would be (6
5500) þ (7
1490) þ (0
125) ¼ 43,430.
The absorption (A280 nm) of a 10–5 M solution would be 0.43. A somewhat
more useful number is the A280 nm of a 1 mg/ml solution of that protein.
1 mg=ml
This is sometimes called E280 nm . You divide the molar extinction coeffi-
cient by the MW (e.g., if the protein above is 30,000 Da, then 43,430/
30,000 ¼ 1.45). Then a careful measurement of the A280 nm with appropri-
ate buffer absorption controls will give you the protein concentration (e.g.,
if the measured A280 nm is 0.75, then the protein concentration of the
measured protein solution is 0.75/1.45 ¼ 0.52 mg/ml).
It should be stressed that this method is not valid if the protein contains any
contaminating nucleic acid, additional 280 nm-absorbing moieties (such as
bound heme, iron–sulfur [Fe–S] centers, or nucleotide substrates or cofac-
tors), or fluorescent modifications (such as with green fluorescent protein).

1.4. Cysteine content

I always look to see if a protein I am trying to purify contains cysteine. If not,
then I can omit reducing agents such as dithiothreitol (DTT) from my buffers.
If the protein is from E. coli and has cysteines, then I assume that there are no
disulfides in the native protein since the environment in the cytoplasm is
reducing making it unlikely that the protein is naturally in an oxidized
24 Richard R. Burgess

condition unless it is located in the periplasm. In general when my protein has

cysteines, I include DTT in my buffers to prevent unwanted intra- or inter-
molecular disulfide formation. It is harder to predict whether proteins from
other sources and that contain multiple cysteines are likely to form disulfides.

1.5. Secondary structure

There are several quite reasonable methods to predict regions of the protein
that are likely to be in the form of alpha helical or beta strand secondary
structures. However, this knowledge is rarely of use in designing a protein
purification scheme.

1.6. Stability
It is possible, using ProtParam (see below) to estimate the in vivo half-life and
the instability index of a protein from its sequence. The in vivo half-life is
calculated based on the N-end rule (Varshavsky, 1997) and the N-terminal
amino acid of the protein in question. Approximate half-life estimates are
given for the protein expressed in mammalian cells, yeast and E. coli. The
instability index provides an estimate of the stability of your protein in vitro
and is based on the analysis of dipeptide occurrences in your protein com-
pared to those of a set of test proteins that are known to be unstable or stable
(Guruprasad et al., 1990). This information might be useful since a protein
predicted to be unstable in vitro might warrant special care in maintaining low
temperatures during purification and perhaps in the use of protease inhibitors.

1.7. Hydrophobicity and membrane-spanning regions

It is possible to analyze sequence information to determine regions that are
particularly hydrophobic or hydrophilic. One such method is that of Kyte
and Doolittle (1982) that allows one to plot the hydropathy value along a
sequence. This can allow one to recognize potential membrane-spanning
regions and, for a protein of unknown function, predict that it is a mem-
brane protein. A membrane-spanning region is usually a stretch of 23
hydrophobic amino acids that form an alpha helix. If you know that the
protein of interest is a membrane protein, it will certainly affect how you
design a purification scheme.

1.8. Sequence similarity suggests homology

and possible cofactor affinity
If your protein is of unknown function, it is possible to do a search to
determine if it is highly similar to other known proteins in the protein
database. If the sequence similarity is great enough to infer homology and if
Bioinformatics in Protein Purification 25

your protein is related to other homologous family members that have been
studied, then perhaps you know, for example, that it usually occurs as a
homodimer. This will help in predicting its behavior on gel filtration
column chromatography. This information might also be used to devise a
suitable assay for your protein.
Again, as above, sequence can identify if it belongs to a protein family all
of which are known to bind to a particular cofactor or substrate. For example,
if it is a member of the AAA ATPase family, it is likely that it will bind to an
affinity column that has an immobilized ATP analog. Such an affinity purifi-
cation step can aid greatly in a purification scheme (see Chapter 26).

1.9. Potential modification sites

It is now possible to identify in amino acid sequences, short amino acid
stretches or ‘‘motifs’’ that are commonly sites of posttranslational modifica-
tion. A few of these motifs are: glycosylation site (NXS or NXT); biotiny-
lation site (AMKM); zinc finger (metal binding site) (F/YXCX2–4
CX3FX5LX2HX3–4HX5), heart muscle protein kinase recognition site
(RRASV). This information can be highly useful. For example, if the
protein is glycosylated, it might bind to a lectin affinity column (see
Chapter 35). Many posttranslational modifications become handles by
which a protein can be fractionated away from many other proteins. The
problems are that one cannot predict if the site is on the surface of the
protein of interest or if it is in fact modified and to what extent.

1.10. Solubility on overexpression in E. coli

Several studies have proposed that one can predict from the protein
sequence the likelihood that a protein, if overproduced in E. coli will be
soluble or form insoluble inclusion bodies (Idicula-Thomas and Balaji,
2005; Wilkinson and Harrison, 1991). Obviously, this is useful information,
but is most easily obtained by simply overproducing a protein, breaking the
cells, centrifuging the lysate to separate soluble from insoluble, and then
analyzing the two resulting fractions by SDS–PAGE. The fact that the
proportion of an overproduced protein that is soluble can be significantly
altered by manipulating cell growth conditions suggests that solubility
predictions based on sequence may be useful but have severe limitations.

2. What You Cannot yet Predict

Unfortunately, most of the information needed to design a purifica-
tion scheme is not predictable from a protein’s amino acid sequence. Some
of the most critical problems are listed below.
26 Richard R. Burgess

2.1. Three-dimension structure; shape, surface features

While a huge amount of time and effort has gone into finding a practical
method of predicting three-dimensional structure from amino acid
sequence, it is presently not yet possible to do so with any confidence. In
the case where a target protein shows high similarity to a protein whose
structure has been determined, it is possible to ‘‘thread’’ the sequence into
the known structure and come up with a reasonable approximation of the
structure of the target protein.
Without accurate structural information, one cannot predict shape, or
detailed surface features such as hydrophobic patches, charge distribution,
and antigenic sites. Therefore, one cannot easily predict behavior on hydro-
phobic interaction chromatography (HIC) or ion exchange chromatogra-
phy. Since the shape affects the Stokes radius, one cannot predict behavior
during gel filtration chromatography (a spherical protein will appear smaller
than an asymmetric or cigar-shaped protein of the same MW during gel
filtration).

2.2. Multisubunit features; homomultimers,

heteromultimers?
Even if one could accurately predict the three-dimensional structure of a
protein, it would still be impossible to predict if it exists in solution as a
multimer (e.g., a hexamer) or as a monomer. That lack of knowledge
precludes any reasonable prediction of its behavior on sizing columns or
on ion exchange columns. Even more problematic is the fact that many
proteins exist as parts of multisubunit complexes and one cannot predict
whether a given protein will exist as a complex whose purification proper-
ties may be determined largely by its binding partners.

2.3. Precipitation properties

Different proteins vary in their solubility properties in ways that are not yet
understood. Therefore, it is not presently possible to predict what ammo-
nium sulfate concentration to use in precipitating a given protein from
solution. This will be discussed further in Chapter 20.

3. Conclusion
I think it is very clear that one cannot yet use amino acid sequence to
predict the behavior of a given protein in most of the primary fractionation
methods used by protein purifiers. By far the best approach still is to
Bioinformatics in Protein Purification 27

fractionate an extract by ammonium sulfate precipitation, by ion exchange

and gel filtration chromatography and determine how the protein of interest
behaves. Therefore, although we have tremendous knowledge of protein
sequence, protein purification is still very empirical! Don’t be afraid to
simply do the experiment!

3.1. Protein bioinformatic resources

One of the most used sites for obtaining sequence data and using it to
compute various physicochemical properties of proteins is the ProtParam
feature of ExPASy (Expert Protein Analysis Software) Web site (http://www.
expasy.org/tools/protparam). ProtParam calculates many of the protein para-
meters including MW, theoretical pI, amino acid composition, atomic com-
position, extinction coefficient, estimated half-life, instability index, aliphatic
index, and grand average of hydropathicity (Gasteiger et al., 2005).
This is very simple to use. Go to ProtParam tools, enter a Swiss-Prot/
TrEMBL accession number (AC) or a sequence identifier (ID), or you can
paste your own sequence in the box, hit button that says ‘‘compute
parameters,’’ and print out the results.
One commercial package available through DNASTAR (http://www.
dnastar.com), called Lasergene, version 8.0, contains a section on protein
sequence analysis called ‘‘Protean.’’ This allows sequence analysis similar to
ProtParam and in addition, predicts antigenicity, surface probability, maps
proteolytic digestion sites, and has very nice graphical displays.

3.2. Purification in the denatured state

After all that was discussed above, it is only fair to point out that one can do
much better at predicting purification properties of a given protein is you
are willing to carry out the purification in the denatured state (Knuth and
Burgess, 1987). If you do not have to worry about possible multimeric
states, then one can simply denature in urea, fractionate somewhat predict-
ably by ion exchange and gel filtration, and then refold the protein to its
native structure (see Chapter 17).

REFERENCES
Gasteiger, E., Hoogland, C., Gattiker, A., Duvaud, S., Wilkins, M. R., Appel, R. D., and
Bairoch, A. (2005). Protein identification and analysis tools on the ExPASy server.
In ‘‘The Proteomics Protocols Handbook’’, ( J. M. Walker, ed.), pp. 571–607. Humana
Press, Totowa.
Gill, S. C., and von Hippel, P. H. (1989). Calculation of protein extinction coefficients from
amino acid sequence data. Anal. Biochem. 182, 319–326.
28 Richard R. Burgess

Guruprasad, K., Reddy, B., and Pandit, M. W. (1990). Correlation between stability of a
protein and its dipeptide composition: A novel approach for predicting in vivo stability of
a protein from it primary sequence. Protein Eng. 4, 155–161.
Idicula-Thomas, S., and Balaji, P. V. (2005). Understanding the relationship between the
primary structure of proteins and its propensity to be soluble on overexpression in E. coli.
Protein Sci. 14, 582–592.
Knuth, M. W., and Burgess, R. R. (1987). Purification of proteins in the denatured state.
In ‘‘Protein Purification: Micro to Macro’’, (R. R. Burgess, ed.), pp. 279–305. Alan R.
Liss, New York.
Kyte, J., and Doolittle, R. F. (1982). A simple method for displaying the hydrophobic
character of a protein. J. Mol. Biol. 157, 105–132.
Pace, C. N., Vajdos, F., Fee, L., Grimsely, G., and Gray, T. (1995). How to measure and
predict the molar absorption coefficient of a protein. Protein Sci. 11, 2411–2423.
Varshavsky, A. (1997). The N-end rule pathway of protein degradation. Genes Cells 2,
13–28.
Wilkinson, D. L., and Harrison, R. G. (1991). Predicting the solubility of recombinant
proteins in E. coli. Biotechnology 9, 443–448.
 C H A P T E R F O U R

Preparing a Purification
Summary Table
Richard R. Burgess

Contents
1. Introduction 29
2. The Importance of Footnotes 32
3. The Value of an SDS–Polyacrylamide Gel Analysis on
Main Protein Fractions 32
4. Some Common Mistakes and Problems 32

Abstract
Once a protein purification scheme has been developed, the purification, char-
acterization, and use/structure of a target protein are usually published. It is
highly desirable to present the major steps in the purification and the
corresponding features of the protein at each step summarized in the form of
a purification summary table. In considering whether to repeat a published
protein purification, a reader needs this information to evaluate the purification,
and to decide if it is worth following or if it needs major modifications. In this
chapter, I discuss the main characteristics of a useful purification summary
table and point out common mistakes and problems I see in many such tables.

1. Introduction
As an executive editor and editor-in-chief of the journal, Protein
Expression and Purification, I have reviewed on the order of 100 protein
purification papers a year for over 18 years. It is remarkable how many
manuscripts I receive where there is either no purification summary or one
that is severely lacking in necessary information and accuracy. The essentials
of a reasonable purification table are illustrated by the example below.
Suppose one set out to purify an enzyme from the bacterium E. coli
starting with 10 g of wet weight cell pellet from a 4-l culture (10 g of wet
weight cells typically would contain about 2 g of dry weight and about

McArdle Laboratory for Cancer Research, University of Wisconsin-Madison, Madison, Wisconsin, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63004-4 All rights reserved.

29
30 Richard R. Burgess

1200 mg of total protein). The cells are lysed by sonication to give a crude
lysate and the debris is removed by centrifugation to give a crude extract.
A 45–50% saturated ammonium sulfate cut was prepared. The 50%
saturated ammonium sulfate pellet was dissolved in buffer and diluted to
low salt and applied to a DEAE anion exchange column. The column was
washed at low salt and then eluted with a linear salt gradient from 0.1 to
0.6 M NaCl, the peak of activity eluting at about 0.25 M NaCl. The peak
was pooled and applied to a Sephacryl S-300 gel filtration column and
eluted with a buffer at constant salt (isocratically). The fractions of peak
activity were pooled and shown by SDS–PAGE with Coomassie blue
staining to be a single band. The specific activity of the final material is
the same as that of a known pure reference sample. The main fractions were
all assayed for enzyme activity and protein determinations were carried out.
The resulting data are given in Table 4.1.
A purification summary table should allow a reader to evaluate easily the
procedure and readily detect particularly effective and ineffective purifica-
tion steps. It should be easy to see if large losses occurred at a particular step.
A suitable table will contain the following columns:
1. Major steps in the purification. These typically include steps like:
Crude lysate (the result of cell or tissue disruption). This step is often
omitted since assays may be difficult but it is useful and even essential
when much of the expressed recombinant target protein is in an
insoluble inclusion body.
Crude extract (the lysate after any insoluble material has been removed
by centrifugation)

Table 4.1 A typical purification summary table

Total Total Specific

protein activity activity Yield Purity
Step (mg)b (units)c (units/mg) (%) (%)
Crude lysatea 1200 120 0.10 100 0.8
Crude extract 1000 110 0.11 92 0.9
Ammonium 180 75 0.42 62 3.4
sulfate
45–55% cutd
DEAE column 24 60 2.5 50 20
(pooled peak)
Sephacryl column 3.6 46 12.5 38 100
(pooled peak)
a
From 10 g of wet weight E. coli cell pellet (from 4 l of bacterial culture).
b
Protein concentration determined by Bradford assay using BSA as a standard protein.
c
Enzyme activity measured as described in the methods section.
d
Crude extract material that is soluble at 45% but precipitates at 55% saturated ammonium sulfate.
Preparing a Purification Summary Table 31

Ammonium sulfate cut

Pooled peak from an ion exchange column
Pooled peak from a gel filtration column
Pooled peak from an affinity column
Concentrated and dialyzed final product
Solubilized inclusion bodies. This step and the following two are often
used in the case where an expressed recombinant protein is produced
as insoluble inclusion bodies; see Chapter 17.
Washed inclusion bodies
Refolded, centrifuged, and concentrated material
2. Amount of total protein (mg). This is usually determined by a standard
protein assay. Most commonly these days a Bradford dye-binding assay
or a bicinchoninic acid (BCA) assay is used (see Chapter 8). It is
important to indicate in the methods section what protein is used as
the protein standard (typically BSA). Once the protein is purified it can
also sometimes be quantified by measuring its absorbance at 280 nm and
the use of an appropriate molar extinction coefficient.
3. Amount of target protein or total activity (mg or units). If there is a suitable
enzyme assay for the target protein, it should be carried out on material
from each major step. If the protein is not an enzyme or there is no
quantitative assay, and if the protein is visible on a Coomassie blue
stained SDS–PAGE, then often the stained gel is scanned and the
amount of protein in the target protein band is determined. In other
words, purity is determined and multiplied by the total protein to give an
estimate of the total target protein.
4. Specific activity (units/mg). If enzyme activity assays are possible, then the
total activity (units) is divided by the total protein (mg) to give specific
activity as units/mg.
5. Overall yield (%). The yield at a step in the procedure is the amount of
target (either total target protein or total activity) at that step divided by
the amount of target in the first step (defined as 100%).
6. Purity of target protein (%). Purity is often determined by scanning a
stained SDS–PAGE and measuring the amount of the stain associated
with the target band as a fraction of the stain associated with all the bands
on the gel. If one has a reliable assay, then if the final material is pure, its
specific activity can be used to define purity. For example, if an earlier
step has a specific activity 10% of the final pure material, then the purity
at that step would be 10%.
7. Relative or fold purification. This is not essential since it can be calculated
from the other values above, but it is often useful. This is merely setting
the initial purity at a value of one and then giving the purity at each step
relative to that of the first step. For example, in Table 4.1, the final
step represents an overall fold purification of 125.
32 Richard R. Burgess

2. The Importance of Footnotes

Every protein and purification is different and footnotes are needed
to help the reader understand what has been done. There should be a
footnote that indicates the amount of raw material used in the preparation
being summarized. For recombinant protein expressed in bacteria, for
example, one should always give the number of grams of wet weight
cell pellet used in the preparation. It is also useful to know the volume of
the bacterial culture used, but that in itself is not enough since, depending
on the growth media and conditions, the yield of wet weight cell pellet
can range from 1 to 80 g/L. Another useful footnote indicates how the
protein amount was determined (e.g., sometimes a Bradford assay is used
on the early steps, but absorbance and extinction coefficient is used on the
final product).

3. The Value of an SDS–Polyacrylamide Gel

Analysis on Main Protein Fractions
I find that an SDS–polyacrylamide gel image is a very valuable com-
plement to the purification summary table. If the same samples that repre-
sent the various steps in the purification table are also shown in a gel photo,
then it is particularly easy to see the progress of the various fractionation
steps toward production of a purified final target protein. The most useful
gels are ones on which an equal proportion of the material at each step is
loaded on the gel lanes.

4. Some Common Mistakes and Problems

1. Use too many significant figures. This is one of my pet peeves. I suspect that
the concept of significant figures is no longer taught, because I find that a
good 75% of the purification papers I review give protein amounts like
235.052 mg and yields like 46.72%. Just because a calculator or com-
puter can divide two numbers and give one the result to eight figures
does not mean that value is what one should enter into the table. Just
remember that most protein quantification methods or enzyme assays are
not accurate to better than 5–10%. When one writes 23.47 mg, one
implies that it is not 23.46 or 23.48, but 23.47. In other words, one is
implying that it is accurate to one part in over 2000 when it is not even
Preparing a Purification Summary Table 33

clear that it is accurate to one part in 20 (is it 22, 23, or 24?). No numbers
should be given to more than three significant figures and in general
most percentages can be given to two significant figures. Also, remember
that any number resulting from the division of two numbers each
accurate to
10% will only be accurate to
20%.
2. Calculate values erroneously. Remarkably many tables contain simple
arithmetic errors. All numbers should be checked and rechecked before
submission of a manuscript.
3. Use step yields instead of overall yields. A step yield is the yield from a single
step in the purification procedure, that is, the amount of target protein or
activity after that step compared to that in the previous step of the
procedure. A series of four fractionation steps might all give 60% step
yields, but the overall yield is (0.6)4 ¼ 0.6  0.6  0.6  0.6 ¼ 0.13 or
13%. Overall yield is more useful. A procedure that gives an overall yield
of a few percent may be due to a very lengthy and difficult purification of
a rare or unstable protein, but more likely it indicates that the procedure
has not been optimized very well.
4. Calculate yield as yield of total protein. Often I see a table in which the yield
given is yield of total protein, rather than target protein. This is relatively
useless information. Yield is always recovery of total target protein or
activity.
5. Write up and try to publish the first purification that gives any product. There is
a tendency for readers, especially inexperienced readers, to assume that a
published purification is the result of many cycles of improvement and
optimization, and represents the best way to purify the protein. This is
very often not true. Many times, it is just a series of steps, often chosen
arbitrarily that happens to result in some product. One should look very
carefully at a purification to assess if it is a procedure that is worth trying
to follow if one wants to purify some of the target protein. This is why a
proper purification summary table is so valuable. If huge losses occur at
a particular fractionation step, if the overall yield is very low, if the final
purity is not high, or if similar fractionation steps are used several times,
then perhaps the procedure should merely be used as a beginning point
in designing a better, more effective purification.
6. What to do when purifying a protein where a fusion partner is cleaved off during
the procedure? Very often a recombinant protein is expressed as a fusion
with another protein or tag that aids in its folding or purification
(see Chapter 16). Since most often the desired final product is the target
protein without the fusion partner or tag, especially for structural studies,
the fusion partner must be cleaved off by one of several specific proteases.
Let us say that the target is 20 kDa and the fusion partner is 40 kDa, so the
fusion protein expressed is 60 kDa. At the step where the fusion protein
is cleaved, the yield of target protein seems to decrease by 67% even if
all of it is recovered. How is this indicated in the summary table?
34 Richard R. Burgess

I suggest that the column on target protein amount contain two numbers;
the mg of fusion protein and the calculated mg of the final target protein in
parenthesis. That way at the step where the cleavage has occurred, one can
continue giving just the mg of the cleaved product and the theoretical
amount of cleaved product in the first step can be used to calculate the
overall yield.
 C H A P T E R F I V E

Setting Up a Laboratory
Murray P. Deutscher

Contents
1. Supporting Materials 38
1.1. Glassware and plasticware 38
1.2. Chemicals 38
1.3. Disposables 38
1.4. Small equipment and accessories 39
1.5. Equipment and apparatus 39
2. Detection and Assay Requirements 39
3. Fractionation Requirements 40

The aim of this chapter is to provide some general information on the

basic equipment, chemicals, and supplies that should be present in any labora-
tory undertaking protein purification. Details relevant to individual pieces of
equipment, information on apparatus and chemicals for specialized applica-
tions, useful vendors, etc., can be found in chapters throughout this volume.
Although any laboratory engaged in protein purification may have many
types of equipment, chemicals, and supplies, all these materials basically fall
into three categories, those used for fractionation, those needed for detec-
tion and assay, and those that I call supporting materials. The supporting
materials (e.g., tubes, pipets, baths, stirrers, timers, salts, buffers, and much
more) are common to every biochemical laboratory. They are generally the
least costly, used most frequently, required in largest numbers, and are the
most essential. It is natural in setting up a laboratory to focus on the large,
expensive apparatus, but in practice, available funds should first go to
ensuring an adequate supply of supporting materials. (It obviously makes
no sense to buy a sophisticated fraction collector, and not to have enough
tubes.) Obtain the necessary amount of glassware, chemicals, disposables,
etc., for the number of people who will be working in the laboratory.
A representative (but not complete) list follows.

Department of Biochemistry and Molecular Biology, University of Miami School of Medicine, Miami,
Florida, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63005-6 All rights reserved.

37
38 Murray P. Deutscher

1. Supporting Materials
1.1. Glassware and plasticware
Tubes, beakers, flasks, bottles, cylinders, funnels, and pipets in a wide
range of sizes (disposable materials are often most useful)
Transfer (Pasteur) pipets
Pipettor tips
Baking dishes
Plastic containers
Large carboys and jars
Ice buckets

1.2. Chemicals
High-grade distilled H2O
Salts (generally chlorides or acetates)
Sodium and potassium phosphates
Enzyme-grade ammonium sulfate
Tris and other organic buffers
EDTA
Acids and bases
Reducing agents (2-mercaptoethanol, dithiothreitol, glutathione)
Protease inhibitors
Detergents
Glycerol

1.3. Disposables
Dialysis tubing
Concentrators/fractionators
Weighing paper and boats
Filter paper
pH paper
Aluminum foil
Glass wool
Syringes and needles
Sterile gloves
Marking tape and pens
Paper wipes and towels
Razor blades
Setting Up a Laboratory 39

1.4. Small equipment and accessories

Burners and flints
Timers (including a stopwatch)
Vortex mixers
Magnetic stirrers and stirring bars of various sizes
Variable pipettors of various sizes
Forceps, scissors, spatulas
Small tools
Hot plate
Homogenizers
Thermometers
1.5. Equipment and apparatus
Refrigerator
Freezer ( 20  C) and, if funds available, a low-temperature freezer
Water baths (shaking and standing)
Balances (top loading and analytical)
pH meter, electrode, and standard solutions
Oven
Pump
Microfuge
Accessibility to a cold room, autoclave, ice machine, lyophilizer, dry ice,
or liquid N2.
If funds still remain after filling the above list, obtain other items directly
relevant to protein purification (although some of these could also be
considered supporting materials), that is, those necessary for detection and
assay and for fractionation. In these areas, some of the equipment could be
quite costly and sophisticated. A great deal of thought should be given to the
planned usage of such equipment to determine your actual needs. In some
cases it might be essential to have the item in your immediate laboratory.
However, in others, if only occasional use is contemplated, you might get
by with nearby access to the piece of equipment. With limited funds, and
the cost of some equipment, a priority list is very helpful. In some instances,
duplicating a frequently used item may be more advantageous than pur-
chasing a new piece of equipment that will only be used infrequently. Thus,
in my experience, a lab actively engaged in protein purification never has
enough fraction collectors, columns, and gel electrophoresis apparatus.

2. Detection and Assay Requirements

Probably, the most important detection device in the laboratory is the
spectrophotometer. It can be used for determining protein concentrations,
measuring the growth of bacterial cultures, as well as for a variety of
40 Murray P. Deutscher

enzymatic and colorimetric assays. The spectrophotometer should be

equipped with both UV and visible optics and cover the range from
about 200 to 800 nm to be of most use. Both glass and quartz cuvettes are
necessary to cover the visible and UV range, respectively. It is often useful to
have one set of microcuvettes for analysis of small volumes ( 0.2 ml).
Disposable cuvettes are available, and are best for measuring cell growth.
Most enzymatic assays rely on either spectrophotometry or the use of
radioisotopes. In the latter case, a scintillation counter is a necessity. The use
of a scintillation counter means that the supplies, chemicals, and other
accessories needed for preparing radioactive samples will also be required.
Scintillation counters are quite costly, and often are shared among several
laboratories. Likewise, a phosphorimager and a fluoroimager are essential
for gel assays, but are often shared. If the use of radioactive material is
contemplated, radiation monitors, shielding, and other precautions will be
needed as well.
Two other detection devices that often come in handy are a conductiv-
ity meter and a refractometer. These are used to measure salt gradients on
chromatographic columns, and sucrose, glycerol, or CsCl centrifugal
gradients.

3. Fractionation Requirements
Protein purification means protein fractionation. What distinguishes a
protein purification laboratory from the usual biochemistry or molecular
biology laboratory is largely the number and types of fractionation apparatus
and materials available. Subsequent chapters will discuss these items in
detail; they will be mentioned here only briefly.
Probably, the most frequently used piece of equipment in the laboratory
is the centrifuge. The workhorse of the protein purification laboratory is the
refrigerated high-speed centrifuge which attains speeds up to 20,000 rpm.
The usefulness of such a centrifuge is directly related to the presence in the
laboratory of a wide variety of rotors and centrifuge tubes and bottles.
Rotors are available that hold as small a volume as a few milliliters per
tube to ones that hold 500-ml bottles. The large rotors are invaluable for
handling the large volumes of extracts often encountered in early steps of a
protein purification.
In instances in which one wants to remove or prepare subcellular
organelles, access to an ultracentrifuge is desirable. Instruments are available
that can process reasonably large volumes at speeds as high as 80,000 rpm,
and smaller machines on the market can go even faster. The availability of
this instrumentation has greatly reduced the time required to prepare
microsomal or high-speed supernatant fractions. In view of the cost of
Setting Up a Laboratory 41

these machines, and their relatively infrequent use in most cases, they are
often shared among laboratories.
The advent of many microanalytical techniques has also made the
minifuge or microcentrifuge an essential item. Though not really a fraction-
ation apparatus in a protein purification laboratory, it is a necessary addition.
In this regard, the larger centrifuges are also frequently used for assays of
various types, rather than only for fractionation purposes.
In order to isolate proteins, a means of rupturing cells is required.
Various apparatuses are available for this purpose, including hand-held and
motor-driven homogenizers, blenders, sonicators, pressure cells, etc. These
will be discussed in detail in other chapters. In general, it is desirable to have
a variety of the less costly items in individual laboratories, with the remain-
der available as shared equipment.
Column chromatography is a primary protein purification method in
use in most laboratories. Every laboratory involved in protein fractionation
should have available a large supply of columns of various lengths and
diameters in anticipation of every conceivable need, since they will arise
during the course of developing purification schemes. Columns are avail-
able in various degrees of sophistication (and cost). We have found that
simple, gravity-flow, open-top columns fitted with stoppers and syringe
needles, or tubing, for fluid inlet and control, are satisfactory for most
chromatographic procedures. However, for many purposes, prepacked
columns and pressurized FPLC systems may be the preferred route.
A dependable fraction collector is one of the most important pieces of
equipment in the laboratory. Failure of a fraction collector may result in the
loss of several month’s work. In this instance, extra money spent on a good,
versatile machine is a wise investment. Instruments able to handle a large
number of tubes, of various sizes, in different collection modes, are the most
useful. Many different types of fraction collectors are available. Careful analysis
of the various models, and matching to anticipated requirements, is good
practice prior to purchase. A suitable strategy for many laboratories would
be the purchase of one of the more sophisticated instruments for special needs,
and one or more of the less costly, simple machines for routine use.
A number of other accessories to column chromatography are useful, if
not essential. These include a peristaltic pump, various sizes of gradient
makers, and a UV monitor. Gradient makers can be homemade from flasks
or bottles, if necessary. Following the protein elution profile during a
chromatographic run provides important information. This can be done
by determining the absorbance of individual fractions with a spectropho-
tometer, or automatically with an in-stream UV monitor. Dual-wavelength
models with different size flow cells are the most versatile (and also most
costly).
Finally, every laboratory should also have on hand a basic supply of
chromatographic gels and resins. These should include an anion and cation
42 Murray P. Deutscher

exchanger, various porosity gel filtration media, hydroxyapatite, a hydro-

phobic gel, and probably an immobilized dye resin.
No protein purification laboratory is complete without the presence of
gel electrophoresis equipment. These items are used to monitor a purifica-
tion procedure or for fractionation itself. Generally, a vertical slab-gel
apparatus with various-sized spacers and combs is satisfactory for most
applications. An electrophoresis power supply unit is also required. If only
one is to be purchased, a regulated constant-current (0–50 mA) constant-
voltage (0–200 V) model is useful. In addition to the chemicals necessary for
preparing gels, several protein standards and staining dyes should also be
obtained. The instrumentation used for gel electrophoresis can also be used
for isoelectric focusing.
The equipment, chemicals, and supplies mentioned in this chapter
should allow you to enter the field of protein purification. As you read
through this volume, and actually begin to purify proteins, many other
useful items will become apparent.
 C H A P T E R S I X

Buffers: Principles and Practice1

Vincent S. Stoll and John S. Blanchard

Contents
1. Introduction 43
2. Theory 44
3. Buffer Selection 45
4. Buffer Preparation 48
5. Volatile Buffers 49
6. Broad-Range Buffers 50
7. Recipes for Buffer Stock Solutions 50
References 56

1. Introduction
The necessity for maintaining a stable pH when studying enzymes is
well established (Good and Izawa, this series; Johnson and Metzler, this
series). Biochemical processes can be severely affected by minute changes in
hydrogen ion concentrations. At the same time, many protons may be
consumed or released during an enzymatic reaction. It has become increas-
ingly important to find buffers to stabilize hydrogen ion concentrations
while not interfering with the function of the enzyme being studied. The
development of a series of N-substituted taurine and glycine buffers by
Good et al. (1966) has provided buffers in the physiologically relevant range
(6.1–10.4) of most enzymes, which have limited side effects with most
enzymes (Good et al., 1966). It has been found that these buffers are
nontoxic to cells at 50 mM concentrations and in some cases much higher
(Ferguson et al., 1980).

Department of Biochemistry, Albert Einstein College of Medicine, Bronx, New York

1
Reprinted from Methods in Enzymology, Volume 182 (Academic Press, 1990)

Methods in Enzymology, Volume 463 Copyright # 1990 by Academic Press

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63006-8 All rights reserved.

43
44 Vincent S. Stoll and John S. Blanchard

2. Theory
The observation that partially neutralized solutions of weak acids or
weak bases are resistant to pH changes on the addition of small amounts of
strong acid or strong base leads to the concept of ‘‘buffering’’ (Perrin and
Dempsey, 1974). Buffers consist of an acid and its conjugate base, such as
carbonate and bicarbonate, or acetate and acetic acid. The quality of a buffer
is dependent on its buffering capacity (resistance to change in pH by
addition of strong acid or base), and its ability to maintain a stable pH
upon dilution or addition of neutral salts. Because of the following equili-
bria, additions of small amounts of strong acid or strong base result in the
removal of only small amounts of the weakly acidic or basic species;
therefore, there is little change in the pH:
HAðacidÞ Ð Hþ þ A ðconjugate baseÞ ð6:1Þ
BðbaseÞ þ Hþ Ð BHþ ðconjugate acidÞ ð6:2Þ
The pH of a solution of a weak acid or base may be calculated from the
Henderson–Hasselbalch equation:
½basic species
pH ¼ pKa0 þ log ð6:3Þ
½acidic species
The pKa of a buffer is that pH where the concentrations of basic and
acidic species are equal, and in this basic form the equation is accurate
between the pH range of 3–11. Below pH 3 and above pH 11 the
concentrations of the ionic species of water must be included in the
equation (Perrin and Dempsey, 1974). Since the pH range of interest here
is generally in the pH 3–11 range, this will be ignored.
From the Henderson–Hasselbalch equation an expression for buffer
capacity may be deduced. If at some concentration of buffer, c, the sum
[A] þ [HA] is constant, then the amount of strong acid or base needed to
cause a small change in pH is given by the relationship:
( )
d½B ½Ka0 c½Hþ  K
þ ½Hþ  þ þ
w
¼ 2:303 ð6:4Þ
dpH ðKa0 þ ½Hþ Þ2 H
In this equation, Kw refers to the ionic product of water, and the second
and third terms are only significant below pH 3 or above pH 11. In the pH
range of interest (pH 3–11), this equation yields the following expression:
2:303c
bmax ¼ ¼ 0:576c; ð6:5Þ
4
which represents a maximum value for d[B]/dpH when pH ¼ pKa. The
buffer capacity of any buffer is dependent on the concentration, c, and may
Buffers: Principles and Practice 45

0.025

b
0.015

0.005
−1.0 −0.5 0.0 0.5 1.0
ΔpH

Figure 6.1 Buffer capacity (b) versus DpH over the range  1 pH unit of the pKa for
HEPES (0.05 M). Points calculated using Eq. (6.5), and data from Perrin and Dempsey
(1974).

be calculated over a buffer range of  1 pH unit around the pK to determine

the buffer capacity, as shown in Fig. 6.1 for one of the Good buffers,
HEPES. It can be seen that the buffer capacity is greatest at its pK, and
drops off quickly 1 pH unit on either side of the pK. In practice, buffers
should not be used beyond these values.

3. Buffer Selection
There are many factors that must be considered when choosing a buffer.
When studying an enzyme one must consider the pH optimum of the enzyme,
nonspecific buffer effects on the enzyme, and interactions with substrates or
metals. When purifying a protein, cost becomes an important consideration, as
does the compatibility of the buffer with different purification techniques.
Table 6.1 lists a wide variety of buffers covering a broad pH range.
Determining the pH optimum of a protein is a first step in determining the
best buffer to employ (Blanchard, this series). Since the buffering capacity is
maximal at the pK, buffers should be used close to this value. When determin-
ing the pH optimum for an enzyme, it is useful to use a series of related buffers
that span a wide pH range. Once an optimal pH has been approximated,
different buffers within this pH range can be examined for specific buffer effects.
The Good buffers have been shown to be relatively free of side effects.
However, inorganic buffers do have a high potential for specific buffer
effects. Many enzymes are inhibited by phosphate buffer, including car-
boxypeptidase, urease, as well as many kinases and dehydrogenases
(Blanchard, this series). Borate buffers can form covalent complexes with
mono- and oligosaccharides, the ribose moieties of nucleic acids, pyridine
nucleotides, and other gem-diols. Tris and other primary amine buffers may
form Schiff base adducts with aldehydes and ketones.
46 Vincent S. Stoll and John S. Blanchard

Table 6.1 Selected buffers and their pK values at 25  C

Trivial name Buffer name pKa dpKa/dt

Phosphate (pK1) – 2.15 0.0044
Malate (pK1) – 3.40 –
Formate – 3.75 0.0
Succinate (pK1) – 4.21  0.0018
Citrate (pK2) – 4.76  0.0016
Acetate – 4.76 0.0002
Malate – 5.13 –
Pyridine – 5.23  0.014
Succinate (pK2) – 5.64 0.0
MES 2-(N-Morpholino) 6.10  0.011
ethanesulfonic acid
Cacodylate Dimethylarsinic acid 6.27
Dimethylglutarate 3,3-Dimethylglutarate (pK2) 6.34 0.0060
Carbonate (pK1) – 6.35  0.0055
Citrate (pK3) – 6.40 0.0
Bis–Tris [Bis(2-hydroxyethyl)imino]tris 6.46 0.0
(hydroxymethyl)methane
ADA N-2-Acetamidoiminodiacetic 6.59  0.011
acid
Pyrophosphate – 6.60 –
EDPS (pK1) N,N 0 -Bis(3-sulfopropyl) 6.65 –
ethylenediamine
Bis–Tris propane 1,3-Bis[tris(hydroxymethyl) 6.80 –
methylamino]propane
PIPES Piperazine-N,N 0 -bis(2- 6.76  0.0085
ethanesulfonic acid)
ACES N-2-Acetamido-2- 6.78  0.020
hydroxyethanesulfonic acid
MOPSO 3-(N-Morpholino)-2- 6.95  0.015
hydroxypropanesulfonic acid
Imidazole – 6.95  0.020
BES N,N-Bis-(2-hydroxyethyl)-2- 7.09  0.016
aminoethanesulfonic acid
MOPS 3-(N-Morpholino) 7.20 0.015
propanesulfonic acid
Phosphate (pK2) – 7.20  0.0028
EMTA 3,6-Endomethylene-1,2,3,6- 7.23 –
tetrahydrophthalic acid
TES 2-[Tris(hydroxymethyl) 7.40  0.020
methylamino]ethanesulfonic
acid
x
Buffers: Principles and Practice 47

Table 6.1 (continued)

Trivial name Buffer name pKa dpKa/dt

HEPES N-2-Hydroxyethylpiperazine- 7.48  0.014
N 0 -2-ethanesulfonic acid
DIPSO 3-[N-Bis(hydroxyethyl)amino]- 7.60  0.015
2-hydroxypropanesulfonic
acid
TEA Triethanolamine 7.76  0.020
POPSO Piperazine-N,N 0 -bis(2- 7.85  0.013
hydroxypropanesulfonic acid)
EPPS, HEPPS N-2-Hydroxyethylpiperazine- 8.00 –
N 0 -3-propanesulfonic acid
Tris Tris(hydroxymethyl) 8.06  0.028
aminomethane
Tricine N-[Tris(hydroxymethyl)methyl] 8.05  0.021
glycine
Glycinamide – 8.06  0.029
PIPPS 1,4-Bis(3-sulfopropyl)piperazine 8.10 –
Glycylglycine – 8.25  0.025
Bicine N,N-Bis(2-hydroxyethyl)glycine 8.26  0.018
TAPS 3-{[Tris(hydroxymethyl)methyl] 8.40 0.018
amino}propanesulfonic acid
Morpholine – 8.49 –
PIBS 1,4-Bis(4-sulfobutyl)piperazine 8.60 –
AES 2-Aminoethylsulfonic acid, 9.06  0.022
taurine
Borate – 9.23  0.008
Ammonia – 9.25  0.031
Ethanolamine – 9.50  0.029
CHES Cyclohexylaminoethanesulfonic 9.55 0.029
acid
Glycine (pK2) – 9.78  0.025
EDPS N,N 0 -Bis(3-sulfopropyl) 9.80 –
ethylenediamine
APS 3-Aminopropanesulfonic acid 9.89 –
Carbonate (pK2) – 10.33  0.009
CAPS 3-(Cyclohexylamino) 10.40 0.032
propanesulfonic acid
Piperidine – 11.12 –
Phosphate (pK3) – 12.33  0.026

Buffer complexation with metals may present additional problems. In this

respect, inorganic buffers can prove problematic in that they may remove, by
chelation, metals essential to enzymatic activity (e.g., Mg2þ for kinases,
48 Vincent S. Stoll and John S. Blanchard

Cu2þ or Fe2þ for hydroxylases). Release of protons upon chelation or precip-

itation of metal–buffer complexes may also be a potential problem. Where
metal chelation presents a problem, the Good buffers are useful since they have
been shown to have low metal-binding capabilities (Good et al., 1966).
Once a suitable buffer has been found (noninteracting, with an appro-
priate pK ), a concentration should be chosen. Since high ionic strength may
decrease enzyme activity, the buffer concentration should be as low as
possible (Blanchard, this series). A reasonable way to determine how low
a concentration may be used is to examine the properties (reaction rate or
protein stability) at a low (10–20 mM) concentration of buffer. The pH
prior to, and an adequate time after, addition of protein should not vary
more than  0.05 pH. If the pH changes too drastically (greater than  0.1
pH unit), then the buffer concentration should be raised to 50 mM. In cases
where protons are consumed or released stoichiometrically with substrate
utilization, pH stability becomes increasingly important.
Buffers may be made up in stock solutions, then diluted for use. When
stock solutions are made, it should be done close to the working tempera-
ture, and in glass bottles (plastic bottles can leach UV-absorbing material)
(Perrin and Dempsey, 1974). Buffers have temperature-sensitive pK values,
particularly amine buffers. The carboxylic acid buffers are generally the least
sensitive to temperature, and the Good buffers have only a small inverse
temperature dependence on pK. The effects of dilution of stock solutions,
or addition of salts, on pH should be checked by measurement of the pH
after addition of all components.
Choosing a buffer for protein purification requires some special con-
siderations. Large amounts of buffer will be needed for centrifugation,
chromatographic separations, and dialysis, which makes cost a concern.
Tris and many inorganic buffers are widely used since they are relatively
inexpensive. Although buffers like Tris are inexpensive, and have been
widely used in protein purification, they do have disadvantages. Tris is a
poor buffer below pH 7.5 and its pK is temperature dependent (a solution
made up to pH 8.06 at 25  C will have a pH of 8.85 at 0  C). Many primary
amine buffers such as Tris and glycine (Bradford, 1976) will interfere with
the Bradford dye-binding protein assay. Some of the Good buffers, HEPES,
EPPS, and Bicine, give false-positive colors with Lowry assay.
Spectroscopic measurement of enzyme rates is a commonly applied
method. It may be important to use a buffer that does not absorb appreciably
in the spectral region of interest. The Good buffers, and most buffers listed
in Table 6.1, can be used above 240 nm.

4. Buffer Preparation
Once a suitable buffer has been chosen it must be dissolved and titrated to
the desired pH. Before titrating a buffer solution, the pH meter must be
calibrated. Calibration should be done using commercially available pH
Buffers: Principles and Practice 49

standards, bracketing the desired pH. If monovalent cations interfere, or are

being investigated, then titration with tetramethylammonium hydroxide can
be done to avoid mineral cations. Similarly, the substitution of the most
commonly used counteranion, chloride, with other anions such as acetate,
sulfate, or glutamate, may have significant effects on enzyme activity or
protein–DNA interactions (Leirmo et al., 1987). Stock solutions should be
made with quality water (deionized and double-distilled, preferably) and
filtered through a sterile ultrafiltration system (0.22 mm) to prevent bacterial
or fungal growth, especially with solutions in the pH 6–8 range. To prevent
heavy metals from interfering, EDTA (10–100 mM) may be added to chelate
any contaminating metals.

5. Volatile Buffers
In certain cases, it is necessary to remove a buffer quickly and
completely. Volatile buffers make it possible to remove components that
may interfere in subsequent procedures. Volatile buffers are useful in elec-
trophoresis, ion-exchange chromatography, and digestion of proteins fol-
lowed by separation of peptides or amino acids. Most of the volatile buffers
(Table 6.2) are transparent in the lower UV range except for the buffers

Table 6.2 Types of systems for use as volatile buffersa

System pH range
87 ml glacial acetic acid þ 25 ml 88% HCOOH in 11 l 1.9
25 ml 88% HCOOH in 1 l 2.1
Pyridine–formic acid 2.3–3.5
Trimethylamine–formic acid 3.0–5.0
Triethylamine–formic (or acetic) acid 3–6
5 ml pyridine þ 100 ml glacial acetic acid in 1 l 3.1
5 ml pyridine þ 50 ml glacial acetic acid in 1 l 3.5
Trimethylamine–acetic acid 4.0–6.0
25 ml pyridine þ 25 ml glacial acetic acid in 1 l 4.7
Collidine–acetic acid 5.5–7.0
100 ml pyridine þ 4 ml glacial acetic acid in 1 l 6.5
Triethanolamine–HCl 6.8–8.8
Ammonia–formic (or acetic) acid 7.0–10.0
Trimethylamine–CO2 7–12
Triethylamine–CO2 7–12
24 g NH4HCO3 in 1 l 7.9
Ammonium carbonate–ammonia 8.0–10.5
Ethanolamine–HCl 8.5–10.5
20 g (NH4)2CO3 in 1 l 8.9
a
From Perrin and Dempsey (1974).
50 Vincent S. Stoll and John S. Blanchard

containing pyridine (Perrin and Dempsey, 1974). An important consider-

ation is interference in amino acid analysis (i.e., reactions with ninhydrin).
Most volatile buffers will not interfere with ninhydrin if the concentrations
are not too high (e.g., triethanolamine less than 0.1 M does not interfere).

6. Broad-Range Buffers
There may be occasions where a single buffer system is desired that can
span a wide pH range of perhaps 5 or more pH units. One method would be
a mixture of buffers that sufficiently covers the pH range of interest. This
may lead to nonspecific buffer interactions for which corrections must be
made. Another common approach is to use a series of structurally related
buffers that have evenly spaced pK values such that each pK is separated by
approximately  1 pH unit (the limit of buffering capacity). The Good
buffers are ideal for this approach since they are structurally related and have
relatively evenly spaced pK values. As the pH passes the pK of one buffer it
becomes nonparticipatory and therefore has no further function. These
nonparticipating buffer components may show nonspecific buffer effects
as well as raising the ionic strength with potential deleterious effects.
A detailed description of buffer mixtures which provide a wide range of
buffering capacity with constant ionic strength is available (Ellis and
Morrison, this series).

7. Recipes for Buffer Stock Solutions

1. Glycine–HCl buffer (Sorensen, 1909a,b) stock solutions
 A: 0.2 M solution of glycine (15.01 g in 1000 ml)
 B: 0.2 M HCl
50 ml of A þ x ml of B, diluted to a total of 200 ml:

x pH x pH
5.0 3.6 16.8 2.8
6.4 3.4 24.2 2.6
8.2 3.2 32.4 –

2. Citrate buffer (Lillie, 1948) stock solutions

 A: 0.1 M solution of citric acid (21.01 g in 1000 ml)
 B: 0.1 M solution of sodium citrate (29.41 g C6H5O7Na3  2H2O in
1000 ml)
x ml of A þ y ml of B, diluted to a total of 100 ml:
Buffers: Principles and Practice 51

x y pH
46.5 3.5 3.0
43.7 6.3 3.2
40.0 10.0 3.4
37.0 13.0 3.6
35.0 15.0 3.8
33.0 17.0 4.0
31.5 18.5 4.2
28.0 22.0 4.4
25.5 24.5 4.6
23.0 27.0 4.8
20.5 29.5 5.0
18.0 32.0 5.2
16.0 34.0 5.4
13.7 36.3 5.6
11.8 38.2 5.8
9.5 41.5 6.0
7.2 42.8 6.2

3. Acetate buffer (Walpole, 1914) stock solutions

 A: 0.2 M solution of acetic acid (11.55 ml in 1000 ml)
 B: 0.2 M solution of sodium acetate (16.4 g of C2H3O2Na or 27.2 g
of C2H3O2Na  3H2O in 1000 ml)
x ml of A þ y ml of B, diluted to a total of 100 ml:

x y pH
46.3 3.7 3.6
44.0 6.0 3.8
41.0 9.0 4.0
36.8 13.2 4.2
30.5 19.5 4.4
25.5 24.5 4.6
14.8 35.2 5.0
10.5 39.5 5.2
8.8 41.2 5.4
4.8 45.2 5.6

4. Citrate–phosphate buffer (McIlvaine, 1921) stock solutions

 A: 0.1 M solution of citric acid (19.21 g in 1000 ml)
 B: 0.2 M solution of dibasic sodium phosphate (53.65 g of
Na2HPO4  7H2O or 71.7 g of Na2HPO4  12H2O in 1000 ml)
x ml of A þ y ml of B, diluted to a total of 100 ml:
52 Vincent S. Stoll and John S. Blanchard

x y pH
44.6 5.4 2.6
42.2 7.8 2.8
39.8 10.2 3.0
37.7 12.3 3.2
35.9 14.1 3.4
33.9 16.1 3.6
32.3 17.7 3.8
30.7 19.3 4.0
29.4 20.6 4.2
27.8 22.2 4.4
26.7 23.3 4.6
25.2 24.8 4.8
24.3 25.7 5.0
23.3 26.7 5.2
22.2 27.8 5.4
21.0 29.0 5.6
19.7 30.3 5.8
17.9 32.1 6.0
16.9 33.1 6.2
15.4 34.6 6.4
13.6 36.4 6.6
9.1 40.9 6.8
6.5 43.6 7.0
5. Succinate buffer (G. Gomori, unpublished observation) stock solutions
 A: 0.2 M solution of succinic acid (23.6 g in 1000 ml)
 B: 0.2 M NaOH
25 ml of A þ x ml of B, diluted to a total of 100 ml:

x pH x pH
7.5 3.8 26.7 5.0
10.0 4.0 30.3 5.2
13.3 4.2 34.2 5.4
16.7 4.4 37.5 5.6
20.0 4.6 40.7 5.8
23.5 4.8 43.5 6.0
6. Cacodylate buffer (Plumel, 1949) stock solutions
 A: 0.2 M solution of sodium cacodylate (42.8 g of Na(CH3)2AsO2 
3H2O in 1000 ml)
 B: 0.2 M NaOH
50 ml of A þ x ml of B, diluted to a total of 200 ml:
Buffers: Principles and Practice 53

x pH x pH
2.7 7.4 29.6 6.0
4.2 7.2 34.8 5.8
6.3 7.0 39.2 5.6
9.3 6.8 43.0 5.4
13.3 6.6 45.0 5.2
18.3 6.4 47.0 5.0
13.8 6.2
7. Phosphate buffer (Sorensen, 1909a,b) stock solutions
 A: 0.2 M solution of monobasic sodium phosphate (27.8 g in 1000 ml)
 B: 0.2 M solution of dibasic sodium phosphate (53.65 g of Na2HPO4
 7H2O or 71.7 g of Na2HPO4  12H2O in 1000 ml)
x ml of A þ y ml of B, diluted to a total of 200 ml:

x y pH x y pH
93.5 6.5 5.7 45.0 55.0 6.9
92.0 8.0 5.8 39.0 61.0 7.0
90.0 10.0 5.9 33.0 67.0 7.1
87.7 12.3 6.0 28.0 72.0 7.2
85.0 15.0 6.1 23.0 77.0 7.3
81.5 18.5 6.2 19.0 81.0 7.4
77.5 22.5 6.3 16.0 84.0 7.5
73.5 26.5 6.4 13.0 87.0 7.6
68.5 31.5 6.5 10.5 90.5 7.7
62.5 37.5 6.6 8.5 91.5 7.8
56.5 43.5 6.7 7.0 93.0 7.9
51.0 49.0 6.8 5.3 94.7 8.0

8. Barbital buffer (Michaelis, 1930) stock solutions

 A: 0.2 M solution of sodium barbital (veronal) (41.2 g in 1000 ml)
 B: 0.2 M HCl
50 ml of A þ x ml of B, diluted to a total of 200 ml:

x pH x pH
1.5 9.2 22.5 7.8
2.5 9.0 27.5 7.6
4.0 8.8 32.5 7.4
6.0 8.6 39.0 7.2
9.0 8.4 43.0 7.0
2.7 8.2 45.0 6.8
17.5 8.0
54 Vincent S. Stoll and John S. Blanchard

Solutions more concentrated than 0.05 M may crystallize on standing,

especially in the cold.
9. Tris(hydroxymethyl)aminomethane (Tris) buffer (Hayaishi, this series)
stock solutions
 A: 0.2 M solution of tris(hydroxymethyl)aminomethane (24.2 g in
1000 ml)
 B: 0.2 M HCl
50 ml of A þ x ml of B, diluted to a total of 200 ml:

x pH
5.0 9.0
8.1 8.8
12.2 8.6
16.5 8.4
21.9 8.4
26.8 8.0
32.5 7.8
38.4 7.6
41.4 7.4
44.2 7.2
10. Boric acid–borax buffer (Holmes, 1943) stock solutions
 A: 0.2 M solution of boric acid (12.4 g in 1000 ml)
 B: 0.05 M solution of borax (19.05 g in 1000 ml; 0.2 M in terms of
sodium borate)
50 ml of A þ x ml of B, diluted to a total of 200 ml:

x pH x pH
2.0 7.6 22.5 8.7
3.1 7.8 30.0 8.8
4.9 8.0 42.5 8.9
7.3 8.2 59.0 9.0
11.5 8.4 83.0 9.1
17.5 8.6 115.0 9.2
11. 2-Amino-2-methyl-1,3-propanediol (Ammediol) buffer (Gomori,
1946) stock solutions
 A: 0.2 M solution of 2-amino-2-methyl-1,3-propanediol (21.03 g in
1000 ml)
 B: 0.2 M HCl
50 ml of A þ x ml of B, diluted to a total of 200 ml:
Buffers: Principles and Practice 55

x pH x pH
2.0 10.0 22.0 8.8
3.7 9.8 29.5 8.6
5.7 9.6 34.0 8.4
8.5 9.4 37.7 8.2
12.5 9.2 41.0 8.0
16.7 9.0 43.5 7.8
12. Glycine–NaOH buffer (Sorensen, 1909a,b) stock solutions
 A: 0.2 M solution of glycine (15.01 g in 1000 ml)
 B: 0.2 M NaOH
50 ml of A þ x ml of B, diluted to a total of 200 ml:

x pH x pH
4.0 8.6 22.4 9.6
6.0 8.8 27.2 9.8
8.8 9.0 32.0 10.0
12.0 9.2 38.6 10.4
16.8 9.4 45.5 10.6
13. Borax–NaOH buffer (Clark and Lubs, 1917) stock solutions
 A: 0.05 M solution of borax (19.05 g in 1000 ml; 0.02 M in terms of
sodium borate)
 B: 0.2 M NaOH
50 ml of A þ x ml of B, diluted to a total of 200 ml:

x pH
0.0 9.28
7.0 9.35
11.0 9.4
17.6 9.5
23.0 9.6
29.0 9.7
34.0 9.8
38.6 9.9
43.0 10.0
46.0 10.1
14. Carbonate–bicarbonate buffer (Delory and King, 1945) stock solutions
 A: 0.2 M solution of anhydrous sodium carbonate (21.2 g in
1000 ml)
 B: 0.2 M solution of sodium bicarbonate (16.8 g in 1000 ml)
x ml of A þ y ml of B, diluted to a total of 200 ml:
56 Vincent S. Stoll and John S. Blanchard

x y pH
4.0 46.0 9.2
7.5 42.5 9.3
9.5 40.5 9.4
13.0 37.0 9.5
16.0 34.0 9.6
19.5 30.5 9.7
22.0 28.0 9.8
25.0 25.0 9.9
27.5 22.5 10.0
30.0 20.0 10.1
33.0 17.0 10.2
35.5 14.5 10.3
38.5 11.5 10.4
40.5 9.5 10.5
42.5 7.5 10.6
45.0 5.0 10.7

REFERENCES
Blanchard, J. S. (this series). Vol. 104, p. 404.
Bradford, M. M. (1976). Anal. Biochem. 22, 248.
Clark, W. M., and Lubs, H. A. (1917). J. Bacteriol. 2, 1.
Delory, G. E., and King, E. J. (1945). Biochem. J. 39, 245.
Ellis, K. J., and Morrison, J. F. (this series). Vol. 87, p. 405.
Ferguson, W. J., et al. (1980). Anal. Biochem. 104, 300.
Gomori, G. (1946). Proc. Soc. Exp. Biol. Med. 62, 33.
Gomori, G. Unpublished observations.
Good, N. E., and Izawa, S. (this series). Vol. 24, p. 53.
Good, N. E., Winget, G. D., Winter, W., Connolly, T. N., Izawa, S., and Singh, R. M. M.
(1966). Biochemistry 5, 467.
Hayaishi, O. (this series). Vol. 1, p. 144.
Holmes, W. (1943). Anat. Rec. 86, 163.
Johnson, R. J., and Metzler, D. E. (this series). Vol. 22, p. 3.
Leirmo, S., Harrison, C., Cayley, D. S., Burgess, R. R., and Record, M. T. (1987).
Biochemistry 26, 2095.
Lillie, R. D. (1948). Histopathologic Technique. Blakiston, Philadelphia, PA.
McIlvaine, T. C. (1921). J. Biol. Chem. 49, 183.
Michaelis, L. (1930). J. Biol. Chem. 87, 33.
Perrin, D. D., and Dempsey, B. (1974). Buffers for pH and Metal Ion Control.
Chapman & Hall, London.
Plumel, M. (1949). Bull. Soc. Chim. Biol. 30, 129.
Sorensen, S. P. L. (1909a). Biochem. Z. 21, 131.
Sorensen, S. P. L. (1909b). Biochem. Z. 22, 352.
Walpole, G. S. (1914). J. Chem. Soc. 105, 2501.
 C H A P T E R S E V E N

Measurement of Enzyme Activity

T. K. Harris and M. M. Keshwani

Contents
1. Introduction 58
2. Principles of Catalytic Activity 58
2.1. Chemical kinetics 58
2.2. Basic enzyme kinetics 62
3. Measurement of Enzyme Activity 64
3.1. Continuous assays 65
3.2. Discontinuous assays 67
4. Formulation of Reaction Assay Mixtures 69
5. Discussion 71
Acknowledgments 71
References 71

Abstract
To study and understand the nature of living cells, scientists have continually
employed traditional biochemical techniques aimed to fractionate and charac-
terize a designated network of macromolecular components required to carry
out a particular cellular function. At the most rudimentary level, cellular func-
tions ultimately entail rapid chemical transformations that otherwise would not
occur in the physiological environment of the cell. The term enzyme is used to
singularly designate a macromolecular gene product that specifically and
greatly enhances the rate of a chemical transformation. Purification and char-
acterization of individual and collective groups of enzymes has been and will
remain essential toward advancement of the molecular biological sciences; and
developing and utilizing enzyme reaction assays is central to this mission. First,
basic kinetic principles are described for understanding chemical reaction rates
and the catalytic effects of enzymes on such rates. Then, a number of methods
are described for measuring enzyme-catalyzed reaction rates, which mainly
differ with regard to techniques used to detect and quantify concentration
changes of given reactants or products. Finally, short commentary is given
toward formulation of reaction mixtures used to measure enzyme activity.
Whereas a comprehensive treatment of enzymatic reaction assays is not within

Department of Biochemistry and Molecular Biology, Miller School of Medicine, University of Miami,
Miami, Florida, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63007-X All rights reserved.

57
58 T. K. Harris and M. M. Keshwani

the scope of this chapter, the very core principles that are presented should
enable new researchers to better understand the logic and utility of any given
enzymatic assay that becomes of interest.

1. Introduction
One primary goal of biological sciences is to deduce the molecular
bases of all chemical processes that take place in living organisms. It is well
established that the vast majority of biochemical reactions are governed by
protein gene products called enzymes. In this sense, enzymes behave as
finely tuned chemical catalysts, which greatly enhance reaction rates with
both temporal and spatial resolution. Thus, elucidation of biochemical
pathways involving biosynthesis, modification, and degradation of macro-
molecular, metabolic, and signaling molecules remains ultimately tied to
experimental methods for purification, reconstitution, and direct demon-
stration of enzymes to catalyze specific chemical reactions. The focus of this
chapter is to explain the most fundamental principles that must be consid-
ered in developing effective methods for measuring the progress of enzyme-
catalyzed reactions. First, kinetic formulation is described for chemical and
enzymatic reaction processes. Next, instructive commentary is given on
various experimental methods that are commonly used for measuring
enzyme activity; and consideration is given toward numerous procedures
and conditions that must be optimized.

2. Principles of Catalytic Activity

2.1. Chemical kinetics
Chemical kinetics involves the experimental study of reaction rates in order
to infer about the kinetic mechanisms for chemical conversion of reactants
(R) into products (P) (Fig. 7.1) (House, 2007; Laidler, 1987). For any given
chemical reaction, (i) the mechanism refers to the sequence of elementary
steps by which overall chemical change occurs and (ii) an elementary step
refers to the passing of a reactant or reaction complex through a single-
transition state to a chemical form with a defined and detectable lifetime.
If chemical conversion of reactants (R) to products (P) involves more than
one elementary step, then the chemical structures that exist after each
elementary step preceding final product formation are defined as reaction
intermediates (I). Figure 7.1 also shows that each elementary step leading to
product is characterized by a forward microscopic rate constant (e.g., kþ 1
and kþ 2). If a particular elementary step is reversible, then a reverse rate
Measurement of Enzyme Activity 59

A k+1 k+2
R1 + R2 I P1 + P2
k−1

Reaction coordinate
B k+1 k+2
E+S ES E +P
k−1

Figure 7.1 (A) Basic kinetic mechanism for chemical conversion of two molecular
reactants (R1 and R2) into two molecular products (P1 and P2). In this mechanism, an
intermediate complex (I) is reversibly formed from reactants (kþ 1 and k 1) but is
irreversibly converted to products (kþ 2). (B) Basic kinetic mechanism for enzyme
(E)-catalyzed conversion of a molecular substrate (S) into a molecular product (P). In
this mechanism, an ES intermediate complex is reversibly formed (kþ 1 and k 1) but is
irreversibly converted to product (kþ 2). In this case, the enzyme (E) is regenerated and
can undergo subsequent catalytic cycles. For both mechanisms, a free-energy diagram
depicts typical energy changes that occur along the reaction coordinate leading to
intermediates and products. Ground-state energies for reactants, intermediates, and
products are depicted as valleys, whereas transition-state energies for chemical changes
are depicted by peaks.

constant is designated (e.g., k 1). The principle of microscopic reversibility

states that only one transition state exists for reversible chemical conversion
between the species associated with one elementary step (i.e., the transition
state for the forward reaction is the same as for the reverse reaction).
For any given unidirectional conversion within an elementary step, a rate
law is used to describe the mathematical relationship between the reaction
rate or velocity and the concentration of each given reactant. For a given
chemical reaction step occurring in solution, the instantaneous reaction
velocity (v) is defined as the derivative of chemical concentration with respect
to time (t); and it is expressed either as (i) a decreasing concentration of a
given reactant (v ¼  d[R]/dt) or (ii) an increasing concentration of a given
product (v ¼ þ d[P]/dt). The standard international units of reaction velo-
city are M s 1. The rate law for a single-unidirectional step shows the
instantaneous velocity to depend on variable reactant concentration values
according to parameter values of (i) a proportionality constant (k) and
(ii) the exponent of each concentration term (Table 7.1). The proportion-
ality constant is defined as the forward microscopic rate constant for a given
chemical conversion, and its dimensions depend on the overall molecularity
or order of the reaction. For the rate laws defined in Table 7.1, the molecu-
larity with respect to a given reactant is given by the exponent in the given
concentration term and represents the stoichiometric quantity of reactant
involved in the reaction step. The kinetic order of a simple chemical reaction
60 T. K. Harris and M. M. Keshwani

Table 7.1 Relationship between rate law, order, and the rate constant, k

Rate lawa Order Units of k

v¼k Zero M s 1
v ¼ k[R1] First order with respect to R1 s 1
First order overall
v ¼ k[R1]2 Second order with respect to R1 M 1 s 1
Second order overall
v ¼ k[R1][R2] First order with respect to R1 s 1
First order with respect to R2
Second order overall
a
For each rate law, the units of the reaction velocity (v) will always be M s 1. M
represents moles per liter (mol l 1).

is usually the same as molecularity, but it is important to emphasize that

order refers to the exponent value that relates the experimentally measured
reaction rate’s dependence on reactant concentration. The overall molecu-
larity or order of a reaction step is the sum of all the concentration term
exponents.
In order to more fully understand the dependence of instantaneous
velocity on reactant concentration, we will first consider a simple first-order
reaction in which one molecule of reactant R forms one molecule of
product P (R ! P) according to the rate law given by Eq. (7.1).

d½R d½P
v¼ ¼ ¼ k½R ð7:1Þ
dt dt

In Eq. (7.1), the instantaneous reaction velocity (v ¼  d[R]/dt ¼ þ d

[P]/dt) is directly proportional to the concentration of R raised to the first
power. If the velocity for either loss of R or gain of P was measured
immediately after initiation of the reaction (Fig. 7.2A), then a secondary
plot of this initial velocity versus different initial or starting concentrations of
[R] would be linear with a slope equal to the first-order rate constant,
k (s 1) (Fig. 7.2B). For such analyses, great care must be taken to ensure
measurement of initial velocities. With increasing times past initiation of the
reaction, the measured velocity decreases proportionately with depletion of
[R]. Thus, it is common and good practice to measure initial velocities for
either loss of reactant or gain of product at times corresponding to 5% of
product conversion.
To better illustrate the decreasing reaction velocity with time, the
differential rate Eq. (7.1) can be integrated with respect to [R] to yield
the first-order or single-exponential decay function given by Eq. (7.2). The
limits of integration correspond to the initial concentration of reactant
Measurement of Enzyme Activity 61

A Slopes = v = d [P]/dt, M s−1 B Slope = k = dv/d [R]0, s−1

3[R]0

v (d[P]/dt)
2[R]0
[P]

[R]0

Time [R]0

Figure 7.2 Determination of a first-order rate constant from measuring the initial
velocity of product formation at differing concentrations of a reactant. (A) Initial
velocities of product formation (slopes ¼ v ¼ d[P]/dt, M s 1) are measured for different
starting concentrations of reactant (i.e., [R]0, 2[R]0, or 3[R]0). (B) The measured initial
velocities are then plotted versus the different starting concentrations of reactant. The
slope of this dependence yields the first-order rate constant for conversion of reactant
into product (slope ¼ k ¼ dv/d[R]0, s 1).

ð ½R ðt
d½R
¼ k dt
½R0 ½R 0

 
½R
ln ¼ kt
½R0

½R ¼ ½R0 ekt ð7:2Þ

when the reaction is initiated (i.e., [R] ¼ [R]0 at t ¼ 0) and the concentra-
tion of reactant at a given time after the reaction has started. Figure 7.3A
shows the relationship between measuring instantaneous velocity (i.e.,
v ¼  d[R]/dt, where Dt is very small) according to Eq. (7.1) and measuring
overall time progress (i.e., either [R] or [P] versus t, where Dt is very large)
according to Eq. (7.2) for the simple first-order conversion of reactant to
product. The derivative in Eq. (7.1) is simply the slope of the tangent for the
concentration curve at a specific time.
If no product is present at t ¼ 0, the sum of reactant and product
concentrations at any time must be equal to [R]0 (i.e., [R] þ [P] ¼ [R]0).
Using this relationship, ([R]0  [P]) may be substituted for [R] in Eq. (7.2)
and rearranged to yield the first-order rate Eq. (7.3) that describes the
corresponding increasing product concentration with time (Fig. 7.3B).

½P ¼ ½R0 ð1  ekt Þ ð7:3Þ

62 T. K. Harris and M. M. Keshwani

A B

[R]

[P]
Time Time

Figure 7.3 (A) Time progress for decreasing reactant concentration [R] in a first-order
reaction. It can be seen that the instantaneous velocity at any given reactant concentration
(lines) decreases with decreasing reactant concentration [R]. (B) Corresponding time
progress curve for increasing product concentration [P] in the same first-order reaction.

2.2. Basic enzyme kinetics

Measurement of enzyme activity is actually a measurement of catalytic activity
(Cook and Cleland, 2007; Cornish-Bowden, 1995; Segel, 1975). An enzyme
or catalyst participates in chemical reactions by increasing the reaction
velocity without itself appearing in the end products. In this case, an enzyme
(E) first combines with or binds one or more chemical reactants or substrates (S)
(Fig. 7.1B, kþ 1 and k 1). The resulting enzyme–substrate (ES) complex is an
intermediate species from which catalytic conversion of substrate to product
(P) takes place, ultimately releasing the intact enzyme so that it may catalyze a
subsequent reaction (Fig. 7.1B, kþ 2). Since the kþ 2 proportionality constant
typically comprises all processes or chemical steps involving conversion of the
ES complex to release of products, it is a ‘‘pseudo’’ first-order rate constant
(s 1). It is more commonly referred to as the turnover number, kcat, which is
the maximum number of substrate molecules converted to product per active
site per unit time. Thus, the instantaneous enzyme-catalyzed reaction veloc-
ity (v ¼ þ d[P]/dt) is directly proportional to the concentration of the ES
complex according to kcat in Eq. (7.4).

d½P
v¼ ¼ kcat ½ES ð7:4Þ
dt

Equation (7.4) is rendered impractical to the experimentalist, because

the differential rate expression that describes the time-dependent concen-
tration of ES complex includes terms for both its formation (kþ 1[E][S]) and
decay (k 1[ES] þ kcat[ES]) according to Eq. (7.5). Furthermore,

d½ES
¼ kþ1 ½E½S  k1 ½ES  kcat ½ES ð7:5Þ
dt
Measurement of Enzyme Activity 63

Equation (7.5) requires expressions for the time-dependent concentra-

tions of E and S. To overcome the difficulty of integrating all differential
rate expressions, the very useful steady-state approximation is applied to the
concentration of ES. In cases where the reaction velocity is measured to be
approximately constant over a time interval (e.g., measuring the initial
velocity), then it follows that the concentration of ES does not vary
appreciably. Thus, Eq. (7.5) is equated to zero and rearranged to obtain
the Eq. (7.6) expression for [ES] in terms of [E] and [S].

kþ1 ½E½S ½E½S

½ES ¼ ¼ ð7:6Þ
k1 þ kcat Km
In Eq. (7.6), the rate constants are grouped to form a composite con-
stant [Km ¼ (k 1 þ kcat)/kþ 1], which is referred to as the Michaelis constant.
When the initial velocity of an enzyme-catalyzed reaction is measured under
steady-state conditions, the time-dependent [S] in Eq. (7.6) is approximated
by its initial concentration, [S]0, since the amount of product formed is
substantially less than the amount of remaining substrate (i.e., [S]  [S]0
since ½P  ½S). Since the time-dependent [E] cannot be approximated by
its initial concentration, [E]0, the law of conservation of mass is used,
whereby [E] ¼ [E]0  [ES] is further substituted into Eq. (7.6) to yield
Eq. (7.7), which is rearranged to give the Eq. (7.8) expression for [ES].
Now, this expression is substituted back into Eq. (7.4) to yield the familiar
Michaelis–Menten Eq. (7.9),

ð½E0  ½ESÞ½S0
½ES ¼ ð7:7Þ
Km
½E0 ½S0
½ES ¼ ð7:8Þ
Km þ ½S0

d½P kcat ½E0 ½S0 Vmax ½S0

v¼ ¼ kcat ½ES ¼ ¼ ð7:9Þ
dt Km þ ½S0 Km þ ½S0

which shows the initial velocity of product formation to be directly pro-

portional to the total concentration of enzyme, [E]0. Equation (9) further
shows the initial velocity of product formation to vary in a hyperbolic-
dependent manner with regard to the initial concentration of substrate
according to the composite Michaelis constant, Km. Here, the Km ¼ (k 1 þ
kcat)/kþ 1 values represents the substrate concentration at which half-
maximum initial velocity is attained, and Vmax (M s 1) ¼ kcat[E]0 represents
the maximum initial velocity of product formation. In the limiting case
where catalytic conversion to products is much slower than dissociation of
64 T. K. Harris and M. M. Keshwani

the substrate (i.e., kcat  k1 ), the Km value approximates the true dissoci-
ation constant for the ES complex, Kd ¼ k 1/kþ 1; and the system is
described as being in rapid equilibrium.
If the measured initial velocity is then normalized to amount of enzyme
in the reaction mixture, then Eq. (7.9) takes the form of Eq. (7.10), whereby
the activity of the enzyme is defined

v kcat ½S0
Activity ðs1 Þ ¼ ¼ ð7:10Þ
½E0 Km þ ½S0

as the instantaneous enzyme-catalyzed reaction velocity normalized to the

amount of enzyme (v/[E]0). In this case, enzyme activity is measured as a
function of substrate concentration to obtain values of the turnover number,
kcat, and the Michaelis constant, Km. The ratio of kcat/Km (M 1 s 1) is termed
the specificity constant, which is an apparent second-order rate constant that
refers to the properties and reactions of the free enzyme leading to the first
irreversible step.
It is important to point out that the above activity definition pertains to
cases in which the enzyme active site concentration (moles per volume) in the
reaction mixture is known (e.g., assay of a purified homogeneous form of
the enzyme). When enzyme activity measurements are carried out in the
presence of other proteins such as crude cell lysates or partially purified
enzyme preparations, the measured initial velocity is normalized to the total
protein concentration in the assay. In this case, total protein concentration is
measured and expressed as weight per volume. As such, when the molar
concentration units of initial velocity (mol l 1 s 1) are divided by weight
concentration units of total protein (g l 1), the volume terms cancel so that
specific enzyme activity is expressed as the number of moles of product
converted per unit time per weight of protein. Specific activity expressed
in this manner is most often reported as mmol min 1 mg 1. When one unit
(U) is defined as conversion of 1 mmol of substrate per minute, then enzyme
activity is also commonly expressed as units per milligram of protein
(U mg 1). The following sections describe a number of initial considera-
tions and experimental methods for proper development of enzymatic
assays.

3. Measurement of Enzyme Activity

In order to measure an enzyme’s catalytic activity, it is essential to first
identify the chemical changes involving conversion of substrate(s) to prod-
uct(s). To date, the textual accounting of the incredible variety of enzymatic
activities is most typically managed by grouping enzymes according to the
Measurement of Enzyme Activity 65

types of chemical reactions they catalyze (e.g., oxidation–reductions, group

transfers, eliminations, isomerizations, rearrangements, condensations, car-
boxylations, etc.) (Frey and Hegeman, 2007). Further, considerations
should take into account whether mechanistically similar chemistry
(i) involves either small molecule substrates or larger macromolecular pro-
teins or nucleic acids, (ii) occurs readily in solution or requires membrane-
bound components, and (iii) proceeds to any extent in the reverse direction.
In any case, the enzymatic assay is designed around the ability to distinguish
between the physicochemical properties of a given substrate with respect to
those of a given product in a quantifiable manner (Eisenthal and Danson,
2002; Rossomando, 1990). To a large degree, similar methods of measuring
activity have been applied to various enzymes categorized within such
subgroups; and it would be prudent to first consider devising assays on the basis
of those already well established for either (i) homologous enzymes from different
organisms or (ii) different enzymes that catalyze a related chemical reaction. The
central theme among all enzyme activity measurements is to ensure measure-
ment of initial velocities. To reiterate this all important point, it is common and
good practice to measure initial velocities for either loss of reactant or gain
of product at times corresponding to 5% of product conversion. This gives
the best approximation for the initial velocity at the starting or initial
substrate concentration.

3.1. Continuous assays

After surveying the literature for applicable enzymatic assays, one may find a
number of suitable methods, which may be initially distinguished according
to whether a ‘‘continuous’’ or a ‘‘discontinuous’’ assay is employed. Con-
tinuous assays are defined by the ability to continuously monitor either
disappearance or appearance of a given substrate or product, respectively;
and they most often rely on spectroscopic techniques such as electronic
ultraviolet–visible absorption and fluorescence emission. For absorption
spectroscopy, molar absorbance extinction coefficients (e) can be gravimet-
rically determined for any number of compounds with conjugated bond
systems. According to the Lambert–Beer relationship (A ¼ ecl ), absorbance
(A, unitless) is directly proportional to (i) the molar extinction coefficient
(e, M 1 cm 1), (ii) the concentration of the compound (c, M), and (iii) the
path length of the cuvette (l, cm) used for the measurement (Segal, 1976).
Thus, the molar change in concentration with time or initial velocity
(v ¼  d[S]/dt or d[P]/dt) is directly calculated from the slope relating the
change in spectroscopic signal at a designated wavelength with respect to
time (dA/dt) on dividing by el; and the enzyme activity is ultimately obtained
by dividing the initial velocity by either the enzyme molar concentration
(M) or the weight concentration of total protein in the assay (mg ml 1)
according to either Eq. (7.11) or (7.12), respectively.
66 T. K. Harris and M. M. Keshwani

v ðdA=dt; s1 Þ
Activity ðs1 Þ ¼ ¼ ð7:11Þ
½E0 ðe; M1 cm1 Þðl; cmÞð½E0 ; MÞ
v
Activity ðmmol min1 mg1 Þ ¼
½E0
ð7:12Þ
ðdA=dt; min1 Þ
¼
ðe; ml mmol 1 cm1 Þðl; cmÞð½E0 ; mg ml1 Þ

Although ultraviolet–visible absorption spectroscopy is (i) convenient

for continuous monitoring and (ii) accurate with respect to determining
concentration changes, it is limited with respect to both (i) the number of
ESs and products that can be detected and (ii) the range of concentrations
that can be accurately detected. Alternatively, fluorescence and phospho-
rescence spectroscopies can be used for continuously monitoring reaction
progress; and in most cases provide significantly enhanced sensitivity. How-
ever, it must be pointed out that the concentration of the fluorophore
cannot be calculated from the measured fluorescence emission by applica-
tion of a universal constant equivalent to the molar extinction coefficient.
Rather, relative changes in fluorescence emission with time are compared.
Regardless, fluorescence and phosphorescence spectroscopies are highly
suited for use in high-throughput platforms utilized for screening very
large numbers of compounds for effects on enzyme activity. For a large
number of enzyme drug targets (e.g., particularly protein kinases and
proteases), artificial peptide substrates have been chemically modified to
contain one or more small molecule fluorophores; and such substrates
undergo fluorescence emission changes occur on either phosphorylation
or cleavage. Before employing fluorescence- or phosphorescence-based
assays, researchers should thoroughly review literature pertaining to such
principles, as numerous properties must be considered (e.g., fluorescence
lifetime, quantum yields, and inner-filter effects) (Lakowicz, 2004).
Finally, it should be pointed out that numerous ‘‘continuous’’ enzymatic
assays can and have been developed for reactions involving substrate–
product pairs with no spectroscopic properties. In these cases, either one
or more additional enzymes are included in the reaction mixture, which
serve to ‘‘couple’’ a given product to a reaction that does yield a desirable
spectroscopic property (Eisenthal and Danson, 2002). For the most part,
such coupled continuous assays have been applied to metabolic enzymes for
which a given product may either be oxidized or reduced by chromogenic
coenzymes such as NADH. If the product is not a direct substrate for such a
reaction, then an additional enzyme may be included to convert the first
product to a second product that undergoes the chromogenic reaction. In
any case, it must be ensured that the reaction mixture contains adequate
Measurement of Enzyme Activity 67

amounts of all additional components to ensure that the reaction velocity is

solely determined by the enzyme of interest. In other words, the measured
reaction velocity should be (i) dependent only on the target substrate(s) and
enzyme concentration and (ii) independent or zero order with respect to the
concentration of each and every component added to facilitate the coupled
reaction.

3.2. Discontinuous assays

Discontinuous assays are defined by the inability to continuously and selec-
tively monitor concentration changes of either substrate or product, such as
when a given substrate–product pair exhibits either similar or no spectro-
scopic properties. Rather, the enzymatic reaction must be manually stopped
or ‘‘quenched’’ at different times. The quenched samples are then subjected
to some method whereby the product can be efficiently separated from the
substrate so that the change in concentration of either can be determined for
each individual time point. In setting up a discontinuous enzymatic assay,
it is crucial to devise methods for both (i) efficiently stopping or ‘‘quench-
ing’’ an enzyme-catalyzed reaction at designated time points and (ii) effi-
ciently separating the large amount of unreacted substrate (95%) from the
very small amount of formed product (5%) required for measuring initial
velocities. These two methodological points must be considered as a pair,
since the ‘‘quenched’’ reaction solution conditions must be amenable to
subsequent procedures required for fractionation of substrate and product
molecules. Since a large number of enzymes can be rapidly inactivated
under a wide variety of conditions (e.g., addition of acid, base, or metal
chelators), substrate–product fractionation is foremost considered.
High-performance liquid chromatography (HPLC) offers extensive
ranges in both resolution capabilities and detection sensitivity (McMaster,
2007). For example, a great variety of mixtures containing either small
molecules or macromolecules can be resolved on the basis of differential
retention when passed over a column containing a given stationary phase
(e.g., ion-exchange, size-exclusion, or reversed-phase chromatography).
In addition, HPLC systems can be fixed with any number of detectors so
that a compound with a given spectroscopic property can be detected as it
elutes from the column, which results in a spectroscopic peak. To gauge the
amount of product formed, the area of the resolved product peak is
integrated and compared with a standard curve prepared from integrated
peak areas obtained for range of product concentrations. In cases where
the designated product does not exhibit any spectroscopic properties, one
may consider possibilities of either (i) chemically modifying the substrate to
contain a spectroscopic active compound or (ii) modifying the product
in the quenched reaction mixture. In the latter case, it must be established
that essentially all product undergoes modification and that the modified
68 T. K. Harris and M. M. Keshwani

product can be readily resolved from any remaining unreacted spectroscopic

active compound. The determination of enzyme activity is carried out
exactly as in Eqs. (7.11) and (7.12) except that dA/dt is obtained as DA/D
t (i.e., a single-point determination rather than a slope determined from many
points).
In cases when a given substrate–product pair exhibits no useful spectro-
scopic properties, the amount of product at a given time point in a discon-
tinuous assay can be determined by radiometric analysis (Eisenthal and
Danson, 2002). In fact, radiometric assays provide the highest degree of
sensitivity, as well as exhibit the highest degree of broad scale utility. Almost
any conceivable enzyme substrate can be synthesized to contain a radioac-
tive isotopic atom (or radioisotope) in a position that undergoes transfer
from the parent substrate, which can be detected in very small amounts. The
most commonly used radioisotopes for enzymatic assays are 3H, 14C, 32P,
and 35S. In all four cases, the unstable nucleus undergoes slow first-order
decay to form a stable isotope with an atomic number one higher and an
atomic weight identical to the original radioisotope through a process called
beta-particle emission. Scintillation counting of beta-particle emission gives
very sensitive determination of the amount of radioactivity present in a
sample, which is directly quantified in counts per minute (cpm). The half-
life (t1/2) of a radioisotope is the time required for half of the original
number of atoms to decay. Half-life values of 3H (t1/2 ¼ 12.3 years), 14C
(t1/2 ¼ 5700 years), 32P (t1/2 ¼ 14.3 days), and 35S (t1/2 ¼ 87.1 days) have
been well determined. From these values, it can be seen that the amount of
radioactivity emitted by 3H and 14C does not appreciably change over the
time course of routine laboratory work; whereas the amount of radioactivity
emitted by 32P and 35S significantly decreases over such time.
Commercial manufacturers incorporate such radioisotopes into desig-
nated positions of target molecules to yield radiolabeled substrate molecules,
which have a defined specific (radio)activity (SA) (Segal, 1976). The specific
activity refers to the amount of radioactivity (cpm) per unit amount of
molecule (mol). Typically, one adds a small amount of the highly radioac-
tive ‘‘hot’’ substrate molecule into a much larger amount of the nonradio-
active or ‘‘cold’’ substrate. A small volume of a known concentration is then
subjected to scintillation counting to determine the specific (radio)activity
of the substrate (SAS, cpm/mol 1). By controlling the mixture of hot and
cold substrate molecules, a very wide range of specific (radio)activities can
be generated so that enzyme activity measurements are possible over an
equal range of substrate concentrations. For example, a radiolabeled sub-
strate with SA ¼ 1000 cpm mmol 1 would enable efficient detection for
transfer of label to form 1 mmol of product (1000 cpm), whereas generation
of a radiolabeled substrate with SA ¼ 1000 cpm pmol 1 would enable
efficient detection of 1 pmol of product!
Measurement of Enzyme Activity 69

Most commonly, ion-exchange methods are used in which the product

form of the molecule is retained on solid support (e.g., ion-exchange resin,
paper, or disk), while the substrate form is not retained. Other methods of
separation may include either selective precipitation, solvent extraction, or
gel electrophoresis. Before carrying out the enzyme-catalyzed reaction, it is
essential to establish that (i) the radiolabeled product is highly resolved from
radioactive substrate and (ii) the amount of resolved product closely approx-
imates the amount subjected for resolution. Typically, one finds a point of
compromise, whereby a certain minimum amount of ‘‘background’’ sub-
strate radiation is tolerated in the medium or location where product is
resolved. For most accurate analysis, the medium containing the purified
product is directly analyzed by scintillation counting (e.g., ion-exchange
resin, paper, or disk). Although selected gel regions after electrophoresis
may be removed for direct counting, it has been more common practice to
quantify spatial radioactivity across an entire gel using scintillation plates and
computer densitometry analysis.
The amount of enzyme-catalyzed product (mol) detected in the aliquot
removed for analysis is calculated by dividing the amount of radioactivity
detected in the product fraction (cpm) by the specific (radio)activity of the
substrate (cpm/mol 1); the concentration of product, [P], is obtained by
dividing by the volume of the aliquot (l); and the initial velocity, d[P]/dt, is
obtained by further dividing [P] by the time of the reaction (s). Thus,
enzyme activity is ultimately obtained by dividing the initial velocity by
either (i) the enzyme molar concentration according to Eq. (7.13) or (ii)
the total protein concentration according to Eq. (7.14).

d½P cpm
Activityðs1 Þ ¼ ¼
dt½E0 ðSA ; cpm=mmol Þðvol; mlÞðt; sÞð½E; mmol=ml1 Þ
S 1

ð7:13Þ

d½P
Activityðmmolmin1 mg1 Þ ¼
dt½E0
ð7:14Þ
cpm
¼
ðSAS ; cpm=mmol1 Þðvol; mlÞðt; minÞð½E; mg=ml1 Þ

4. Formulation of Reaction Assay Mixtures

Having established effective methods for resolution, detection, and
quantification of reaction species, it is next time to consider preparation of
the reaction mixture itself. For continuous spectroscopic assays, total
volumes should be selected to accommodate cuvettes made for the given
70 T. K. Harris and M. M. Keshwani

spectrometer. For visible absorption (325 nm), 1 ml plastic disposable

cuvettes are most widely used due to convenience. Since plastic cuvettes
increasingly absorb light of lower wavelengths, quartz cuvettes should be
used in this region. Of increasing availability and utility are fluorometers
fitted with microplate readers, which simultaneously obtain accurate
emission readings from a large number of individual small reactions
(50 ml) on a single microtiter plate. For discontinuous enzyme assays,
a total volume of the reaction mixture should be selected in which several
(3) designated fixed amount aliquots can be removed at varying times for
reaction quenching and product analysis.
For all enzyme assays, stock concentrations of suitable buffer, substrate(s),
and enzyme are prepared so that dilution of each component to its desired
concentration can be achieved by adding volumetric amounts that sum to
less than the total volume. The remaining volume is occupied by water. The
buffer component should have a pKa value  1 of the desired reaction pH,
and its concentration should be chosen to exceed any other component that
has ionizable groups. The stock buffer solution may also contain additional
compounds that may be necessary to maintain enzyme stability and activity
such as salts (e.g., NaCl or KCl), osmolytes (e.g., polyethylene glycol,
glycerol, or sucrose), and reducing agents (e.g., 2-mercaptoethanol, dithio-
threitol, or TCEP). It is very important to include such components in the
preparation of the stock buffer, and the final stock buffer mixture should
be titrated to its desired pH at the temperature in which the reaction will be
carried out. Pay special attention to use of (i) nitrogen containing buffers
(such as Tris) whose pKa values are very sensitive to temperature and (ii)
phosphate buffers whose pKa values are very sensitive to ionic strength. In
fact, dilution of phosphate buffer will change its pH value, so the pH of its
stock concentration must be adjusted accordingly.
It is best to generate the reaction mixture by first adding nonenzyme
components in the following order: (i) designated volume of water,
(ii) designated volume of stock buffer mixture, and (iii) designated volume
of stock substrate. It is best to prepare a stock concentration of enzyme so that
only a relatively small amount (10% volume) is added to bring the mixture
to the designated total volume and initiate the reaction. Immediately, after the
enzyme has been added to initiate a continuous assay, the cuvette(s) or
microtiter plate must be inserted into a spectrometer that is programmed to
begin collecting absorption or emission data at the designated wavelength(s).
For discontinuous assays, one must be prepared to remove a precise volume
of the reaction mixture and mix with a precise volume of the quench reagent;
and this must be repeated at precise times. In all enzyme assays, one should
always formulate an additional control reaction mixture that does not contain
enzyme to determine the ‘‘background’’ amounts of either (i) noncatalyzed
product formation or (ii) signal retained from incomplete resolution. Any
such background amounts must be subtracted from the signal detected in the
presence of enzyme.
Measurement of Enzyme Activity 71

5. Discussion
In this chapter, the most fundamental principles of measuring enzyme
reaction rates were presented, and it is in no way a complete treatment on
the subject. Nevertheless, clear understanding of these principles is abso-
lutely prerequisite toward either (i) effective utilization of well-developed
assays or (ii) efficient development of assays for newly discovered enzymes.
Over the past century, countless research articles, review articles, book
chapters, and even entire books have reported on the development, utiliza-
tion, and interpretation of enzyme activity measurements. The selection of
books and articles cited in this article provide more comprehensive treat-
ments of the various topics related to measuring enzyme activity.

ACKNOWLEDGMENTS
This work was supported by NIGMS, National Institutes of Health Grant GM69868 to
T. K. H. and a Maytag Fellowship to M. M. K.

REFERENCES
Cook, P. F., and Cleland, W. W. (2007). Enzyme Kinetics and Mechanism. Garland
Science, Hamden, CT.
Cornish-Bowden, A. (1995). Fundamentals of Enzyme Kinetics. Portland Press, Ltd.,
London, UK.
Eisenthal, R., and Danson, M. (2002). Enzyme Assays: A Practical Approach. Oxford
University Press, Inc., New York, NY.
Frey, P. A., and Hegeman, A. D. (2007). Enzymatic Reaction Mechanisms. Oxford
University Press, Inc., New York, NY.
House, J. E. (2007). Principles of Chemical Kinetics. 2nd edn. Academic Press, Burlington,
MA.
Laidler, K. J. (1987). Chemical Kinetics. 3rd edn. Prentice Hall, Upper Saddle River, NJ.
Lakowicz, J. R. (2004). Principles of Fluorescence Spectroscopy. 2nd edn. Springer
ScienceþBusiness Media, Inc., New York, NY.
McMaster, M. C. (2007). HPLC, a Practical Users Guide. 2nd edn. John Wiley & Sons, Inc.,
Hoboken, NJ.
Rossomando, E. F. (1990). Measurement of enzyme activity. Methods Enzymol. 182, 38–49.
Segel, I. H. (1975). Enzyme Kinetics. John Wiley & Sons, Inc., New York, NY.
Segal, I. H. (1976). Biochemical Calculations. 2nd edn. John Wiley & Sons, Inc.,
New York, NY.
 C H A P T E R E I G H T

Quantitation of Protein
James E. Noble* and Marc J. A. Bailey†

Contents
1. Introduction 74
2. General Instructions for Reagent Preparation 75
3. Ultraviolet Absorption Spectroscopy 80
3.1. Ultraviolet absorbance at 280 nm (Range: 20–3000 mg) 80
3.2. Method 81
3.3. Comments 81
3.4. Ultraviolet absorbance at 205 nm (Range: 1–100 mg) 82
3.5. Calculation of the extinction coefficient 82
4. Dye-Based Protein Assays 83
4.1. Protein concentration standards 83
5. Coomassie Blue (Bradford) Protein Assay (Range: 1–50 mg) 85
5.1. Reagents 85
5.2. Procedure 85
5.3. Comments 86
6. Lowry (Alkaline Copper Reduction Assays) (Range: 5–100 mg) 86
6.1. Reagents 87
6.2. Procedure 87
6.3. Comments 88
7. Bicinchoninic Acid (BCA) (Range: 0.2–50 mg) 88
7.1. Reagents 88
7.2. Procedure 89
7.3. Comments 89
8. Amine Derivatization (Range: 0.05–25 mg) 89
8.1. Reagents 90
8.2. Procedure 90
8.3. Comments 90
9. Detergent-Based Fluorescent Detection (Range: 0.02–2 mg) 91

* Analytical Science, National Physical Laboratory, Teddington, Middlesex, United Kingdom

{
Nokia Research Centre - Eurolab, University of Cambridge, Cambridge, United Kingdom

Methods in Enzymology, Volume 463 Crown Copyright # 2009. Published by Elsevier Inc.
ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63008-1 All rights reserved.

73
74 James E. Noble and Marc J. A. Bailey

10. General Instructions 91

10.1. Cuvettes 91
10.2. Microwell plates 92
10.3. Interfering substrates 93
Acknowledgment 94
References 94

Abstract
The measurement of protein concentration in an aqueous sample is an impor-
tant assay in biochemistry research and development labs for applications
ranging from enzymatic studies to providing data for biopharmaceutical lot
release. Spectrophotometric protein quantitation assays are methods that use
UV and visible spectroscopy to rapidly determine the concentration of protein,
relative to a standard, or using an assigned extinction coefficient. Methods are
described to provide information on how to analyze protein concentration using
UV protein spectroscopy measurements, traditional dye-based absorbance
measurements; BCA, Lowry, and Bradford assays and the fluorescent dye-
based assays; amine derivatization and detergent partition assays. The obser-
vation that no single assay dominates the market is due to specific limitations of
certain methods that investigators need to consider before selecting the most
appropriate assay for their sample. Many of the dye-based assays have unique
chemical mechanisms that are prone to interference from chemicals prevalent
in many biological buffer preparations. A discussion of which assays are prone
to interference and the selection of alternative methods is included.

1. Introduction
The quantity of protein is an important metric to measure during
protein purification, for calculating yields or the mass balance, or determin-
ing the specific activity/potency of the target protein. Various platforms and
methods are available to quantitate proteins and will be described elsewhere
in this volume; however, for this chapter, we will concentrate on spectro-
photometric assays of protein in solution that do not require either enzy-
matic/chemical digestion or separation of the mixture prior to analysis.
The spectrophotometric assays described are UV absorbance methods
and dye-binding assays using colorimetric and fluorescent-based detection.
In comparison to other methods, these assays can be run at a high through-
put, using inexpensive reagents with equipment found in the majority of
biochemical laboratories. These spectrophotometric assays require an
appropriate protein standard or constituent amino acid sequence informa-
tion to make a good estimate of concentration. The choice of method used
to determine the concentration of a protein or peptide in solution is
dependent on many factors that will be discussed. The majority of methods
Quantitation of Protein 75

require a soluble analyte such as peptides, proteins, posttranslationally mod-

ified protein (e.g., glycosylated), or chemically modified proteins (e.g.,
PEGylated). For common protein purification procedures, the flow chart
in Fig. 8.1 describes the process of selecting the most appropriate assay,
based on key criteria.
Where other protein purification techniques are available or complex
buffer systems are present in the sample, refer to Table 8.1, or other reviews
(Olson and Markwell, 2007) for assay selection. Other criteria that need to
be considered when selecting an assay include:

Sample volume: The amount of material available to analyze, typically
fluorescent-based assays display the best sensitivity and dynamic range
(see Fig. 8.2). Microplate assays (by using lower assay volumes, thereby
less protein sample) show improved sensitivity, typically up to 10-fold
when compared with cuvette-based assays.

Sample recovery: If the sample is limited, a nondestructive method, for
example, UV spectroscopy may be more appropriate.

Throughput: If multiple samples are to be analyzed, a microplate-compatible
rapid one step assay should be considered.

Robustness: The absorbance-based dye-binding assays appear to display
enhanced repeatability and robustness when compared to fluorescent
assays.

Chemical modification: Covalent modification, for example glycosylation
(de Moreno et al., 1986; Fountoulakis et al., 1992) or PEGylation (Noble
et al., 2007), can interfere with specific assays.

Protein aggregation: The solubility of a protein in solution, often a problem
for membrane proteins, or proteins prone to aggregation can alter the
expected response for many assays.
Other protein quantitation methods are becoming more commonly
employed in biochemistry laboratories due to automation, regulatory, and
sensitivity requirements. Alternative methods not detailed in this chapter
include isotope dilution mass spectrometry (ID IC MS/MS) (Burkitt et al.,
2008), Kjeldahl nitrogen method, amino acid analysis (Ozols, 1990), gravi-
metric determination (Blakeley and Zerner, 1975), immunological, and
quantitative gel electrophoresis with fluorescent staining.

2. General Instructions for Reagent

Preparation
For the methods detailed, reagents should be used at the highest purity
available and dyes should be obtained at spectroscopy grade where available.
Ideally deionized, filtered water should be used at a minimum quality of
 Start: protein sample, with known purity

Is a reference standard protein available No Use an appropriate standard, for

(matched protein)? example BSA, or IgG. Avoid Bradford
and amine derivatization methods
Yes

Is the protein free of post-translational No Avoid using the Bradford, and Lowry
modifications, for example
assay for glycosylated proteins
glycosylation?
Yes

Does the sample protein/peptide have No

a Mw > 8 kDa, and/multiple Tyr/Trp Avoid UVAbs280 nm and Bradford
amino acids? methods

Yes

Avoid BCA,
Are thiol, or reducing agents Yes
Is the protein solution free of interfering No Lowry and
present, e.g. DTT from GST
compounds? CBQCATM
affinity purification?
methods
Yes

Are chelating agents Yes Avoid BCA

Any of the assays described would be appropriate, present, e.g. EDTA for His- and Lowry
selection should be based on criteria detailed in tag affinity purification? methods
text and available equipment

Are detergents present, e.g. Yes Check detergent

used for protein compatibility data
solubilization? (Table 8.1)

Check compatibility
Are high concentrations of Yes with BCA, Lowry
salts, or acids used for
and Bradford
precipitation present?
methods

Avoid BCA, Lowry

Are amines, or ammoniun Yes and amine
ions present in the protein derivatization
diluent? methods

Figure 8.1 Flow chart for the selection of assays for quantitation or proteins in common protein purification procedures. The chart assumes
that the sample for analysis is relatively pure, that is the analyte for quantitation is the major component, for example fractions from affinity
chromatography, or extraction from inclusion bodies. The ‘‘reference standard protein’’ refers to a standard that is the same protein that is
being quantitated in the same, or similar matrix that is ‘‘matched.’’
Table 8.1 Substance compatibility table

Compatible concentrationa
Amine Fluorescent
Substance BCAb Lowryc Bradfordd derivatizatione detergent f UV Abs280 nmg
Acids/bases
HCl 0.1 M na 0.1 M na 10 mM >1 M
NaOH 0.1 M na 0.1 M na 10 mM >1 M
Perchloric acid <1% >1.25% na na na 10%
Trichloroacetic acid <1% >1.25% na na na 10%
Buffers/salts
Ammonium sulfate 1.5 M >28 mM 1.0 M 10 mM 10–50 mM >50%
Borate 10 mM Undiluted Undiluted Undiluted Undiluted Undiluted
Glycine 1 mM 1 mM 100 mM – na 1M
HEPES 100 mM 1 mM 100 mM na 10–50 mM na
Imidazole 50 mM 25 mM 200 mM na na na
Potassium chloride <10 mM 30 mM 1.0 M na 20–200 mM 100 mM
PBS Undiluted Undiluted Undiluted Undiluted Undiluted Undiluted
Sodium acetate 200 mM 200 mM 180 mM na na na
Sodium azide 0.2% 0.5% 0.5% 0.1% 10 mM na
Sodium chloride 1.0 M 1.0 M 5.0 M na 20–200 mM >1 M
Triethanolamine 25 mM 100 mM na na na na
Tris 250 mM 10 mM 2.0 M 10 mM na 0.5 M
Detergents
Brij 35 5% 0.031% 0.125% na na 1%
CHAPS 5% 0.0625% 5% na na 10%
(continued)
Table 8.1 (continued)

Compatible concentrationa
Amine Fluorescent
Substance BCAb Lowryc Bradfordd derivatizatione detergent f UV Abs280 nmg
Deoxycholic acid 5% 625 mg/ml 0.05% na na 0.3%
Nonidet P-40 5% 0.016% 0.5% na na na
SDS 5% 1% 0.125% – 0.01–0.1% 0.1%
Triton X-100 5% 0.031% 0.125% na 0.001% 0.02%
Tween-20 5% 0.062% 0.062% 0.1% 0.001% 0.3%
Reducing agents
Cysteine na 1 mM 10 mM na na na
DTT 1 mM 0.05 mM 5–1000 mM 0.1 mM 10–100 mM 3 mM
2-Mercaptoethanol 0.01% 1 mM 1.0 M 0.1 mM 10–100 mM 10 mM
Thimerosal 0.01% 0.01% 0.01% na na na
Chelators
EDTA 10 mM 1 mM 100 mM na 5–10 mM 30 mM
EGTA na 1 mM 2 mM na na na
Solvents
DMSO 10% 10% 10% na na 20%
Ethanol 10% 10% 10% na na na
Glycerol 10% 10% 10% 10% 10% 40%
 Guanidine–HCl 4.0 M 0.1 M 3.5 M na na na
Methanol 10% 10% 10% na na na
PMSF 1 mM 1 mM 1 mM na na na
Sucrose 40% 7.5% 10% 10% 10–500 mM 2M
Urea 3.0 M 3.0 M 3.0 M na na >1.0 M
Miscellaneous
DNA 0.1 mg 0.2 mg 0.25 mg na 50–100 mg/ml 1 mg

Values relate to the maximum concentration of interfering compound within the protein sample that does not result in significant loss in assay performance. The guide is an
updated version of that prepared by Stoscheck to inform of any issues related to assay interference. Concentrations were obtained from product inserts and references
(Bradford, 1976; Peterson, 1979; Smith et al., 1985; Stoscheck, 1990), where there is not a consensus of values a range is given. Changing the protein-to-dye ratios, or
formulation of many of the dye-based assays can alter the maximum concentration of compound permissible. Interfering compounds have been selected to represent those
commonly encountered in protein purification and enzymology.
a
na indicates the reagent was not tested. A blank indicates that the reagent is not compatible with the assay at the reagent concentrations analyzed. A figure preceded by
(<) or (>) symbols indicates the tolerable limit is unknown but is respectively, less than or greater than the amount shown.
b
Figures indicate the concentration in a 0.1-ml sample using a final reaction volume of 2.1 ml.
c
Figures indicate the concentration in a 0.2-ml sample using a final reaction volume of 1.3 ml.
d
Figures indicate the concentration in a 0.05-ml sample using a final reaction volume of 1.55 ml.
e
Figures indicate interference concentrations with the CBQCATM assay (You et al., 1997) in a 90-ml sample using a final reaction volume of 100 ml.
f
Figures indicate interference concentrations with the NanoOrangeTM and Quant-iTTM assays ((Hammer and Nagel, 1986) and Quant-iTTM product insert) in a 40- and
20-ml sample (respectively) using a final reaction volume of 200 ml.
g
Figures indicate the concentration of the chemical that does not produce an absorbance of 0.5 over water (Stoscheck, 1990).
80 James E. Noble and Marc J. A. Bailey

Quant-IT

Fluorescamine

Lowry

CBQCA

Bradford

BCA

100 101 102 103 104 105

Quantitation range (ng protein)

Figure 8.2 The markers designate the upper and lower values for the quantitation
range of dye-based protein assay performed in a microplate format. The quantitation
range was defined as the range of protein amounts (ng) that displayed good precision
and did not show any deviation from the fitted response curve. Figure used with
permission from Noble et al. (2007).

18 MO cm and a total organic carbon of below 6 ppb. All buffer prepara-

tions should be filtered using 0.2 mm filtration (Millipore, Sartorius) devices
upon preparation to remove bacteria and fines. If precipitation occurs
during storage, the reagent should be discarded, unless stated in the method.

3. Ultraviolet Absorption Spectroscopy

3.1. Ultraviolet absorbance at 280 nm (Range: 20–3000 mg)
Proteins display a characteristic ultraviolet (UV) absorption spectrum around
280 nm predominately from the aromatic amino acids tyrosine and tryptophan.
If the primary sequence contains no or few of these amino acids then this
method will give erroneous results. Quartz crystal cuvettes are routinely used
for measurement as plastic materials can leach plasticizers, and are not UV
transparent. Similarly, buffer components with strong UV absorbance such as
some detergents especially Triton X-100 should be avoided (Table 8.1) and
‘‘blank’’ samples should be measured using the sample buffer solution but with
Quantitation of Protein 81

no protein present. UV absorbance is routinely used to give an estimate of

protein concentration but if the molar extinction coefficient of the protein is
known then the Beer–Lambert law can be used to accurately quantitate
amount of protein by UV absorbance, assuming the protein is pure and
contains no UV absorbing nonprotein components such as bound nucleotide
cofactors, heme, or iron–sulfur centers.
Beer–Lambert (molar absorption coefficient):

A ¼ am cl ð8:1Þ

where am is the molar extinction coefficient, c the concentration of analyte,

and l the path length in cm.

3.2. Method
For the measurement of a protein with unknown extinction coefficient,
using a protein standard:
1. Add blank buffer to a clean quartz cuvette and use to zero the
spectrophotometer.
2. Either using a fresh identical cuvette or replace the buffer with the
sample, then measure the absorbance at 280 nm. If the signal is outside
the linear range of the instrument (typically an absorbance greater than
2.0), then dilute the protein in buffer and remeasure.
3. After measurement of the sample remeasure the blank buffer to correct
for any instrument drift.
4. Determine the unknown concentrations from the linear standard
response.

3.3. Comments
The determination of the absorbance coefficient for a protein is discussed
below but if a stock of the protein at known concentration is available then
this can be used as a standard. Very rough estimates can be made from the
relationship that if the cuvette has a path length of 1 cm, and the sample volume
is 1 ml then concentration (mg/ml) ¼ absorbance of protein at 280 nm.
Light scattering from either turbid protein samples or particles suspended
in the sample with a comparable size to the incident wavelength (250–
300 nm) can reduce the amount of light reaching the detector leading to an
increase in apparent absorbance. Filtration using 0.2 mm filter units (that do
not adsorb proteins), or centrifugation can be performed prior to analysis to
reduce light scattering. Corrections for light scattering can be performed by
measuring absorbance at lower energies (320, 325, 330, 335, 340, 345, and
350 nm), assuming the protein does not display significant absorbance at
these wavelengths. A log–log plot of absorbance versus wavelength should
82 James E. Noble and Marc J. A. Bailey

generate a linear response that can be extrapolated back to 280 nm, the
resulting antilog of which will give the scattering contribution at this
wavelength (Leach and Scheraga, 1960).
Nucleic acids absorb strongly at 280 nm and are a common contaminant
of protein preparations. A pure protein preparation is estimated to give a
ratio of A280 to A260 of 2.0 while, if nucleic acid is present, the protein
concentration can be derived by the following formula (Groves et al., 1968).

Protein concentration ðmg=mlÞ ¼ 1:55A280  0:76A260 ð8:2Þ

3.4. Ultraviolet absorbance at 205 nm (Range: 1–100 mg)

The peptide bond absorbs photons at a maximum wavelength below 210 nm.
However, the broad absorption peak of the peptide bond allows measurements
at longer wavelengths, which can have many practical advantages in terms of
instrumentation and measurement accuracy. Due to interference from solvents
and components of biological buffers, absorbance at 214 and 220 nm is often
used as an alternative to measure proteins and peptides.
The large number of peptide bonds within proteins can make Abs205 nm
measurements more sensitive and display less protein-to-protein variability
than Abs280 nm measurements. Most proteins have extinction coefficients at
Abs205 nm for a 1-mg/ml solution of between 30 and 35; however, an
improved estimate can be obtained using Eq. (8.3) that takes into account
variations in tryptophan and tyrosine content of the protein to be quanti-
tated (Scopes, 1974). Absorbance at 205 nm is used to quantitate dilute
solutions, or for short path length applications, for example, continuous
measurement in column chromatography, or for analysis of peptides where
there are few, if any aromatic amino acids.
 
1 mg=ml A280 nm
e205 nm ¼ 27:0 þ 120  ð8:3Þ
A205 nm

3.5. Calculation of the extinction coefficient

The extinction coefficient (e) at a set wavelength describes the summation
of all the photon absorbing species present within the molecule at a defined
wavelength; the molar extinction coefficient is defined in Eq. (8.1). The
extinction (absorption) coefficient is commonly expressed either in terms of
molarity (M 1 cm 1) or as a percentage of the mass e1% (% 1 cm 1), where
e1% is defined as the absorbance value of a 1% protein solution.
Deviation from experimentally derived values for e, and those derived by
sequence data can be due to the influence of salts and buffers within the
Quantitation of Protein 83

protein sample. The absorbance spectra from various amino acids are
environmentally sensitive; therefore, e derived for a protein in a set buffer
may not be the same for another buffer system if gross changes in pH (tyrosine
ionization at pH 10.9), or solvent polarity (denaturing agents) occur.
To determine e280, the amino acid composition or sequence of the
protein is required. From the protein sequence, e280 can be calculated from
first principles using a standard formula (Gill and von Hippel, 1989), which
has been refined (Pace et al., 1995). Such models use the absorption coeffi-
cients for specific amino acids (Trp, Tyr, and disulfide bond) to generate a
good estimate of e280 where these amino acids are in abundance. However,
where there is a low abundance of these amino acids (e.g., insulin), the model
can display deviations of up to 15% from that determined by physical methods
(Pace et al., 1995). Physical (empirical) methods to determine extinction
coefficient include amino acid analysis (AAA) via acid hydrolysis and chro-
matographic separation of resulting amino acids (Sittampalam et al., 1988)
and Kjeldahl and gravimetric analysis (Kupke and Dorrier, 1978).

4. Dye-Based Protein Assays

Methods to prepare the established (nonproprietary) protein quantita-
tion assays are described. These reagents can be economically prepared in
bulk and stored for prolonged periods. The majority of such assays are
available from commercial suppliers such as Sigma-Aldrich, Bio-Rad, Nova-
gen, and Pierce. It should be noted that suppliers can have different prepara-
tions of such reagents and these can perform differently with specific proteins.
The use of commercial reagents can improve the long-term repeatability and
performance of the assay and for microplate-based assays is reported to reduce
issues with dye precipitation after long-term storage of reagents (Stoscheck,
1990). The majority of the spectrophotometric protein quantitation methods
described can be adapted to a microplate format (typically 96-well plate), we
have highlighted where changes in the assay formulations are required.

4.1. Protein concentration standards

The ideal protein standard to use in a quantitative assay is the exact same
protein in a matched matrix/solution that has been assigned using a higher
order method, for example AAA (Sittampalam et al., 1988) or gravimetric
analysis (Blakeley and Zerner, 1975). Gravimetric analysis is prone to errors
due to the extensive dialysis and drying to remove water and salts from
commercial preparations. Prepared standards should be redissolved at a high
concentration in water and stored at  20  C for long-term storage.
84 James E. Noble and Marc J. A. Bailey

In practice, there is not always a matched protein standard available;

however, some commercially available standards may be suitable for use, the
most common being BSA, bovine gamma globulins, or immunoglobulins
(used for antibody quantitation). The use of a BSA standard is known to
give misleading results in many assays, especially those methods that are
sensitive to the protein sequence, that is where the signal is generated by
specific amino acids (Fig. 8.3). Assays with a low protein sequence depen-
dence will give better estimates when BSA calibration is compared to AAA
assignment (Alterman et al., 2003). AAA assignment quantitates the amount

3
IgG
Insulin
Insulin-PEG
2.5 Lysozyme
Ratio of protein concentration estimates:

Lyso-PEG
RNase A
2 RNase B
(BSA standard/AAA)

1.5

0.5

0
BCA Bradford CBQCA Lowry Fluorescamine Quant-iT

Figure 8.3 A comparison of the accuracy of the BSA standard in estimating the
concentration of a protein using dye-based protein quantitation assays. AAA was used
to determine the concentration of the model proteins; from these estimates a calibration
curve for each protein was prepared using the dye-based assays. In the same plate, a
calibration curve using the BSA standard (Pierce; concentration defined by manufac-
turer) was also prepared and the response of this was compared to that of the model
proteins to see how well the BSA standard estimated the true concentration of the
model proteins. The ‘‘ratio of concentration estimations’’ refers to the concentration of
protein derived using the BSA standard when compared to the ‘‘true’’ value using
AAA, where a ratio of 1 indicates the two methods gave the same value. The variation
‘‘% CV’’ associated with the dye-based protein concentration assays ranged from 2% to
8%, dependent on the assay. AAA concentration assignment typically displayed 5% CV
values, dependent on the protein analyzed. Figure adapted with permission from Noble
et al. (2007).
Quantitation of Protein 85

of specific amino acids present following protein hydrolysis and separation,

using peptide sequence information the amount of target protein can then
be calculated.

5. Coomassie Blue (Bradford) Protein Assay

(Range: 1–50 mg)
The Bradford assay encompasses various preparations of the dye Coo-
massie Brilliant Blue G-250 used for protein quantitation purposes, and was
first described by Bradford (1976). The basic mechanism of the assay is the
binding of the dye at acidic pH to arginine, histidine, phenylalanine,
tryptophan and tyrosine residues (de Moreno et al., 1986), and hydrophobic
interactions (Fountoulakis et al., 1992). The exact mechanism is however
still not fully understood (Sapan et al., 1999). Upon binding protein, a
metachromatic shift from 465 to 595 nm is observed due to stabilization
of the anionic form of the dye. The majority of the observed signal is due to
the interaction with arginine residues, resulting in the wide protein-to-
protein variation characteristic of Bradford assays (Fig. 8.3).

5.1. Reagents
Dissolve 100 mg Coomassie Brilliant Blue G-250 in 50 ml of 95% ethanol
and add 100 ml of 85% phosphoric acid while stirring continuously. When
the dye has dissolved dilute to 1 l in water. The reagent is stable for up to a
month at room temperature; however, for long-term storage keep at 4  C,
if precipitation occurs filter before use.

5.2. Procedure
1. Prepare standards in the range 100–1500 mg/ml in a Bradford-compatible
buffer. For more dilute samples the sensitivity can be extended by increas-
ing the ratio of sample to reagent volumes (Micro Bradford assay: 1–25
mg/ml). If the ratio of the sample to dye is too high, the pH of the reaction
mixture could increase leading to higher background responses.
2. Add the standard and unknown samples to disposable cuvettes (plastic
disposable cuvettes and microplates should be used as the dye sticks to
various surfaces).
3. Allow the Bradford reagent to warm to room temperature. Add 1 ml of
the dye solution to 25 ml of the protein sample, mix and incubate for 10
min at room temperature.
86 James E. Noble and Marc J. A. Bailey

4. Measure the absorbance at 450 and 595 nm (for filter-based instruments

a range from 570 to 610 nm can be used without significant loss of assay
performance).
5. Plot either the 595 nm data or for improved precision at lower response
values the ratio 595 nm/450 nm. The standard response curve can be fit
to a polynomial response, from which unknown protein estimates can be
calculated.

5.3. Comments
The advantages of the Bradford assay include the ease of use, sensitivity and
low cost of the reagents. For microplate-based assays the reagent volumes
can be decreased giving a total volume of 300 ml. Due to the path of the
light source on the majority of microplate spectrophotometers, it is recom-
mended to use commercial sources of Bradford reagent that are less predis-
posed to precipitation during prolonged storage.
We have observed significant variation in response between various
commercial suppliers of Bradford preparations (Noble et al., 2007). This
appears to be most pronounced when analyzing low-molecular-weight
proteins or peptides. Indeed the assay is reported to display a molecular
weight cutoff ‘‘threshold’’; requiring a certain number of residues for full
signal development (de Moreno et al., 1986). Changes in the formulation of
the Bradford reagent are reported to change the response generated from
specific proteins; therefore, care should be taken when comparing Bradford
data from different suppliers or preparations (Chan et al., 1995; Friedenauer
and Berlet, 1989; Lopez et al., 1993; Read and Northcote, 1981).
The Bradford assay is sensitive to interferences from various reagents
detailed in Table 8.1 that include most ionic and nonionic detergents and
glycosylated proteins. If precipitation of the reaction mixture occurs, for
example hydrophobic or membrane proteins, the reaction can be supple-
mented with 1 M NaOH at 5–10% (v/v) to aid solubilization.

6. Lowry (Alkaline Copper Reduction Assays)

(Range: 5–100 mg)
The Lowry assay (Lowry et al., 1951) and other preparations with
enhanced assay performance are based on a two-step procedure. Initially,
the Biuret reaction involves the reduction of copper (Cu2þ to Cuþ) by
proteins in alkaline solutions, followed by the enhancement stage, the
reduction of the Folin–Ciocalteu reagent (phosphomolybdate and phos-
photungstate) (Peterson, 1979) producing a characteristic blue color with
absorbance maxima at 750 nm. The assay displays protein sequence
Quantitation of Protein 87

variation, as color development is due not only to the reduced copper–

amide bond complex but also to tyrosine, tryptophan, and to a lesser extent
cystine, cysteine, and histidine residues (Peterson, 1977; Wu et al., 1978).
The Lowry assay has been modified to reduce its sensitivity to interfering
agents, increase the dynamic range and increase the speed and resulting
stability of the color formation (Peterson, 1979). There are many commer-
cial sources of the modified Lowry assay (Roche, Pierce, Bio-Rad, and
Sigma), but different preparations may not give equal responses when using
the same standard, dilution buffer, or interfering compounds.

6.1. Reagents
6.1.1. Folin and Ciocalteu’s reagent
The preparation of this reagent has been described (Lowry et al., 1951);
however, the solution can be obtained from commercial sources (Sigma).
Mix 10 ml of Folin–Ciocalteu’s Phenol reagent to 50 ml of water.

6.1.2. Copper sulfate reagent

100 mg CuSO4  5H2O and 200 mg of sodium tartrate dissolved in 50 ml of
water. Dissolve 10 g of sodium carbonate into 50 ml of water, then pour
slowly while mixing to the copper sulfate solution, prepare fresh daily.

6.1.3. Alkaline copper reagent

Mix one-part copper sulfate solution, one-part 5% SDS (w/v) and two-parts
3.2% sodium hydroxide (w/v). This solution can be stored at room tem-
perature for up to 2 weeks, discard if a precipitate forms.

6.2. Procedure
1. To 1 ml of sample and protein standards  5–100 mg/ml, add 1 ml of the
alkaline copper reagent, mix and allow to stand for 10 min.
2. Add 0.5 ml of Folin–Ciocalteu’s reagent mix, vortex thoroughly and
incubate for 30 min.
3. After incubation vortex again and measure the absorbance at 750 nm.
Absorbance can be read from 650 to 750 nm depending on the avail-
ability of appropriate filters (microplate readers), or if the signal is too
high, without significant loss in assay performance. Lowry is not an
endpoint assay, so samples should be staggered to obtain more accurate
estimates.
4. The response observed will be linear over a limited range of standards.
Polynomial, exponential, and logarithmic models can be used to fit the
data to extend the dynamic range of the response curve.
88 James E. Noble and Marc J. A. Bailey

6.3. Comments
The Lowry method above can be adapted to a microplate format by
reducing the volume of reactants added, resulting in a dynamic range
 50–500 mg/ml. The Lowry assay has been largely superseded by the
BCA assay due to sensitivity, linearity, and improved methodology.
The Lowry protein assay is sensitive to many interfering compounds
(Table 8.1), which may not generate a linear response (making extrapola-
tions of interfering data complex). Formation of precipitates can occur with
detergents, lipids, potassium ions, and sodium phosphate.

7. Bicinchoninic Acid (BCA) (Range: 0.2–50 mg)

The BCA assay replaces the Folin–Cioalteu’s reagent as described for
the Lowry method with Bicinchoninic acid that results in a protein assay
with improved sensitivity and tolerance to interfering compounds (Smith
et al., 1985). The BCA reaction forms an intense purple complex with
cuprous ions (Cuþ) resulting from the reaction of protein and alkaline
Cu2þ. The residues that contribute to the reduction of Cu2þ include the
cysteine, cystine, tryptophan, tyrosine, and the peptide bonds (Smith et al.,
1985). The chemical reaction is temperature dependent with different
functional groups displaying a different reactivity at elevated temperatures,
which result in less protein variability (Wiechelman et al., 1988). At elevated
temperatures (60  C compared to 37  C), more color formation is observed
due to the higher reactivity of tryptophan, tyrosine, and peptide bonds.
Most of the commercial preparations are formulated close to the original
preparation described by Smith et al. (1985), which is described in the
following subsections. Variations have been employed to improve the
sensitivity of the assay and can be obtained from commercial sources
(Pierce, Novagen). The sample-to-working reagent ratio can be varied to
maximize signal, or reduce assay interference, typically ratios of 8–20-fold
excess of BCA working reagent are added to the protein sample.

7.1. Reagents
Reagent A: 1 g sodium bicinchoninate, 2 g Na2CO3, 0.16 g sodium tartrate,
0.4 g NaOH, and 0.95 g NaHCO3, made up to 100 ml and the pH
adjusted to 11.25 with either solid or concentrated NaOH.
Reagent B: 0.4 g CuSO4  5H2O dissolved in 10 ml water. Both reagent
A and B are stable indefinitely at room temperature.
The working solution is prepared by mixing 100 parts of reagent A with
two parts reagent B to form a green solution that is stable for up to a week.
Quantitation of Protein 89

7.2. Procedure
1. Cuvette analysis can be performed with 50–150 ml of protein and 3 ml of
BCA working reagent, whereas microplate assay can use 25 ml of protein
and 200 ml of BCA working reagent, that is a lower reagent to protein
ratio.
2. Incubate the sample and standards  5–250 mg/ml at either 37 or 60  C
for 30 min (longer incubations at 37  C will improve protein-to-protein
variability) and allow the sample to equilibrate to room temperature
before reading. Microplates should be covered during incubation to
avoid evaporation of the sample. For cuvette analysis at 37  C, samples
should be staggered to ensure equal incubation times.
3. Measure absorbance at 562 nm, for filter-based plate readers wavelengths
in the range of 540–590 nm can be used instead without a significant loss
in assay performance.
4. The BCA assay will produce a linear response over a wide concentration
range; however, to extend the dynamic range of the data analysis a
quadratic response can be used to model the data.

7.3. Comments
A microbased BCA assay can be used to improve the sensitivity of the
procedure (1–25 mg/ml). The microbased assay uses a more concentrated
working solution and can be prone to precipitation; again commercial
sources of this modified BCA assay are available (Pierce). The BCA assay
is sensitive to either copper chelators (e.g., EDTA) or reagents that can also
reduce Cu2þ (e.g., DTT), a summary of the maximum tolerances can be
found in Table 8.1.

8. Amine Derivatization (Range: 0.05–25 mg)

Amine-labeling ‘‘derivatization’’ using various fluorescent probes is a
common technique to quantitate amino acid mixtures in AAA. The same
technique can be used to quantitate proteins and peptides containing either
lysine or a free N-terminus, both of which need to be accessible to the dye.
Upon reaction with amines, the dyes display a large increase in fluorescence
that for part of the dynamic range will generate a linear response with
increasing protein concentration. Three dyes that have been used to quanti-
tate proteins, or amino acids in a microplate format include o-phthalaldehyde
(OPA) (Hammer and Nagel, 1986), Fluorescamine (Lorenzen and Kennedy,
1993), and 3-(4-carboxybenzoyl)quinoline-2-carboxyaldehyde (CBQCATM)
(Asermely et al., 1997; Bantan-Polak et al., 2001; You et al., 1997).
90 James E. Noble and Marc J. A. Bailey

Fluorescamine reacts directly with the amine functional group, whereas OPA
and CBQCATM require the addition of a thiol (2-mercaptoethanol) or cyanide
(CBQCATM). A cuvette-based format is described for the OPA assay, which
can be converted to a microplate format by adjusting the volume of reactants
and NaOH.

8.1. Reagents
OPA stock: Dissolve 120 mg of o-phthalaldehyde (high purity grade from
Sigma or Invitrogen) in methanol, then dilute to 100 ml in 1 M boric acid,
pH 10.4 (pH adjusted with potassium hydroxide). Add 0.6 ml of polyox-
yethylene (23) lauryl ether and mix. The stock is stable for 3 weeks at room
temperature.

8.2. Procedure
1. At least 30 min before analysis, add 15 ml of 2-mercaptoethanol to 5 ml
of OPA stock, this reagent is stable for a day. Protect all fluorescent
samples and reactions from light at all times.
2. Protein standards (0.2–10 mg/ml) and unknown samples need to be
adjusted to a pH between 8.0 and 10.5 before analysis. Mix 10 ml of test
sample with 100 ml of OPA stock (supplemented with 2-mercaptoethanol)
and incubate at room temperature for 15 min.
3. Add 3 ml of 0.5 N NaOH and mix.
4. Read fluorescence at excitation 340 nm and emission from 440 to
455 nm in a fluorescent cuvette.
5. The relationship between protein concentration and fluorescence
should be linear over the dynamic range of the assay and can be used
to estimate unknown samples.

8.3. Comments
All three dyes offer improved sensitivity and dynamic range when compared
with absorbance-based protein quantitation assays. OPA is generally pre-
ferred over fluorescamine due to its enhanced solubility and stability in
aqueous buffers.
The use of amine-derivatization agents for protein quantitation is limited
as the assay displays a large protein-to-protein variability due to variation in
the number of lysine residues in proteins, requiring the need for a
‘‘matched’’ standard. Assay interference from glycine and amine containing
buffers, ammonium ions, and thiols common in many biological-buffering
systems limit the application of such assays (Table 8.1). The reproducibility
of the assay is dependent on the pH of the reaction, protein samples that
Quantitation of Protein 91

contain residual acids, for example from precipitation steps could reduce the
rate of amine derivatization (You et al., 1997).
A noncovalent amine reactive dye epicocconone can also be used for
total protein assays in solution (Sigma), for which the mechanism has been
reported (Bell and Karuso, 2003; Coghlan et al., 2005).

9. Detergent-Based Fluorescent Detection

(Range: 0.02–2 mg)
The development of fluorescent probes whose quantum yields are
enhanced significantly when binding at the detergent–protein interface
have been used to quantitate proteins within gels and in solution-based
assays (Daban et al., 1991; Jones et al., 2003). Two commercial preparations
of these assays are available NanoOrangeTM and Quant-iTTM (Invitrogen);
however, limited independent testing of the respective reagents prevents a
full critical analysis. The NanoOrangeTM assay is limited by the need to heat
samples to 90  C to denature the proteins thereby reducing the protein-
to-protein variability ( Jones et al., 2003). Both assays are sensitive to
detergents and high salt concentrations (Table 8.1), which presumably
disrupt the protein-dye-detergent interface. The Quant-iTTM assay displays
good sensitivity and dynamic range compared to other dye-based assays
(Fig. 8.2), and a relatively low protein-to-protein variability (Fig. 8.3).

10. General Instructions

The choice of measurement format used will depend on the through-
put, sensitivity, and precision required of the assay, and concerns about assay
interferences that can be reduced by dilution in cuvette-based assays. From
our experience, both in industry and academia, plate-based assays are
replacing cuvette assays due to increases in throughput. For all spectropho-
tometric techniques, the instrument should be warmed up for 15 min prior
to measurement and any calibration programs run before sample analysis.
Samples and reagents should be equilibrated to room temperature before
analysis to avoid condensation on optical surfaces.

10.1. Cuvettes
Traditionally, cuvettes have been used for the majority of spectrophoto-
metric protein assays. Quartz cuvettes can be costly, therefore glass cuvettes
are preferred; however, both of these may have to be washed between
measurements to remove dye and adsorbed protein. Disposable plastic
92 James E. Noble and Marc J. A. Bailey

cuvettes are available and can be used to increase the throughput where
many samples have to be measured, or the reagent is prone to sticking to the
cuvette surface, for example Bradford reagent. Staggering of sample analysis
is especially important if the signal is not stable, or does not run to comple-
tion within the time frame of the assay, for example BCA or Lowry assays.
The best precision is obtained from a two-beam instrument incorporating a
reference cell to account for instrument drift. Replacing the cuvette in the
holder between each measurement due to cleaning, or the use of disposable
cuvettes can result in changes in alignment, resulting in significant changes
in amount of light reaching the detector. This is especially important if low-
volume cuvettes are being used where the transmission window is reduced
in size. Care should also be taken with low-volume cuvettes to ensure the
sample covers the entire transmission window.
Care should be taken when handling and cleaning cuvettes. Prevent
fingerprints from contaminating the transmitting surfaces. Cuvettes should
be washed with either water or an appropriate solvent between runs and
dried using a stream of nitrogen gas. If smearing of the transmitting surface is
observed, the cuvette can be rewashed in water, ethanol, and finally ace-
tone, or removed using ethanol and lintless lens tissue. If protein deposition
is a recurring problem, cuvettes can be washed overnight in nitric acid and
thoroughly washed before use.

10.2. Microwell plates

The majority of protein assays have been adapted for use in microwell
plates, typically 96-well plates to enhance speed, throughput and lower
sample and reagent usage. Many of the commercial fluorescent assays are
specifically designed for plate formats. The plate reader format also offers the
advantage of being able to read multiple samples within a short period
(typically 25 s) reducing potential timing differences in reactions that do
not go to completion, or are unstable.
Protein UV measurements can be made in a plate format; however, the
effective path length can be difficult to calculate due to meniscus formation
for concentrated protein solutions (many commercial plate readers can
estimate effective path-length and thereby improving protein quantitation
calculations). Quartz 96-well plates tend to be expensive, difficult to clean
and prone to scratches that can affect light transmission.
Care should be taken in the preparation of protein assays in plate formats.
The use of lower volume samples (down to 5 ml for some assays) can
increase the relative pipetting errors of high viscosity solutions. Well-
to-well contamination should be avoided by using fresh pipette tips for
each sample and reagent. Regular calibration of the instrument should be
performed using either optical standards or solid phase fluorescent standard
plates (Matech) to ensure equal transmission/light detection from all wells.
Quantitation of Protein 93

Many of the 96-well plates conform to a standard geometry; however, in

our experience, it is worth analyzing plate geometry in the plate-reader,
especially if a different plate supplier is used to ensure equal illumination,
and detection for fluorescent-based measurements. Plate-based assays can
also be more sensitive to sample precipitation (common in the Lowry and
Bradford assays) when compared to cuvette-based assay due to the detection
geometry.
Recently, spectrophotometers that can measure low microliter samples
(typically 1–2 ml), without the need for a cuvette or microplates have
become commercially available (Tecan and Thermo Scientific), further
minimizing sample usage.

10.3. Interfering substrates

Interfering substances for many protein preparations will be variable from
batch-to-batch and can be difficult to adequately control for when standards
are formulated differently. The choice of assay used should take into
account interfering contaminants in the protein preparation, either used as
stabilizers or as a result of purification that cannot be replaced, or substituted
with a suitable alternative, for example reducing agents or chelators. Inclu-
sion of an interfering substance can be accommodated using a matched
standard; however, this can result in loss of dynamic range and poor assay
performance and is therefore not recommended. The concentration or
amount of interfering substance that can be tolerated is often quoted with
the assay instructions; however, this can be dependent on the formulation of
the assay, the maximum tolerated concentrations are summarized in
Table 8.1.
Interfering substances can be removed prior to concentration determi-
nation, however, this adds additional steps to the procedure and can often
result in dilution, or incomplete recovery of the original sample leading to
errors in the concentration estimate. Changes in sample recovery can be
compensated for by comparing the recovery of the standard that has been
subjected to interference removal steps.
Precipitation of protein followed by separation and resuspension proba-
bly offers the most accurate method to remove interfering substances where
they cannot be avoided. Buffer components, detergents, and lipids can be
removed by precipitating the protein with trichloroacetic acid (TCA),
perchloric acid (PCA), or acetone (Olson and Markwell, 2007); however
Triton X-100 can coprecipitate with TCA and PCA.
In addition to precipitation techniques, specific interferences can be
removed through chemical treatment, for example reducing agents (iodoa-
cetic acid treatment), lipids through chloroform extraction, volatility, or
neutralization of strong acids/bases.
94 James E. Noble and Marc J. A. Bailey

ACKNOWLEDGMENT
We thank A. Hills for comments and help in preparing this review.

REFERENCES
Alterman, M., Chin, D., Harris, R., Hunziker, P., Le, A., Linskens, S., Packman, L., and
Schaller, J. AAARG2003 study: Quantitation of proteins by amino acid analysis and
colorimetric assays. Poster available for download from www.abrf.org (http://www.abrf.
org/ResearchGroups/AminoAcidAnalysis/EPosters/aaa2003_poster_print.pdf ).
Asermely, K. E., Broomfield, C. A., Nowakowski, J., Courtney, B. C., and Adler, M.
(1997). Identification of a recombinant synaptobrevin-thioredoxin fusion protein by
capillary zone electrophoresis using laser-induced fluorescence detection. J. Chromatogr.
B Biomed. Sci. Appl. 695, 67–75.
Bantan-Polak, T., Kassai, M., and Grant, K. B. (2001). A comparison of fluorescamine and
naphthalene-2, 3-dicarboxaldehyde fluorogenic reagents for microplate-based detection
of amino acids. Anal. Biochem. 297, 128–136.
Bell, P. J., and Karuso, P. (2003). Epicocconone, a novel fluorescent compound from the
fungus epicoccumnigrum. J. Am. Chem. Soc. 125, 9304–9305.
Blakeley, R. L., and Zerner, B. (1975). An accurate gravimetric determination of the
concentration of pure proteins and the specific activity of enzymes. Methods Enzymol.
35, 221–226.
Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram
quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72,
248–254.
Burkitt, W. I., Pritchard, C., Arsene, C., Henrion, A., Bunk, D., and O’Connor, G. (2008).
Toward Systeme International d’Unite-traceable protein quantification: From amino
acids to proteins. Anal. Biochem. 376, 242–251.
Chan, J. K., Thompson, J. W., and Gill, T. A. (1995). Quantitative determination of
protamines by coomassie blue G assay. Anal. Biochem. 226, 191–193.
Coghlan, D. R., Mackintosh, J. A., and Karuso, P. (2005). Mechanism of reversible
fluorescent staining of protein with epicocconone. Org. Lett. 7, 2401–2404.
Daban, J. R., Bartolome, S., and Samso, M. (1991). Use of the hydrophobic probe Nile red
for the fluorescent staining of protein bands in sodium dodecyl sulfate-polyacrylamide
gels. Anal. Biochem. 199, 169–174.
de Moreno, M. R., Smith, J. F., and Smith, R. V. (1986). Mechanism studies of coomassie
blue and silver staining of proteins. J. Pharm. Sci. 75, 907–911.
Fountoulakis, M., Juranville, J. F., and Manneberg, M. (1992). Comparison of the
Coomassie brilliant blue, bicinchoninic acid and Lowry quantitation assays, using non-
glycosylated and glycosylated proteins. J. Biochem. Biophys. Methods 24, 265–274.
Friedenauer, S., and Berlet, H. H. (1989). Sensitivity and variability of the Bradford protein
assay in the presence of detergents. Anal. Biochem. 178, 263–268.
Gill, S. C., and von Hippel, P. H. (1989). Calculation of protein extinction coefficients from
amino acid sequence data. Anal. Biochem. 182, 319–326.
Groves, W. E., Davis, F. C. Jr., and Sells, B. H. (1968). Spectrophotometric determination
of microgram quantities of protein without nucleic acid interference. Anal. Biochem. 22,
195–210.
Hammer, K. D., and Nagel, K. (1986). An automated fluorescence assay for subnanogram
quantities of protein in the presence of interfering material. Anal. Biochem. 155, 308–314.
Quantitation of Protein 95

Jones, L. J., Haugland, R. P., and Singer, V. L. (2003). Development and characterization of
the NanoOrange protein quantitation assay: A fluorescence-based assay of proteins in
solution. Biotechniques 34, 850–854, 856, 858 passim.
Kupke, D. W., and Dorrier, T. E. (1978). Protein concentration measurements: The dry
weight. Methods Enzymol. 48, 155–162.
Leach, S. J., and Scheraga, H. A. (1960). Effect of light scattering on ultraviolet difference
spectra. J. Am. Chem. Soc. 82, 4790–4792.
Lopez, J. M., Imperial, S., Valderrama, R., and Navarro, S. (1993). An improved Bradford
protein assay for collagen proteins. Clin. Chim. Acta 220, 91–100.
Lorenzen, A., and Kennedy, S. W. (1993). A fluorescence-based protein assay for use with a
microplate reader. Anal. Biochem. 214, 346–348.
Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951). Protein measure-
ment with the Folin phenol reagent. J. Biol. Chem. 193, 265–275.
Noble, J. E., Knight, A. E., Reason, A. J., Di Matola, A., and Bailey, M. J. (2007). A
comparison of protein quantitation assays for biopharmaceutical applications. Mol. Bio-
technol. 37, 99–111.
Olson, B. J., and Markwell, J. (2007). Assays for determination of protein concentration.
Curr. Protoc. Protein Sci. Chapter 3, Unit 3.4.
Ozols, J. (1990). Amino acid analysis. Methods Enzymol. 182, 587–601.
Pace, C. N., Vajdos, F., Fee, L., Grimsley, G., and Gray, T. (1995). How to measure and
predict the molar absorption coefficient of a protein. Protein Sci. 4, 2411–2423.
Peterson, G. L. (1977). A simplification of the protein assay method of Lowry et al. which is
more generally applicable. Anal. Biochem. 83, 346–356.
Peterson, G. L. (1979). Review of the Folin phenol protein quantitation method of Lowry,
Rosebrough, Farr and Randall. Anal. Biochem. 100, 201–220.
Read, S. M., and Northcote, D. H. (1981). Minimization of variation in the response to
different proteins of the Coomassie blue G dye-binding assay for protein. Anal. Biochem.
116, 53–64.
Sapan, C. V., Lundblad, R. L., and Price, N. C. (1999). Colorimetric protein assay
techniques. Biotechnol. Appl. Biochem. 29(Pt 2), 99–108.
Scopes, R. K. (1974). Measurement of protein by spectrophotometry at 205 nm. Anal.
Biochem. 59, 277–282.
Sittampalam, G. S., Ellis, R. M., Miner, D. J., Rickard, E. C., and Clodfelter, D. K. (1988).
Evaluation of amino acid analysis as reference method to quantitate highly purified
proteins. J. Assoc. Off. Anal. Chem. 71, 833–838.
Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H.,
Provenzano, M. D., Fujimoto, E. K., Goeke, N. M., Olson, B. J., and Klenk, D. C.
(1985). Measurement of protein using bicinchoninic acid. Anal. Biochem. 150, 76–85.
Stoscheck, C. M. (1990). Quantitation of protein. Methods Enzymol. 182, 50–68.
Wiechelman, K. J., Braun, R. D., and Fitzpatrick, J. D. (1988). Investigation of the
bicinchoninic acid protein assay: Identification of the groups responsible for color
formation. Anal. Biochem. 175, 231–237.
Wu, A. M., Wu, J. C., and Herp, A. (1978). Polypeptide linkages and resulting structural
features as powerful chromogenic factors in the Lowry phenol reaction. Studies on a
glycoprotein containing no Lowry phenol-reactive amino acids and on its desialylated
and deglycosylated products. Biochem. J. 175, 47–51.
You, W. W., Haugland, R. P., Ryan, D. K., and Haugland, R. P. (1997). 3-(4-Carbox-
ybenzoyl)quinoline-2-carboxaldehyde, a reagent with broad dynamic range for the assay
of proteins and lipoproteins in solution. Anal. Biochem. 244, 277–282.
 C H A P T E R N I N E

Concentration of Proteins
and Removal of Solutes
David R. H. Evans,* Jonathan K. Romero,* and Matthew Westoby†

Contents
1. Chromatography 98
1.1. Gel filtration 99
1.2. Ion-exchange chromatography 99
1.3. Reversed phase chromatography 100
1.4. Hydrophobic and affinity chromatography 100
1.5. Manufacturing-scale chromatographic applications 102
2. Electrophoresis 103
3. Dialysis 104
3.1. Concentration and affinity binding applications 107
4. Ultrafiltration 107
4.1. Ultrafiltration membranes 109
4.2. Ultrafiltration devices 110
4.3. Purification applications 112
5. Lyophilization 113
6. Precipitation 116
7. Crystallization 118
References 118

Abstract
The dramatic advances in recombinant DNA and proteomics technology over the
past decade have supported a tremendous growth in biologics applied to
diagnostics, biomarkers, and commercial therapeutic markets. In particular,
antibodies and fusion proteins have now become a main focus for a broad
number of clinical indications, including neurology, oncology, and infectious
diseases with projected increase in novel first-class molecules and biosimilar
entities over the next several years. In line with these advances are the
improved analytical, development, and small-scale preparative methods

* Process Biochemistry, Biopharmaceutical Development, Biogen Idec. Inc., Cambridge, Massachusetts, USA
{
Process Biochemistry, Biopharmaceutical Development, Biogen Idec. Inc., San Diego, California, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63009-3 All rights reserved.

97
98 David R. H. Evans et al.

employed to elucidate biologic structure, function, and interaction. A number of

established methods are used for solvent removal, including lyophilization,
reverse extraction, solute precipitation, and dialysis (solvent exchange), ultra-
filtration, and chromatographic techniques. Notably, advances in the miniaturi-
zation and throughput of protein analysis have been supported by the
development of a plethora of microscale extraction procedures and devices
that exploit a wide array of modes for small-scale sample preparation, including
the concentration and desalting of protein samples prior to further analysis.
Furthermore, advances in process handling and data monitoring at microscale
have dramatically improved complex control and product recovery of small
quantities of biologics using techniques such as lyophilization and precipita-
tion. In contrast, the efficient concentration of feed streams during preparative
chromatography has been enhanced by improvements to protein binding capac-
ity achieved through advanced bead and ligand design.
The objective of solvent removal may be to prepare or concentrate solutes for
analysis, or to facilitate their production or modification. Here, we describe the
most recent advances in these techniques, particularly focusing on improved
capabilities for bench-scale preparative methods.

1. Chromatography
Over the past two decades, the innovation of chromatographic tech-
niques for protein concentration and desalting has been driven by the
increasing application of proteomics technologies. Proteomics generally
requires an enrichment of proteins from small amounts of complex mixtures
including the selective concentration of specific species. The high sensitivity
of detection and detailed biochemical characterization provided by mass
spectrometry (MS) and antibody-based assays has found utility in diverse
applications including clinical diagnostics, biologic drug characterization,
and basic research into protein structure–function. These applications
require the analysis of drug targets and biomarkers present at low amounts
in biological fluids, or the acquisition of detailed information on the
structure and posttranslational modification of proteins expressed from
heterologous systems. Samples for proteomics analysis are routinely gener-
ated in aqueous solutions that may contain a range of nonvolatile
salts, solvents, dyes, detergents, chaotropes, DNA/RNA, and lipids.
These contaminants can impair the performance of sensitive analytical
techniques, notably electrospray ionization (ESI) matrix-assisted laser
desorption/ionization (MALDI) MS, where they cause a detrimental effect
on resolution and sensitivity by interfering with sample ionization, adduct
formation, and ion-source fouling. As a result, an excess of microscale solid
phase extraction procedures and devices have been developed by multiple
Concentration of Proteins and Removal of Solutes 99

vendors to concentrate target proteins and purify them from contaminating

molecules with high yield.

1.1. Gel filtration

Gel filtration matrices have been widely used for desalting proteins under
nondenaturing conditions. Unlike small molecules, macromolecular pro-
teins are excluded from the pores of these resins and are desalted during
flow. Typically, a bed volume of 4–20 times the volume of the sample and a
column length-to-diameter ratio of between 5 and 10 provide optimal
resolution. A minimum sample dilution of 1.25-fold is generally obtained
during desalting but, with an appropriate selection of eluant, protein losses
can be minimal even with sample concentrations in the mg/ml range.
Polymeric carbohydrate-based sephadex media offered by various vendors
have been extensively used for desalting and a range Bio-Gel P acrylamide-
based matrices (Bio-Rad, Hercules, CA) permit the user to tailor the gel
exclusion limit of the matrix to the molecular weight of the target protein.
Added convenience is provided by Zeba Spin columns (Thermo/Pierce,
Rockford, IL) which permit rapid desalting of protein samples of molecular
weight >7000 Da by centrifugation without the requirement for column
equilibration or gravity flow fractionation. Through appropriate choice of
column size, sample volumes in the range 2–4000 ml can be accommodated
with high recovery at protein concentrations as low as 25 mg/ml and with
95% retention of molecules of molecular weight <1000 Da. The Zeba Spin
matrix is also available in 96-well plate format facilitating high-throughput
nondenaturing desalting of multiple protein samples in parallel.

1.2. Ion-exchange chromatography

An alternative to gel filtration for protein desalting is the use of mixed bed
ion-exchange (IEX) resins. Large bead size and surface modification of
mixed bed IEX sorbents can permit effective capture of small anions and
cations (<1000 Da) with minimal retention of high-molecular-weight
protein molecules. Desalting may be achieved in batch mode or via a packed
bed of the matrix. However, an undesirable consequence of this technique
is that the exchange of ions that occurs on a mixed bed of conventional IEX
sorbents can cause acidification of the protein sample and precipitation.
In contrast, ion retardation matrices, which have been used to desalt milk
and whey, undergo salt uptake by absorption rather than by ion exchange
and may therefore achieve the required protein desalting without significant
pH change. A commercially available example is AG11 A8 resin (Bio-Rad,
Hercules, CA). Similar to gel filtration, a drawback to these techniques is
that the protein sample is diluted during the desalting process and may
require subsequent concentration.
100 David R. H. Evans et al.

1.3. Reversed phase chromatography

Reversed phase (RP) chromatography is a popular technique for the sepa-
ration, desalting, and concentration of proteins, in part because the sample is
concentrated in a small volume of volatile solvent that can be removed by
evaporation. With the advent of routine analysis by MS, a broad array of RP
extraction formats have been commercially developed improving the speed
and convenience of the technique. These formats include microspin col-
umns, capillaries, micropipette tips, and plates for rapid desalting of as little
as femtomole quantities of proteins with recovery in microlitre volumes of
eluate. Typically, optimal protein binding is achieved at pH <4 in the
presence of an ion-pairing agent such as 0.1–1.0% TFA. Reversible binding
may be promoted by the presence of an organic component, typically 15%
acetonitrile or methanol, or a chaotropic agent such as guanidine hydro-
chloride. Dilution of contaminating detergents may be required to prevent
interference with protein binding to the sorbent. Desalting may require the
resin to be washed, for example with 5% methanol, 0.1% TFA, prior to
elution while selective elution may be achieved by experimenting with
alternative ion-pairing agents. The choice of RP sorbent is based on protein
size, with aliphatic C4, C8, C12, and C18 chemistries being routinely
employed. Examples of micropipette-based products are ZipTips
(Millipore, Billerica, MA) which are packed with RP beads or NuTips
(Glygen, Columbia, MD) in which the interior walls of the tip are coated
with the binding material thereby reducing backpressure and minimizing
nonspecific interactions with the tip surfaces. For high-throughput desalting
of multiple samples in parallel, RP resin can also be obtained in 96-well
plates, an example being the ZipPlate format (Millipore, Billerica, MA).

1.4. Hydrophobic and affinity chromatography

In addition to the commonly used gel filtration, IEX, and RP sorbents,
additional chemistries are available as batch resin or in microscale device
format that support various modes of chromatography for protein concen-
tration and desalting. These include hydrophobic, hydrophilic, metal che-
lating, and protein affinity resins. Notably, hydrophilic media provide a
useful alternative to RP chemistry when the protein sample is suspended in
an organic solvent and removal of contaminating salts or detergents is
required. An example is PAGE-Prep media (Thermo/Pierce, Rockford, IL)
that can be used to desalt proteins prior to SDS–PAGE analysis when bound
to the resin in the presence of 50% dimethyl sulfoxide. Alternatively,
Glygen (Columbia, MD) offers hydrophilic media in a micropipette tip
format (NuTip). Similar to RP media, a potential disadvantage of hydro-
philic resins is that the protein is generally recovered in a denatured form. In
contrast, an important use of immobilized metal affinity chromatography
Concentration of Proteins and Removal of Solutes 101

(IMAC) sorbents and, more recently, media-containing TiO2 has been to

selectively concentrate and desalt phosphopeptides prior to MS analysis.
This has found particular application in the area of signal transduction
research where the detection of low-abundance phosphopeptides is critical
to the understanding of cellular responses. A recent report (Hsieh et al.,
2007) described the preparation of a TiO2 nanoparticle matrix capable of
selectively binding phosphopeptides at pH 2.0 in 0.1% TFA and 50%
acetonitrile to prevent nonspecific binding. In this case, elution was per-
formed with 100 mM ammonium phosphate pH 8.5, which was compatible
with detection of phosphorylated peptides by MALDI-MS analysis without
further desalting.
The routine preparation of protein samples for MS analysis using HPLC
techniques has driven vendors to provide their sorbents in an increasing
array of column, cartridge, and capillary formats. This in turn has permitted
the streamlining of sample preparation via the use of column switching
procedures. In this approach a short precolumn is used to rapidly concen-
trate and desalt a protein prior to a subsequent, high-resolution separation
via a longer column on the same HPLC system. As an example, PepMap
microscale precolumns (Dionex, Sunnyvale, CA) can be used to rapidly
concentrate and desalt protein samples over a packed bed length of only
5 mm. With a choice of resins available, the efficiency of analyte trapping is
governed by the choice of packing material while the reduction in sample
volume improves the resolution of the subsequent chromatography step. In
this context, it is also worth comment that monolithic separation materials
are being increasingly employed for protein concentration and desalting
(Schley et al., 2006). The effective pore size of up to 20 mm permits rapid
convective transport of solutes through the monolithic matrix. This con-
trasts with the slower diffusion-mediated mass transport achieved within the
submicron pores of bead-based sorbents and enables monoliths to support
efficient separations at high flow rates with low backpressure, which is
particularly useful for rapid desalting by column switching. Nevertheless,
in a more direct approach, Mass-Prep columns (Waters, Milford, MA) have
been used for in-line desalting of proteins on a hydrophobic matrix fol-
lowed by direct injection into an ESI mass spectrometer (Wheat et al.,
2007). Furthermore, sample preparation for MALDI-MS analysis may be
performed without the requirement for a microextraction device. Thus, a
wide range of sorbents have been used for desalting proteins in batch mode
prior to spotting onto a MALDI plate. A recently developed example is a
ZnO-poly(methyl methacrylate) nanobead adsorbent capable of clearing
saturating levels of NaCl (6.2 M ) or NH4HCO3 (2.6 M ) or 1 M urea
from a protein sample prior to MALDI-MS (Shen et al., 2008). Similar
desalting has been achieved by applying a hydrophobic matrix directly to
the wells of a MALDI plate ( Jia et al., 2007). In this case, the target protein
was captured on the matrix and, following drying, desalting of the sample
102 David R. H. Evans et al.

was achieved by applying a minimal on-plate wash which permitted

MALDI-MS analysis without further purification.

1.5. Manufacturing-scale chromatographic applications

Similar to microscale extractions used in analytical applications, protein recov-
ery at preparative-scale and manufacturing-scale frequently involves concen-
tration and desalting. Column chromatography remains a central technique for
the purification of proteins due to the high-selectivity achievable combined
with the convenience and robustness of a packed resin bed to which the feed is
delivered under flow. Commercial manufacturing of enzymes and biologic
drugs commonly requires the processing of thousands of liters of a starting
feedstock. To minimize operating complexity and processing time while
controlling costs, it may be essential to concentrate the feed and remove the
bulk of impurities and unwanted salts early in the downstream process. This can
be achieved by the use of large resin beds of up to 2 m in column diameter for
bind-and-elute chromatography that generates an eluate fraction smaller than
the column load volume. Furthermore, the selectivity of the resin may be
exploited to elute a purified product. A notable example is the use of protein A
affinity resins for the production of antibody-based proteins expressed from
mammalian cell culture. During affinity capture, which is mediated by a
specific epitope present in the Fc region of the target molecule, the bulk of
the media components, salts, and impurities flow through the column. Product
recovery is typically achieved by elution at low pH and is generally associated
with a 90% reduction in volume. The latest generation of Protein A affinity
resins have been optimized for ligand density, bead structure, and mass transfer
characteristics to maximize dynamic binding capacity while minimizing
column cycling. This in turn minimizes intermediate volumes and effectively
concentrates the stream at the first column step in the process. Examples of
resins are MabSelect (GE Healthcare, Piscataway, NJ) and ProSep Ultra Plus
(Millipore, Billerica, MA) for which binding capacities in the range of
20–50 g/l of resin can be achieved depending on the target molecule and its
contact time with the resin during column loading. Conventional chromatog-
raphy resins, including IEX and hydrophobic interaction chromatography
(HIC) sorbents, have been similarly optimized for particle size, pore size, and
ligand density to achieve highest capacity binding of molecules within a
particular molecular weight range but with efficient elution. For example,
ToyoPearl GigaCap S-650M (Tosoh Biosciences, Tokyo, Japan) is a cation-
exchange resin that has been optimized for the capture of IgG1 molecules
with a molecular weight of approximately 150 kDa at high dynamic binding
capacity (>100 g/l of resin) while permitting recovery in a minimal elution
volume, typically 2–3 column volumes. Finally, mixed mode sorbents have
been developed that permit IEX chromatography to be performed with high-
capacity binding at relatively high conductivity. This has been achieved
Concentration of Proteins and Removal of Solutes 103

through the combination of an IEX and hydrophobic moiety on the same

ligand. An example is Capto MMC resin (GE Healthcare, Piscataway, NJ)
which combines a cation exchange and hydrophobic group and binds
bovine serum albumin with a capacity of 45 mg/ml of resin at a conductivity
of 30 mS/cm. This property can be useful because, while optimal protein
purification is generally achieved via the application of multiple orthogonal
modes of chromatography performed as separate steps, it can be challenging to
maintain the feed conductivity at a level compatible with a conventional IEX
step without dilution. Therefore, although this type of sorbent may not
concentrate the process stream directly, it may obviate the need for
a reduction in conductivity by dilution during downstream processing.
A further attraction of mixed mode resins is that they may provide unique
selectivities which may enhance the overall purification process.

2. Electrophoresis
Concentration of proteins by gel electrophoresis is generally employed
to enhance the sensitivity of a subsequent analytical technique. In one
example, the sensitivity of Western blot analysis was enhanced by applying
up to 250 ml of a protein sample over five consecutive loadings on a 1–2 cm
stacking gel in a mini-gel system (Sheen and Ali-Khan, 2005). The
approach permitted effective sample concentration with vertical band
broadening as the only artifact. A similar concentration of a dilute sample
prior to western blotting has been achieved via a funnel shaped loading well.
Concentration of protein samples by stacking has been achieved by capillary
electrophoresis (CE) (e.g., see Shihabi, 2002). Moreover, a recent report has
described conditions for CE that permit simultaneous concentration and
desalting of proteins. This is a notable advance, as desalting is normally
required prior to CE because ionic salts cause band broadening through
Joule heating and electrodispersion. In addition, a desirable stacking effect
can be achieved when samples for CE are loaded at a lower ionic strength
than the electrolyte, leading to protein enrichment and improved sensitivity
in analytical applications. Desalting of proteins prior to CE can be achieved
by standard RP extractions. However, while solid phase mini-beds that
permit desalting in line with CE have been described, the approach may be
technically challenging and generally the desalting step is performed as a
separate operation. Thus, to overcome these drawbacks a novel technique
termed capillary isoelectric trapping was developed to simultaneously con-
centrate and desalt protein samples during CE (Booker and Yeung, 2008).
The approach utilizes a discontinuous buffer system to establish a stable pH
boundary within the capillary. The stationary boundary traps zwitterionic
proteins at their pI while removing contaminating ionic salts, which
104 David R. H. Evans et al.

continue to migrate due to their constant charge. This technique facilitates

the recovery of concentrated and desalted protein samples for MS or,
following a switch to an alternative buffering system, permits the analysis
of the concentrated protein by capillary zone electrophoresis. Finally,
microfluidic devices are finding increasing utility for sample preparation
prior to biochemical analyses, in part due to their ability to rapidly concen-
trate and desalt small protein samples via the application of an electric field
(e.g., see Yu et al., 2008).

3. Dialysis
Dialysis is the size-based separation of molecules in solution by
selective diffusion through a semipermeable membrane. This technique
is considered the most popular method employed for removal of low-
molecular-weight solutes from large protein molecules. In particular, the
nondenaturing desalting technique permits buffer exchange under benign
or physiological conditions with minimum risk of impacting the target
protein’s properties.
The mechanism of dialysis has not changed, but the techniques and tools
for employing dialysis have improved over the past decade. In particular,
optimized techniques have allowed for higher throughput implementation
(reduced processing times), increased flexibility for sample volume proces-
sing (10 ml ! 100 ml), improved membrane morphology for reduced protein
adsorption and loss, and enhanced convenience for sample handling.
Typically, the volume of the target buffer solution (dialysate) will be several
orders of magnitude greater than the protein sample volume. The sample is
transferred to a sealed compartment and exposed to dialysis membrane which
functions as a semipermeable barrier. The membrane typically possesses a
specific pore size, termed the molecular weight cutoff (MWCO) limit, and
permits the free passage of molecules smaller than this in both directions while
retaining large macromolecules such as proteins. The buffer concentration
gradient across the membrane drives the diffusion of solutes from regions of
high concentration to regions of low concentration (see Fig. 9.1). The diffu-
sional transport or flux is described using Fick’s law for diffusion:

@C
J ¼ D
@x

where J is the diffusion flux in dimensions of [(amount of substance) L 2 T 1],

D is the diffusion coefficient or diffusivity in dimensions of [L2 T 1], c is the
concentration in dimensions of [(amount of substance) L 3], and x is the
position [L]. The diffusion coefficient is proportional to the squared velocity
 t=0 t = t1 t = t2

Sample Dialysate Sample Dialysate

Sample Dialysate

Dialysis membrane Dialysis membrane Dialysis membrane

Sample Dialysate Sample Dialysate Sample Dialysate
Ci CA,o Ci Ci
CB,o CB,t1 CB,t = final
Concentration

Concentration

Concentration
CA,t1

CP,o CP,o
CP,o
CA,t = final
0 0 0
Volume of sample (VS) is placed in Since protein P can not permeate Since VS << VD solute A concentration will
a dialysis device containing a membrane or dialysis device, drop to near undetectable levels while
membrane pore size that retains concentration profile remains solute B in sample device will attain
protein of interest (P) and device unchanged. Solute A concentration in concentrations levels near CB,o. Replacing
placed in the dialysate (VD) where sample drops as A diffuses through dialysate with new buffer will ensure
VS<<VD membrane an into dialysate while solute B CB,0 = CB,tfinal. Equilibrium is reached when
diffuses from dialysate into sample solute concentrations are equivalent on
each side of membrane, respectively

Macromolecule, P Solute A Solute B VD Volume dialysate VS Volume sample

Figure 9.1 Schematic of the dialysis process. Top circular diagrams provide graphic representation of solute distributions within ‘‘sample’’
and ‘‘dialysate’’ pools throughout dialysis process. The bottom graphs show the respective concentration profiles from each pool.
106 David R. H. Evans et al.

of the diffusing particles, which depends on the temperature, viscosity of the

fluid, and the size of the particles according to the Stokes–Einstein relation
(Cussler, 1997). Dialysis is most efficient when the dialysate volume is several
orders of magnitude larger than the sample volume (see Fig. 9.1). This volume
difference maintains a concentration differential across the membrane and
ensures that solute being transferred from the sample device to the dialysate is
diluted to near undetectable concentrations. Figure 9.1 shows the separation of
small (e.g., salts) and large (e.g., macromolecules, proteins) molecules in
solution by selective diffusion through a semipermeable dialysis membrane
including the respective concentration profiles at various time points. Once the
liquid–membrane–liquid interface is established, including concentration dif-
ferences across the semipermeable membrane (t ¼ 0), molecules will diffuse in
the direction of decreasing concentration across the membrane (t > 0). The
rate of diffusion is directly proportional to the diffusion coefficient ‘‘D’’ and the
concentration gradient @C/@x. Assuming D is constant, @C/@x decreases with
time moving from left to right in Fig. 9.1, with the rate of dialysis decreasing as
equilibrium is approached (Cussler, 1997). The starting dialysate can be
refreshed to recreate the concentration differential and drive the buffer
exchange to effective completion.
There are several aspects of dialysis which can influence recovery of a stable
product. These include the time required for buffer exchange, the design of
the dialysis system, and membrane chemistry and morphological properties. The
time required to complete the required buffer exchange will depend on several
factors. The inherent diffusion of molecules across a semipermeable membrane
can be enhanced to minimize process time. Based on the equation shown
previously, transport can be enhanced by increasing the membrane area or the
solute flux. It may be easier to achieve a large ratio of sample to dialysate when
sample volume is small. However, the impact of increased membrane area or
increased time (for small membrane area) needs to be evaluated in the context of
protein loss due to adsorption on the membrane. Since, solute diffusion coeffi-
cients are temperature dependent, an increase in temperature will drive faster
molecular motion, increase intrinsic diffusion coefficients, and result in faster
transport. However, increasing temperature can adversely impact sample stabil-
ity. Therefore, since sample integrity is generally of highest importance, it may
be necessary to predetermine the highest temperature compatible with stability.
The dialysis system design can also dictate the time required for effective
dialysis and efficiency of product recovery. Solutes diffusing away from the
sample permeate enter a boundary layer on the dialysate side of the membrane.
The solute concentration in this boundary layer is higher in relation to the rest
of the dialysate thus lessening the concentration gradient across the membrane
and inhibiting diffusion. However, the boundary layer can be reduced by
inducing convective flow or mixing on the dialysate side of the membrane.
Finally, the mechanical design of current dialysis membranes and
cartridges allows for the processing of a wide range of sample volumes
Concentration of Proteins and Removal of Solutes 107

(10 ml ! 100 ml) with negligible product loss. The traditional dialysis
tubing is difficult to prepare before it can be used. Pierce Biotech (Rockford,
IL) offers SnakeSkinÒ dialysis tubing to simplify large sample dialysis. This
technology comes in a pleated regenerated cellulose membrane format (from
3.5 to 10 kDa MWCO) allowing rapid preparation. In the case with small
sample volume dialysis, Pierce offers compact Slide-A-LyzerÒ Mini-dialysis
Units (10–100 ml), and Cassettes (0.1–30 ml) to process limited small biological
samples. These units are disposable cups made of polypropylene and regener-
ated cellulose allowing sample to be easily removed using standard lab pipettes.
The new designs allow for improved recovery of valuable small sample
volumes.

3.1. Concentration and affinity binding applications

Protein buffer-exchange applications have dominated the use of dialysis
membranes. However, there is increased application of dialysis in the area of
concentration and macromolecular affinity binding interactions.
The use of dialysis membranes for protein concentration is similar to
aqueous two-phase extraction where reagents (e.g., polyethylene glycol,
dextran, polymers) are used to generate a two-phase system with a dialysis
membrane separating the phases (Waziri et al., 2004). The system is
designed such that protein can freely pass through the membrane and
interact with the precipitating agent which can then be subsequently con-
centrated and product recovered.
The technique termed ‘‘equilibrium dialysis’’ has also been utilized for
characterization of a candidate drug in serum binding assays, studying
antigen–antibody interactions and evaluating low-affinity interactions that
are undetectable using other methods. In equilibrium dialysis, a membrane
is selected such that the desired protein or ligand can pass freely through but
with the receptor molecule located only on one side and unable to perme-
ate. As the protein diffuses across the membrane some of it will bind to the
receptor and some will remain free in solution. Diffusion of the ligand across
the membrane and binding to the receptor continues until equilibrium has
been reached. The concentration of free ligand in the ligand chamber can
then be used to determine the binding characteristics of the sample (Waters
et al., 2008)

4. Ultrafiltration
Ultrafiltration (UF) exploits a separation mechanism essentially equiv-
alent to that employed for dialysis. Both techniques employ a semiperme-
able barrier or membrane to separate sample (containing desired product or
108 David R. H. Evans et al.

solute) from sample-free solution. The membrane is designed to permit

low-molecular-weight solutes (e.g., salts) to permeate while completely
retaining the desired product. Ultrafiltration membranes are well suited
for the concentration of biologics since they can operate at wide tempera-
tures ranging from 2 to 26  C and involve no phase changes or chemical
additives thereby minimizing the extent of denaturation and/or degradation
of labile biological products. The features that distinguish UF from dialysis
are threefold: (1) application, (2) mode of operation, (3) membrane
construction. First, dialysis is primarily used for solute or buffer exchange,
whereas UF is employed for either buffer exchange or product
concentration. In addition, dialysis operates via diffusional mass transfer
driven by concentration differences across the semipermeable membrane,
whereas UF works primarily by conventional mass transfer driven by
pressure differences applied across the membrane. As a result of the
different driving forces used for solute or solvent transfer, membranes
employed for ultrafiltration are mechanically stronger than those for dialysis.
Pressure-driven UF membranes are designed to withstand high pressure
drops exceeding 75 psi.
In general, three modeling approaches have been used to describe the
actual mechanism of solute transport via ultrafiltration (Denisov, 1994).
A description of one transport mechanism by ultrafiltration is shown in
Fig. 9.2. The pressure-driven fluid flow promotes convective transport of
solutes and solvent toward the upstream surface of the membrane where the
applied pressure drop is termed the transmembrane pressure (DPTM). If
the membrane is completely retentive to a given solute (e.g., protein), as
in the case of concentration by UF, Cfiltrate ¼ 0 and the retained solute will
accumulate at the upstream surface of the membrane causing an increase
concentration within a boundary layer thickness, d at the membrane sur-
face. This phenomenon is referred to as ‘‘concentration polarization.’’
The accumulation of solutes at the membrane surface can affect solvent
filtrate flux, Jv, by several mechanisms. First, the accumulated solute can
generate an osmotic pressure drop across the membrane driving fluid from
the filtrate to the feed or retentate side, thereby reducing the net rate of
solvent transport. The flux can be described as shown below and in Fig. 9.2:

Jv ¼ L½DPTM  so ðPW  PF Þ

where L is the hydraulic permeability, so is the osmotic reflection

coefficient, and PW and PF are the solute osmotic pressures at the mem-
brane wall and filtrate, respectively (Denisov, 1994). This model assumes
the filtrate flux is a result of the difference between the applied pressure and
induced osmotic pressure and that any membrane fouling or resistance to
flow is negligible. The gel polarization model (Cheryan, 1986), however,
assumes that as DPTM the high concentration of retained solute at the
Concentration of Proteins and Removal of Solutes 109

Pressure driving force

Pretentate

Convective flow ΔPTM = Pretentate - Pfiltrate

Cwall

Pfiltrate

Cbulk
d
Cfiltrate

Back diffusive flow PW

PF

Osmotic pressure
driving force

Jv = L[ΔPTM - so (ΠW - ΠF)] Membrane

Figure 9.2 Schematic representation of convective and diffusive solute transport for
ultrafiltration processes.

membrane surface forms a gel layer providing an additional hydraulic

resistance to filtrate flow. The third modeling approach assumes solute at
the membrane surface can irreversibly foul the membrane, thereby reducing
its hydraulic permeability (Cheryan, 1986).
Overall, these models allow the user to better understand the impact of
UF processing parameters on solute and solvent flux so that conditions can
be optimized to improve membrane throughout. This is probably most
important for developing and designing UF processes to concentrate expen-
sive product such as protein therapeutics at large scale (>100 l) where
processing productivity and recovery are critical. These factors are not as
critical for ultrafiltration at the lab bench but are considered in designing UF
systems for very small-scale preparations.

4.1. Ultrafiltration membranes

In general, most UF membranes consist of a strong substructure (e.g., TyvekÒ )
upon which a very thin skin polymeric layer is attached. This thin layer
essentially provides the desired permselective membrane properties and
dictates the flow resistance. The technologies employed to generate the
membranes for ultrafiltration have not changed over the past few decades.
However, there have been significant improvements in the control of
110 David R. H. Evans et al.

membrane pore size distributions, membrane morphology, and membrane

modifications techniques. The three main casting technologies are air, immer-
sion, and melt casting. They differ primarily in the desolvation methods and
equipment (Zeman and Zydney, 1996). Current membrane compositions can
accommodate multiple polymers that are grafted, blended, or applied in the
form of copolymers to improve chemical and mechanical stability. The cellu-
lose-type polymers offer low protein binding that is required by many bio-
technology applications. However, cellulosic-type membranes generally
degrade upon exposure to alkaline hypochlorite cleaning solutions. New
cellulose composite membranes (e.g., UltracelÒ from Millipore Corp.) offer
low fouling and low protein binding for excellent product retention, recovery,
and higher yields. These Ultracel membranes are constructed of regenerated
cellulose membrane cast on a microporous substrate for defect-free membranes
with superior robustness compared to conventional membranes. The compos-
ite technology offers a mechanically robust design able to withstand extreme
operating conditions. In addition, polysulfone (PS) and polyethersulfone (PES)
polymeric membranes have sufficient alkaline resistance for applications where
NaOH cleaning is required, but are more susceptible to fouling by biologics-
containing solutions. Surface modification has been proposed to be the most
versatile method to reduce fouling on these membrane types. Membrane
modification is performed using negatively charged monomers (Rong et al.,
2006) or by hydrophilization of polymeric membranes with neutral polymers.
Several key membrane vendors such as Millipore, Pall Corporation, Gelman
Sciences, Sartorius, and Sterlitech provide a wide variety of modified
polymeric and ceramic membranes for use in both small- and large-scale
biologic applications.

4.2. Ultrafiltration devices

As mentioned previously, current ultrafiltration membrane and related
devices can be used for solute or buffer exchange and concentration of
the desired protein. Most small-scale UF membrane devices are designed to
operate only in a dead-end mode whereas larger scale UF systems employ
the crossflow mode of operation (see Fig. 9.3).
Dead-end devices employ a feed and permeate stream, each with the
same flow rate. There have been a number of modules developed specifi-
cally for dead-end applications. The foremost approaches include:
1. Vacuum filtration: Membrane disk is placed between a support plate and
an open glass or plastic funnel. The feed solution is placed in the funnel
and drawn through the membrane into a collection container by appli-
cation of a vacuum.
2. Syringe-end filters: UF membranes are bonded to plastic support plate and
sealed into a small-disk shaped unit equipped with luer connectors that
Concentration of Proteins and Removal of Solutes 111

Cross-flow
Dead-end operation (tangential flow)
operation
Feed or retentate

Feed Retentate

Ultrafiltration
membrane

Permeate Permeate
(filtrate) (filtrate)

Figure 9.3 Schematic representation of dead-end and crossflow ultrafiltration

operating modes (adapted from Zeman and Zydney, 1996).

can be attached to a syringe. The syringe is filled with sample and

manually driven through the syringe filter to concentrate sample.
3. Centrifugal filters: A membrane disk is placed inside a small centrifuge tube
and the permeate is collected in a space at the bottom of the tube beneath
the membrane. The feed sample is driven across the UF filter by centrifu-
gal force leaving concentrated sample above the membrane. Pall Corpo-
ration offers a wide range of centrifugal UF devices for processing samples
< 1 ml up to 60 ml (NanosepÒ , MicroSepTM, MacroSepÒ , and Jumbo-
SepTM) containing OmegaTM modified PES membranes with MWCO
ranging from 10 to 100 K. Millipore offers disposable CentriplusÒ centrif-
ugal filter devices used for the concentration (100-fold) and desalting of
2–15 ml of sample through Amicon’s low-adsorptive, hydrophilic,
membranes. These devices are designed for use in most centrifuges that
can accommodate 50 ml centrifuge tubes, with swinging-bucket or fixed-
angle rotors. Finally, Millipore’s Amicon Ultra-4 centrifuge filters are
designed to process and recover extremely valuable biologics by accurately
concentrating (80 to 100-fold) small sample volumes (1–4 ml).
4. Multiwell filter plate: UF membranes are incorporated into 96- or 384-
multiwell filter plates for high-throughput concentration or buffer
exchange of small samples (80–300 ml). Systems can be designed to
contain membranes of various molecular weight cutoff (10–100 kDa)
and provide greater than 90% recovery. The plates are constructed to
eliminate lateral flow and mixing between wells and contain outlet tips
and splash guards that ensure clean filtrate collection. Pall Corporation
112 David R. H. Evans et al.

offers the The AcroPrepTM filter plates (Pall Corporation, Meriden, CT)
in 96- or 384-multiwell Filter Plate format.
5. Stirred-cell device: A flat sheet UF membrane disk is placed on an appro-
priate support plate and sealed in place at the bottom of a cylindrical
housing. Fluid in the stirred device is agitated using a magnetic stirring
bar suspended from the top of the housing with the device designed to
operate with air to provide pressure driving force. Both Millipore and
Sterlitech provide a variety of device sizes with sample volumes ranging
from 3 to 400 ml. These devices are designed to simulate crossflow/
tangential filtration using different ultrafiltration membranes.
Even though the first four devices described above are probably the
easiest to operate and can process very small sample volumes, they are not
designed to provide precise control of concentration or concentration
polarization on the membrane. In contrast, the stirred-cell device supports
improved mass transport and concentration control due to its mixing
capabilities and capability to perform buffer exchange and product concen-
tration simultaneously. The process of buffer exchange using ultrafiltration
is termed ‘‘diafiltration.’’ In diafiltration, the exchange buffer is added to the
retentate or feed during the filtration, with the UF membrane-permeating
buffer species removed from the feed as the excess fluid is filtered through
the membrane. In contrast to dialysis, in which differences in solute con-
centration across the membrane drive buffer exchange, the UF buffer-
exchange process is faster due to convective flow of buffer across the
membrane. This can be beneficial for processing proteins or biologics that
are unstable and that require immediate buffer exchange. Diafiltration can
be performed in one of two modes. In continuous diafiltration, the wash
(exchange) buffer is added continuously throughout the filtration, with the
system often designed so that the total feed volume remains constant during
the process. In discontinuous diafiltration, the feed volume is first reduced
by a predetermined amount using a standard ultrafiltration. The wash buffer
is then added to the feed reservoir and the filtration continued. This process
can be repeated to obtain desired final solute concentration and volume
reduction.

4.3. Purification applications

The most recent advances in ultrafiltration membranes and systems tech-
nology have expanded the capability of UF from concentration and buffer-
exchange applications to more complex biological separations. These new
types of membrane-based ‘‘purification’’ applications have been driven by
their low operating costs and ease of use compared to traditional bind-and-
elute column chromatography. Recent work has shown that electrostatic
interactions can be used to enhance traditional sized-based, membrane-
Concentration of Proteins and Removal of Solutes 113

based processes and the approach has been termed high-performance tan-
gential-flow filtration (Sakena and Zydney, 1994; van Reis et al., 1997).
These membranes are designed to separate biological molecules of identical
size but differing in isoelectric points (pI) by exploiting electrostatic inter-
actions between the molecules and the membrane pores. In addition, there
have been significant advances in the area of membrane chromatography (or
affinity membrane chromatography) using membranes with larger pore
distributions (microfiltration membranes) than ultrafiltration membranes.
Membrane adsorbers have become accepted as an alternative to final polish-
ing chromatography steps for removal of trace impurities (Ghosh, 2002).
There are several commercial adsorbers available from Sartorius and Pall
Corporation that serve as anion and cation exchangers which have designed
to overcome the diffusion mass transfer limitations associated with bead-
based chromatography. Typically, the charged ligand is immobilized in the
membrane pores while convective flow delivers the solute molecules in
close proximity to the ligand thus minimizing diffusional limitations. None-
theless, the adoption of this technology has been slow because membrane
chromatography has been limited by a lower binding capacity than that
achieved by conventional chromatography resins even though the high-flux
advantages provided by membrane adsorbers offer the potential for high
productivity.

5. Lyophilization
Concentration of solutes in solution can be achieved thermodynami-
cally by driving solvent into a gaseous head space (evaporation) or
by boiling. Unfortunately, labile solutes such as protein molecules can be
quickly degraded by the heat required to eliminate solvents. Since the
boiling point of a solution is the temperature at which the vapor pressure
of the solution equals the atmospheric pressure (headspace air pressure),
solvents can boil by increasing solution temperature or by lowering the
atmospheric pressure (e.g., applying vacuum). This latter technique is
termed vacuum concentration. For the process of lyophilization the sample
temperature is lowered to the point where the solution freezes and solvents
are removed by sublimation. The freezing and vacuum step can be per-
formed simultaneously when both heat and headspace gases are removed to
concentrate the product of interest. The liquid protein product to be
lyophilized typically contains the active biologic ingredient, a solvent sys-
tem and several bulking and stabilizing agents. The bulking agents are used
to provide support structure for the active molecule while the stabilizing
agent will play a significant role in maintaining the activity of the target
protein in liquid form prior to the lyophilization process and after sample
114 David R. H. Evans et al.

reconstitution. Lyophilization generally consists of three major steps

(Costantino and Pikal, 2004; Jennings, 1999; Przic et al., 2004):
1. Freezing: The rate of freezing can affect final product quality. For
example, slow freezing can cause ice crystals to form which enhances
freeze-drying, but may denature labile proteins. Conversely, when
sample freezing is performed rapidly small ice crystals form which can
impede freeze-drying but protect protein stability.
2. Primary drying: During primary drying, greater than 80% of the solvent is
directly converted from solid to vapor phase through sublimation. The
residual solvent remains adsorbed on the product as moisture. Similar to
freezing rates, drying rates can affect the structure and morphology of the
lyophilized ‘‘cake’’ with rates designed to eliminate sample thaw.
3. Secondary drying: Residual solvent is desorbed during secondary drying to
attain a moisture level low enough to inhibit microbial growth or
chemical reactions, while still preserving the activity and integrity of
the freeze-dried product. The product’s physical properties determine
the duration of secondary drying. For example, proteins generally
require the presence of residual water to maintain structural integrity
and biological activity. The requisite residual moisture is product depen-
dent and must be determined empirically.
Figure 9.4 summarizes the key constituents and operations of a lyophi-
lizer. Although lyophilizer design varies dependent on the operational scale,
the key components and systems requirements do not change. Innovations
over the last decade have improved the efficiency, cost-effectiveness,
accuracy, and convenience of lyophilization equipment at all scales. These
improvements have been augmented by computer-controlled automation.
In particular, robotic sample loading/unloading systems have made the
process more accurate by minimizing user handling. The application of
enhanced real time in-process controls and monitoring such as SMART
Freeze DryerTM Temperature Monitoring technology (SP Industries,
Warminister, PA) allows improved temperature control during critical
freeze-drying cycles. Techniques such as nuclear magnetic resonance
(NMR) weighing technology eliminates traditional sample weighing
through noninvasive, noncontact check weigh technology.
In spite of these improvements, inherent limitations of the lyophilization
process remain. One of the most difficult challenges is achieving the target
moisture content in a specific product. While residual solute-bound water is
removed from the product during secondary drying, overdrying can result
in reduced product stability. Future freeze dry systems will incorporate
systems for precise monitoring of water removal throughout a freeze-dry
cycle using improved in-process mass flow meter technology.
Commensurate with the improved capabilities of current pilot and
large-scale lyophilizers, an expansion of lab-scale systems has occurred
 Lyophilization process and key components

Specialized containers: vials with Vacuum chamber: contains shelves and Heat transfer fluid circulates
stoppers, ampules or syringes that formulation in vials or glass containers to inside shelves to drive heat
contain formulation be lyophilized transfer away from the product

Condenser

Crygenic
refrigerator

Condenser removes sublimated and

desorbed solvent from vapor phase by
condensing or freezing it to maintain
adequate vacuum inside the vacuum
chamber

Cryogenic refrigerator: provides

refrigeration to shelves and condenser

Heater Heater: heats transfer fluid to sublimate

Vacuum pump: pulls vacuum
ice and desorb solvent from frozen
in the chamber and condenser
cake on shelves

Figure 9.4 Schematic of the lyophilization process and key system components.
116 David R. H. Evans et al.

permitting relatively small volumes to be processed. SP Industries, Lab-

conco and Martin Christ corporations have specialized in the development
of laboratory-scale systems. Novel benchtop systems provide a simple,
economical device for freeze-drying via dry ice. For example, particular
systems are equipped with a center well to accommodate dry ice and solvent
and which can serve both as a water vapor collector and a convenient
prefreezing bath. Flasks, serum bottles, and ampules may be frozen by
dipping and rotating them in the well. In addition, compact benchtop
freeze-dryers have been designed with rapid drying rate, an unrestricted
vapor path for maximum efficiency, more efficient condensers, and multiple
port manifold with valves. Overall, both the current and future lab-bench
and commercial lyophilizers are being designed to include microprocessor
control reducing operator error and minimizing handling to improve over-
all system performance.

6. Precipitation
Precipitation is a common purification technique utilized to concen-
trate and desalt proteins. Protein precipitation is the process of separating a
protein from a solution as a solid by altering the protein solubility with
addition of a reagent. A detailed discussion of protein precipitation is
presented in Chapter 20. Precipitation is generally inexpensive and scales
easily. In addition, most precipitations can often be performed in crude feed
streams where impurities such as DNA, lipids, and contaminant proteins are
present. There exist several methods or systems by which one can conduct
precipitation. The challenge is to determine which method is most suitable
to accomplish the desired goals. Precipitation may be induced by the
addition of neutral salts, organic solvents, nonionic hydrophilic polymers,
polyvalent metal ions, or by an acid or base to induce isoelectric point
precipitation (Harrison et al., 2003). The most important factors that influ-
ence protein solubility are structure, size, charge, and the solvent (Ladisch,
2007). Once a protein is precipitated, it can be separated from the solution
by subsequent centrifugation or filtration. The precipitate can then be
resolubilized in a smaller volume to concentrate the protein or washed
and resolubilized in a different solution to change the solution conditions.
A number of commercially available filtration plates and tubes are
available in 96-multiwell plate formats (Anachem, UK; Millipore, Bedford,
MA; Pall, Meriden, CT) which can be useful in screening precipitation
conditions with minimal resources. These plates can also enable precipita-
tion techniques to be automated, further reducing resources. Crystallization
screening kits, which contain ready-to-use solutions reflecting potential
precipitation/crystallization conditions, can also be useful for exploring a
Concentration of Proteins and Removal of Solutes 117

broad range of buffers, pH, and precipitants (Hampton Research, Aliso

Viejo, CA; Sigma-Aldrich, St Louis, MO).
Many precipitation methods, including precipitation with organic sol-
vents and isoelectric precipitation can result in the denaturation of proteins
or reduction of a protein’s bioactivity. Although detrimental to the function
of the protein, these precipitation methods are commonly used to eliminate
components contained in biological fluids that can interfere with down-
stream applications and analyses. Several analytical techniques are routinely
performed on denatured proteins and do not require a protein to be
biologically active and others, such as SDS–PAGE, denature proteins as
part of an analytical method. For these applications, precipitations that
denature proteins are still appropriate. Some common precipitants utilized
to concentrate proteins and remove interfering species for analytical analysis
include ethanol, acetone, chloroform/methanol, and trichloroacetic acid.
Protein precipitation using acetone can be accomplished by adding a 4:1
ratio of cold acetone ( 20  C) to a protein sample. The solution is then
mixed, centrifuged, and the supernatant is discarded. This process may be
repeated in order to achieve better removal of salts and other impurities.
The precipitate is then dried and resolubilized in the desired solution.
A similar procedure can be employed using a 9:1 volumetric ratio of cold
ethanol ( 20  C) to protein sample. An advantage of using ethanol is it is
less flammable than acetone. Wessel and Flugge (1984) proposed a
chloroform–methanol precipitation system with improved recoveries for
dilute protein solutions.
TCA precipitation is considered a very efficient protocol for precipitat-
ing proteins from dilute solutions. An equal volume of 20% TCA is added to
a protein sample and incubated on ice for 30 min. The solution is centri-
fuged at  4  C and decanted. The precipitate pellet is then washed with
cold acetone ( 20  C), centrifuged again at 4  C, decanted, and dried. The
precipitated pellet can then be dissolved in the desired buffer. Additional
acetone washes of the precipitate or neutralization of the resolubilized
sample may be necessary to reduce the acidity prior to further analysis.
Sivaraman, et al. (1997) demonstrated that protein precipitation is optimized
at a TCA concentration of roughly 15%.
In many cases, it is desirable to maintain the activity and structure of the
protein after precipitation. This can be accomplished by neutral salt addition
(salting out) or by the addition of nonionic hydrophilic polymer such as
polyethylene glycol. Hofmeister (1887, 1890) first described the use of
neutral salts to precipitate protein. The most commonly used salt is ammo-
nium sulfate, as it is very water soluble and has no adverse effects upon the
bioactivity of proteins. Ammonium sulfate precipitation is generally accom-
plished by adding a saturated ammonium sulfate solution (41.22 g/100 g
water at 25  C) slowly to a protein sample to a desired concentration.
Protein solubilities differ greatly in ammonium sulfate so it may be
118 David R. H. Evans et al.

beneficial to perform initial experiments to determine the ammonium

sulfate concentration at which the protein of interest is precipitated from
solution. This property may enable the ability to purify a protein from other
proteins and contaminants through fractional precipitation. Hydrophilic
polymers such as polyethylene glycol (PEG) can also be utilized to precipi-
tate proteins while preserving bioactivity. Proteins are generally easier to
precipitate using PEGs with molecular weights greater than 4000. Atha and
Ingham (1981) demonstrate the precipitation of several proteins in the
presence of PEG 4000 at concentrations of 3–30% (w/v).

7. Crystallization
Crystallization is another purification technique that can be used to
concentrate and desalt proteins similarly to precipitation. Protein crystalli-
zation is a powerful separation technology because it simultaneously con-
centrates, purifies, and stabilizes the product (Etzel, 2006). Crystallization is
a controlled precipitation from an aqueous solution with four main variables
that control crystal morphology and recovery: protein concentration, pre-
cipitant concentration, pH, and temperature. It is similar to precipitation in
that solid particles are formed from solution; however, precipitates have
poorly defined morphology and are characterized by small particle size,
whereas crystals are highly ordered with generally larger particle sizes. High-
throughput screening systems are being developed in multiwell plate and
smaller formats that, combined with robotic liquid handling and novel
analysis methods, are enabling the rapid screening of broad ranges of con-
ditions with nanoliter amounts (Brown et al., 2003). In addition, applica-
tions for microfluidic chips have been developed for biomolecule
crystallization (Teragenics, Watertown, MA). In practice, it is usually
more difficult to crystallize than precipitate a protein and for purposes of
concentrating and desalting a protein, precipitation is generally a preferable
method. In addition, as crystallization processes are scaled up from the lab
bench, mixing dynamics can significantly impact robustness.

REFERENCES
Atha, D. H., and Ingham, K. C. (1981). Mechanism of precipitation of proteins by
polyethylene glycols. J. Biol. Chem. 256, 12108–12117.
Booker, C. J., and Yeung, K. K. C. (2008). In-capillary protein enrichment and removal of
nonbuffering salts using capillary electrophoresis with discontinuous buffers. Anal. Chem.
80, 8598–8604.
Brown, J., Walter, T. S., Carter, L., Abrescia, N. G. A, Aricescu, A. R., Batuwangala, T. D.,
Bird, L. E., Brown, N., Chamberlain, P. P., and Davis, S. (2003). A procedure for setting
Concentration of Proteins and Removal of Solutes 119

up high-throughput nanolitre crystallization experiments. II. Crystallization results.

J. Appl. Crystallogr. 36, 315–318.
Cheryan, M. (1986). Ultrafiltration Handbook. Technomic, Lancaster, PA.
Costantino, H. R., and Pikal, M. J. (2004). Lyophilization of Biopharmaceuticals. AAPS
Press.
Cussler, E. L. (1997). Diffusion Mass Transfer in Fluid Systems: Mass Transfer in Fluid
Systems. 2nd edn. Cambridge University Press.
Denisov, G. A. (1994). Theory of concentration polarization in cross-flow ultrafiltration:
Gel-layer model and osmotic–pressure model. J. Membr. Sci. 91, 173–187.
Etzel, M. (2006). Bulk protein crystallization—Principles and methods. In ‘‘Process Scale
Bioseparations for the Biopharmaceutical Industry’’ (A. Shukla, M. Etzel, and S. Gadam,
eds.). Taylor & Francis, Boca Raton, FL.
Ghosh, R. (2002). Protein separation using membrane chromatography: Opportunities and
challenges. J. Chromatogr. A 952, 13–27.
Harrison, R. G., et al. (2003). Bioseparations Science and Engineering. Oxford University
Press, New York, NY.
Hsieh, H.-C., Sheu, C., Shi, F. K., and Li, D.-T. (2007). Development of a titanium dioxide
nanoparticle pipette-tip for the selective enrichment of phosphorylated peptides.
J. Chromatogr. A 1165, 128–135.
Hofmeister, F. (1887). Zur Lehre von der Wirkung der Sake. Arch. Exp. Path. Pharm. 24,
247.
Hofmeister, F. (1890). Uber die Darstellung von krystallisirtem Eiralbumin und die Krys-
tallisirbarkeit colloider Stoffe. Z. Physiol. Chem. 14, 165.
Jennings, T. (1999). Lyophilization: Introduction and Basic Principles. Interpharm Press.
Jia, W., Wu, H., Lu, H., Li, N., Zhang, Y., Cai, R., and Yang, P. (2007). Rapid and
automatic on-plate desalting protocol for MALDI-MS: Using imprinted hydrophobic
polymer template. Proteomics 7, 2497–2506.
Ladisch, M. (2007). Bioseparations Engineering: Principles, Practice, and Economics. John
Wiley & Sons Inc.
Przic, D. S., Ruzic, D. S., Nenad, N. L., and Petrovic, S. D. (2004). Lyophilization—The
process and industrial use. Chem. Ind. 58, 552.
Rong, W., Zhansheng, L., and Fane, A. G. (2006). Development of a novel electrophoresis-
UV grafting technique to modify PES UF membranes used for NOM removal. J. Membr.
Sci. 273, 47–57.
Sakena, S., and Zydney, A. L. (1994). Effect of solution pH and ionic strength on the
separation of albumin from immunoglobulins (IgG) by selective filtration. Biotechnol.
Bioeng. 43, 960–968.
Schley, C., Swart, R., and Huber, C. G. (2006). Capillary scale monolithic trap column for
desalting and preconcentration of peptides and proteins in one- and two-dimensional
separations. J. Chromatogr. A 1136, 210–220.
Shen, W., Xiong, H., Xu, Y., Cai, S., Lu, H., and Yang, P. (2008). ZnO-poly(methyl
methacrylate) nanobeads for enriching and desalting low-abundant proteins followed by
directly MALDI-TOF MS analysis. Anal. Chem. 80, 6758–6763.
Sheen, H., and Ali-Khan, Z. (2005). Protein sample concentration by repeated loading onto
SDS-PAGE. Anal. Biochem. 343, 338–340.
Sivaraman, T., et al. (1997). The mechanism of 2, 2, 2-trichloroacetic acid-induced protein
precipitation. J. Protein Chem. 16, 291–297.
Shihabi, Z. K. (2002). Transient pseudo-isotachophoresis for sample concentration in
capillary electrophoresis. Electrophoresis 23, 1612–1617.
van Reis, R., Gadam, S., Frautschy, L. N., Orlando, S., Goodrich, E. M., Saksena, S.,
Kuriyel, R., Simpson, C. M., Pearl, S., and Zydney, A. L. (1997). High performance
tangential flow filtration. Biotechnol. Bioeng 56, 71–82.
120 David R. H. Evans et al.

Waters, N. J., et al. (2008). Validation of a rapid equilibrium dialysis approach for the
measurement of plasma protein binding. J. Pharm. Sci. 97(10), 4586–4595.
Waziri, S. M., Abu-Sharkh, B. F., and Ali, S. A. (2004). Protein partitioning in aqueous
two-phase systems composed of a pH-responsive copolymer and poly(ethylene glycol).
Biotechnol. Prog. 20, 526–532.
Wessel, D., and Flugge, U. I. (1984). A method for the quantitative recovery of protein in
dilute solution in the presence of detergents and lipids. Anal. Biochem. 138, 141–143.
Wheat, T. E., Rainville, P. R., Gillece-Castro, B. L., Lu, Z., Cravello, L., and Mazzeo, J. R.
(2007). Fast on-line desalting of proteins for determination of structural information
using exact mass spectrometry. Bioprocess. J. 6, 55–59.
Yu, H., Lu, Y., Zhou, Y.-G., Wang, F.-B., He, F.-Y., and Xia, X.-H. (2008). A disposable
microfluidic device for rapid protein concentration and purification via direct-printing.
Lab Chip 8, 1496–1501.
Zeman, L., and Zydney, A. (1996). Microfiltration and Ultrafiltration. Marcel Dekker Inc.
 C H A P T E R T E N

Maintaining Protein Stability

Murray P. Deutscher

Contents
1. Causes of Protein Inactivation 121
2. General Handling Procedures 122
3. Concentration and Solvent Conditions 122
4. Stability Trials and Storage Conditions 123
5. Proteolysis and Protease Inhibitors 124
6. Loss of Activity 125

Abstract
Proteins are fragile molecules that often require great care during purification
to ensure that they remain intact and fully active. Nowadays, many proteins are
also purified in small amounts under denaturing conditions by various gel
electrophoretic techniques, such that inactive proteins are obtained. But even
here, it is usually advantageous to maintain the protein in an intact form. In the
case of enzymes, and other proteins with assayable biological activities, main-
tenance of activity is generally of prime importance, both for following the
protein during purification and for subsequent studies of function. This chapter
will focus on the major points to keep in mind with regard to maintaining the
stability of a protein during purification and storage. Various other chapters
describe in detail stabilization procedures for specific biological systems and
specific classes of proteins.

1. Causes of Protein Inactivation

Removal of proteins from the cellular environment subjects them to a
variety of conditions and processes that can lead to loss of activity or
alteration of structure. These include dilution, change in solution condi-
tions, exposure to degradative enzymes, oxygen, heavy metals, and surfaces,
and change in physical condition (e.g., freezing and thawing). Awareness
that any one of these situations could affect the protein, and knowledge of

Department of Biochemistry and Molecular Biology, University of Miami School of Medicine, Miami,
Florida, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63010-X All rights reserved.

121
122 Murray P. Deutscher

precautions that could be taken to minimize their effect, will go a long way
toward ensuring a successful purification protocol. If the protein of interest is
lost or inactivated during the course of any procedure, determination of the
reason for this loss will often suggest a simple solution. Thus, if possible,
examine whether the loss of activity is accompanied by loss of the protein or
changes in its structure, or whether the protein remains, but is now inactive.
Distinguishing among these different possibilities might indicate what type
of process is behind the problem and, thus, what an appropriate solution
might be.

2. General Handling Procedures

Obviously, to maintain the stability of a protein, avoid treatments that
will denature it. Thus, protein solutions should generally not be stirred
vigorously or vortexed since this may lead to oxidation or surface denatur-
ation. Protein solutions should not be exposed to extremes of pH, high
temperatures, organic solvents, or any other condition that might promote
denaturation. Likewise, if one is storing a protein solution for extended
periods of time in an unfrozen state, bacterial and fungal growth can become
a problem. In these situations sterile solutions and antibacterial or antifungal
agents may be necessary. Finally, it is best to make up all solutions that will
come in contact with the protein with glass–distilled water, and to store the
water in containers that do not have algal growth.

3. Concentration and Solvent Conditions

Extraction of proteins from cells inevitably leads to a change in their
environment. Since proteins are generally stable in vivo, the theoretical goal
is to try to reproduce the cellular milieu as closely as possible. This would
mean very high protein concentrations, close to neutral pH, moderate ionic
strengths, reducing conditions, etc. In practice, some of these conditions are
compatible with protein purification, and some are not. As a first approxi-
mation, it is generally a good practice to keep the protein concentration
high (>1 mg/ml). This would help to maintain protein complexes, possibly
minimize the effects of deleterious contaminants and surface adsorption, and
provide a general stabilizing environment for the protein of interest. It is
relatively easy to maintain high protein concentrations early in a purification
scheme, but this becomes more difficult as the protein is purified unless one
resorts to concentration procedures after each step. Since these latter pro-
cedures often have their own problems, one may have to settle for more
dilute solutions unless particular stability problems become obvious. It may
be helpful to alternate purification steps between ones that concentrate
Maintaining Protein Stability 123

proteins with ones that dilute them. For example, elution of proteins from a
column to which they are bound using a batchwise procedure will tend to
concentrate the eluted proteins, whereas gradient elution will tend to dilute
them. Columns to which proteins bind will tend to concentrate, whereas
gel filtration will dilute. By judicious arrangement of purification steps, one
may be able to avoid extensive dilution.
The solution conditions are also extremely important. Although it is not
possible to describe a universal stabilizing solvent applicable to every pro-
tein, the addition of certain components is generally helpful. These include
a buffer, usually around neutrality, to avoid unnecessary pH changes.
Careful attention should be given to the buffer anion since in many cases
Cl may be harmful. EDTA is usually added at about 0.1 mM to chelate
heavy metal ions that could interact with the protein or promote oxidation.
A reducing agent such as 2-mercaptoethanol or dithiothreitol is often
present to counteract oxidative effects, particularly of cysteine residues.
The use of dithiothreitol at about 0.1–1 mM is preferred because it does
not form mixed disulfides with proteins, as 2-mercaptoethanol does. Suffi-
cient reducing agent should be present since it can oxidize relatively rapidly
and lead to protein oxidation. In some cases, salts are also added to maintain
a certain ionic strength, but only if they are compatible with the next
purification or analytical step. Likewise, glycerol at 5–20% often helps to
maintain stability and is compatible with most purification steps at these
concentrations. Occasionally, low levels of a nonionic detergent are added
to prevent aggregation or the adsorption of proteins to surfaces, such
as glassware. Finally, it is good practice to include protease inhibitors,
particularly at early steps.

4. Stability Trials and Storage Conditions

One of the most important studies that can be performed during the
course of a new protein purification is a stability and storage study. What
this means is that after every step of the purification procedure the stability
and storage properties of the protein of interest should be determined.
Although rapid purification of a protein is desirable, the situation will
often arise, especially during a new purification, when it becomes necessary
to keep a protein for some length of time prior to the next step. For this
purpose you will need to know how stable it is under different storage
conditions. The simplest way to test this is to take small portions of the
protein solution, store them under a variety of conditions (e.g., in ice,
frozen, at room temperature, with and without different stabilizing agents),
and then assay the activity of the protein after different periods of time.
Again, keep in mind what the next step in the purification procedure will be.
Some storage conditions may be fine for stability, but not useful for further
124 Murray P. Deutscher

purification. A case in point is storage at 20  C in 50% glycerol (v/v).

This is often a useful condition for maintaining stability, but terrible if one
plans further purification. Sometimes, it may be necessary to use such a
procedure (the glycerol could be removed by dialysis), but generally it should
be a last resort.
A different situation arises when one has completed a purification
procedure and wants to store the purified protein for long periods of
time. Here, the primary concern is long-term stability, and many conditions
that might be impractical during the course of purification could be used.
These might include addition of high concentrations of glycerol, addition of
stabilizing substrates, even addition of an extraneous protein such as serum
albumin. The choice of storage condition depends on what is effective for
stabilization, and for what the purified protein will be used. If one is
primarily interested in studying enzyme activity, the presence of serum
albumin may not matter. In contrast, one would not want an extraneous
protein present if structural studies are planned. If one is unsure, the best
course may be to store portions of the protein under a variety of conditions.
Related to the question of storage is the problem of freezing and thawing
solutions of purified proteins. One way to avoid repeated cycles of freezing
and thawing is to store the purified protein in small portions and to thaw
individual samples once, as needed; alternatively, the protein may be stored
under conditions in which it does not freeze, such as 50% glycerol at 20  C.
If repeated freezing and thawing is necessary, it is best to use quick freezing
and thawing procedures. During freezing, solutes are concentrated and the
protein could be exposed to unusually harsh conditions. We routinely quick
freeze protein solutions in dry ice–ethanol baths to avoid this problem.
Likewise, in thawing protein solutions, this should be done rapidly with
gentle mixing in lukewarm water until only a small amount of ice is present;
the solution is then placed in ice or kept at room temperature during use. The
final thawed solution should be mixed gently, or inverted if in a tube, to
ensure even distribution of solutes.

5. Proteolysis and Protease Inhibitors

Proteolysis is a major problem for the purification of proteins. It is a
particularly insidious problem because in many cases the protein of interest is
only partially degraded and may retain biological activity. This can result in
erroneous conclusions about the size and structure of the protein. Proteoly-
sis can be a problem at any stage of a purification procedure. Total proteo-
lytic activity is generally highest in the initial crude extract since purification
will tend to eliminate these contaminating activities; however, there are also
more proteins present in the extract that could act to protect the protein of
Maintaining Protein Stability 125

Table 10.1 Common protease inhibitors

Protease class Concentration

Protease inhibitor inhibited used
PMSF (phenylmethylsulfonyl Serine proteases 0.1–1 mM
fluoride)
EDTA and EGTA Metalloproteases 0.1–1 mM
Benzamidine Serine proteases 1 mM
Pepstatin A Acid proteases 1 mg/ml
Leupeptin Thiol proteases 1 mg/ml
Aprotinin Serine proteases 5 mg/ml
Antipain Thiol proteases l mg/ml

interest. As purification proceeds, even a small contamination with a prote-

ase could have a large effect because a larger fraction of the available
protein substrate will be the one with which you are working.
How can you tell if proteolysis is a problem in your particular situation?
The simplest test is to incubate the extract or partially purified protein at a
moderate temperature (e.g., 30  C), withdrawing portions at intervals, and
assaying for biological activity. Although this method is not foolproof because
there may be other reasons for loss of activity, most proteins will not be heat
inactivated under these conditions. If activity is lost, the addition of protease
inhibitors is recommended since even if proteins are kept at 0–4  C through-
out purification, some cleavages will occur unless the proteases are inactivated.
Cells contain a variety of different classes of proteases. Fortunately, protease
inhibitors are available that can act on the various proteases. A list of some
commonly used inhibitors is presented in Table 10.1. Protease inhibitors useful
for particular systems or situations are described in other chapters of this volume.
Probably the best approach in working with a new protein is to use a mixture of
inhibitors that affect different classes of proteases. Such mixtures are commer-
cially available. Once conditions for maintaining the protein of interest in a
stable form are obtained, inhibitors can be removed one by one to determine
which are really necessary. Some trial and error will be involved, as well as for
deciding which inhibitors, if any, are needed as the purification proceeds. Note
that protease inhibitors can be toxic and/or unstable under certain conditions.
They should not be used without first learning their properties.

6. Loss of Activity
The most commonly heard lament during a protein purification is,
‘‘I’ve lost my activity.’’ When this happens, a careful analysis of the situation
is required to determine the cause. Most importantly, one should have a
126 Murray P. Deutscher

careful accounting of enzyme units to evaluate the extent of the activity loss.
For many purification steps, percentage losses of as much as 50% are not
unusual, but of course, these vary with each individual protein. Generally,
purification methods that involve binding of a protein to a matrix, and
which may require conformational changes during binding, have a greater
effect on activity than a procedure such as gel filtration.
If activity is totally lost during a particular purification step, other
possibilities need to be considered. In some cases proteins may bind very
tightly to columns, and require more extreme procedures for elution.
Depending on the type of chromatography (see Sections VI and VII of
this volume), this may require increased ionic strength, use of a chaotropic
salt (e.g., KBr), or inclusion of detergent or ethylene glycol in the elution
buffer.
A second possibility is that more than one component is required for the
activity of the protein, and these components (e.g., another subunit or a
cofactor) are separated during the fractionation step. Thus, either compo-
nent by itself would be inactive or all have to be present to observe activity.
To test for this possibility, all the fractions from the previous step are mixed
back together, and activity measured. In some cases, it may be necessary to
concentrate the mixture back to the original volume in order to observe
activity. If mixing of all the fractions results in the appearance of activity,
one could then test fractions or groups of fractions in a pairwise fashion.
Often it is found that a little activity remains and that the second component
is needed for optimal activity. In this case, one of the required fractions is
already known, and the other fractions can then be tested for their stimulat-
ing activity.
Sometimes activity may be lost between purification steps, such as
during dialysis or concentration, or even during storage. In the former
situations, one should again test for removal of a possible required compo-
nent. The possibility also exists that the protein has stuck to the dialysis
tubing or the concentration membrane. Here, washing the tubing or
membrane with buffer containing some detergent may be helpful. Problems
of stability during storage have been discussed above.
The most frustrating situation is if none of the above possibilities is the
cause of the loss of activity. Under these circumstances the most likely
explanation is actual inactivation of the protein due to denaturation, prote-
olysis, etc. If an independent measure for the protein is available (e.g., a
Western blot), this can be shown directly. If not, an answer to the inactiva-
tion problem may require trial and error experiments to test various condi-
tions. Sometimes, the best solution is simply to avoid that particular
purification step.
Micropurification (purification of very small amounts of protein in the
submilligram range) presents a special set of problems. In particular, one has
to be especially mindful of the possibility of protein loss due to adsorption
Maintaining Protein Stability 127

on surfaces. Small amounts of protein bind tightly to surfaces such as glass

and many plastics, often on the order of 1 mg/cm2. During micropurifica-
tion procedures in which the amount of protein present is small to begin
with, these adsorptive losses can represent a significant percentage of the
protein present. Certainly, one should avoid containers made of polystyrene
since this material (e.g., a microtiter dish) has a high protein binding
capacity. In these situations, tubes should be carefully rinsed to ensure
maximum recovery of protein. In addition, the presence of low concentra-
tions of a nonionic detergent may help to prevent binding of protein to
the surface. Careful accounting of protein throughout the procedure is
important and may alert the investigator to the occurrence of unwanted
protein loss.
 C H A P T E R E L E V E N

Selecting an Appropriate Method for

Expressing a Recombinant Protein
William H. Brondyk

Contents
1. Introduction 132
2. Escherichia coli 133
2.1. E. coli: Temperature and molecular chaperones 133
2.2. E. coli: Fusion partners 134
2.3. E. coli: Disulfide bond formation 134
2.4. E. coli: Posttranslational modifications 134
3. Pichia pastoris 135
4. Baculovirus/Insect Cells 136
5. Mammalian Cells 138
6. Protein Characteristics 139
6.1. Protein characteristics: E. coli and codon usage 141
6.2. Protein characteristics: Cytoplasmic proteins 141
6.3. Protein characteristics: Secreted proteins 142
6.4. Protein characteristics: Membrane proteins 142
6.5. Protein characteristics: Toxic proteins 142
7. Recombinant Protein Applications 143
8. Conclusion 144
References 144

Abstract
Recombinant proteins are important tools for studying biological processes.
Generating a recombinant protein requires the use of an expression system.
Selection of an appropriate expression system is dependent on the character-
istics and intended application of the recombinant protein and is essential to
produce sufficient quantities of the protein. Over the last 30 years, there have
been considerable advances in the technologies for expressing recombinant
proteins. In this chapter the unique characteristics of four commonly used
expression systems, Escherichia coli, Pichia pastoris, baculovirus/insect cell,
and mammalian cells are described. The E. coli system is a rapid method for

Genzyme Corporation, Framingham, Massachusetts, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63011-1 All rights reserved.

131
132 William H. Brondyk

expressing proteins but lacks many of the posttranslational modifications found

in eukaryotes. The capacity of E. coli for protein folding and forming disulfide
bonds is not sufficient for many recombinant proteins although there are a
number of tools developed to overcome these limitations. In contrast to E. coli,
the eukaryotic P. pastoris, baculovirus/insect cell, and mammalian systems
promote good protein folding and many posttranslational modifications. How
the characteristics and the downstream application of a recombinant protein
can influence the choice of an expression system is then reviewed.

1. Introduction
Choosing an appropriate method for expressing a recombinant protein
is a critical factor in obtaining the desired yields and quality of a recombinant
protein in a timely fashion. Selecting a wrong expression host can result in
the protein being misfolded or poorly expressed, lacking the necessary
posttranslational modifications or containing inappropriate modifications.
Factors to consider when selecting an expression system include the mass of
the protein and number of disulfide bonds, type of posttranslational mod-
ifications desired on the expressed protein, and the destination of the
expressed protein. The intended application of the purified recombinant
protein is also critical in the decision-making process and the applications can
be categorized into four broad areas: structural studies, in vitro activity assays,
antigens for antibody generation, and in vivo studies. The purpose of this
chapter is to help guide the investigator in the decision-making process for
choosing an appropriate expression system. However, even with the
described guidelines there are many circumstances when it is not obvious a
priori which expression system is the best choice, and the use of multiple
expressions systems must be attempted before an optimal system is identified.
Numerous expression systems are currently being used in academic and
industrial settings. Some of these systems are too new and insufficiently
tested to comment on their utility. In addition, some established systems for
expressing recombinant proteins, such as transgenic animals, are too tech-
nically challenging, time consuming and prohibitively expensive to be a
viable option for the average laboratory. For the purpose of this chapter,
only Escherichia coli, Pichia pastoris, baculovirus/insect cell, and mammalian
expression systems will be considered (for more detailed coverage of these
systems, see Chapters 12–15). These four systems have straightforward
protocols, are readily accessible either from colleagues or from research
product companies (e.g., Invitrogen, EMD-Novagen, Stratagene, and
Promega), and are relatively inexpensive for small-scale production. The
characteristics and available options of these expression systems will be
briefly reviewed with the focus on the differences between the systems.
Strategies will then be presented to help guide the investigator in making
the best choice for an expression system.
Selecting an Expression System 133

2. Escherichia coli
The bacteria E. coli was the first host used to express recombinant
proteins and is still considered to be the workhorse in the field. Using the
E. coli system offers a rapid and simple method for expressing recombinant
proteins due to its short doubling time. Consequently, the assessment of
recombinant gene expression in E. coli can take less than a week. The growth
media for E. coli are inexpensive and there are relatively straightforward
methods to scale-up bioproduction (see Chapter 12).
In E. coli, recombinant proteins are normally either directed to the
cytoplasm or to the periplasm and, to a lesser extent, secreted. Proteins
directed to the cytoplasm are the most efficiently expressed, giving yields of
up to 30% of the biomass ( Jana and Deb, 2005). However, the high
expression of recombinant protein can often lead to the accumulation of
aggregated, insoluble protein that forms inclusions bodies. Inclusion bodies
have been observed not only with eukaryotic proteins but also to a lesser
extent with overexpressed proteins from prokaryotes including E. coli.
The rate of translation and folding in E. coli is almost 10-fold higher than
that observed in eukaryotic cells, and this presumably contributes to the
inclusion body formation of eukaryotic proteins (Andersson et al., 1982;
Goustin and Wilt, 1982). Inclusion bodies can be a significant hindrance in
obtaining soluble, active protein in some situations. However, in some
cases, inclusion bodies are advantageous because they are resistant to prote-
olysis, easy to concentrate by centrifugation, minimally contaminated with
other proteins, and, with some effort, able to be refolded to form active,
soluble proteins (see Chapter 17 for details).

2.1. E. coli: Temperature and molecular chaperones

Several methods have been described for maximizing the formation of
soluble, properly folded proteins in the cytoplasm and minimizing inclusion
body formation. The most straightforward method involves lowering the
temperature to 15–30  C during the expression period (Sahdev et al., 2008).
Presumably, the reduced temperature slows the rate of transcription, trans-
lation, and refolding, thereby allowing for proper folding (Vera et al., 2006).
In addition, lower temperature has been shown to decrease heat shock
protease activity (Spiess et al., 1999). Some investigators have coexpressed
molecular chaperones in the cytoplasm along with the recombinant protein
for promoting protein solubility (Young et al., 2004). The utility of this
approach appears to be quite protein-specific, and therefore needs to be
tested individually for each recombinant protein of interest (Baneyx and
Mujacic, 2004).
134 William H. Brondyk

2.2. E. coli: Fusion partners

Alternatively, a method that promotes solubility with many proteins is to
fuse the recombinant protein at either the N-terminus or C-terminus to a
soluble fusion tag (Esposito and Chatterjee, 2006). Fusion partners that have
been shown to increase solubility of recombinant proteins include glutathi-
one-S-transferase (GST), thioredoxin, maltose-binding protein (MBP),
small ubiquitin-modifier (SUMO), and N-utilization substrate (NusA).
Both GST and MBP have the added advantage of also being an affinity
purification tag. Unfortunately, no single tag appears to work for all recom-
binant proteins, and multiple fusion partners may need to be evaluated for
promoting soluble expression. Fusion tags can be removed from the recom-
binant protein by several strategies, and a widespread approach involves
adding a protease site between the fusion partner and the recombinant
protein that can be cleaved with the specific protease. This approach must
be carefully tested since removal of the fusion tag can, in some cases, render
the recombinant protein insoluble (Waugh, 2005).

2.3. E. coli: Disulfide bond formation

E. coli is normally inefficient in promoting the correct formation of disulfide
bonds when recombinant proteins are expressed in the cytoplasm; normally
disulfide bond formation occurs only in the periplasm where it is catalyzed
by the Dsb system (Andersen et al., 1997; Bardwell, 1994). Consequently, if
disulfide bond formation is needed, the recombinant protein can be directed
to the periplasm via a cleavable signal peptide (e.g., pelB). However, a major
disadvantage of periplasmic expression is the significant reduction in pro-
duction yields. Through engineering of the E. coli genome, a more suitable
environment for disulfide bond generation in the cytoplasm can be induced
by disrupting the thioredoxin reductase (trxB) and glutathione reductase
( gor) genes in the Dsb system which in turn enables thioredoxin and
glutaredoxin to promote cytoplasmic reduction of cysteines (Bessette
et al., 1999). These engineered strains are commercially available through
EMD-Novagen (Origami). If additional disulfide bond formation is still
needed, the recombinant protein can be fused to thioredoxin, and the
fusion protein expressed in a trxB /gor E. coli strain (LaVallie et al., 1993).

2.4. E. coli: Posttranslational modifications

Finally, it is important to recognize that E. coli has a limited capacity for
posttranslational modifications compared to eukaryotic organisms. For
example, E. coli does not support enzyme-mediated N-linked glycosylation,
O-linked glycosylation, amidation, hydroxylation, myristoylation, palmita-
tion, or sulfation.
Selecting an Expression System 135

3. Pichia pastoris
Yeast is another traditional, powerful tool for expressing recombinant
proteins and has been used successfully to express a multitude of proteins.
Yeast has many of the advantageous features of E. coli such as a short
doubling time and a readily manipulated genome, but also has the additional
benefits of a eukaryote that includes improved folding and most posttrans-
lational modifications. The first yeast routinely used for recombinant pro-
tein expression was Saccharomyces cerevisiae (Strausberg and Strausberg,
2001). However, in the last 15 years, P. pastoris has become the yeast of
choice because it typically permits higher levels of recombinant protein
expression than does S. cerevisiae (see Chapter 13). P. pastoris is a methyltropic
yeast, and can use methanol as its only carbon source (Cregg et al., 1985). The
growth of P. pastoris in methanol-containing medium results in the dramatic
transcriptional induction of the genes for alcohol oxidase (AOX) and dihy-
droxyacetone synthase (Cregg, 2007). After induction, these proteins com-
prise up to 30% of the P. pastoris biomass. Investigators have exploited this
methanol-dependent gene induction by incorporating the strong, yet tightly
regulated, promoter of the alcohol oxidase I (AOX1) gene into the majority
of vectors for expressing recombinant proteins (Daly and Hearn, 2005). The
P. pastoris expression vectors integrate in the genome whereas by contrast, S.
cerevisiae vectors use the more unstable method of replicating episomally. The
length of time to assess recombinant gene expression with the P. pastoris
method is approximately 3–4 weeks which includes the transformation of
yeast, screening the transformants for integration, and an expression time-
course. An appealing feature of P. pastoris is the extremely high cell densities
achievable under appropriate culture conditions (Cregg, 2007). Using inex-
pensive medium, the P. pastoris culture can reach 120 g/l of dry cell weight
density. An important caveat is that the induction medium requires a low
percentage of methanol. In large-scale cultures, the amount of methanol
becomes a fire hazard requiring a new level of safety conditions.
P. pastoris has been used to obtain both intracellular and secreted recom-
binant proteins. Like other eukaryotes, it efficiently generates disulfide
bonds and has successfully been used to express proteins containing many
disulfide bonds. To facilitate secretion, the recombinant protein must be
engineered to carry a signal sequence. The most commonly used signal
sequence is the pre-pro sequence from S. cerevisiae a-mating factor (Daly
and Hearn, 2005). Because P. pastoris secretes few endogenous proteins,
purification of the recombinant protein from the medium is a relatively
simple task. If proteolysis of the recombinant protein is a concern, expres-
sion can be completed using the pep4 protease-deficient strain of P. pastoris
136 William H. Brondyk

(Gleeson et al., 1998). This strain has reduced vacuole peptidase A activity
which is responsible for activation of carboxypeptidase Y and protease B1.
Yeast has the posttranslational capacity to add glycans at both specific
asparagine residues (N-linked) and serine/threonine residues (O-linked).
These glycan structures are substantially different from the modifications
added by insect and mammalian cells. In P. pastoris the N-linked glycan is a
high mannose type and usually contains 8–17 mannoses, which is quite
different from S. cerevisiae structures that consist of approximately 50–150
mannose residues (Celik et al., 2007; Gemmill and Trimble, 1999). Similar
to insect and mammalian cells, the consensus sequence for N-linked glycans in
yeast is Asn-Xaa-Ser/Thr. Two groups have completed extensive engineering
to create P. pastoris strains that produce complex N-linked glycan structures
comparable to those produced by mammalian cells (Hamilton and Gerngross,
2007). However, only the strains developed by Roland Contreras’ group are
available to investigators and must be licensed through Research Corporation
Technologies. The O-linked structures in P. pastoris have not been studied
comprehensively but are known to be formed by the addition of one to four
mannose residues to serines/threonines (Goto, 2007; Trimble et al., 2004).
Several reports have indicated that expression of certain proteins in P. pastoris
resulted in the addition O-linked glycans not observed when the protein was
expressed endogenously in mammalian cells (Daly and Hearn, 2005).

4. Baculovirus/Insect Cells
Baculovirus-mediated expression in insect cells offers another useful
tool for generating recombinant proteins (see Chapter 14). Baculovirus is a
lytic, large (130 kb), double-stranded DNA virus, and the Autographa
californica virus is the most commonly used baculovirus isolate for recombi-
nant expression. Baculovirus is routinely amplified in insect cell lines
derived from the fall armyworm Spodoptera frugiperda (Sf 9, Sf 21), and
recombinant protein expression is completed either in the aforementioned
lines or in a line derived from the cabbage looper Trichoplusia ni (High-Five)
(Kost et al., 2005). Originally, creating recombinant baculoviruses involved
cotransfecting the gene of interest flanked by baculovirus sequence with
baculovirus DNA into insect cells, and screening for rare homologous
recombination events. Recombinants were identified by screening plaques
with a modified morphology, and often additional rounds of plaque screen-
ing were required to ensure that the recombinant viral preparation was
not contaminated with wild-type virus. This lengthy and laborious process
for generating recombinant viruses has been largely replaced by using
site-specific transposition (Bac-to-Bac or BaculoDirect, Invitrogen) or an
improved homologous recombination method with an engineered
Selecting an Expression System 137

baculovirus containing a lethal mutation in orf1629 (flashBAC from Oxford

Expression Technologies or BacMagic from EMD-Novagen). Both of
these approaches overcome the requirement to isolate plaques because the
efficiency of recombination is 100%. Following one or two rounds of
amplifying the recombinant baculovirus, the investigator can quantify the
baculovirus concentration stock either by the plaque assay or by using the
newer, more rapid real-time PCR or antibody-based assays (Hitchman
et al., 2007; Kwon et al., 2002). The improvements in creating and quanti-
fying recombinant baculoviruses have dramatically reduced the time for
evaluating baculovirus expression to approximately 3 weeks, including a
time-course study for optimizing expression.
The most common promoters used with baculovirus expression are the
polH and p10 promoters, both of which induce a high level of expression in
the very late phase of the baculovirus infection (Ikonomou et al., 2003).
During this phase, cells undergo cell death with the concomitant release of
proteases, which can result in degradation of the expressed recombinant
protein. To reduce proteolysis of the recombinant protein, promoters active
in earlier phases of the lytic cycle such as the basic promoter have been used
(Ikonomou et al., 2003). Alternatively proteolytic activity can be minimized
by using constructs deleted in the chiA and v-cath genes, which encode
chitinase and a cathepsin protease, respectively (Monsma and Scott, 1997).
Baculovirus-mediated expression is routinely used to generate both
cytoplasmic and secreted recombinant proteins. Efficient secretion generally
requires the presence of a signal peptide. Both insect and mammalian signal
sequences can promote entry into the insect cell secretory pathway. Insect
cells were originally grown in serum-containing medium which compli-
cated purification of the secreted proteins. Recent advances in media
development permit the replacement of serum with protein hydrolysates
derived from either animal tissues or plants, thereby greatly simplifying
protein purification (Ikonomou et al., 2003). However, the high cost of
this specialized media can limit its use for large-scale bioproduction.
Insect cells efficiently generate disulfide bonds in recombinant proteins.
They also produce the majority of the posttranslational modifications found
in mammalian cells. However, the N-linked glycan structure formed in
most insect cells is the predominantly fucosylated paucimannose structures
(Man3GlcNAc2-N-Asn) (Harrison and Jarvis, 2006; Jenkins et al., 1996).
This finding has prompted the recent generation of insect cell lines that
produce glycoproteins with the complex N-linked glycans normally found
in mammalian cells (Harrison and Jarvis, 2006). A transgenic Sf-9 insect line
expressing several glycosyltransferases is commercially available (Mimic cell
line, Invitrogen) and produces N-linked glycans containing a biantennary,
sialylated structure. There are only a few reports describing the O-linked
glycans structures generated by insect cells (Chen et al., 1991; Sugyiama
et al., 1993; Thomsen et al., 2004).
138 William H. Brondyk

5. Mammalian Cells
Mammalian expression methods have conventionally been considered
to be the least efficient vehicle for expressing recombinant proteins. How-
ever, recent advances have significantly improved the expression levels from
mammalian cell lines (see Chapter 15). For example, stably transfected
Chinese hamster ovary (CHO) cells have been reported to express recom-
binant antibodies up to a level of a few grams per liter (Figueroa et al., 2007;
Wurm, 2004). While many cell lines and expression strategies have been
tested, this chapter will focus on transient transfection in human embryonic
kidney (HEK293) cells and stable transfection with CHO cells.
The HEK293 cell line was derived from human embryonic kidney cells
transformed with adenovirus. HEK293 cells can be transiently transfected
with a high efficiency (>80%) using certain cationic lipids, calcium phos-
phate, or polyethyleneimine as transfection reagents (Durocher et al., 2002;
Jordan et al., 1996). For large-scale transient transfections (>100 ml), cal-
cium phosphate or polyethyleneimine reagents are more cost-effective
options when compared to cationic lipids (Baldi et al., 2007). Transient
transfections have been performed at even the bioreactor level but for most
laboratories this scale is technically challenging (Girard et al., 2002). The
transient transfection method is relatively easy, and the evaluation for a
given recombinant protein can be made in less than 2 weeks.
CHO cells are commonly used for mammalian expression when large
quantities of recombinant protein are needed. For example, most therapeu-
tic antibodies currently on the market are manufactured using this method.
The standard method for stable CHO expression involves transfecting
dihydrofolate reductase (DHFR)-deficient CHO cells with a DHFR selec-
tion cassette along with an expression cassette containing the gene of
interest (Wurm, 2004). Dihydrofolate reductase converts dihydrofolate
into tetrahydrofolate which is required for the de novo synthesis of purines,
certain amino acids, and thymidylic acid. Methotrexate, which binds and
inhibits DHFR, is used as a selection agent and only those cells that have
integrated the DHFR selection cassette will survive. Sequentially increasing
the concentration of methotrexate will result in amplification of the DHFR
gene along with the linked gene of interest. Following at least one round of
selection with the drug methotrexate, the stably transfected pools are sub-
cloned using limiting dilution cloning into multiwell plates. Typically only a
small percentage of the screened subclones will be expressing the recombi-
nant gene at a high level since in the majority of the clones, the expression
cassette has integrated into the heterochromatin region which is transcrip-
tionally inactive. Unfortunately, the entire selection and screening process
takes at least 2–3 months, making this the major drawback of the CHO
Selecting an Expression System 139

method. However, recent high-throughput methods based on flow cyto-

metry or automation have increased the ease in rapidly screening and
selecting high expressing clones (Browne and Al-Rubeai, 2007). Another
development has been to use specific cis-acting DNA elements flanking the
recombinant gene cassette that confer active transcription to integration
sites (Kwaks and Otte, 2006). Unfortunately, the majority of these DNA
elements is owned by companies and must be licensed for use in the
laboratory, and, even with the aforementioned advances in CHO expres-
sion, the timelines for generating a high expressing CHO clone have not
changed considerably.
Mammalian expression systems are used primarily to generate secreted
rather than intracellular recombinant proteins. Serum-free media have been
developed for both the CHO and HEK293 cell lines, which simplifies the
purification of secreted recombinant proteins. However, the cost of the
media is quite high, making large-scale bioproduction rather costly.
Mammalian cells contain the most superior folding and disulfide bond
formation when compared to other expression hosts. The N-linked and
O-linked glycan structures formed by mammalian cells are extremely varied
and are not only dependent on the protein but also on the mammalian cell
type used as the expression host ( Jenkins et al., 1996). Furthermore, the cell
culture conditions such as nutrient content, pH, temperature, oxygen levels
and ammonia concentration can significantly affect the glycosylation profile
(Butler, 2006). N-linked glycosylation can result in oligomannose, hybrid,
and complex structures, and the structures all contain the Man3GlcNac2
core (Bhatia and Mukhopadhyay, 1998). The oligomannose glycans can
have two to six additional mannoses and the mannoses can be phosphory-
lated or sulfated. The most common complex structures have two to four
Gal b1,4-GlcNac2 attached to the mannoses which result in bi-, tri-, and
tetra-antennary branches. The branches can terminate with sialic acid, and
fucose can also be attached to the structures. Hybrid structures contain
features of both the oligomannose and complex structures. O-Glycosylation
structures can be classified into eight types based on their core
structures: O-GalNAc-type glycosylation, O-GlcNAc-type glycosylation,
O-fucosylation, O-mannosylation, O-glucosylation, phosphoglycosylation,
O-glycosaminoglycan-type glycosylation, and collagen-type glycosylation
(Peter-Katalinic, 2005).

6. Protein Characteristics
When choosing an expression system, one can easily survey the
literature to determine if the recombinant protein has previously been
expressed in the past and then assess the success of the published strategy.
140 William H. Brondyk

Table 11.1 Summary of expression methods

Expression
systems Advantages Disadvantages
E. coli Rapid expression Limited capacity for
method (days) posttranslational
modifications
Inexpensive Difficult to produce
bioproduction media some proteins in a
and high density soluble, properly
biomass folded state
Simple process scale-up
Well characterized
genetics
P. pastoris Moderately rapid N-linked glycan
expression method structures different
(weeks) from mammalian
forms
Inexpensive Enhanced safety
bioproduction media precautions needed
and high density for large-scale
biomass bioproduction due to
Most posttranslational methanol in induction
modifications and media
high folding capacity
Baculovirus/ Moderately rapid N-linked glycan
insect cell expression method structures different
(weeks) from mammalian
forms
Most posttranslational Low density biomass and
modifications and expensive
high folding capacity bioproduction media
Difficult process scale-up
Mammalian— Moderately rapid Low density biomass and
transient expression method expensive
expression (weeks) bioproduction media
All posttranslational Difficult process scale-up
modifications and
high folding capacity
Mammalian— All posttranslational Lengthy expression
stable modifications and method (months)
expression high folding capacity Low density biomass and
expensive
bioproduction media
Difficult process scale-up
Selecting an Expression System 141

In drawing from reports in the literature, it is important to consider whether

the downstream application of the recombinant protein in the report is
similar or compatible with your intended application. Lacking that infor-
mation, reports about orthologs that have been expressed and described in
detail can prove useful. The number of proteins being expressed has
increased dramatically over the last decade which is in part due to the
sequencing of many genomes, the development of high-throughput expres-
sion methods, and the large Protein Structure Initiative (PSI). This expres-
sion trend will likely continue, which will mean that eventually the amount
of prior data will make the choice of an expression host a more straightfor-
ward decision.

6.1. Protein characteristics: E. coli and codon usage

In general, when considering the recombinant protein-host system options,
prokaryotic proteins should be expressed with only E. coli and not with
eukaryotic systems since usually the posttranslational modifications and
improved folding found in eukaryotes are either not necessary or perhaps
not desired for a prokaryotic protein. For eukaryotic proteins the circum-
stance is different since there is a preponderance of examples where eukary-
otic proteins are successfully expressed in E. coli (Sahdev et al., 2008). When
using this strategy, an important consideration is that E. coli, like all organ-
isms, has a bias for codon usage and the abundance of the tRNA pools in
E. coli mirror this bias (Gustafsson et al., 2004; Marin, 2008). Expressing a
eukaryotic protein containing several codons that are rare in E. coli can be
inefficient when the pool of corresponding tRNAs is limiting. The shortage
of tRNAs can lead to frameshifts in translation, misincorporation of amino
acids, or premature termination of translation. This problem is most evident
when rare codons are grouped together at the N-terminus (Kane, 1995).
However, the problem can be avoided by synthesizing a codon-optimized
gene or by using engineered E. coli strains containing increased pools of rare
tRNAs, which are available commercially (e.g., Rosetta strains, EMD-
Novagen). In most cases, codon bias should also be corrected in
those circumstances where the recombinant gene is being expressed in a
phylogenetically distant organism.

6.2. Protein characteristics: Cytoplasmic proteins

For a cytoplasmic protein, the optimal choice of an expression system
depends on the protein mass and the number of disulfide bonds in the
protein. For proteins between 10 and 50 kDa and containing few disulfide
bonds, E. coli is a good option for soluble protein expression (Dyson et al.,
2004). For larger proteins or those with many disulfide bonds, if soluble
expression is desired then usually either baculovirus or yeast is the preferred
142 William H. Brondyk

choices. Successful expression of proteins smaller than 10 kDa, with few or

no disulfide bonds, has been achieved through fusion with soluble tags and
expression in E. coli (Esposito and Chatterjee, 2006). Alternatively, expres-
sion of these small proteins can be directed into the secretory pathway of
P. pastoris (Daly and Hearn, 2005). However, in this pathway, care must be
taken to monitor potential inadvertent glycosylation of the normally cyto-
solic protein when it is forced into the secretory pathway. This can be
achieved by inspecting the sequence for the lack of the consensus N-linked
glycosylation sites. Unfortunately, for O-linked glycosylation there are no
consensus sequences, so the secreted recombinant protein must be analyzed
to ensure the lack of O-linked glycans.

6.3. Protein characteristics: Secreted proteins

Any of the expression hosts can be used to produce secreted proteins.
However, as described earlier, E. coli lacks most of the posttranslational
capabilities found in eukaryotic hosts. Consequently, E. coli may be subop-
timal for expressing secreted eukaryotic proteins but this is highly depen-
dent on the downstream application.

6.4. Protein characteristics: Membrane proteins

Membrane proteins represent an extremely challenging class of proteins to
express in large quantities. For some purposes, production of just the
soluble, hydrophilic portion will suffice, in which case the membrane-
spanning domain can be removed and the desired soluble portion can be
expressed. There are no clear guidelines on choosing the best system to
express intact membrane proteins (Sarramegna et al., 2003). However, for
most eukaryotic membrane proteins E. coli as an expression method is
generally not a good option because of its limited capacity for folding and
posttranslational modifications. By contrast, researchers have reported mod-
est success with expressing G protein-coupled receptors using the baculo-
virus and yeast methods (Sarramegna et al., 2003).

6.5. Protein characteristics: Toxic proteins

Recombinant proteins that are toxic to the expression host can be challeng-
ing to produce but this obstacle can usually be overcome. If the recombi-
nant protein is toxic, it is often useful to determine whether the problem is
host cell specific. If so, then the protein can be expressed in a more
compatible expression host. Another option is to use a tightly regulated,
inducible expression system such as those available for E. coli and P. pastoris.
For example, several elaborate inducible expression systems have been
developed for E. coli (Saida, 2007). In these systems, expression of the
Selecting an Expression System 143

recombinant gene is regulated by an inducible promoter, transcription

terminators, control of the plasmid copy number, or modification of the
coding sequence of the recombinant gene. In the available P. pastoris system,
the AOX1 promoter is tightly regulated by the combination of an induction
mechanism as well as a repression/derepression method (Daly and Hearn,
2005). Alternatively, several studies have demonstrated that the baculo-
virus/insect system can be used to express toxic proteins (Aguiar et al.,
2006; Korth and Levings, 1993). Lastly, with mammalian systems the easiest
option is to use transient expression. The several inducible systems compat-
ible with CHO cells take a considerable amount of time to complete the
necessary cell engineering and also with these systems it is difficult to obtain
tight regulation of gene expression which is required to prevent cell death
(Rossi and Blau, 1998).

7. Recombinant Protein Applications

Recombinant protein expression for structural studies requires proper
folding, correct formation of disulfide bonds and homogeneity of the recom-
binant product. The intrinsic cellular capacity for folding and disulfide bond
formation has been described for each of the expression systems. Potential
sources of heterogeneity include phosphorylation, inefficient cleavage of the
initiator methionine by methionine aminopeptidase, and glycosylation.
Unfortunately, phosphorylation heterogeneity is often observed with recom-
binant kinases and has been reported for each of the expression hosts. In these
cases, homogeneity can be attained by removing the phosphates on the
recombinant protein through treatment with phosphatase. N-terminal methi-
onine heterogeneity can occur when the recombinant protein is expressed in
the cytoplasm. Prokaryotic and eukaryotic organisms contain methionine
aminopeptidases, and the efficiency of methionine cleavage is influenced by
the amino acid adjacent to the initiator methionine (Giglione et al., 2004). In
general the size of the side chain of the adjacent amino acid inversely affects the
efficiency of methionine cleavage. However, in E. coli the high expression
level of recombinant proteins can saturate the enzyme and alter the rules for
efficient cleavage (Dong et al., 1996). This heterogeneity can be avoided by
carefully selecting the amino acid adjacent to the initiator methionine or by
using a cleavable tag at the N-terminus. Glycosylation heterogeneity occurs
with glycoproteins expressed in all eukaryotic organisms and normally less
heterogeneity is observed with insect and yeast hosts ( Jenkins et al., 1996).
Regardless of the expression host, the glycans are usually removed before
attempting to crystallize the recombinant protein.
Normally, when generating recombinant proteins for use as immunizing
antigens, any of the expression hosts can be used for this application. With
144 William H. Brondyk

glycoproteins it is not always clear whether the presence of glycans will alter
the immunogenicity of the recombinant antigen (Bhatia and Mukhopadhyay,
1998; Prasad et al., 1995).
Producing recombinant proteins suitable for in vitro activity studies as
well as for in vivo experiments requires appropriate protein folding and
disulfide bond formation. For glycoproteins, the presence of N-linked
glycans along with the glycan structure can have a significant impact on
both applications, and therefore must be considered when selecting a
eukaryotic expression host. N-linked glycosylation has been shown to
positively influence protein structure and increase protein stability (Bhatia
and Mukhopadhyay, 1998). In vitro, the structure of the N-linked glycan on
certain protein ligands has been demonstrated to affect the affinity for
receptor binding and signal transduction. With recombinant immunoglo-
bulins, the conserved N-linked glycan present in the Fc region influences
in vitro effector activity (Presta, 2008). For instance, the presence of fucose
on the N-linked glycan in human IgG1 reduces antibody-dependent cell
cytotoxicity activity (Shinkawa et al., 2003). In vivo, the N-linked glycan
structure on proteins dramatically impacts metabolic clearance and biodis-
tribution. For example, N-linked glycans that are not capped with sialic acid
are cleared by hepatic receptors, including the asialoglycoprotein and man-
nose receptors (Weigel and Yik, 2002). The importance of O-linked
glycosylation has not yet been defined. If the effect of glycosylating the
recombinant protein is unknown then a safer strategy involves choosing an
expression host similar to the source of the recombinant gene.

8. Conclusion
The E. coli, P. pastoris, baculovirus/insect cells and mammalian systems
each have both advantages and disadvantages for expressing recombinant
proteins (Table 11.1). Whether a given system will express a protein at a high
level and generate a quality product is largely protein dependent. A careful
evaluation of the characteristics of the recombinant protein along with the
downstream application must be considered when selecting an expression
method. Unfortunately, there will be circumstances when the expression
choice will not be obvious and several expression hosts must be evaluated.

REFERENCES
Aguiar, R. W., Martins, E. S., Valicente, F. H., Carneiro, N. P., Batista, A. C.,
Melatti, V. M., Monnerat, R. G., and Ribeiro, B. M. (2006). A recombinant truncated
Cry1Ca protein is toxic to lepidopteran insects and forms large cuboidal crystals in insect
cells. Curr. Microbiol. 53, 287–292.
Selecting an Expression System 145

Andersen, C. L., Matthey-Dupraz, A., Missiakas, D., and Raina, S. (1997). A new Escherichia
coli gene, dsbG, encodes a periplasmic protein involved in disulphide bond formation,
required for recycling DsbA/DsbB and DsbC redox proteins. Mol. Microbiol. 26,
121–132.
Andersson, D. I., Boham, K., Isaksson, L. A., and Kurland, C. G. (1982). Translation rates
and misreading characteristics of rpsD mutants in Escherichia coli. Mol. Gen. Genet. 187,
467–472.
Baldi, L., Hacker, D. L., Adam, M., and Wurm, F. M. (2007). Recombinant protein
production by large-scale transient gene expression in mammalian cells: state of the art
and future perspectives. Biotechnol. Lett. 29, 677–684.
Baneyx, F., and Mujacic, M. (2004). Recombinant protein folding and misfolding in
Escherichia coli. Nat. Biotechnol. 22, 1399–1408.
Bardwell, J. C. (1994). Building bridges: Disulphide bond formation in the cell. Mol.
Microbiol. 14, 199–205.
Bessette, P. H., Aslund, F., Beckwith, J., and Georgiou, G. (1999). Efficient folding of
proteins with multiple disulfide bonds in the Escherichia coli cytoplasm. Proc. Natl. Acad.
Sci. USA 96, 13703–13708.
Bhatia, P. K., and Mukhopadhyay, A. (1998). Protein glycosylation: Implications for in vivo
functions and therapeutic applications. Adv. Biochem. Eng. Biotechnol. 64, 155–201.
Browne, S. M., and Al-Rubeai, M. (2007). Selection methods for high-producing mamma-
lian cell lines. Trends Biotechnol. 25, 425–432.
Butler, M. (2006). Optimisation of the cellular metabolism of glycosylation for recombinant
proteins produced by mammalian cell systems. Cytotechnology 50, 57–76.
Celik, E., Calik, P., Halloran, S. M., and Oliver, S. G. (2007). Production of recombinant
human erythropoietin from Pichia pastoris and its structural analysis. J. Appl. Microbiol.
103, 2084–2094.
Chen, W., Shen, Q., and Bahl, O. P. (1991). Carbohydrate variant of the recombinant
b-subunit of human choriogonadotropin expressed in baculovirus expression system.
J. Biol. Chem. 266, 4081–4087.
Cregg, J. M. (2007). Introduction: Distinctions between Pichia pastoris and other expression
systems. Methods Mol. Biol. 389, 1–10.
Cregg, J. M., Barringer, K. J., Hessler, A. Y., and Madden, K. R. (1985). Pichia pastoris as a
host system for transformation. Mol. Cell. Biol. 5, 3376–3385.
Daly, R., and Hearn, M. T. W. (2005). Expression of heterologous proteins in Pichia pastoris:
A useful experimental tool in protein engineering and production. J. Mol. Recognit. 18,
119–138.
Dong, M. S., Bell, L. C., Guo, Z., Phillips, D. R., Blair, I. A., and Guengerich, F. P. (1996).
Identification of retained N-formylmethionine in bacterial recombinant mammalian
cytochrome P450 proteins with the N-terminal sequence MALLLAVFL. . .: Roles of
residues 3–5 in retention and membrane topology. Biochemistry 35, 10031–10041.
Durocher, Y., Perret, S., and Kamen, A. (2002). High-level and high-throughput recombi-
nant protein production by transient transfection of suspension-growing human
293-EBNA1 cells. Nucleic Acids Res. 30, E9.
Dyson, M. R., Shadbolt, S. P., Vincent, K. J., Perera, R. L., and McCafferty, J. (2004).
Production of soluble mammalian proteins in Escherichia coli: Identification of protein
features that correlate with successful expression. BMC Biotechnol. 4, 32–49.
Esposito, D., and Chatterjee, D. K. (2006). Enhancement of soluble protein expression
through the use of fusion tags. Curr. Opin. Biotechnol. 17, 353–358.
Figueroa, B., Ailor, E., Osborne, D., Hardwick, J. M., Reff, M., and Betenbaugh, M. J.
(2007). Enhanced cell culture performance using inducible anti-apoptotic genes E1B–
19 K and Aven in the production of a monoclonal antibody with Chinese Hamster Ovary
cells. Biotechnol. Bioeng. 97, 877–892.
146 William H. Brondyk

Gemmill, T. R., and Trimble, R. B. (1999). Overview of N- and O-linked oligosaccharide

structures found in various yeast species. Biochim. Biophys. Acta 1426, 227–237.
Giglione, C., Boularot, A., and Meinnel, T. (2004). Review: Protein N-terminal methio-
nine excision. Cell. Mol. Life Sci. 61, 1455–1474.
Girard, P., Derousazi, M., Baumgartner, G., Bourgeois, M., Jordan, M., Jacko, B., and
Wurm, F. M. (2002). 100-liter transient transfection. Cytotechnology 38, 15–21.
Gleeson, M. A., White, C. E., Meininger, D. P., and Komives, E. A. (1998). Generation of
protease-deficient strains and their use in heterologous protein expression. Methods Mol.
Biol. 103, 81–94.
Goto, M. (2007). Protein O-glycosylation in fungi: Diverse structures and multiple func-
tions. Biosci. Biotechnol. Biochem. 71, 1415–1427.
Goustin, A. S., and Wilt, F. H. (1982). Direct measurement of histone peptide elongation
rate in cleaving sea urchin embryos. Biochim. Biophys. Acta 699, 22–27.
Gustafsson, C., Govindarajan, S., and Minshull, J. (2004). Codon bias and heterologous
protein expression. Trends Biotechnol. 22, 346–353.
Hamilton, S. R., and Gerngross, T. U. (2007). Glycosylation engineering in yeast: The
advent of fully humanized yeast. Curr. Opin. Biotechnol. 18, 387–392.
Harrison, R. L., and Jarvis, D. L. (2006). Protein N-glycosylation in the baculovirus-insect
cell expression system and engineering of insect cells to produce ‘‘mammalianized’’
recombinant glycoproteins. Adv. Virus Res. 68, 159–191.
Hitchman, R. B., Siaterli, E. A., Nixon, C. P., and King, L. A. (2007). Quantitative real-time
PCR for rapid and accurate titration of recombinant baculovirus particles. Biotechnol.
Bioeng. 96, 810–814.
Ikonomou, L., Schneider, Y., and Agathos, S. N. (2003). Insect cell culture for industrial
production of recombinant proteins. Appl. Microbiol. Biotechnol. 62, 1–20.
Jana, S., and Deb, J. K. (2005). Strategies for efficient production of heterologous proteins in
Escherichia coli. Appl. Microbiol. Biotechnol. 67, 289–298.
Jenkins, N., Parekh, R. B., and James, D. C. (1996). Getting the glycosylation right:
implications for the biotechnology industry. Nat. Biotechnol. 14, 975–981.
Jordan, M., Schallhorn, A., and Wurm, F. M. (1996). Transfecting mammalian cells:
Optimization of critical parameters affecting calcium-phosphate precipitate formation.
Nucleic Acids Res. 24, 596–601.
Kane, J. F. (1995). Effects of rare codon clusters on high-level expression of heterologous
proteins in Escherichia coli. Curr. Opin. Biotechnol. 6, 494–500.
Korth, K. L., and Levings, C. S. (1993). Baculovirus expression of the maize mitochondrial
protein URF13 confers insecticidal activity in cell cultures and larvae. Proc. Natl. Acad.
Sci. USA 90, 3388–3392.
Kost, T. A., Condreay, J. P., and Jarvis, D. L. (2005). Baculovirus as versatile vectors for
protein expression in insect and mammalian cells. Nat. Biotechnol. 23, 567–575.
Kwaks, T. H. J., and Otte, A. P. (2006). Employing epigenetics to augment the expression of
therapeutic proteins in mammalian cells. Trends Biotechnol. 24, 137–142.
Kwon, M. S., Dojima, T., Toriyama, M., and Park, E. Y. (2002). Development of an
antibody-based assay for determination of baculovirus titers in 10 hours. Biotechnol. Prog.
18, 647–651.
LaVallie, E. R., DiBlasio, E. A., Kovacic, S., Grant, K. L., Schendel, P. F., and McCoy, J. M.
(1993). A thioredoxin gene fusion expression system that circumvents inclusion body
formation in the E. coli cytoplasm. Biotechnology (NY) 11, 187–193.
Marin, M. (2008). Folding at the rhythm of the rare codon beat. Biotechnol. J. 3, 1047–1057.
Monsma, S. A., and Scott, M. (1997). BacVector-3000: An engineered baculovirus designed
for greater protein stability. Innovations 16–19.
Peter-Katalinic, J. (2005). O-glycosylation of proteins. Methods Enzymol. 405, 139–171.
Selecting an Expression System 147

Prasad, S. V., Mujtaba, S., Lee, V. H., and Dunbar, B. S. (1995). Immunogenicity enhance-
ment of recombinant rabbit 55-kilodalton zona pellucida protein expressed using the
baculovirus expression system. Biol. Reprod. 52, 1167–1178.
Presta, L. G. (2008). Molecular engineering and design of therapeutic antibodies. Curr. Opin.
Immunol. 20, 460–470.
Rossi, F. M., and Blau, H. M. (1998). Recent advances in inducible gene expression systems.
Curr. Opin. Biotechnol. 9, 451–456.
Sahdev, S., Khattar, S. K., and Saini, K. S. (2008). Production of active eukaryotic proteins
through bacterial expression systems: A review of existing biotechnology strategies. Mol.
Cell Biochem. 307, 249–264.
Saida, F. (2007). Overview on the expression of toxic gene products in Escherichia coli. Curr.
Protoc. Protein Sci. 5.19.1–5.19.13.
Sarramegna, V., Talmont, F., Demange, P., and Milon, A. (2003). Heterologous expression
of G-protein-coupled receptors: Comparison of expression systems from the standpoint
of large-scale production and purification. Cell. Mol. Life Sci. 60, 1529–1546.
Shinkawa, T., Nakamura, K., Yamane, N., Shoji-Hosaka, E., Kanda, Y., Sakurada, M.,
Uchida, K., Anazawa, H., Satoh, M., Yamasaki, M., Hanai, N., and Shitara, K. (2003).
The absence of fucose but not the presence of galactose or bisecting N-acetylglucosamine
of human IgG1 complex-type oligosaccharides shows the critical role of enhancing
antibody-dependent cellular cytotoxicity. J. Biol. Chem. 278, 3466–3473.
Spiess, C., Beil, A., and Ehrmann, M. (1999). A temperature-dependent switch from
chaperone to protease in a widely conserved heat shock protein. Cell 97, 339–347.
Strausberg, R. L., and Strausberg, S. L. (2001). Overview of protein expression in
Saccharomyces cerevisiae. Curr. Protoc. Protein Sci. Chapter 5, Unit 5.6.
Sugyiama, K., Ahorn, H., Maurer-Fogy, I., and Voss, T. (1993). Expression of human
interferon-a2 in Sf 9 cells: Characterization of O-linked glycosylation and protein het-
erogeneities. Eur. J. Biochem. 217, 921–927.
Thomsen, D. R., Post, L. E., and Elhammer, A. P. (2004). Structure of O-glycosidically
linked oligosaccharides synthesized by the insect cell line Sf 9. J. Cell. Biochem. 43, 67–79.
Trimble, R. B., Lubowski, C., Hauer, C. R., Stack, R., McNaughton, L., Gemmill, T. R.,
and Kumar, S. A. (2004). Characterization of N- and O-linked glycosylation of recom-
binant human bile salt-stimulated lipase secreted by Pichia pastoris. Glycobiology 14,
265–274.
Vera, A., Gonzalez-Montalban, N., Aris, A., and Villaverde, A. (2006). The conformational
quality of insoluble recombinant proteins is enhanced at low growth temperatures.
Biotechnol. Bioeng. 96, 1101–1106.
Waugh, D. S. (2005). Making the most of affinity tags. Trends Biotechnol. 23, 316–320.
Weigel, P. H., and Yik, J. H. N. (2002). Glycans as endocytosis signals: The cases of the
asialoglycoprotein and hyaluronan/chondroitin sulfate receptors. Biochim. Biophys. Acta
1372, 341–363.
Wurm, F. M. (2004). Production of recombinant protein therapeutics in cultivated mam-
malian cells. Nature Biotechnol. 22, 1393–1398.
Young, J. C., Agashe, V. R., Siegers, K., and Hartl, P. U. (2004). Pathways of chaperone-
mediated protein folding in the cytosol. Nat. Rev. Mol. Cell Biol. 5, 781–791.
 C H A P T E R T W E LV E

Bacterial Systems for Production

of Heterologous Proteins
Sarah Zerbs, Ashley M. Frank, and Frank R. Collart

Contents
1.Introduction 150
2.Heterologous Protein Production Using Escherichia coli 150
3.Planning a Bacterial Expression Project 151
4.Evaluation of Project Requirements 152
5.Target Analysis 152
6.Cloning 153
7.Preparation of T4 DNA Polymerase-Treated DNA Fragments 155
8.Expression in the E. coli Cytoplasm 156
9.Expression of Cytoplasmic Targets in E. coli 157
9.1. Analysis of protein expression and solubility 157
10. Analysis of Heterologous Protein Expression in E. coli 157
10.1. Analysis of protein solubility 158
10.2. Analysis of protein expression results 159
10.3. Autoinduction method for protein expression 159
11. Small-Scale Expression Cultures in Autoinduction Media Protocol 160
12. Periplasmic Expression of Proteins 160
13. Expression of Periplasmic Targets in E. coli 161
14. Small-Scale Osmotic Shock Protocol 162
15. Alternative Bacterial Systems for Heterologous Protein Production 164
16. Alternative Vector and Induction Conditions 165
17. Production Scale 166
Acknowledgment 166
References 166

Abstract
Proteins are the working molecules of all biological systems and participate in a
majority of cellular chemical reactions and biological processes. Knowledge of the
properties and function of these molecules is central to an understanding of
chemical and biological processes. In this context, purified proteins are a starting
point for biophysical and biochemical characterization methods that can assist in

Biosciences Division, Argonne National Laboratory, Lemont, Illinois, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63012-3 All rights reserved.

149
150 Sarah Zerbs et al.

the elucidation of function. The challenge for production of proteins at the scale and
quality required for experimental, therapeutic and commercial applications has led
to the development of a diverse set of methods for heterologous protein produc-
tion. Bacterial expression systems are commonly used for protein production as
these systems provide an economical route for protein production and require
minimal technical expertise to establish a laboratory protein production system.

1. Introduction
Bacterial expression systems and Escherichia coli in particular are frequently
used for production of heterologous proteins at laboratory and industrial
process scales (Baneyx, 1999; Terpe, 2006). The widespread use of bacterial
systems for protein production arises primarily from the nominal cost and
minimal technical requirements for implementation in a laboratory scale
environment. A variety of vector and host options are available and the short
doubling time of most engineered strains enables rapid evaluation of experi-
mental outcome and reduces the stringency of requirements for sterile tech-
nique and facilities. Many of these intrinsic advantages at the laboratory scale
can be easily modified to accommodate automation of the process and imple-
mentation in a high throughput setting (Dieckman et al., 2002). In most cases,
the bacterial expression systems commonly used for benchtop scale processing
can also be adapted to large-scale projects that require production and screen-
ing of hundreds of expression clones or large-scale production of selected
proteins (Baneyx, 1999; Klock et al., 2005; Terpe, 2006; Yokoyama, 2003).
In spite of these advantages, bacterial systems have a number of important
limitations for expression of heterologous proteins that should be considered
in the development of a protein production expression strategy. These limita-
tions are especially apparent for eukaryotic proteins (Dyson et al., 2004) or
proteins that require the coexpression of maturation proteins (Londer et al.,
2008). This is not unexpected since the biological and chemical characteristics
of the bacterial cellular compartments differ from eukaryotic organisms. In
particular, accessory proteins such as chaperones, posttranslational modifica-
tion proteins, or maturation proteins vary widely between eukaryotic and
bacterial organisms. In some cases, these limitations can be circumvented by
the use of genetically modified bacterial expression strains but the selection of
an alternative expression system must often be considered.

2. Heterologous Protein Production Using

Escherichia coli
There are an increasing number of bacterial systems available for the
production of heterologous proteins. The selection of a particular system is
influenced by the nature of the target protein, the experience of the user,
Protein Expression in Bacterial Systems 151

and the intended use of the product. E. coli is the most commonly used
bacterial system for production of heterologous proteins. Over a century of
intensive study of E. coli has provided a great deal of information about
regulatory mechanisms and the function of the host accessory proteins that
may impact expression outcome. In addition, there is an extensive resource
of methodological and technological materials in support of protein pro-
duction in a laboratory or commercial setting. For many protein production
projects, the availability of these resources and the minimal technical
requirements make E. coli a host of first choice for preliminary protein
expression screening. Consequently, we use this organism as a model to
illustrate specific approaches and methods for protein production. The
description of specific methods for protein production using E. coli will be
followed by a discussion of alternative bacterial expression systems. Many of
the core concepts and techniques described for protein production using
E. coli are directly applicable to other bacterial systems.

3. Planning a Bacterial Expression Project

The successful implementation of a system for bacterial expression is
dependent on a series of sequential steps that are illustrated in chronological
order in Table 12.1. The initial stages of the process are not experimental
but rely on defining a set of project requirements and analysis of the

Table 12.1 Scheme for heterologous protein production using a bacterial host

Stage Impact on selection of expression system

Outline project requirements Define requirements such as scale,
functionality, resources, and intended
application to select an appropriate
expression system
Target analysis Utilize coding region sequence features,
available experimental data or historical
data to define and optimize expression
strategies
Cloning Cloning options include vectors, selection
strategy, fusion tags to enhance
solubility or facilitate purification,
inducer, or cellular targeting sequences
Selection of expression strain(s) Rare codons, protease, accessory proteins
Characterization of the Expression, solubility, functionality
expression product
152 Sarah Zerbs et al.

sequence features and biochemical characteristics of the expression target.

These steps are critical to the ultimate success of the project and can
contribute to a realistic assessment of project costs and likelihood for success.
We present a brief review of these topics to provide continuity with the
experimental components as some of these topics are addressed in greater
detail in chapters included in this volume.

4. Evaluation of Project Requirements

The intended use of the protein product can have a major impact in
the selection and prioritization of expression systems. Preservation of the
native function and characteristics of the protein are essential for some
applications (e.g., functional assignment or drug screening) but may be of
less concern if the target is to be used as an immunogen. Other practical
considerations include protein yield, time constraints, or production costs.
A clear definition of the project requirements is vital to ensure the selection
of an appropriate expression system and the strategy for cloning and pro-
duction of the desired product (Dieckman et al., 2006).

5. Target Analysis
The biochemical and biological attributes of the target are essential
considerations for selection of an expression strategy and are primary pre-
dictors of expression and solubility outcomes. One set of attributes can be
assembled by analysis of the primary sequence for prediction of secondary
structure, biological localization (membrane, cytoplasmic, periplasmic, or
extracellular), classification into families (fold and domains) and inference
of biochemical properties (pI, disordered regions, ligands). Features such as
a high probability for membrane spanning helices may dictate the use of a
system designed for the expression of membrane proteins or the use of
a domain cloning strategy for production of the soluble component of the
protein. Use of the target sequence features to guide the selection of the
expression host and vector construct will contribute to production of a
mature and appropriately localized protein. Sequence information can be
supplemented with experimental or historical data on the target protein or a
homolog to provide further insight into expression system optimization.
For example, proteins known to require a prosthetic group for proper
function may necessitate selection of a specific host to ensure production
of a fully functional protein. A case in point is the production of type c
cytochromes which is facilitated by the use of genetically engineered strains
of E. coli containing accessory proteins for covalent attachment of hemes to
Protein Expression in Bacterial Systems 153

polypeptide chains of apocytochromes (Londer et al., 2008). In practice, it is

often difficult to predict the outcome of a protein expression trial in the
absence of historical expression data for the target protein or a similar
protein. A common experimental approach is to utilize several systems,
beginning with the simplest and most cost effective method and invoke
more complex systems for targets that fail.

6. Cloning
A variety of options are available for cloning the target sequence
which can be generally categorized into serial and parallel systems. Serial
systems, such as those that utilize restriction enzymes to generate compatible
target and vector termini, are ubiquitous and provide several options to
generate the target/vector compatible ends required to generate an expres-
sion construct. A disadvantage of this approach is the requirement of
validating restriction enzyme cleavage strategies for each target and vector.
There is growing interest in the use of parallel systems or universal cloning
methods that enable easy transfer of the target to multiple vectors and
expression systems regardless of the target sequence (Table 12.2). Examples
of these systems include the Gateway (Esposito et al., 2009), Infusion (Zhu
et al., 2007) or ligation independent cloning (LIC) methods (Aslanidis and
de Jong, 1990; Haun et al., 1992). An advantage of this approach is the
ability to clone a target in multiple vectors and to simultaneously evaluate
multiple expression strategies in a cost effective manner.
LIC is a cost effective method particularly suited for bacterial expression
as the cloning reagents are nonproprietary and available from several com-
mercial suppliers. In the LIC method, specific nucleotide sequences are
appended to the PCR primers and allow any gene to be cloned regardless of
DNA sequence. Compatible ends in both the vector and PCR fragments
are generated by treatment with T4 DNA polymerase in the presence of a
specific nucleoside triphosphate. The procedure generates complementary
10–15 bp overhangs in the vector and PCR fragment that anneal with
sufficient strength to permit transformation without ligation. The process
allows consistent design of PCR primers, is directional, and results in high
cloning efficiency. Although commercial vectors are available, the approach
has been used in several large-scale cloning projects which provide vector
resources to individual investigators. The procedure described below was
developed for use with the suite of vectors designed at the Midwest Center
for Structural Genomics (Eschenfeldt et al., 2009). The method is generally
applicable to other LIC compatible systems provided adjustments are made
for the specific sequences appended to the amplification primers.
154 Sarah Zerbs et al.

Table 12.2 Commonly used E. coli expression hosts and vectorsa

Strain Feature References or Suppliers

HB101 DH5a Common strains for Boyer and Roulland-
subcloning and preparation Dussoix (1969) and
of plasmid DNA. Strains Woodcock et al.
have reduced recombination (1989)
and restriction capabilities
that aid in plasmid stability
and improved quality of
plasmid DNA
BL21(DE3) Widely used expression strain Studier and Moffatt
which lacks the lon and (1986), multiple
opmT proteases and contains suppliers
a copy of the T7 RNA
polymerase gene under the
control of the lacUV5
promoter. These
modifications enable stable
expression of proteins using
T7 promoter driven
constructs
C41 C43 Effective for expression of Miroux and Walker
toxic and membrane (1996), Lucigen
proteins
ABLE strains Strains enable copy number Stratagene
control of ColE1 derived
plasmids
Origami K-12 derivatives with Derman et al. (1993),
mutations in the trxB and gor EMD Biosciences
genes which enhance
disulfide bond formation in
the cytoplasm
Vector
pET series Widely used expression EMD Biosciences
systems for inducible
expression of proteins using
a T7 promoter construct
pBAD series Tightly regulated expression Guzman et al. (1995)
system with expression
controlled by the presence
of arabinose
His GST Common purification fusion Hengen (1995) and
tags Smith and Johnson
(1988)
x
Protein Expression in Bacterial Systems 155

Table 12.2 (continued)

Strain Feature References or Suppliers

MBP Nus Common solubility enhancing De Marco et al. (2004)
fusion tags and Kapust and
Waugh (1999)
Gateway A series of vectors that use a Hartley et al. (2000),
vectors recombinational strategy to Invitrogen
enable transfer of DNA
fragments between different
cloning vectors
Ligation A series of vectors that use an Aslanidis and de Jong,
independent anneal strategy to enable (1990) and Haun et al.
cloning parallel cloning of coding (1992)
vectors region fragments
pET26b Vector contains the pelB leader Makrides (1996) and
sequence that targets Matthey et al. (1999)
expression to the
periplasmic space
a
The NIH funded PSI Materials Repository (http://www.hip.harvard.edu/PSIMR/index.htm) and
Protein Expression Laboratory (http://web.ncifcrf.gov/atp/) maintain an extensive collection of
expression strains and vectors that are available to the general public.

7. Preparation of T4 DNA Polymerase-Treated

DNA Fragments
1. Append the appropriate LIC specific nucleotide sequences (e.g., forward
primer: TACTTCCAATCCAATGCC, reverse primer with added stop:
TTATCCACTTCCAATGTTA) to the target specific PCR primers.
2. The specified method is scaled to microwell plates (96 targets) but can be
adjusted to any number of targets. Make LIC Reaction mix sufficient for
one 96-well plate by combining the following reagents:
a. 465 mL of 10
T4 DNA polymerase buffer. T4 DNA polymerase
reaction buffers supplied by most vendors can be substituted for the
LIC reaction buffer described in this section. Our comparison of
various common T4 DNA polymerase reaction buffers shows less
than a 25% difference in the cloning efficiency of the final product.
b. 465 mL of 25 mM dCTP, molecular biology grade (Promega cat. no.
U1221).
c. 228 mL of 100 mM dithiothreitol (DTT) solution (Novagen cat. no.
70099).
d. 60 mL of water.
e. 250 units of T4 DNA Polymerase (LIC quality,  2.5 units/mL,
EMD Biosciences/Novagen).
156 Sarah Zerbs et al.

3. Keep this mixture on ice and add the T4 DNA polymerase just before
use. Pipette up and down several times to uniformly distribute the
enzyme in the reaction mix.
4. Array 10.4 mL of the LIC reaction mix into a polypropylene 96-well
plate.
5. Add 30 mL (40–100 ng) of purified PCR fragment to the LIC reaction
mix and pipette up and down several times to mix. Incubate at room
temperature for 30 min. Our studies of various fragment-to-vector ratios
(Dieckman et al., 2006) indicate a wide tolerance for variation in the
amount of target DNA fragment on the annealing reaction.
6. Incubate on a heat block at 75  C for 20 min to inactivate the T4 DNA
polymerase.
7. In another 96-well plate, anneal 1–2 mL of the T4-Polymerase-treated
PCR LIC fragments with 4 mL (20–50 ng) T4-Polymerase-treated LIC
vector.
8. Incubate the annealing reaction 5–10 min at room temperature.
9. Use the entire annealing reaction to transform  50 mL competent E. coli
cells and select for transformed colonies (Sambrook and Russell, 2001).
Following the heating process the LIC plates are stored in the refrigera-
tor at 4  C until needed. The preparation of LIC compatible vector is
similar to the above procedure but uses the complementary base dGTP
(Eschenfeldt et al., 2009). The constructs can be validated by sequence
analysis at this stage or the analysis can be performed after analysis for
expression and solubility.

8. Expression in the E. coli Cytoplasm

The majority of vectors available for expression in E. coli are designed
for cytoplasmic expression. These vectors have been engineered with an
assortment of selectable markers, bacterial promoters, plasmid replication
origins, localization signals, and fusion tags (Table 12.2). The protocol we
describe is adapted for the T7 promoter (Novagen, pET vector series)
which can overexpress a target at a level representative of the most abundant
native proteins in E. coli (Studier and Moffatt, 1986). An E. coli strain that
expresses the T7 RNA polymerase (e.g., BL21(DE3)) is required for protein
expression using the pET family of vectors. Variants of these strains are
available to coexpress rare tRNAs (Carstens, 2003), additional cofactors
necessary for protein folding (Baneyx and Palumbo, 2003), or proteins that
enable disulfide bond formation and promote protein folding activity in the
cytoplasm (Prinz et al., 1997). This protocol summarizes the process for use
of IPTG in conjunction with an inducible system for production of the
target protein.
Protein Expression in Bacterial Systems 157

9. Expression of Cytoplasmic Targets in E. coli

1. Use a single colony to inoculate a 2-mL culture of media with appro-
priate antibiotic. Use a culture tube that will hold at least 5
the final
culture volume for sufficient aeration.
2. Incubate samples at 37  C, 250 rpm until the OD600 reaches 0.4–0.8.
The culture should appear cloudy but not completely turbid. In general,
this requires 3–4 h of growth with BL21(DE3) cells.
3. Add 20 mL 100 mM IPTG to each culture (1 mM final concentration).
Return induced cultures to the incubator set at 37  C, 250 rpm for 4 h.
The cultures should be turbid after 4 h of induction.
4. Cultures should show significant expression after 3–8 h of induction at
37  C. Remove sample and analyze as described below.

9.1. Analysis of protein expression and solubility

Validation of protein expression typically involves an assessment of expres-
sion and solubility of the target protein and a qualitative verification of the
expected protein size. Most of the E. coli systems used for heterologous
expression produce sufficient level of the target protein to enable assessment
of expression/solubility outcome by denaturing gel electrophoresis
(Fig. 12.1). The method is low cost, relatively easy, and returns results
quickly. Proteins expressed at low levels or that may be marginally soluble
may require a more sensitive detection method such as Western blotting.

10. Analysis of Heterologous Protein

Expression in E. coli
1. Remove 200 mL of induced culture from each sample to a clean micro-
centrifuge tube. Centrifuge this portion at 14,000 rpm for 1 min. The
cells should form a dense pellet with a clear supernatant. The remaining
culture is used for assessment of target protein solubility described in
Section 10.1.
2. Decant or aspirate spent media, being careful to retain the cell pellet.
Add 50 mL of 2
SDS loading dye and use repeated pipeting to resus-
pend cells.
3. Boil samples for 5 min and allow to cool slightly before loading on an
SDS–PAGE gel. For a 17-well 8
10 cm gel stained with Coomassie,
5 mL of sample is usually sufficient.
158 Sarah Zerbs et al.

ZF356 ZF384 ZF203 ZF254 ZF137

M A B C A B C A B C A B C A B C

Expression analysis

Solubility analysis

Figure 12.1 Expression and solubility for a set of zebrafish proteins. The top figure
represents a Coomassie-stained gel displaying total expression products for proteins
expressed in three different vector systems. The bottom figure is a Coomassie-stained
gel with the soluble fractions of the same constructs. In the second gel, the order of
ZF356 and the molecular weight marker is reversed from the top gel. Vector ‘‘A’’ is the
pMCSG7 vector and produces an N-terminal fusion containing a TEV protease cleav-
able His tag. Vector ‘‘B’’ targets the protein to the periplasm and produces a N-terminal
fusion similar to that described for Vector A. Vector ‘‘C’’ is pMCSG19 which contains a
maltose binding protein (MBP) fusion sequence (Donnelly et al., 2006). Protein acces-
sion numbers are as follows: ZF356; AAH56726.1, ZF384; AAH58296.1, ZF203;
AAH46038.1, ZF254; AAH47843.1, ZF137; AAH67155.1.

10.1. Analysis of protein solubility

1. Pellet the remaining cell culture by centrifugation at 3500 rpm for
10 min. Carefully decant spent media and blot remaining liquid on a
paper towel.
2. Freeze pellets at  80  C. Prepare enough lysis buffer (final concentra-
tion: 300 mM NaCl, 50 mM Na2PO4, protease inhibitor cocktail
(Sigma), 120 kU/mL rLysozyme (Novagen) and 25 U/mL Benzonase
(Novagen)), for all samples. Alternatively, Bugbuster (Novagen) or
B-PER (Pierce) can be substituted for lysis buffer. Use volumes indi-
cated for a 2-mL culture in reagent instructions.
Protein Expression in Bacterial Systems 159

3. Remove samples from freezer and thaw slightly. Add 180 mL of lysis buffer
to each sample, cover tightly and vortex to completely resuspend pellet.
4. Return plate to  80  C for 5 min, then remove and incubate at room
temperature until the ice is completely melted. Repeat this freeze–thaw
cycle once more. Samples may turn clear or they may remain cloudy.
5. Pellet cell debris at 3500 rpm for 10–15 min.
6. Remove a 50 mL portion from the top of each sample being careful not
to remove any of the cell debris with the supernatant.
7. Add 60 mL of 2
SDS loading dye to the supernatant and boil for 5 min.
Analyze the samples by denaturing gel electrophoresis. For a 17-well
8
10 cm gel stained with Coomassie, 5 mL of sample is usually sufficient.
8. [Optional] To examine the insoluble fraction of the cell lysate, discard
all of the remaining supernatant from the cell debris in step 7. Be careful
not to disturb or remove the pellet.
9. Add 300 mL of 1
SDS loading dye to the pellet and cover tightly.
Vortex sample until entire pellet is resuspended.
10. Boil sample for 5 min and cool slightly before loading on an SDS–
PAGE gel. For a 17-well 8
10 cm gel stained with Coomassie, 3 mL is
usually enough to visualize expression.

10.2. Analysis of protein expression results

Use of the same culture for assessment of expression and solubility can
compensate for the variance in expression level. Expression rates in E. coli
typically exceed 80% but solubility can be a limiting factor and is target
dependent (Fig. 12.1). Targets are scored as ‘‘no expression’’ or ‘‘insoluble’’
based on the absence of a detectable stained protein band of the correct
molecular weight observed after SDS–PAGE analysis. Targets can be scored
as positive for expression or solubility based on observation of a protein
stained band of correct molecular weight. This is a qualitative assessment
and additional validation such as sequence analysis and/or mass spectrome-
try is necessary to assure identity of the polypeptide. We use a relative
ranking scale that compares staining of the target relative to the general
intensity of native E. coli proteins. Target bands that are visible but with
intensity level less than most of the E. coli proteins are scored as level 1 or
low expression/solubility. Levels 2 and 3 (moderate and high expression/
solubility) have staining intensity comparable to that of highly expressed
E. coli proteins or more prominent than any E. coli protein, respectively.

10.3. Autoinduction method for protein expression

A disadvantage of induction by IPTG is the requirement to monitor
bacterial growth to achieve optimal induction conditions. This attribute is
especially apparent in experiments which survey expression in large clone
160 Sarah Zerbs et al.

sets. In this scenario, differences in growth rates will be observed and

optimal induction conditions are unlikely to be achieved for all expression
clones. Autoinduction systems provide an approach to circumvent these
difficulties and simplify the expression protocol (Studier, 2005). Several
systems have been described which provide high-level protein expression
with pET and other IPTG-inducible bacterial expression systems (Blommel
et al., 2007).

11. Small-Scale Expression Cultures in

Autoinduction Media Protocol
1. A transformation plate or a fresh streak of a frozen stock may be used for
this experiment. The expression cell line must contain genes for lac
permease (lacY) and ß-galactosidase (lacZ) as well as T7 RNA polymerase.
2. Prepare autoinduction media (e.g., Overnight Express Kit, Novagen)
and add appropriate antibiotic to media.
3. Use a single colony to inoculate 2-mL cultures of autoinduction media.
Use tubes that hold at least 5
the final volume for adequate aeration.
4. Incubate tubes at 37  C and 250 rpm for at least 16 h. Alternatively,
incubate cultures at 25  C and 250 rpm for at least 20 h. The culture
must reach stationary phase for robust induction of expression.
Expression and solubility outcome of individual targets is assessed as
described in the previous section for IPTG-induced cultures. Expression
yield of heterologous proteins is similar for autoinduced and IPTG-induced
cultures but individual variation in target protein solubility may be observed
(Fig. 12.2). In some studies, autoinduction produced 5–20 times as much
target protein per volume of culture as conventional IPTG induction
(Studier, 2005).

12. Periplasmic Expression of Proteins

Approximately 8–12% of the bacterial proteome is not destined for the
cytoplasm. Heterologous expression strategies for this group include ampli-
fication of the nonsignal component of the coding region and processing
these targets through the standard E. coli cloning and expression pipeline
using a cytoplasmic or periplasmic targeting vector with outcomes similar to
those observed for cytoplasmic proteins. This approach is also successful for
some eukaryotic proteins containing disulfide bonds. Proteins are routed to
the E. coli periplasmic space by addition of an appropriate targeting signal
(e.g., the E. coli PelB signal sequence) to the N-terminus of the target
Protein Expression in Bacterial Systems 161

M 1 2 3 4 5 6 7 8 9 10 11 12
Autoinduction

IPTG induction

Figure 12.2 Coomassie-stained gel analysis of target solubility outcome for autoin-
duced and IPTG-induced cultures. Targets were amplified from Shewanella oneidensis
genomic DNA and cloned into the cytoplasmic expression vector pMCSG7 (targets
1–6) or the periplasmic expression vector pBH31 SA (targets 7–12). Soluble fractions
for analysis were prepared as described in the text. Targets are as follows: 1 ¼ SO 3070;
2 ¼ SO 2454; 3 ¼ SO 2444; 4 ¼ SO 1503; 5 ¼ SO 1190; 6 ¼ SO1560; 7 ¼ SO 0809;
8 ¼ SO 0837; 9 ¼ SO 4048; 10 ¼ SO 1503; 11 ¼ SO 1190; 12 ¼ SO 1560.

protein. As the periplasm accounts for 20–40% of the total volume of the
cell, overall expression yields are typically lower when compared to cyto-
plasmic expression. This is apparent from a direct comparison of the expres-
sion levels for targets cloned into a cytoplasmic expression vector (Fig. 12.2,
targets 4–6) and targets cloned into a periplasmic expression vector
(Fig. 12.2, targets 10–12).

13. Expression of Periplasmic Targets in E. coli

1. Use a single colony to inoculate a 2-mL culture of media with appro-
priate antibiotic. Use a culture tube that will hold at least 5
the final
culture volume for sufficient aeration.
162 Sarah Zerbs et al.

2. Incubate samples at 30  C, 250 rpm until the OD600 reaches 0.4–0.8.

The culture should appear cloudy but not completely turbid. In general,
this requires 3–5 h of growth with BL21(DE3) cells.
3. Drop the temperature of the incubator to 19  C and incubate the
cultures for at least 30 min to allow cells to equilibrate. The OD of the
culture should continue to increase during the temperature shift.
4. Add 20 mL 100 mM IPTG to each culture (1 mM final concentration).
Return induced cultures to the incubator set at 19  C, 250 rpm over-
night. The cultures should be turbid after the overnight induction.
5. Cultures should show some expression after 4 h and significant expres-
sion after 12–16 h of induction at 19  C.

14. Small-Scale Osmotic Shock Protocol

If the localization of a target to the periplasm must be confirmed, an
osmotic shock can be performed to analyze the contents of the periplasm
separately from the cytoplasm. Many methods for releasing proteins from
the outer compartment of bacteria are available. Addition of chloroform
(Ames et al., 1984) or Polymyxin B (Dixon and Chopra, 1986) releases
proteins from the periplasm. The method described below is inexpensive
and uses common laboratory reagents (Neu and Heppel, 1965; Nossal and
Heppel, 1966). Lysozyme can be added to the osmotic shock SET (sucrose,
EDTA, Tris–HCl) buffer to remove more of the outer cell wall and
components (Birdsell and Cota-Robles, 1967; Malamy and Horecker,
1964); the spheroplasts made during this procedure are very prone to lysis
and cytoplasmic proteins may contaminate the periplasmic fraction. For
cells that lyse easily the addition of 0.5 mM MgCl2 to the cold water shock
can stabilize the spheroplasts (Neu and Chou, 1967). The method we
describe works well for a majority of proteins targeted to the periplasm
but may need to be optimized for some targets. Comparison of the amount
of target protein in the SET and water fractions in Fig. 12.3 illustrates the
variance obtained with individual proteins. However, the proportion of
soluble heterologous proteins is enriched in the shock fractions compared to
background host proteins (Fig. 12.3). This method can be scaled up as a first
purification step to reduce background proteins and cell debris.
1. Use a 5-mL culture where protein expression has been induced for 4 h
to overnight.
2. Pellet the induced culture at 3500 rpm for 5 min. Decant or aspirate
spent media.
3. Wash cells twice with 5 mL of 10 mM Tris–HCl, pH 8.0. Centrifuge
suspensions at 3500 rpm for 5 min, decant washes after pelleting cells.
A vortexer may be used to resuspend the cell pellet.
Protein Expression in Bacterial Systems 163

SO 1190 SO 0809

H2O

H2O
SET

SET
WC

WC
SP

SP
M

Figure 12.3 An example of osmotic shock fractions for targets from Shewanella
oneidensis cloned into the periplasmic expression vector pBH31 SA. Fractions are
labeled as follows: WC ¼ whole cell expression; SET ¼ SET buffer fraction; H2O ¼
water shock fraction; SP ¼ spheroplast fraction; M ¼ molecular weight markers.

4. Resuspend cells in 1 mL 10 mM Tris–HCl and transfer to a 1.5 mL

microcentrifuge tube. Pellet cells by centrifuging for 30 s at 14,000 rpm
and aspirate all remaining Tris–HCl wash.
5. Resuspend pellet completely in 300 mL SET buffer (50 mM Tris–HCl,
pH 7.5, 20% (w/v) sucrose, 1 mM EDTA). Incubate cells and buffer on
ice for 10 min.
6. Centrifuge samples at 14,000 rpm for 30 s to 1 min. Some preps may
require longer centrifugation times, as the cells do not pellet easily in
SET buffer. Remove a 50 mL portion of supernatant from the top of the
sample to avoid picking up cell debris. Discard remaining liquid, but
retain the cell pellet.
7. Add 300 mL ice-cold sterile water to the pellet and quickly resuspend
cells. Incubate cells in water on ice for 5 min, without shaking.
8. Centrifuge sample at 14,000 rpm for 20–30 s. Immediately remove a
50 mL portion from the top of the supernatant to avoid picking up debris.
9. Steps 6–9 can be repeated with remaining pellet to release additional
material from the periplasm.
10. Add 50 mL 2
SDS loading dye to the SET buffer fraction and H2O
shock fraction and mix well. Boil samples 5 min and cool slightly before
loading on gel. For a 17-well 8
10 cm gel stained with Coomassie,
10 mL of each sample was usually sufficient to visualize bands.
164 Sarah Zerbs et al.

11. [Optional] To examine the spheroplast fractions after the shock proce-
dure remove all remaining liquid from the cell pellet. Add 500 mL
2
SDS loading dye directly to the pellet and resuspend. Boil sample
for 5 min and cool slightly before loading on the gel. For a 17-well,
8
10 cm gel stained with Coomassie, 3 mL was usually sufficient. This
sample can be extremely viscous and may require additional SDS
loading buffer.

15. Alternative Bacterial Systems for

Heterologous Protein Production
A variety of alternative bacterial systems are available for heterologous
protein expression. In many instances, the selection of a bacterial system is
motivated by characteristics of the target proteins. Although E. coli can be
utilized for the heterologous production of multiheme cytochromes, an
alternative host is often preferred for complex cytochromes with multiple
heme groups. Shewanella oneidensis is a Gram-negative bacterium often used
for the production of this particular protein class. The genome of this
organism encodes a large number of predicted cytochrome c genes and
contains accessory proteins that promote correct processing of the apoc-
ytochromes into the mature protein (Takayama and Akutsu, 2007).
A heterologous expression system that uses a genetically engineered Pseudo-
monas fluorescens strain (Pfenex, DOW Chemical Co.) has been developed
for large-batch fermentation as well as high-throughput small-scale screen-
ing of target proteins. This modified Gram-negative organism is metaboli-
cally versatile and contains various microbial secretion systems that enable
production in the cytoplasm, periplasm, or extracellular space.
Protein production in nonpathogenic Gram-positive Bacillus species
provides an alternative to the Gram-negative host strains (Schmidt, 2004).
These organisms contain a naturally efficient secretion system to direct
expressed proteins into the culture supernatant often resulting in improved
yield and ease of purification. The most frequently used species are modified
B. subtilis strains which lack genes for both extracellular lipolytic enzymes
and proteases, resulting in improved stability of heterologous proteins.
Bacillus megaterium has several large plasmids and is known for the stable
replication and maintenance of these plasmids. Expression strains typically
have low intrinsic protease activity and a combination of features that result
in stable high-yield protein expression.
Membrane proteins comprise a significant fraction of the proteins
encoded by the genome but represent a challenge for protein production.
In addition to E. coli (Neophytou et al., 2007; Shaw and Miroux, 2003),
several other membrane protein expression systems have been developed
Protein Expression in Bacterial Systems 165

for production of integral membrane proteins. The Gram-positive lactic

acid bacterium Lactococcus lactis grows to high densities and is suited for
large-scale overproduction of membrane proteins (Kunji et al., 2003).
Several auxotrophic strains are available as well as an inducible expression
system regulated by polycyclic peptide nisin. Functional screens for the
characterization of the membrane protein can be performed with whole
cells because ligands can act directly on the cytoplasmic membrane in which
the membrane proteins are expressed.
The photosynthetic bacterium Rhodobacter has been engineered for
heterologous production of full-length membrane proteins (Laible et al.,
2004). This system is unique in that it incorporates foreign membrane
proteins into its own system of intracytoplasmic membranes by synthesizing
new membrane coordinately with the expression of foreign target mem-
brane proteins. These new membranes form as invaginations of the cyto-
plasmic membrane and allow integration of heterologous membrane
proteins. One of the key strengths in using the Rhodobacter membrane
expression system is that proteins produced are localized to the membranes
nearly quantitatively in all cases. Membrane protein expression and purifi-
cation is challenging and the functionality of proteins produced in heterol-
ogous expression systems must be carefully validated.

16. Alternative Vector and Induction

Conditions
Aside from employing different bacterial hosts, the heterologous
expression system may also be optimized by manipulating vector and
induction conditions. One such method that has demonstrated recent
success in protein expression and solubility is a cold-shock expression
system using a pCold vector (Qing et al., 2004). This system uses the
knowledge of E. coli cell cold-shock response to inhibit endogenous protein
production while enhancing target protein expression. It has been suggested
that when the cspA mRNA is truncated and expressed under cold-shock
conditions (37  C dropped to 15  C for 36 h), polysomes are occupied with
translation of the truncated cspA gene and cannot adapt to ribosomal form
III to translate non-cold-shock proteins ( Jiang et al., 1996). Consequently,
only the genes in the cspA mRNA will be translated, while endogenous
protein expression is inhibited. The pCold vectors include a cloning region
after the cspA promoter for a target gene insert that is overexpressed upon
15  C cold-shock induction and derepression with 1 mM IPTG. Due to the
minimal level of expressed host background proteins, this technology often
eliminates the need to purify the heterologously expressed protein, greatly
cutting production costs.
166 Sarah Zerbs et al.

17. Production Scale

The goal of most bacterial expression projects is to produce large
amounts of soluble, correctly folded and active proteins. In general, the
expression and solubility levels of a target protein decrease in larger cultures.
However, targets that perform well in small-scale experiments are more
likely to produce satisfactory results during scale-up (Moy et al., 2004).
A common approach to improve the final production yield is to select
several of the most soluble constructs obtained from the pilot studies for
evaluation in large-scale production processes. The second component of
the strategy is optimization of growth and culture parameters of the large-
scale process to improve the quality and yield of the final product. The
ability to generate and screen multiple bacterial expression products in a
timely manner contributes to the utility of bacterial expression systems for
protein production.

ACKNOWLEDGMENT
The submitted manuscript has been created by UChicago Argonne, LLC, Operator of
Argonne National Laboratory (‘‘Argonne’’). Argonne, a U.S. Department of Energy Office
of Science laboratory, is operated under Contract No. DE-AC02-06CH11357. The U.S.
Government retains for itself, and others acting on its behalf, a paid-up nonexclusive,
irrevocable worldwide license in said article to reproduce, prepare derivative works, distrib-
ute copies to the public, and perform publicly and display publicly, by or on behalf of the
Government.

REFERENCES
Ames, G. F., Prody, C., and Kustu, S. (1984). Simple, rapid, and quantitative release of
periplasmic proteins by chloroform. J. Bacteriol. 160, 1181–1183.
Aslanidis, C., and de Jong, P. J. (1990). Ligation-independent cloning of PCR products
(LIC-PCR). Nucleic Acids Res. 18, 6069–6074.
Baneyx, F. (1999). Recombinant protein expression in Escherichia coli. Curr. Opin. Bio-
technol. 10, 411–421.
Baneyx, F., and Palumbo, J. L. (2003). Improving heterologous protein folding via molecu-
lar chaperone and foldase co-expression. Methods Mol. Biol. 205, 171–197.
Birdsell, D. C., and Cota-Robles, E. H. (1967). Production and ultrastructure of lysozyme
and ethylenediaminetetraacetate-lysozyme spheroplasts of Escherichia coli. J. Bacteriol.
93, 427–437.
Boyer, H. W., and Roulland-Dussoix, D. (1969). A complementation analysis of the
restriction and modification of DNA in Escherichia coli. J. Mol. Biol. 41, 459–472.
Carstens, C. P. (2003). Use of tRNA-supplemented host strains for expression of heterolo-
gous genes in E. coli. Methods Mol. Biol. 205, 225–233.
Protein Expression in Bacterial Systems 167

De Marco, V., Stier, G., Blandin, S., and de Marco, A. (2004). The solubility and stability of
recombinant proteins are increased by their fusion to NusA. Biochem. Biophys. Res.
Commun. 322, 766–771.
Derman, A. I., Prinz, W. A., Belin, D., and Beckwith, J. (1993). Mutations that allow
disulfide bond formation in the cytoplasm of Escherichia coli. Science 262, 1744–1747.
Dieckman, L., Gu, M., Stols, L., Donnelly, M. I., and Collart, F. R. (2002). High
throughput methods for gene cloning and expression. Protein Expr. Purif. 25, 1–7.
Dieckman, L. J., Hanly, W. C., and Collart, F. R. (2006). Strategies for high-throughput
gene cloning and expression. Genet. Eng. (NY) 27, 179–190.
Dixon, R. A., and Chopra, I. (1986). Leakage of periplasmic proteins from Escherichia coli
mediated by polymyxin B nonapeptide. Antimicrob. Agents Chemother. 29, 781–788.
Donnelly, M. I., Zhou, M., Millard, C. S., Clancy, S., Stols, L., Eschenfeldt, W. H.,
Collart, F. R., and Joachimiak, A. (2006). An expression vector tailored for large-scale,
high-throughput purification of recombinant proteins. Protein Expr. Purif. 47, 446–454.
Dyson, M. R., Shadbolt, S. P., Vincent, K. J., Perera, R. L., and McCafferty, J. (2004).
Production of soluble mammalian proteins in Escherichia coli: Identification of protein
features that correlate with successful expression. BMC Biotechnol. 4, 32.
Eschenfeldt, W. H., Lucy, S., Millard, C. S., Joachimiak, A., and Mark, I. D. (2009).
A family of LIC vectors for high-throughput cloning and purification of proteins.
Methods Mol. Biol. 498, 105–115.
Esposito, D., Garvey, L. A., and Chakiath, C. S. (2009). Gateway cloning for protein
expression. Methods Mol. Biol. 498, 31–54.
Guzman, L. M., Belin, D., Carson, M. J., and Beckwith, J. (1995). Tight regulation,
modulation, and high-level expression by vectors containing the arabinose PBAD pro-
moter. J. Bacteriol. 177, 4121–4130.
Hartley, J. L., Temple, G. F., and Brasch, M. A. (2000). DNA cloning using in vitro site-
specific recombination. Genome Res. 10, 1788–1795.
Haun, R. S., Serventi, I. M., and Moss, J. (1992). Rapid, reliable ligation-independent
cloning of PCR products using modified plasmid vectors. Biotechniques 13, 515–518.
Hengen, P. (1995). Purification of His-Tag fusion proteins from Escherichia coli. Trends
Biochem Sci. 20, 285–286.
Jiang, W., Fang, L., and Inouye, M. (1996). The role of the 5’-end untranslated region of the
mRNA for CspA, the major cold-shock protein of Escherichia coli, in cold-shock
adaptation. J. Bacteriol. 178, 4919–4925.
Kapust, R. B., and Waugh, D. S. (1999). Escherichia coli maltose-binding protein is
uncommonly effective at promoting the solubility of polypeptides to which it is fused.
Protein Sci. 8, 1668–1674.
Klock, H. E., White, A., Koesema, E., and Lesley, S. A. (2005). Methods and results for
semi-automated cloning using integrated robotics. J. Struct. Funct. Genomics 6, 89–94.
Kunji, E. R., Slotboom, D. J., and Poolman, B. (2003). Lactococcus lactis as host for
overproduction of functional membrane proteins. Biochim. Biophys. Acta 1610, 97–108.
Laible, P. D., Scott, H. N., Henry, L., and Hanson, D. K. (2004). Towards higher-
throughput membrane protein production for structural genomics initiatives. J. Struct.
Funct. Genomics 5, 167–172.
Londer, Y. Y., Giuliani, S. E., Peppler, T., and Collart, F. R. (2008). Addressing Shewanella
oneidensis ‘‘cytochromome’’: The first step towards high-throughput expression of cyto-
chromes c. Protein Expr. Purif. 62, 128–137.
Makrides, S. C. (1996). Strategies for achieving high-level expression of genes in Escherichia
coli. Microbiol. Rev. 60, 512–538.
Malamy, M. H., and Horecker, B. L. (1964). Release of alkaline phosphatase from cells of
Escherichia coli upon lysozyme spheroplast formation. Biochemistry 3, 1889–1893.
168 Sarah Zerbs et al.

Matthey, B., Engert, A., Klimka, A., Diehl, V., and Barth, S. (1999). A new series of pET-
derived vectors for high efficiency expression of Pseudomonas exotoxin-based fusion
proteins. Gene 229, 145–153.
Miroux, B., and Walker, J. E. (1996). Over-production of proteins in Escherichia coli: Mutant
hosts that allow synthesis of some membrane proteins and globular proteins at high levels.
J. Mol. Biol. 260, 289–298.
Moy, S., Dieckman, L., Schiffer, M., Maltsev, N., Yu, G. X., and Collart, A. F. (2004).
Genome-scale expression of proteins from Bacillus subtilis. J. Struct. Funct. Genomics 5,
103–109.
Neophytou, I., Harvey, R., Lawrence, J., Marsh, P., Panaretou, B., and Barlow, D. (2007).
Eukaryotic integral membrane protein expression utilizing the Escherichia coli
glycerol-conducting channel protein (GlpF). Appl. Microbiol. Biotechnol. 77, 375–381.
Neu, H. C., and Chou, J. (1967). Release of surface enzymes in Enterobacteriaceae by
osmotic shock. J. Bacteriol. 94, 1934–1945.
Neu, H. C., and Heppel, L. A. (1965). The release of enzymes from Escherichia coli by
osmotic shock and during the formation of spheroplasts. J. Biol. Chem. 240, 3685–3692.
Nossal, N. G., and Heppel, L. A. (1966). The release of enzymes by osmotic shock from
Escherichia coli in exponential phase. J. Biol. Chem. 241, 3055–3062.
Prinz, W. A., Aslund, F., Holmgren, A., and Beckwith, J. (1997). The role of the thior-
edoxin and glutaredoxin pathways in reducing protein disulfide bonds in the Escherichia
coli cytoplasm. J. Biol. Chem. 272, 15661–15667.
Qing, G., Ma, L. C., Khorchid, A., Swapna, G. V., Mal, T. K., Takayama, M. M., Xia, B.,
Phadtare, S., Ke, H., Acton, T., Montelione, G. T., Ikura, M., et al. (2004). Cold-shock
induced high-yield protein production in Escherichia coli. Nat. Biotechnol. 22, 877–882.
Sambrook, J., and Russell, D. W. (2001). Molecular cloning: A laboratory manual. Cold
Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Schmidt, F. R. (2004). Recombinant expression systems in the pharmaceutical industry.
Appl. Microbiol. Biotechnol. 65, 363–372.
Shaw, A. Z., and Miroux, B. (2003). A general approach for heterologous membrane protein
expression in Escherichia coli: the uncoupling protein, UCP1, as an example. Methods
Mol. Biol. 228, 23–35.
Smith, D. B., and Johnson, K. S. (1988). Single-step purification of polypeptides expressed in
Escherichia coli as fusions with glutathione S-transferase. Gene. 67, 31–40.
Studier, F. W. (2005). Protein production by auto-induction in high density shaking
cultures. Protein Expr. Purif. 41, 207–234.
Studier, F. W., and Moffatt, B. A. (1986). Use of bacteriophage T7 RNA polymerase to
direct selective high-level expression of cloned genes. J. Mol. Biol. 189, 113–130.
Takayama, Y., and Akutsu, H. (2007). Expression in periplasmic space of Shewanella
oneidensis. Protein Expr. Purif. 56, 80–84.
Terpe, K. (2006). Overview of bacterial expression systems for heterologous protein pro-
duction: from molecular and biochemical fundamentals to commercial systems. Appl.
Microbiol. Biotechnol. 72, 211–222.
Woodcock, D. M., Crowther, P. J., Doherty, J., Jefferson, S., DeCruz, E., Noyer-
Weidner, M., Smith, S. S., Michael, M. Z., and Graham, M. W. (1989). Quantitative
evaluation of Escherichia coli host strains for tolerance to cytosine methylation in plasmid
and phage recombinants. Nucleic Acids Res. 17, 3469–3478.
Yokoyama, S. (2003). Protein expression systems for structural genomics and proteomics.
Curr. Opin. Chem. Biol. 7, 39–43.
Zhu, B., Cai, G., Hall, E. O., and Freeman, G. J. (2007). In-fusion assembly: Seamless
engineering of multidomain fusion proteins, modular vectors, and mutations. Biotechniques
43, 354–359.
 C H A P T E R T H I R T E E N

Expression in the Yeast

Pichia pastoris
James M. Cregg,*,† Ilya Tolstorukov,*,† Anasua Kusari,*
Jay Sunga,* Knut Madden,† and Thomas Chappell†

Contents
1.Introduction 170
2.Other Fungal Expression Systems 170
3.Culture Media and Microbial Manipulation Techniques 171
4.Genetic Strain Construction 172
4.1. Mating, creating diploids 172
4.2. Random spore analysis 173
5. Gene Preparation and Vector Selection 174
6. Transformation by Electroporation 176
7. DNA Preparation 176
8. Examination of Strains for Recombinant Protein Production 178
9. Assay Development—The Yeastern Blot 182
10. Posttranslational Modification of the Recombinant Protein
(Proteinases and Glycosylation) 184
11. Selection for Multiple Copies of an Expression Cassette 185
References 187

Abstract
The yeast Pichia pastoris has become the premier example of yeast species
used for the production of recombinant proteins. Advantages of this yeast for
expression include tightly regulated and efficient promoters and a strong
tendency for respiratory growth as opposed to fermentative growth. This chapter
assumes the reader is proficient in molecular biology and details the more
yeast specific procedures involved in utilizing the P. pastoris system for gene
expression. Procedures to be found here include: strain construction by classical
yeast genetics, the logic in selection of a vector and strain, preparation of
electrocompetent yeast cells and transformation by electroporation, and the
yeast colony western blot or Yeastern blot method for visualizing secreted
proteins around yeast colonies.

* Keck Graduate Institute of Applied Life Sciences, Claremont, California, USA

{
Biogrammatics, Inc., Carlsbad, California, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63013-5 All rights reserved.

169
170 James M. Cregg et al.

1. Introduction
Relative to other expression systems, yeast got off to a slow start in the
early 1980s, primarily due to poor results with several proteins in baker’s
yeast Saccharomyces cerevisiae (Romanos et al., 1992). Promising results in
shake flask culture, more often than not, led to disappointing yields when
scaled up in fermentor cultures. In addition, yeast were not likely to be of
value in producing human proteins containing N-linked carbohydrates as
injectable pharmaceutical drugs, a major goal of many biotechnology and
pharmaceutical companies, because their sugars were of a high mannose
type in composition and configuration, quite different from that of humans.
As a result, recombinant glycoproteins made in a yeast system have a typical
fungal-like N-glycosylation pattern, which a human immune system recog-
nizes as foreign and rejects. This results in the rapid clearance of yeast
products from the blood and a strong immune response from a patient
that can possibly result in death. Since the 1980s, these problems have been
addressed and several yeasts have become productive alternative systems for
recombinant protein production. As a eukaryotic microbial expression
system, yeast are a good alternative for proteins for which expression in a
bacterial system leads to the synthesis of improperly folded, and inactive
protein aggregates or inclusion bodies.
This review will focus primarily on the most popular of these new yeast
expression systems, Pichia pastoris. Only details of procedures that are specific
or peculiar to expression in P. pastoris will be covered; for more common
methods (e.g., agarose gel electrophoresis, immunoblotting, general recom-
binant DNA methodology, etc.), readers are referred to the many excellent
books describing these methods including: (Sambrook et al., 1989) or the
series by (Ausubel et al., 2001).

2. Other Fungal Expression Systems

In addition to P. pastoris, several other yeast species have been devel-
oped for expression; among them, Hansenula polymorpha, Pichia methanolica,
Kluyveromyces lactis, Arxula adeninivorans, and Yarrowia lipolytica are the best
known and developed (Gellissen, 2005). Many of the reasons for using one
of these alternative yeast expression systems, as well as the methods needed
to construct recombinant strains, are similar to each other and to P. pastoris.
Like P. pastoris, H. polymorpha and P. methanolica are methylotrophic
yeast and virtually all of the advantages cited for P. pastoris are also true for
these other species (Gellissen, 2002). In particular, all three have the potent
promoter regulating the expression of the alcohol oxidase gene from their
Expression in the Yeast Pichia pastoris 171

respective species available for controlling recombinant protein expression.

However, both H. polymorpha and P. methanolica are different from
P. pastoris in that the construction of expression strains is often somewhat
more laborious. With H. polymorpha, this extra labor can be rewarded with
strains harboring 100 or more copies of an expression cassette. In addition,
H. polymorpha can readily grow at temperatures up to 50
C providing the
potential of significantly decreasing bioprocess times and thereby reducing
costs.
K. lactis was the first yeast species after S. cerevisiae to be developed for
recombinant protein expression and is well known for its use in production
of rennin for cheese processing (van Ooyen et al., 2006). An advantage of
K. lactis for some purposes is that, like S. cerevisiae, it is officially on the
generally recognized as safe (‘‘GRAS’’) list of microorganisms.
A. adeninivorans is a dimorphic yeast with advantages for secreting recom-
binant proteins (Boer et al., 2005). This yeast has both a vegetative yeast like
growth state and a mycelial growth state in which the cells send out mycelia
like filamentous fungi. During the mycelial growth phase, the synthesis
of secreted proteins is enhanced relative to its vegetative growth phase.
In addition, A. adeninivorans does not O-glycosylate secreted proteins in its
mycelial growth phase. Finally, A. adenininvorans has a relatively high growth
temperature of up to 48
C and osmotolerance up to 3.4 M (10%) NaCl.
Y. lipolytica, like A. adeninivorans is a dimorphic yeast (Madzak et al.,
2005). Most recombinant genes are expressed in Y. lipolytica off the XPR2
promoter which, in its native state, expresses the gene for alkaline extracellular
protease. However, several other promoters, all of which are constitutive, are
also available for the yeast.
Finally, certain filamentous fungi, such as Neurospora crassa, various
Aspergillus species and Sordaria macrospora, have proved to be effective
expression systems for certain recombinant products, particularly secreted
proteins. However, the techniques for dealing with filamentous fungi
are very different from yeast and will not be dealt with here. Readers
interested in expression in filamentous fungi are referred to one of several
excellent reviews on these systems (Heerikhuisen et al., 2005; Kuck and
Poggeler, 2005).

3. Culture Media and Microbial

Manipulation Techniques
Techniques for culturing P. pastoris (and most other yeast species) at
the bench level are identical to those used for Escherichia coli and S. cerevisiae.
The most common rich medium for cultivation is YPD (1% yeast extract,
172 James M. Cregg et al.

2% peptone, 2% dextrose) and defined medium is YNB (0.67% yeast

nitrogen base with ammonium sulfate and without amino acids, 2% dex-
trose plus any amino acids or nucleotides required for growth at  50 ug/ml
each). Growth of P. pastoris on methanol requires that the dextrose is
replaced with methanol to 0.5%. The recipe for mating/sporulation
medium is 0.5% sodium acetate, 1% KCl 1% glucose. For Petri plates the
media are prepared with 2% agar. Incubations are typically done at 30
C. In
liquid YPD, P. pastoris has a generation time of approximately 90 min and a
generation time of approximately 3 h in defined medium. With methanol as
sole carbon source and a defined culture medium, the generation time is
around 5 h.

4. Genetic Strain Construction

Although a number of markers of various types exist for P. pastoris, the
right combination for your purposes may not exist. Therefore, it may be
necessary to construct a new host strain with the optimal set of genetic
markers. The first step in strain construction is the mating and selection of
diploid strains (Tolstorukov and Cregg, 2008). Because P. pastoris is func-
tionally homothallic, the mating type of a strain is not a consideration in
planning a genetic cross as cells of the same strain will also mate; However,
the mating efficiency between P. pastoris cells is low. Therefore, it is
essential that strains to be crossed contain complementary markers that
allow for selective growth of crossed diploids, and against the growth of
self-mated diploids and parental strains. Auxotrophic markers are generally
most convenient for this purpose, but mutations in any gene that affect the
growth or other phenotype of P. pastoris such as genes required for utiliza-
tion of methanol or a nitrogen source (e.g., methylamine) can be used as
well. Following are the steps needed to mate P. pastoris:

4.1. Mating, creating diploids

1. To begin a mating experiment, select a fresh colony from each strain to
be mated from YPD plate (no more than one week old) and streak each
across the length of an independent YPD plate.
2. After overnight incubation at 30
C, transfer the cell streaks from both
plates onto a single sterile velvet such that the streaks from one plate are
perpendicular to those on the other.
3. Transfer the cross streaks from the velvet to a mating/sporulation plate
and incubate for 2–3 days at room temperature to initiate mating.
4. After incubation, replica plate to an appropriate agar medium for the
selection of complementing diploid cells. Diploid colonies will grow at
Expression in the Yeast Pichia pastoris 173

the junctions of the streaks after approximately 2–3 days of incubation at

30
C. Diploid P. pastoris cells are approximately twice as large as haploid
cells and easily distinguished by examination under a light microscope by
their size and their high efficiency of sporulation.
5. Purify diploid colonies by streaking at least once for single colonies on
diploid selection medium.
6. Diploid P. pastoris strains efficiently undergo meiosis and sporulation in
response to nitrogen limitation. To initiate this phase of the life cycle,
transfer freshly grown diploid colonies from a YPD or YNB plus glucose
plate to a mating/sporulation plate either by replica-plating or with an
inoculation loop; incubate the plate for 3–4 days at room temperature.
Sporulation in all Pichia species correlates with accumulation of a red
pigment in the ascus; therefore, sporulated diploid samples are easily
distinguished by their tan color relative to the white color of haploid
cultures. Diploids can also be distinguished by a high number of asci in
the cell culture as observed by normal or phase-contrast microscopy.

4.2. Random spore analysis

P. pastoris spores are small and adhere to one another making tetrad dissec-
tion via micromanipulation difficult. Therefore, spore products are analyzed
using a random spore analysis (RSA), as follows:
1. Transfer an inoculation loop full of sporulated P. pastoris cells (from Step 6
above) to a 1.5 ml microcentrifuge tube containing 0.25 ml of sterile
water, vortex the mixture.
2. In a fume hood, add 0.25 ml of diethyl ether to the spore preparation and
vortex thoroughly for approximately 5 min at room temperature. The
ether treatment selectively kills vegetative cells remaining in spore
preparations.
3. While still in the hood, centrifuge the cells for 2 min, remove the ether
(upper phase) and resuspend the pellet in the remaining water. Spread
10 and 100 ul of the suspension onto a nonselective YPD plate.
4. After 4–5 days incubation at 30
C, streak out single colonies onto a
fresh YPD plate as a master plate for further analysis.
5. Replica plate the master plate onto a set of plates containing suitable
diagnostic media. Alternatively/additionally, the initial plates with
100–600 colonies can also be replica plated. For example, spore products
from a his4 and arg4 cross would be analyzed on YNB plus glucose
supplemented with:
a. No amino acids
b. Arginine
c. Histidine
d. Arginine and Histidine
174 James M. Cregg et al.

6. Compare the phenotype of individual colonies on each of the diagnostic

plates to identify strains with the desired phenotype(s).
7. Select several colonies that appear to have the appropriate phenotype
and streak for single colonies onto a nonselective medium plate such as
YPD; then, retest a single colony from each streak on the same set of
diagnostic plates. This step is important since P. pastoris spores adhere
tightly to one another and colonies resulting from spore germination
frequently contain cells derived from more than one spore. Another
consequence of spore clumping is that markers appear not to segregate
1:1 but to be biased toward the dominant or wild-type phenotype. For
example, in ahis4  arg4 cross as described above, more HisþArgþ spore
products will be apparent than the 25% expected in the population and
HisArg spore products will appear to be underrepresented.

5. Gene Preparation and Vector Selection

The first consideration in the selection of a P. pastoris expression vector
is whether you intend to secrete a protein product or produce it intracellu-
larly. A general rule of thumb is to produce a recombinant protein in the
same way it is expressed in its native host: if a protein is produced intracel-
lularly by its native host, one should also produce it intracellularly in the
yeast host; if the protein is secreted from its native host, secrete it from the
yeast system. Although there have been exceptions to this general rule, it is
generally best to follow it since the intracellular and secretory environments
are very different from each other and synthesis of a protein in the wrong
compartment may result in a protein that is improperly folded and inactive.
A number of vectors have been constructed for P. pastoris expression; a list
and detailed discussion of these vectors can be found in (Lin-Cereghino and
Lin Cereghino, 2008) and at http://www.biogrammatics.com (Fig. 13.1).
To clone a gene into a P. pastoris expression vector, an available template
can be amplified with appropriate primers by the polymerase chain reaction
(PCR), or the gene can be synthesized. De novo synthesis can facilitate cases
where DNA optimization or gene modification is required. In any case,
suitable restriction sites at the termini can be generated to facilitate the
cloning. For example, an EcoRI site can be added to the 50 (ATG-contain-
ing) end of the gene and a NotI site to the 30 end of the gene, to facilitate
cloning into the pPICZ series of vectors (Invitrogen, Carlsbad, CA), as long
as sites for these enzymes do not exist within the sequence of the gene. For
P. pastoris expression vectors from Biogrammatics, Inc., an appropriate type
IIS restriction enzyme site and ‘‘seamless’’ cloning sequences can be added
to flank a gene for cloning into an optimal expression context.
Expression in the Yeast Pichia pastoris 175

A Bsa I Bsa I AOX1 TT

Alpha pre- pro-

A. gossypii TEF promoter

AOX1 promoter
Pme I
ZeoR

pJAZ-aMF

pMB origin
ampR

Sfi I

B GGCT ORF
ATTA

TTGGACAAGAGAGAGGCTGA TAATCAAGAGGATGT
AACCTGTTCTCTCTCCGACTCCGA GTTCTCCTACA
L D K R E A E A

Figure 13.1 Map of Biogrammatics expression vector, pJAZ-aMF.

Cloning a gene whose product is to be secreted can be a little trickier.

Extracellular, or secreted expression using a genes native secretion signal
sequence will follow the same procedures as above for intracellular expres-
sion; however, several options exist when using a foreign signal sequence,
such as the S. cerevisiae alpha mating factor secretion signal sequence (aMF),
encoded in the expression vector. In this case, the subcloning procedure
must place the gene of interest in frame with the aMF codons. This aMF
secretion signal is very commonly utilized for secretion because it has
proved to be very good at secreting recombinant proteins of many types.
Although the recombinant protein may be successfully secreted using the
aMF signal, proper processing for the aMF at the NH2-terminus of
the desired protein may not occur and modifications of the aMF signal,
or the use of an alternative secretion signal sequence, may ultimately be
necessary to obtain a properly processed protein. In this regard, the aMF
signal sequence, may include two Glu–Ala repeats at the junction between
the signal peptide and the NH2-terminus of the mature protein of interest.
The Biogrammatics vector, pJAZ-aMF, is designed for making a construction
176 James M. Cregg et al.

with these Glu–Ala repeats (Fig. 13.1). Another set of Alternative Biogram-
matics vectors such as pJAZ-aMF-KR do not contain the Glu–Ala repeats in
the cloning site. To utilize the aMF signal in a pPICZ alpha vector one must
add sequences to the 50 of the gene such that when ligated to the portion of
aMF present in the vector it results in the reconstruction of a functional aMF
signal sequence (either with or without the sequences encoding the Glu–Ala
repeats in aMF). Typically, XhoI is used to cut the pPICZ vector which cuts it
inside the aMF signal encoding sequences. To restore these sequences, an
oligonucleotide with the sequence:
XhoI
TCT CTC GAG AAA AGA GAG GCT GAA GCT-ABC-DEF. . ..
Ser Leu Glu Lys Arg Glu Ala Glu Ala
is synthesized where ‘‘ABC DEF. . .’’ denotes the nucleotide sequences
encoding the first amino acids of the mature recombinant protein. This
oligonucleotide will hybridize appropriately to the 50 end of the gene
encoding the mature protein and result in the incorporation of the missing
portion of the aMF sequence. Similarly, the ‘‘seamless’’ cloning scheme
used to clone genes into the Biogrammatics vectors utilize type IIS restric-
tion sites to join the ABC DEF nucleotides of a gene of interest to the last
Alanine in the aMF by creating a four base ‘‘sticky’’ end comprising the last
nucleotide in a Glu codon and an entire Ala codon (Fig. 13.1).

6. Transformation by Electroporation
At least four different procedures to introduce foreign plasmid DNA
into P. pastoris have been developed using: spheroplast-generation, LiCl,
polyethelene glycol1000 and electroporation. The electroporation proce-
dure is most commonly used and therefore a modified version of that
described by (Becker and Guarante, 1991) will be outlined in detail. For
the other procedures, readers are referred to either of the volumes of
Methods in Molecular Biology: Pichia Protocols (Cregg, 2008; Higgins
and Cregg, 1998).

7. DNA Preparation
For all transformation methods, linear plasmid DNA is most commonly
transformed into P. pastoris for integration into the yeast genome. The DNA
sequence at the ends of the linear plasmid DNA stimulate integration by a
single crossover recombination event into the locus shared by vector and host
genome. Therefore, linearization of an expression plasmid is performed in
Expression in the Yeast Pichia pastoris 177

Pichia DNA in the plasmid, such as in the promoter (i.e., a PmeI site in the
AOXI promoter). The final vector, prepared in E. coli, is cut with a restriction
enzyme that linearizes the vector, and then the DNA is purified and concen-
trated to at least 100 ng/ul in water prior to transformation. At this point, the
vector is ready for transformation into P. pastoris.
Procedure for preparation of electrocompetent cells (Lin-Cereghino
et al., 2005).
Prepare the following (all solutions should be autoclaved except for the
DTT and HEPES solutions, which should be filter sterilized):
1. 500 ml liquid YPD media in a 2.8 l Fernbach culture shaking flask.
2. H2O (1 l).
3. 1 M sorbitol (100 ml).
4. Appropriate selective agar plates.
5. 1 M DTT (2.5 ml).
6. BEDS solution (9 ml): 10 mM Bicine–NaOH, pH 8.3, 3% ethylene
glycol, 5% DMSO (dimethyl sulfoxide), 1 M sorbitol and 0.1 M DTT.
7. 1 M HEPES buffer, pH 8.0 (50 ml).
8. Sterile 250 ml centrifuge tubes.
9. Sterile electroporation cuvettes.
10. Electroporation instrument: BTX Electro Cell Manipulator 600 (BTX,
San Diego, CA); Bio-Rad Gene Pulser (Bio-Rad, Hercules, CA);
Electroporator II (Invitrogen, San Diego, CA). Parameters for electro-
poration with different instruments vary with the instrument. Be sure
to check instructions for each type of instrument. (Becker and
Guarante, 1991; Grey and Brendel, 1995; Pichia Expression Kit
Instruction Manual; Stowers et al., 1995).
Protocol:
1. Inoculate 10 ml YPD media with a single fresh P. pastoris colony of the
strain to be transformed from an agar plate and grow overnight shaking at
30
C.
2. Use the overnight culture to inoculate a 500 ml YPD culture in a 2.8 l
Fernbach culture flask to a starting OD600 of 0.01 and grow to an OD600 of
1.0 ( 12 h).
3. Harvest the cells by centrifugation at 2000g at 4
C, discard the superna-
tant and resuspend the cells in 100 ml of fresh YPD medium plus HEPES
(pH 8.0, 200 mM ) in a sterile 250 ml centrifuge tube.
4. Add 2.5 ml of 1 M DTT and gently mix.
5. Incubate at 30
C for 15 min with slow rotating.
6. Add 150 ml cold water to the culture and wash by centrifugation at 4
C
with an additional 250 ml of cold water. At this stage and from here on,
keep the cells ice cold and do not vortex the cells to resuspend them
(slow pippeting is best).
178 James M. Cregg et al.

7. Wash cells a final time in 20 ml of cold 1 M sorbitol, then resuspend in 0.5 ml

of cold 1 M sorbitol (final volume including cells will be 1.0–1.5 mls).
8. Use these cells directly without freezing to achieve the most transformants.
9. To freeze competent cells, distribute in 40 ul aliquots to sterile 1.5 ml
minicentrifuge tubes and place the tubes in a  70
C freezer.
Electroporation procedure:
1. Add up to 1 ug of linearized plasmid DNA sample in no more than 5 ul
of water to a tube containing 40 ul of frozen or fresh competent cells and
then transfer the entire mixture to a 2 mm gap electroporation cuvette
held on ice.
2. Pulse cells according to the parameters suggested for yeast by the manu-
facturer of the specific electroporation instrument (Table 13.1).
3. Immediately add 0.5 ml of cold 1 M sorbitol and 0.5 ml of cold YPD,
then transfer the entire cuvette contents to a sterile 1.5–2.0 ml mini-
centrifuge tube.
4. Incubate for 3.5–4 h at 30
C with slow shaking (100 rpm).
5. Spread aliquots onto selective agar plates and incubate for 2–4 days.
6. To avoid mixed colonies, pick and restreak transformants on selective
medium at least once before proceeding with further analysis.
Rapid preparation of electrocompetent P. pastoris:
1. Grow a 5 ml culture of P. pastoris in YPD overnight, shaking at 30
C.
2. Dilute the overnight culture to an OD600 of 0.15–0.20 in 50 ml YPD
medium in a flask large enough to provide good aeration.
3. Grow to an OD600 of 0.8–1.0 at 30
C with shaking (4–5 h).
4. Centrifuge cells at 500g for 5 min at room temperature and discard
supernatant.
5. Resuspend cells in 9 ml of ice-cold BEDS solution supplemented with
DTT.
6. Incubate the cell suspension for 5 min, shaking at 30
C.
7. Centrifuge cells at 500g for 5 min at room temperature and resuspend in
1 ml of BEDS (without DTT).
8. Perform electroporation as described above, immediately or freeze cells
in small aliquots at  80
C.

8. Examination of Strains for Recombinant

Protein Production
Yeast expression systems have been successful at generating large
quantities of recombinant proteins. For example, production levels of
between 1 and 10 g/l of culture supernatant have been secreted from
Table 13.1 Parameters for electroporation using selected instruments

Cuvette Sample Charging Field Pulse

gap volume voltage Capacitance Resistance strength length
Instrument (mm) (ul) (V) (uF) (O) (kV/cm) ( ms) References
ECM600 (BTX) 2 40 1500 Out 129 7500 5 Becker and
Guarante
(1991)
Electroporator II 2 80 1500 50 200 7500 10 Pichia Manual
(Invitrogen)
Gene-Pulser 2 40 1500 25 200 7500 5 Grey and Brendel
(Bio-Rad) (1995)
Cell-Porator 1.5 20 480 10 Low 2670 NS Lorow-Murray
(BRL) and Jesse (1991)
and Stowers
et al. (1995)
180 James M. Cregg et al.

P. pastoris strains. However, one should not expect to see a band on a

coomassie-stained gel in initial expression experiments. Thus, it is impera-
tive that by the time a recombinant strain is ready to begin expression
studies, one or more sensitive assays for the detection of the targeted protein
is in place. Assays can be based on enzyme activity, an epitope tag fused to
ones product or an antibody against the desired gene product; however, the
point is: serious efforts to develop sensitive detection methods should begin
at the same time or even prior to expression studies. Good assays facilitate
strain development as illustrated in the following examples.
The most convenient way to detect foreign protein expression in yeast is
via a plate activity assay. Plate assays allow one to crudely quantify and
compare productivity of 100s to 1000s of transformants directly on diag-
nostic plates at a single glance using replica-plating or other techniques.
An example of a plate activity assay is shown in Fig. 13.2 for secretion of
the enzyme phytase. This enzyme degrades phytate which results in the
clearing of a zone around the expressing colony on agar plates. The size of
the zone roughly correlates to the amount of phytase being secreted. As a
second example, the expression of bacterial b-lactamase as an intracellular
protein is shown in P. pastoris (Fig. 13.3). The colonies, which are typically
yellow in color, turn pink to purple with the expression of b-lactamase.
The intensity of the purple color roughly indicates the amount of enzyme
expressed and, in this case, the number of copies of the expression cassette in

Figure 13.2 Plate assay for detection of phytase constitutively expressed and secreted
by selected P. pastoris strains. Top spot: negative control; remaining spots show five
transformants secreting various amounts of the enzyme.
Expression in the Yeast Pichia pastoris 181

Figure 13.3 Expression of intracellular b-lactamase in P. pastoris. The two spots at the
top of the plate are non-b-lactamase expressing negative control strains.

each strain. Finally, with good quality polyclonal or monoclonal antibodies

against a tag region or the recombinant protein itself, a plate antibody assay
or ‘‘Yeastern Blot’’ is possible, as described below.
Once colonies of transformed cells have been selected, single cell pur-
ified and collected onto a master plate, samples of multiple strains are
examined for expression of the foreign protein. Before analysis of expres-
sion, one can screen for the presence of the recombinant gene in transfor-
mants by PCR. Simply prepare genomic DNA in a cell-free extract by glass
bead disruption as described below and utilize it as template in a PCR
reaction with primers that are complementary to a portion of the recombi-
nant gene. Different methods of detection for the expression of proteins in
P. pastoris can be applied depending on what kind of promoter is used for
expression (inducible or constitutive), and what kind of vector is used
(intracellular or secretory):
A. If the gene of interest is expressed constitutively, inoculate a sample
colony into YPD medium, grow the culture for 2–3 days with good
aeration and analyze the proteins in samples taken periodically during
this time by any available detection method.
B. For methanol-induced expression, a colony should be grown for
17–24 h in YPD and then transferred to a fresh methanol-containing
medium for induction. The induction can be performed in 15-ml tubes
182 James M. Cregg et al.

with 2 ml medium containing 0.5% methanol with good aeration at a

starting OD600 of about 10. For intracellular proteins an induction time
of 6–12 h in methanol medium is sufficient.
Intracellularly expressed samples require the preparation of cell extracts
for analysis. Culture samples can be harvested after 10–12 h and prepared for
cell breakage by glass bead disruption. Harvest approximately 10–50 OD600
units of cells and resuspend in 150 ul of a breakage buffer. Add an equal
volume of sterile acid washed 0.45 mm diameter glass beads. Vortex the
mixture on the highest setting for 1 min, then place the sample on ice for
1 min; repeat this process at least five times. Alternatively, load samples into
a vortexer with a head made to hold multiple microfuge tubes. Place the
vortexer in a 4
C cold box or cold room and vortex on high for approxi-
mately 10 min. Examine cultures for cell breakage under a microscope,
80–90% of the cells should be disrupted. After disruption, draw off the cell
debris and buffer into a fresh microcentrifuge tube. Rinse the beads with an
additional 100 ul portion of buffer by a brief vortexing and transfer the wash
to the tube with the rest of the cell debris. Centrifuge the samples on high
speed for 5 min at 4
C. Draw off the top liquid phase containing your
protein and transfer to a fresh microcentrifuge tube. This is your crude
protein sample ready for SDS–PAGE, western blotting or enzyme assay.
Secreted proteins build up in the medium much more slowly and
require at least 2 days to reach high levels. Allow the cultures to incubate
with shaking for 2–5 days. Add fresh methanol to a final concentration of
0.5% every 12 h and collect 50–100 ul supernatant samples during this
induction period. The supernatant samples are ready for SDS–PAGE,
western blotting or enzyme assay and can be stored at  20
C.

9. Assay Development—The Yeastern Blot

There is much in common in different antibody assays. One useful and
yeast-specific antibody assay is the yeast colony western blot, sometimes
referred to as the ‘‘Yeastern’’ blot for secreted proteins (Fig. 13.4). For
standard western blot assay procedures, you are referred to (Sambrook et al.,
1989).
The Yeastern blot is a convenient means of qualitatively screening large
numbers of yeast colonies directly on plates for expression of a recombinant
protein. However, readers should be aware that the correlation between the
size of the ‘‘halo’’ surrounding a colony and the amount of recombinant
product is not always linear and all results with this method should be
confirmed with a standard western blot or other assay. The procedure for
Yeastern blotting is as follows:
Expression in the Yeast Pichia pastoris 183

Figure 13.4 ‘‘Yeastern Blot’’ of P. pastoris colonies secreting both heavy and light
chains of IgG antibodies. Negative controls of P. pastoris strain that does not secrete
antibody shown on second row second streak from the left and at first two positions
from the left on bottom row.

1. Transfer freshly grown yeast colonies from the surface of an agar Petri
plate onto a sterile piece of Whatman No. 1 filter paper by a standard
replica-plating method. The filter paper should be cut to a size that
exactly fits within the plate.
2. Place the filter with yeast cells onto the surface of a fresh plate containing
an appropriate induction medium and incubate the plate for 1–2 days.
3. Prepare a piece of nitrocellulose membrane the same size and dimensions
as the filter paper by soaking the membrane for 5 min or more in 15 ml of
transfer buffer (25 mM Tris base, pH 8.5, 0.2 M glycine, 20% methanol).
Soak two additional pieces of cut filter paper in transfer buffer.
4. Prepare a sandwich of the papers as described below:
a. Place one piece of the filter paper on the anode platform of a western
blotter. (It is essential to remove all bubbles between membrane
layers. This can be done by rolling a pipette over their surface.).
b. Place the soaked nitrocellulose filter on top of the filter paper.
c. Place the filter paper with replica-plated yeast cells on top of the
nitrocellulose paper.
d. Place the second piece of soaked filter paper on top of the filter with
cells.
e. Finally, place the cathode plate on top of the sandwich.
5. Transfer proteins to the nitrocellulose membrane with a constant
current (1–4 mA/cm2) for 1 h.
6. Remove the nitrocellulose membrane and wash it for 5 min in 15 ml
TBS buffer (50 mM Tris–HCl, pH 7.6, 150 mM NaCl). Replace the
TBS buffer with 15 ml of TBST (TBST is prepared by adding Tween-
20 to 0.05% to TBS) containing 1–5% bovine serum albumin (BSA).
184 James M. Cregg et al.

Place filter in blocking buffer and rock gently at room temperature

for 1 h.
7. Transfer filter to 15 ml of TBST buffer and wash by rocking at room
temperature for 5 min.
8. Prepare primary antibody dilution in 15 ml of blocking buffer accord-
ing to the vendor’s recommendations (typically, 0.5 ug/ml).
9. Incubate membrane overnight at 2–8
C with rocking action (can be as
short as 1–3 h at room temperature depending on antibody).
10. Wash membrane in at least four changes of TBST buffer (15 ml each,
5 min/wash) then briefly rinse in fresh TBST.
11. Prepare an enzyme-conjugated secondary antibody dilution in block-
ing buffer according to the vendor’s recommendations (typically 3 ul in
15 ml of blocking buffer).
12. Incubate membrane in secondary antibody solution for 1 h, rocking at
room temperature.
13. Repeat Step 10.
14. Treat membrane in the dark room with visualization reagents accord-
ing to the vendor’s recommendations (e.g., Pierce ECL Western Blot-
ting Substrate), remove excess liquid by blotting with filter paper and
place the filter in a plastic protector and expose to X-ray film for
anywhere from 30 s to several minutes. The intensity and size of the
signal around each colony approximately reflects the secretion level of
the recombinant product by the different colonies/strains.

10. Posttranslational Modification of the

Recombinant Protein (Proteinases and
Glycosylation)
Posttranslational problems that affect the quality of recombinant pro-
teins expressed in P. pastoris include proteolysis and glycosylation.
A proteinase deficient strain of P. pastoris can be tested if initial results by
SDS–PAGE analysis (coomassie staining or western blots) suggest proteo-
lytic degradation of the protein. Signs of proteolysis include low recombi-
nant protein levels and active or immunoreactive products that are smaller
than the full-length product. Degraded protein can also run as a ‘‘smear’’
after PAGE, running from approximately the correct size of the product to
smaller sizes. The Pichia strain SMD1168 ( pep4 his4) is deleted for much of
the PEP4 gene (Gleeson et al., 1998). The PEP4 gene product is responsible
for activating many of the proteases in the vacuole of P. pastoris which enter
the vacuole as inactive zymogens and are activated there by the PEP4
product. Although secreted recombinant proteins do not go to the vacuole,
they can contact proteases in the culture medium from the lysis of a small
Expression in the Yeast Pichia pastoris 185

portion of the cells. Due to the extremely high culture densities used with
P. pastoris, the concentration of vacuolar proteases in the medium can be
considerable. To utilize a PEP4 strain, one must transform one of these
strains (i.e., SMD1168 ( pep4 his4) or SMD1168H ( pep4)) with the expres-
sion vector or create deletions in the PEP4 gene in a strain already expres-
sing a protein of interest. To determine if PEP4 deficient strains benefit
the expression of the protein, the recombinant protein should be examined
from wild type and the PEP4 deficient strains after an induction performed
in parallel. Note: PEP4 deficient strains of P. pastoris are not as robust as
wild-type strains. In particular, PEP4 deficient strains die more quickly
when stored on plates, do not transform as efficiently as wild-type strains,
grow more slowly in culture, and are more difficult to induce on methanol.
Furthermore, PEP4 deficient strains are difficult to mate with other non-
PEP4 P. pastoris strains.
If a recombinant protein expressed in P. pastoris is larger than expected
and somewhat heterogenous in size by SDS–PAGE analysis, it may be
glycosylated. One should first examine the amino acid sequence of the
protein for potential glycosylation sites: the signal for addition of N-linked
oligosaccharides is ASN-X-Ser/Thr, and O-linked sugars can be added to
the OH-group of any Ser or Thr residue. Glycosylation can be confirmed by
deglycoslyating suspect protein with PNGase F and examining the product
by SDS–PAGE. A good protocol for deglycoslation of proteins can be found
at: http://www.neb.com/nebecomm/products/productP0704.asp. If this
treatment reduces the apparent molecular size of your protein resulting in a
more homogenous product, the protein is almost certainly glycosylated.

11. Selection for Multiple Copies

of an Expression Cassette
Perhaps the most productive means of increasing the per cell amount
of a recombinant protein using the P. pastoris system is by increasing the
number of copies of the expression cassette in a strain (Brierley, 1998; Thill
et al., 1990). Two general approaches have been developed to create multi-
copy expression strains in P. pastoris. The first approach involves construct-
ing a vector with multiple head-to-tail copies of an expression cassette
(Brierley, 1998). The key to generating this construction is a vector that
has an expression cassette flanked by restriction sites that have complemen-
tary termini (e.g., BamHI–BglII, SalI–XhoI combinations). The process of
repeated cleavage and reinsertion results in the generation of a series of
vectors that contain increasing numbers of expression cassettes. A particular
advantage to this approach, especially in the production of human
186 James M. Cregg et al.

pharmaceuticals, is that the precise number of expression cassettes is known

and can be recovered for direct verification by DNA sequencing.
The second approach utilizes expression vectors that contain a drug
resistance gene as the selectable marker, and selection for strains resistant
to higher levels of the drug (Scorer et al., 1994). Drug resistant genes used in
P. pastoris include the bacterial KanR, ZeoR, and BsdR genes, as well as, the
P. pastoris FLD1 gene (Shen et al., 1998). Each of these genes supports
significant enrichment for strains with increased copy number of the
expression vector with higher levels of drug resistance. For example, selec-
tion of zeocin resistant P. pastoris transformants is performed on plates with
1–2 mg/ml zeocin instead of the standard 100–200 ug/ml zeocin. How-
ever, no matter which drug is used, the vector copy number will still vary
greatly. After selection most transformants still contain only a single vector
copy, even if they are resistant to high drug levels. Thus, 50–100 indepen-
dent transformants selected at the high drug concentration should be ana-
lyzed for copy number and expression level to identify better expressing
strains. By this approach, strains carrying up to 30 copies of an expression
cassette have been isolated (Scorer et al., 1994). Importantly, once isolated,
multicopy strains are stable with standard microbial handling procedures
and do not require continued drug selection on plates or in liquid medium
(i.e., maintain a stock of a given strain selected on medium with the drug
stored frozen at 80
C, then use a working stock kept on plate of
noninducing medium for a limited number of experiments or a single
production run).
One drawback of this selection procedure is the difficulty in obtaining
enough clones to screen for multiple expression cassettes. First, the number
of transformants resistant to high levels of drug is very low, often 0.1–1% of
the number on low levels of drug (100 ug/ml for zeocin). Therefore, unless
ones transformation efficiency is at its peak, there may not be any transfor-
mants resistant to the highest drug levels. Furthermore, 50–100 transfor-
mants are needed to screen for a multicopy strain since only about 1–5% of
the transformants resistant to high levels of the drug are due to added copies
of the resistance gene and most are resistant due to other unknown factors.
In part due to the difficulty in obtaining multicopy strains by direct
selection a new method for obtaining strains with high copy numbers of
vector and elevated recombinant protein expression levels was developed
(Sunga et al., 2008). Briefly, P. pastoris transformants selected on a low level
of drug and containing only one or a few copies of the vector are subse-
quently subjected to higher drug levels to obtain strains with higher num-
bers of copies. Simply streak transformants on agar plates containing higher
and higher levels of zeocin. For example, if the original transformant was
selected on 100 ug/ml of zeocin, streak the strain on plates containing
500 ug/ml of the drug. Collect colonies that are resistant to the higher
level of drug as individual strains and confirm that one or more have
Expression in the Yeast Pichia pastoris 187

elevated recombinant protein expression levels and are resistant due to a

higher number of vector copies by PCR. Once the strain is confirmed as a
better expressing/multicopy strain, the process can be repeated at an even
higher concentration of zeocin (2 mg/ml). Again, high resistance strains are
collected and examined for expression and the number of vector copies.
This iterative process has been termed Posttransformational vector amplifi-
cation (PTVA) and results in strains containing multiple head-to-tail copies
of the entire vector integrated at a single locus in the genome. An analysis of
PTVA-selected clones indicates 40% showed a three- to fivefold increase
in vector copy number. So-called ‘‘jackpot’’ clones, with greater than
10 copies of the expression vector, represented 5–6% of selected clones
and in some cases had a proportional increase in recombinant protein
production.
Although the molecular details of the process(s) by which these amplifi-
cation events occur are not well understood, key observations about the
process have been made. First, the amplification process appears to occur
naturally in a small percentage of cells in virtually any vector containing
strain. Second, the PTVA process leads to a considerable increase in copy
number of the entire vector and not just portions of the vector such as the
resistance gene. This is clearly important if a uniform recombinant product
is desired. Third, Southern blot data demonstrated that all the copies are
inserted into the P. pastoris genome in the same location as the original copy
and in a head-to-tail configuration (Sunga et al., 2008).
Finally, the PTVA method works with other drug resistant selectable
markers in P. pastoris and not just with zeocin vectors. Thus, this amplification
process seems to be a general response to high drug levels in this yeast
species. Given that most yeast species have similar homologous recombination
systems, this technique should work in other yeast species as well.

REFERENCES
Ausubel, E. M., Brent, R., Kingston, E., Moore, D. D., Seidman, J. G., Smith, J. A., and
Struhl, K. (eds.) (2001). In ‘‘Current Protocols in Molecular Biology’’ John Wiley and
Sons, Inc., New York.
Becker, D. M., and Guarante, L. (1991). High-efficiency transformation of yeast by electro-
poration. Methods Enzymol. 194, 182–187.
Boer, E., Gellisson, G., and Kunze, G. (2005). Arxula adeninivorans. In ‘‘Production of
Recombinant Proteins’’ (G. Gellisson, ed.), pp. 89–110. Wiley-VCH Verlag GmbH and
Co. KGaA, Weinheim, Germany. Chapter 5.
Brierley, R. A. (1998). Secretion of recombinant human insulin-like growth factor I (IGF-I).
In ‘‘Methods in Molecular Biology Vol. 103 Pichia Protocols’’ (D. R. Higgins and
J. M. Cregg, eds.), pp. 149–178. Humana Press, Totowa, NJ.
Cregg, J. M. (2008). DNA-mediated transformation. In ‘‘Methods in Molecular Biology:
Pichia Protocols’’ ( J. M. Cregg ed.), 2nd ed. pp. 27–42. Humana Press, Totowa, NJ.
Chapter 3.
188 James M. Cregg et al.

Gellissen, G. (ed.) (2002). In ‘‘Hansenula polymorpha: Biology and Applications’’ Wiley-VCH

Verlag GmbH and Co. KGaA, Weinheim, Germany.
Gellissen, G. (ed.) (2005). In ‘‘Production of Recombinant Proteins’’ Wiley-VCH Verlag
GmbH and Co. KGaA, Weinheim, Germany.
Gleeson, M. A. G., White, C. E., Meininger, D. P., and Komives, E. A. (1998). Generation
of protease-deficient strains and their use in heterologous protein expression.
In ‘‘Methods in Molecular Biology, Vol. 103, Pichia Protocols’’ (D. Higgins and
J. M. Cregg, eds.), pp. 81–94. Humana Press, Totowa, NJ.
Grey, M., and Brendel, M. F. (1995). Ten-minute eletrotransformation of Saccharomyces
cerevisiae. In ‘‘Methods in Molecular Biology, Vol. 47: Electroporation Protocols for
Microorganisms’’ ( J. A. Nickoloff, ed.), pp. 269–272. Humana Press, Totowa, NJ.
Heerikhuisen, M., van den Hondel, C., and Punt, P. (2005). Aspergillus sojae.
In ‘‘Production of Recombinant Proteins’’ (G. Gellissen, ed.), pp. 191–214. Wiley-
VCH Verlag GmbH and Co. KGaA, Weinheim, Germany. Chapter 9.
Higgins, D. R., and Cregg, J. M. (eds.) (1998). Methods in Molecular Biology, Vol. 103,
Pichia Protocols, Humana Press, Totowa, NJ.
Kuck, U., and Poggeler, S. (2005). Sordaria macrospora. In ‘‘Production of Recombinant
Proteins’’ (G. Gellissen, ed.), pp. 215–232. Wiley-VCH Verlag GmbH and Co. KGaA,
Weinheim, Germany. Chapter 10.
Lin-Cereghino, J., and Lin Cereghino, G. P. (2008). Vectors and strains for expression.
In ‘‘Methods in Molecular Biology: Pichia Protocols’’ ( J. M. Cregg, ed.), pp. 111–126.
Humana Press, Totowa, NJ. Chapter 2.
Lin-Cereghino, J., Wong, W. W., Xiong, S., Giang, W., Luong, L. T., Vu, J.,
Johnson, S. D., and Lin-Cereghino, G. P. (2005). Condensed protocol for competent
cell preparation and transformation of the methylotrophic yeast Pichia pastoris. Biotechni-
ques 38, 44–48.
Lorow-Murray, D., and Jesse, J. (1991). High efficiency transformation of Saccharomyces
cerevisiae by electroporation. Focus 13, 65–68.
Madzak, C., Nicaud, J.-M., and Gaillardin, C. (2005). Yarrowia lipolytica. In ‘‘Production
of Recombinant Proteins’’ (G. Gellissen, ed.), pp. 163–190. Wiley-VCH Verlag GmbH
and Co. KGaA, Weinheim, Germany. Chapter 8.
Pichia Expression Kit Instruction Manual p. 63. Version E, Invitrogen, San Diego, CA.
Romanos, M. A., Scorer, C. A., and Clare, J. J. (1992). Foreign gene expression in yeast:
A review. Yeast 8, 423–488.
Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). Molecular Cloning: A Laboratory
Manual. 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Scorer, C. A., Clare, J. J., McCombie, W. R., Romanos, M. A., and Sreekrishna, K. (1994).
Rapid selection using G4118 of high copy number transformants in Pichia pastoris for
high-level foreign gene expression. Bio/Technology 12, 181–184.
Shen, S., Sulter, G., Jeffries, T. W., and Cregg, J. M. (1998). A strong regulated promoter for
controlled expression of foreign genes in the yeast Pichia pastoris. Gene 216, 93–102.
Stowers, L., Gautsch, J., Dana, R., and Hoekstra, M. F. (1995). Yeast transformation and
preparation of frozen spheroplasts for electroporation. In ‘‘Methods in Molecular
Biology, Vol 47: Electroporation Protocols for Microorganisms’’ ( J. A. Nickoloff, ed.),
pp. 261–267. Humana Press, Totowa, NJ.
Sunga, A. J., Tolstorukov, I., and Cregg, J. M. (2008). Post transformational vector amplifi-
cation in the yeast Pichia pastoris. FEMS Yeast Res. 8, 870–876.
Thill, G. P., Davis, G. R., Stillman, C., Holtz, G., Brierley, R., Engel, M., Buckholtz, R.,
Kinney, J., Provow, S., Vedvick, T., and Siegel, R. S. (1990). Positive and negative
effects of multi-copy integrated expression vectors on protein expression in Pichia pastoris.
In ‘‘Proceedings of the 6th International Symposium on Genetics of Microorganisms’’
Expression in the Yeast Pichia pastoris 189

(H. Heslot, J. Davies, J. Florent, L. Bobichon, G. Durand, and L. Penasse, eds.) Vol. II,
pp. 477–490. Societe Francaise de Microbiologie, Paris.
Tolstorukov, I., and Cregg, J. M. (2008). Classical Genetics. In ‘‘Methods in Molecular
Biology: Pichia Protocols’’ ( J. M. Cregg, ed.), 2nd ed. pp. 189–202. Humana Press,
Totowa, NJ. Chapter 14.
van Ooyen, A. J., Dekker, P., Huang, M., Olsthoorn, M. M., Jacobs, D. I., Colussi, P. A.,
and Taron, C. H. (2006). Heterologous protein production in the yeast Kluyveromyces
lactis. FEMS Yeast Res. 6, 381–392.
 C H A P T E R F O U R T E E N

Baculovirus–Insect Cell
Expression Systems
Donald L. Jarvis

Contents
1. Introduction 192
2. A Brief Overview of Baculovirus Biology and Molecular Biology 193
3. Baculovirus Expression Vectors 195
4. Baculovirus Expression Vector Technology—The Early Years 196
5. Baculovirus Expression Vector Technology—Improved 198
6. Baculovirus Transfer Plasmid Modifications 198
7. Parental Baculovirus Genome Modifications 200
8. The Other Half of the Baculovirus–Insect Cell System 210
9. A New Generation of Insect Cell Hosts for Baculovirus
Expression Vectors 212
10. Basic Baculovirus Protocols 214
10.1. Insect cell maintenance 215
10.2. Isolation of baculovirus genomic DNA 215
10.3. Baculovirus plaque assays 216
References 218

Abstract
In the early 1980s, the first-published reports of baculovirus-mediated foreign
gene expression stimulated great interest in the use of baculovirus–insect cell
systems for recombinant protein production. Initially, this system appeared to
be the first that would be able to provide the high production levels associated
with bacterial systems and the eukaryotic protein processing capabilities asso-
ciated with mammalian systems. Experience and an increased understanding of
basic insect cell biology have shown that these early expectations were not
completely realistic. Nevertheless, baculovirus–insect cell expression systems
have the capacity to produce many recombinant proteins at high levels and they
also provide significant eukaryotic protein processing capabilities. Furthermore,
important technological advances over the past 20 years have improved upon

Department of Molecular Biology, University of Wyoming, Laramie, Wyoming, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63014-7 All rights reserved.

191
192 Donald L. Jarvis

the original methods developed for the isolation of baculovirus expression

vectors, which were inefficient, required at least some specialized expertise
and, therefore, induced some frustration among those who used the original
baculovirus–insect cell expression system. Today, virtually any investigator with
basic molecular biology training can relatively quickly and efficiently isolate a
recombinant baculovirus vector and use it to produce their favorite protein in an
insect cell culture. This chapter will begin with background information on the
basic baculovirus–insect cell expression system and will then focus on recent
developments that have greatly facilitated the ability of an average investigator
to take advantage of its attributes.

1. Introduction
It seems that nearly every book chapter or review focusing on the
baculovirus–insect cell expression system begins with a statement emphasiz-
ing the popularity of this system and/or noting its widespread use for
recombinant protein production. While this introduction has become
increasingly redundant, it also has become increasingly accurate over the
past 20 years. An advanced PubMed search using the terms ‘‘baculovirus’’
and ‘‘expression’’ and ‘‘vector’’ at the time of this writing yielded over 2000
hits. While this is a significant number, it vastly underestimates the actual
number of published studies involving recombinant protein production
using the baculovirus–insect cell expression system. Moreover, it does not
include the large number of studies performed behind closed doors in the
biotechnology industry, which are clearly evidenced by the number of oral
presentations given by industrial scientists at various baculovirus–insect cell
technology conferences held over the past two decades.
The previous edition of this Guide to Protein Purification (1990)
included a comprehensive description of the technical details involved in
using the baculovirus–insect cell expression system, as it existed at that time.
This exercise will not be repeated here and the reader should refer to the
previous edition of this book and other sources given below for those
details. In this edition, I will begin by providing the background informa-
tion needed for the reader to understand the basic principles underlying the
original baculovirus–insect cell expression system. I will then focus on the
new tools and approaches developed for the isolation of recombinant
baculoviruses since the publication of the previous edition of this book, as
these have greatly facilitated the ability of virtually any biomedical investi-
gator to utilize the baculovirus–insect cell system, relative to the state of the
art in 1990.
Baculovirus–Insect Cell Expression Systems 193

2. A Brief Overview of Baculovirus Biology

and Molecular Biology
Baculoviridae is a large family of viruses with relatively large, double-
stranded, circular DNA genomes (see Miller, 1997 for a comprehensive
review). The natural hosts for these viruses are arthropods, mainly insects,
and most baculoviruses have a very narrow host range, which is usually
restricted to just one insect species. The most extensively studied and
exploited member of the Baculoviridae is Autographa californica multicapsid
nucleopolyhedrovirus, or AcMNPV. Compared to most baculoviruses, this
isolate has a relatively broad host range, as it can productively infect around
25 different lepidopteran insects and can infect most tissue types within an
individual target insect species. AcMNPV was the isolate used to develop
the original recombinant baculovirus vectors (Pennock et al., 1984; Smith
et al., 1983b) and its genome is still used as the backbone for the production
of most baculovirus vectors today. Thus, while they can be produced using
other types of baculoviruses, most generic references to ‘‘baculovirus vec-
tors’’ actually refer more specifically to recombinant variants of AcMNPV.
Because this is a relatively unimportant detail for most baculovirus–insect
cell expression system users, I will slip into this rather loose language in the
latter parts of this chapter.
With respect to the development of the original baculovirus–insect cell
expression system (see Summers, 2006 for a detailed history), one of the
most important features of AcMNPV and other baculoviruses was the ability
of these viruses to produce ‘‘polyhedra’’ or ‘‘occlusion bodies’’ during
productive viral infections. Polyhedra are large particles that appear in the
nuclei of AcMNPV-infected insect cells near the end of the infectious cycle.
At late times after infection, host cell nuclei are typically packed with 2–3
dozen of these complex particles, which consist, in part, of progeny virions
embedded within a protective, paracrystalline array. Importantly, this para-
crystalline array is composed of a single, virus-encoded protein called
polyhedrin. Thus, it is obvious that AcMNPV and other baculoviruses
must produce extremely large amounts of polyhedrin protein in order to
fulfill their need to produce large numbers of polyhedra. In fact, polyhedrin
is the major protein in AcMNPV-infected cells near the end of the infec-
tious cycle, typically representing over half of the total protein found in
these cells at that time.
The ability of baculoviruses to produce large amounts of polyhedrin is
important because it was one of the fundamental features that spurred the
development of these viruses as vectors for foreign protein production. The
basic notion was that the ability of these viruses to produce extremely large
amounts of polyhedrin could be harnessed to produce large amounts of
194 Donald L. Jarvis

other proteins of far greater interest to the biomedical research community.

This notion was extended by the finding that the polyhedrin protein is
not required for the replication of baculoviruses in cultured insect cells
(Smith et al., 1983a). Thus, it was theoretically possible to create helper-
independent recombinant baculoviruses in which the viral DNA sequence
encoding polyhedrin was replaced by a foreign DNA sequence encoding a
protein of interest. This theory was reduced to practice when the first
recombinant baculoviruses were produced by homologous recombination
between the polyhedrin region in the AcMNPV genome and ‘‘transfer’’
plasmids containing a foreign gene placed under the control of the
polyhedrin promoter (Pennock et al., 1984; Smith et al., 1983b).
The original recombinant baculoviruses were designed to express for-
eign genes under the control of the polyhedrin promoter because the
strength of this promoter is ultimately responsible for the ability of
AcMNPV and other baculoviruses to produce such large amounts of the
polyhedrin protein. However, it is relevant that baculoviruses actually have
three or four distinct classes of genes, which are expressed in a temporally
regulated, sequential fashion. The first to be expressed are the early genes,
which can be subclassified as immediate-early and delayed-early genes
(reviewed in Friesen, 1997). The salient feature of the early genes is that
they have host-like promoters, which can be recognized and transcribed by
host transcriptional machinery. With some important caveats that will not
be discussed here, these promoters can function in the absence of other
baculoviral factors to induce viral gene expression at the very beginning of
the infectious cycle. For this reason, baculovirus early promoters have been
used to drive constitutive foreign gene expression in uninfected lepidop-
teran insect cells (reviewed in Douris et al., 2006; Harrison and Jarvis, 2007a;
Jarvis and Guarino, 1995) and this is an important facet of efforts to develop
transgenic insect cell lines as improved hosts for baculovirus-mediated
foreign protein production (reviewed in Harrison and Jarvis, 2006, 2007b;
Shi and Jarvis, 2007). The next class of genes to be expressed during
baculovirus infection is the late genes, which is a set of genes expressed
after the onset of viral DNA replication (reviewed in Lu and Miller, 1997;
Passarelli and Guarino, 2007). The salient feature of these genes is that they
have virus-specific promoters, which can be recognized and transcribed
only by virus-encoded transcriptional machinery. Thus, these promoters
can function only in the context of baculovirus infection and are activated at
later times of infection, once a virus-specific transcription complex has been
produced. The last class of genes to be expressed during baculovirus infec-
tion is the very late genes, which are expressed after the onset of viral DNA
replication, like the late genes, but later, closer to the end of the infectious
cycle. The very late genes encode proteins such as polyhedrin that are
involved in the production of polyhedra. Like the late genes, the very late
genes have virus-specific promoters that can be recognized and transcribed
Baculovirus–Insect Cell Expression Systems 195

only by a virus-encoded transcription complex and, therefore, can function

only in the context of baculovirus infection. In fact, the very late promoter-
specific transcription complex includes the same proteins as the late pro-
moter-specific complex plus one or more additional proteins (McLachlin
and Miller, 1994; Mistretta and Guarino, 2005). Late and very late promoter
elements also are quite similar, but the very late elements include an
additional downstream ‘‘burst’’ sequence, which leads to the extremely
high levels of transcription (‘‘hypertranscription’’) observed during the
very late phase of baculovirus infection (Ooi et al., 1989).

3. Baculovirus Expression Vectors

The classic baculovirus expression vector, in its simplest form, is a
recombinant baculovirus whose genome contains a foreign nucleic acid
sequence, almost always a cDNA, encoding a protein of interest under
the transcriptional control of the polyhedrin promoter. The chimeric
gene consisting of the polyhedrin promoter and foreign protein coding
sequence is found in the polyhedrin locus of the viral genome in place of
the nonessential, wild-type polyhedrin gene. The recombinant virus can be
used to infect cultured insect cells or larvae (caterpillars) in the laboratory
and this leads to high-level transcription of the foreign cDNA during the
very late phase of infection. The resulting mRNA then can be translated to
produce the protein of interest. The polyhedrin promoter seems to invari-
ably provide extremely high levels of foreign gene expression at the tran-
scriptional level and this leads, in many cases, to high levels of foreign
protein production, as originally anticipated from the ability of baculo-
viruses to produce large amounts of the polyhedrin protein. Indeed, the
potential for high-level recombinant protein production is one of the major
advantages of the baculovirus–insect cell system, with ‘‘high-level’’ defined
rather loosely here as
100 mg of recombinant protein per liter of infected
insect cell culture, or  4 g of cells at the usual density of 1  106 cells/mL.
In addition, high-level production of recombinant proteins in baculovirus-
infected insect cells is rarely associated with inclusion body formation,
which is commonly observed in bacterial systems. However, anyone who
works in the area of recombinant protein production knows that the actual
production and solubility levels achieved in any system are highly depen-
dent upon the specific protein under investigation. Thus, a more useful
generalization that has arisen from 25 years of collective experience with
baculovirus–insect cell systems is that these systems are much more likely to
produce high levels of foreign nuclear and cytoplasmic proteins than secre-
tory pathway proteins. The latter are typically produced at lower levels than
the former, often at the level of single to tens of mg/L ( Jarvis, 1997).
196 Donald L. Jarvis

Another major advantage typically associated with the baculovirus–

insect cell system is its eukaryotic protein processing capabilities, which
include the ability to provide protein modifications such as phosphorylation
and glycosylation, among many others. Again, however, this claim comes
with a caveat, as it is now well recognized that the protein processing
pathways of the lepidopteran insect cell hosts for baculovirus vectors are
not identical to those of higher eukaryotes. An additional complication is
that baculovirus infection can have an adverse impact on host protein
processing functions (Azzouz et al., 2000; Jarvis and Summers, 1989).
Obviously, these are important considerations for anyone interested in
producing recombinant proteins with eukaryotic modifications, particularly
those known to directly or indirectly influence their functions.
Finally, it should be noted that the baculovirus–insect cell system has
proved to be particularly useful for the production of multiprotein subunit
complexes (reviewed in Berger et al., 2004; Kost et al., 2005). The power of
this system for this increasingly important application is exemplified by its
ability to produce virus-like particles composed of multiple virion compo-
nents, which are outstanding vaccine candidates. For example the
baculovirus–insect cell system has been used to produce virus-like particles
consisting of multiple proteins from polio, bluetongue, adeno-associated,
hepatitis C, and papilloma viruses, among others. This has been accom-
plished by using multiple recombinant baculoviruses, each encoding indi-
vidual proteins, or single recombinant viruses encoding multiple proteins to
infect insect cell cultures.

4. Baculovirus Expression Vector

Technology—The Early Years
The first recombinant baculovirus vectors designed to express chime-
ric genes consisting of the polyhedrin promoter and a foreign coding
sequence, as described in the preceding section, were produced using a
basic homologous recombination approach. The methodological details of
this approach were described in the last edition of this volume (Bradley,
1990) and also are available in primary papers and excellent technical
manuals from the original contributors (O’Reilly et al., 1992; Summers
and Smith, 1987). Thus, this exercise will not be repeated here, as indicated
above. However, for background purposes it is important to briefly note that
this general method involved (1) construction of a bacterial ‘‘transfer’’ plasmid
containing the chimeric gene flanked by sequences derived from the polyhe-
drin region of the viral genome (Fig. 14.1) and (2) cotransfection of cultured
insect cells with a mixture of this transfer plasmid DNA and genomic DNA
extracted from purified preparations of wild-type AcMNPV (Fig. 14.2).
Baculovirus–Insect Cell Expression Systems 197

Plasmid backbone
(e.g. pUC-based)

5⬘-viral flanking 3⬘-viral flanking

sequence sequence
Polyhedrin PolyA
promoter signal
cDNA encoding
protein of interest

Figure 14.1 A simple baculovirus transfer plasmid.

Transfer plasmid
Transfer plasmid

Polyhedrin gene
Foreign gene

Polyhedrin gene Homologous recombination Foreign gene

(co-transfected insect cells)

Baculoviral DNA Recombinant baculoviral DNA

Clone polyhedron-negative recombinant (s)

Figure 14.2 Producing a baculovirus expression vector by homologous

recombination.

Homologous recombination between the sequences flanking the chimeric

gene of interest in the transfer plasmid and the sequences upstream and
downstream of the polyhedrin gene in the wild-type AcMNPV genome
produced recombinant viral DNA molecules in these cotransfected insect
cells. A double crossover recombination event was necessary to simultaneously
knockout the polyhedrin gene and knock-in the chimeric gene encoding the
protein of interest. Of course, this was a relatively rare event with an estimated
frequency of  0.1% (Smith et al., 1983a). Thus, it was necessary to separate the
small minority of recombinant virus progeny from the vast majority of parental
viral progeny produced by the cotransfected insect cells by cloning. This was
easily accomplished by baculoviral plaque assays, but then one had to be able to
distinguish recombinant viral plaques from the much larger background of
plaques derived from the parental virus. Initially, this was accomplished using a
198 Donald L. Jarvis

simple visual screen, as the parental viral progeny produced polyhedron-

positive plaques while the recombinant viral progeny, which lacked a func-
tional polyhedrin gene, produced polyhedron-negative plaques.
An investigator with a trained eye could visually identify polyhedron-negative
plaques by examining the assay plate under a dissecting microscope. However,
the trained eye was a key to success and the inability of many investigators to
recognize polyhedron-negative baculoviral plaques was a serious problem that
constrained the use of the baculovirus–insect cell system as a recombinant
protein production tool for several years.

5. Baculovirus Expression Vector

Technology—Improved
Starting around 1990, various investigators began to address the tech-
nical constraints associated with the isolation of baculovirus vectors and
to improve the system in other ways by developing a wide variety of
modifications of the basic approach described above. These modifications
fell into two different categories, including (1) transfer plasmid modifica-
tions and (2) parental baculovirus genome modifications. These will be
considered separately.

6. Baculovirus Transfer Plasmid Modifications

Transfer plasmid modifications were designed to serve two different
purposes. The most important purpose was to facilitate the identification of
recombinant baculovirus plaques by visual screening, which had been a
difficult task for the reasons described above. One general type of modifi-
cation that fell within this category was to incorporate chimeric marker
genes under the control of baculovirus promoters encoding products such as
the E. coli b-galactosidase protein, which could be far more easily recog-
nized by visual screening of plaque assays than the polyhedron-negative
plaque phenotype (Vialard et al., 1990). The introduction of a marker gene
into transfer plasmids was a clever development that served its intended
purpose, but it also was important to be aware of a potential trap associated
with this approach (O’Reilly et al., 1992). While the introduction of a
marker gene could signal a double crossover homologous recombination
event, which was the desired outcome, it also could indicate a single
crossover recombination event between the transfer plasmid and viral DNA.
Single crossover recombination was a far more frequent event that produced
recombinant viral genomes containing the entire transfer plasmid, including
the bacterial replicon, at somewhat random loci. These recombinants were
Baculovirus–Insect Cell Expression Systems 199

genetically unstable and the newly acquired foreign gene typically would be
lost within a round or two of viral replication. Despite this major limitation,
transfer plasmids that incorporated marker genes into the recombinant
baculovirus genome were widely used and particularly useful in the hands
of investigators aware of the potential single crossover recombination trap.
These investigators would simply use incorporation of the marker gene as a
prescreen and then perform additional screening to map the genomic
position of a foreign gene and confirm that a double crossover recombina-
tion event was responsible for transfer of the gene of interest to the
recombinant baculovirus vector.
The second purpose served by transfer plasmid modifications was to
facilitate recombinant protein expression and purification in the
baculovirus–insect cell system. While the specific, individual modifications
that were undertaken are far too numerous to discuss here, three general
types of transfer plasmid modifications that fell within this category included
the addition of sequences encoding secretory signal peptides, the addition
of sequences encoding amino- or carboxy-terminal purification tags (e.g.,
6 HIS), replacement of the polyhedrin promoter with alternate baculo-
virus promoters, and replacement of the polyhedrin promoter with multiple
promoter elements, which could simultaneously drive the expression of
multiple recombinant proteins during baculovirus infection.
A nearly comprehensive list of baculovirus transfer plasmids with the
two general types of modifications discussed above, together with short
descriptions of their specific, functional features was recently published and
this list is a good source of information on this topic (Possee and King,
2007). However, one type of modification that was not included and
deserves further consideration is the ‘‘immediate-early’’ transfer plasmid,
in which the polyhedrin promoter is replaced by a promoter from a
baculovirus immediate-early gene, such as the AcMNPV ie1 gene ( Jarvis
et al., 1996). The idea of using promoters from earlier classes of baculovirus
genes, such as the ie1 gene, might seem counterintuitive because these
promoters are weaker than the polyhedrin promoter. However, there is
evidence to suggest that higher quality products can be obtained using
recombinant baculoviruses that express foreign genes under the control of
earlier classes of promoters, despite their inability to drive equally high levels
of transcription (Chazenbalk and Rapoport, 1995; Hill-Perkins and Possee,
1990; Jarvis et al., 1996; Murphy et al., 1990; Rankl et al., 1994; Sridhar
et al., 1993). In fact, this approach can be particularly useful for secretory
pathway proteins, which are often produced at relatively low levels when
their genes are expressed under polyhedrin control, as mentioned above.
In this circumstance, an investigator can potentially gain the advantage of
initiating foreign gene expression at earlier times of infection, thereby
avoiding the potentially adverse effects of baculovirus infection on host
protein processing pathways, without sacrificing high production levels.
200 Donald L. Jarvis

7. Parental Baculovirus Genome Modifications

Like transfer plasmid modifications, parental baculovirus genome
modifications have been designed to serve a variety of different purposes.
Initially, the most important purpose was to find a way to overcome the low
efficiency of recombinant baculovirus vector production and isolation,
which was a major problem with the original baculovirus–insect cell system.
It was clear that the heart of this problem was the extremely low efficiency
of homologous recombination between the transfer plasmid and parental
baculovirus DNA that occurred in cotransfected insect cells. Ultimately,
various investigators developed both indirect and direct solutions to this
problem.
The first major step toward a solution came when Kitts and Possee
developed the first baculovirus with a linearizable DNA genome (Kitts
et al., 1990). The key feature of this viral DNA genome, which was derived
from a recombinant baculovirus produced using the original approach, was
that it had a novel sequence in the polyhedrin locus that contributed a
unique Bsu36I site (Fig. 14.3). Thus, this genome could be linearized with
Bsu36I and used as the parental viral DNA to be mixed with a transfer
plasmid and used to cotransfect insect cells. The most important feature of
this approach was that the linearized parental viral DNA molecule could not
replicate. Thus, linearization vastly reduced the number of parental progeny
produced by the cotransfected insect cells. On the other hand, the linearized
viral DNA could still undergo homologous recombination with the transfer
plasmid, which would recircularize the viral DNA molecule and restore its
ability to replicate. Together, these factors increased the efficiency of
recombinant baculovirus vector production by cotransfected insect cells
from  0.1 to  10–20% and produced a quantum leap in the simplification
of baculovirus–insect cell technology. This basic approach was improved
when Kitts and Possee created a recombinant baculovirus known as
BakPAK6, which could be gapped with Bsu36I (Kitts and Possee, 1993).
The BakPAK6 genome included an E. coli lacZ gene in the polyhedrin
locus, which contributed one Bsu36I site, and it also had two additional
Bsu36I sites that had been introduced into upstream and downstream viral
genes (Fig. 14.4). Importantly, Bsu36I digestion deleted a portion of
orf 1629, which is an essential viral gene located just downstream of poly-
hedrin that encodes a viral nucleocapsid phosphoprotein (Vialard and
Richardson, 1993). With BakPAK6 as the parental viral DNA, the average
efficiency of recombinant baculovirus vector production by cotransfected
insect cells increased to about 95%. The success of this latter develop-
ment spurred commercialization and widespread dissemination of several
‘‘linearized’’ baculoviral DNAs as starting materials for recombinant
Baculovirus–Insect Cell Expression Systems 201

AcRP6SC viral DNA

Bsu36I
Digest with Bsu36I

Linearized viral DNA

(nonreplicative)

Recombine/repair with
transfer plasmid
(co-transfected insect cells)

Circular recombinant viral

DNA
(replicative)

Clone white plaque recombinant (s)

Figure 14.3 Producing a baculovirus expression vector by homologous recombina-

tion with a linearized parental viral genome.

baculovirus vector production. Among others, these included the original

BakPAK6TM viral DNA, which was commercialized by ClonTech; analo-
gous, prelinearized viral DNAs called BaculoGoldTM (Pharmingen/BD
Biosciences), BacVectorÒ (Novagen), and DiamondBacTM (Sigma-Aldrich);
and a slightly modified, linearized viral DNA/transfer plasmid system known
as Bac-N-BlueTM (Invitrogen), in which the E. coli lacZ marker in the viral
genome is regenerated upon recombination.
At this point in this chapter, it is important to emphasize that the
development of these viral DNA backbones effectively solved the major
problem that had been associated with the production of baculovirus
expression vectors at the time of publication of the first edition of this
book. These linearizable baculovirus DNAs were widely marketed by
various companies and were quickly recognized and adopted by the
202 Donald L. Jarvis

BacPAK6 viral DNA

Orf 603 LacZ Orf 1629

Bsu36I sites
Digest with Bsu36I

Linearized (gapped) viral DNA

(nonreplicative)

Recombine/repair with
transfer plasmid
(co-transfected insect cells)

Circular recombinant viral DNA

(replicative)

Clone white plaque recombinant (s)

Figure 14.4 Producing a baculovirus expression vector by homologous recombina-

tion with a linearized/gapped parental viral genome.

biomedical research community as improved tools for recombinant bacu-

lovirus production. This was important because these and other emerging
tools and approaches (see below) were responsible for making baculovirus
expression vector technology far more accessible to the average laboratory
investigator with a basic background in molecular biology. Ultimately, this
led to more widespread use and increased general recognition of the
baculovirus–insect cell system as an excellent tool for recombinant protein
production.
In parallel with the development of linearizable baculovirus genomes,
Luckow and his colleagues developed a totally different approach for the
isolation of baculovirus expression vectors, which relied upon the process of
genetic transposition, rather than homologous recombination (Luckow
et al., 1993). This new approach allowed these investigators to address the
low homologous recombination efficiency problem directly by using a
Baculovirus–Insect Cell Expression Systems 203

pPolh GOI selectn markr

Tn7R Tn7L
Bacmid
Donor Bacterial
transformation LacZ-MiniF-Att7 Helper

Transposition

Recombinant bacmid

pPolh GOI selectn markr Helper

Selection
Screening
Bacmid extraction
Recombinant baculoviruses

Recombinant bacmid

pPolh GOI selectn markr

Insect cell transfection

Figure 14.5 Producing a baculovirus expression vector by transposition between a

bacmid and a transfer plasmid.

different molecular mechanism to produce recombinant baculovirus DNAs.

A key element of this approach was the creation of a new strain of E. coli
containing an autonomously replicating plasmid, or bacmid, which
included a cloned copy of the entire baculovirus genome (Fig. 14.5).
The polyhedrin region of the bacmid included an E. coli lacZ gene and a
‘‘mini-Att Tn7’’ site, which is an attachment site used in the transposition
process. The new E. coli strain also included a helper plasmid encoding the
transposase. Upon transformation with a baculovirus transfer plasmid con-
taining a polyhedrin-promoted gene of interest flanked by the left and right
204 Donald L. Jarvis

ends of Tn7, the chimeric gene could be efficiently transposed from the
transfer plasmid to the polyhedrin locus of the bacmid. This transposition
event would simultaneously delete the lacZ gene in the bacmid, allowing
bacteria containing recombinant bacmids to be detected by standard blue–
white screening on a selective bacterial medium. Thus, one could simply
isolate the recombinant bacmid DNA encoding the gene of interest from an
E. coli clone with a white colony phenotype and use this DNA to transfect
insect cells. Transfection of the DNA would initiate a viral infection and
lead to the production of recombinant baculovirus progeny. This in vivo
transposition approach, which can be used to produce recombinant
baculovirus DNAs at 100% efficiency, was initially commercialized as the
Bac-to-BacTM system by Life Technologies and is now available from
Invitrogen. The bacteriocentric nature of the Bac-to-BacTM system was
important because it allowed investigators who were less familiar with
virological and more familiar with basic bacteriological and molecular
biological methods to work within their comfort zones. One disadvantage
of this system, however, is the fact that it yields recombinant baculoviruses
that retain the bacterial replicon, which appears to be associated with a
higher level of genetically instability upon serial passage of these viruses in
insect cells, relative to baculoviruses lacking this element (Pijlman et al.,
2003; and see below).
Most recently, the Possee and King groups have described a clever cross-
hybridization of the basic ideas underlying the linearizable baculoviral DNA
and bacmid approaches (Possee et al., 2008). In essence, they have created a
new type of bacmid, which consists of a recombinant baculoviral genome
with a bacterial replicon in the polyhedrin locus and a deletion in the
downstream orf 1629 gene (Fig. 14.6). Due to the presence of the replicon

Transfer plasmid Transfer plasmid

9
orf60 629 orf60 162
3 orf1 3 bac replicon Δorf
Foreign gene

Homologous recombination
bac replicon Δorf1629
(co-transfected insect cells) orf1629
orf603 3
orf60 Foreign gene

FlashBAC viral DNA Recombinant baculoviral DNA

(nonreplicative) (replicative)

Clone white plaque recombinant (s)

Figure 14.6 Producing a baculovirus expression vector using the flashBAC approach.
Baculovirus–Insect Cell Expression Systems 205

and the orf 1629 deletion, this bacmid can replicate in E. coli, but not in
insect cells. Therefore, an E. coli strain carrying this bacmid provides a
convenient source of replication defective parental baculovirus genomic
DNA for recombinant viral vector production. One can simply isolate the
viral DNA from E. coli, mix it with a transfer plasmid, and then use the
mixture to cotransfect insect cells. Homologous recombination between
the viral and transfer plasmid DNAs will simultaneously restore orf 1629,
knockout the bacterial replicon in the polyhedrin locus, and knock-in the
gene of interest at this same site. While this approach does not improve the
efficiency of homologous recombination, it can provide an extremely high
efficiency of recombinant baculovirus production by the cotransfected
insect cells because the parental viral DNA is defective and cannot replicate
on its own. This approach has been commercialized as flashBACTM
by Oxford Expression Technologies. One feature of the flashBACTM sys-
tem that warrants additional emphasis is that the bacterial replicon is deleted
from the bacmid upon homologous recombination with the transfer
plasmid, as mentioned above. This is advantageous because it eliminates
potential problems with the genetic stability of baculovirus expression
vectors produced using conventional bacmids, which retain the bacterial
replicon within their genomes (Pijlman et al., 2003). Another feature that
warrants additional emphasis is that the bacmid used in the flashBACTM
system consists of a defective baculovirus genome that cannot replicate in
insect cells. Clearly, this means that homologous recombination between
the defective parental viral genome and the transfer plasmid is required to
produce helper-independent baculovirus progeny in this system. However,
it is important to recognize that this does not necessarily mean that the
cotransfected insect cells will produce only recombinant baculovirus prog-
eny. These cells also can produce progeny derived from the defective
parental viral genome as a result of genetic complementation. Specifically,
the recombinant virus produced in the cotransfected insect cells could
provide the orf 1629 product as a helper function needed to package defec-
tive viral DNAs in trans. Thus, while Oxford Expression Systems’ commer-
cial literature indicates that it is not necessary, baculovirus vectors produced
using the flashBACTM system should be subjected to at least one round of
plaque purification. Without this step, primary recombinant virus stocks
produced by cotransfected insect cells are likely to be contaminated with
defective, parental viruses that could interfere with downstream vector
replication and foreign gene expression.
Arguably, the simplest way to produce baculovirus vectors is to use a
cross-hybridization of the basic ideas underlying GatewayÒ technology
(Hartley, 2003; Walhout et al., 2000) and the linearized parental viral
DNA approach, which involves in vitro, rather than in vivo recombination
between a GatewayÒ entry plasmid encoding a protein of interest and the
linearized parental viral DNA. In this system, the parental viral DNA is a
206 Donald L. Jarvis

prelinearized baculovirus genome in which the polyhedrin coding sequence

is replaced by an E. coli LacZ gene and a herpes simplex virus thymidine
kinase gene flanked by bacteriophage l site-specific recombination sites
(attR1 and attR2; Fig. 14.7). The LacZ gene contributes a blue plaque
phenotype and the thymidine kinase gene contributes a marker that confers
negative selection against the parental viral DNA in insect cells treated with
certain nucleoside analogs, such as gancyclovir. One can mix this prelinear-
ized viral DNA with an entry plasmid encoding a protein of interest flanked
by lambda virus site-specific recombination sites (attL1 and attL2), add a
purified recombinase (LR ClonaseTM; Invitrogen), and this enzyme will
mediate site-specific recombination between the att sites in the entry plasmid
and parental viral DNA. It is important to note that this in vitro reaction
provides a high, but not quantitative efficiency of recombination and there-
fore yields a mixture of parental and recombinant baculovirus DNAs. This
mixture is subsequently transfected into insect cells, which are cultured in the
presence of gancyclovir to select against replication of the parental viral
DNAs, as noted above. One round of selection gives rise to mostly recom-
binant baculovirus vector progeny, which can be resolved from a relatively
low level of parental progeny by performing a plaque assay. Putative recom-
binants can be easily identified by their white-plaque phenotypes in the
presence of a chromogenic substrate for b-galactosidase. The development
of this in vitro approach for the production of recombinant baculovirus
vectors was initiated by Franke and colleagues at Life Technologies and

BaculoDirect viral DNA Recombinant baculoviral DNA

(linearized, nonreplicative) Foreign gene (replicative)
attL1 attL2

Donor
Site-specific
Polyhedrin
(gateway entry clone)
promoter
recombination
Foreign gene
attR1 IE0-HSV-TK p10-LacZ attR2
(in vitro) Polyhedrin
Bsu36I promoter

Insect cell transfection

gancyclovir selection

Mostly recombinant baculoviruses

Clone white plaque recombinant (s)

Figure 14.7 Producing a baculovirus expression vector using the BaculoDirect

approach.
Baculovirus–Insect Cell Expression Systems 207

subsequently commercialized as the BaculoDirectTM system by Harwood

and colleagues at Invitrogen. This system is extremely rapid, efficient, and
simple to use. However, it is important to note that the BaculoDirectTM
system, like the flashBACTM system, is visibly marketed as one that ‘‘can be
used to clone and express genes from baculovirus without plaque purifica-
tion or selection in bacteria.’’ Again, this is not good advice because insect
cells transfected with the in vitro recombination reaction will produce both
recombinant and parental viral progeny, as mentioned above, despite the
application of negative selection pressure against the parental virus by the
thymidine kinase gene product and gancyclovir. Thus, investigators using
the BaculoDirectTM system should ignore the manufacturer’s claims to the
contrary and perform at least one round of plaque purification to resolve the
recombinant baculovirus vector of interest from contaminating parental viral
progeny. In fact, those who follow this advice and incorporate a chromo-
genic b-galactosidase substrate into the agarose overlay used for the plaque
purification process will typically observe at least some blue plaques, which
will indicate the presence of parental viruses and confirm their wise decision
to plaque purify their recombinant baculovirus vector. This process is
facilitated by the commercial BaculoDirectTM instruction manual (Invitro-
gen), which, in fairness, includes protocols for baculovirus plaque purifica-
tion in the presence of a chromogenic b-galactosidase substrate.
In context of the in vitro approach described above, it should be briefly
noted that there are several published examples of direct in vitro recombinant
DNA approaches that involve digestion of modified baculovirus genomic
DNAs followed by direct ligation of the products with foreign DNA
fragments. One example of this type of approach involved the introduction
of two Bsu36I sites downstream of the polyhedrin promoter in the
AcMNPV genome (Lu and Miller, 1996). These sites had slightly different
recognition sequences and, upon digestion with Bsu36I, produced single-
stranded 50 -TTA-30 overhanging sequences that could be converted to
50 -TT-30 with dTTP and the Klenow fragment of E. coli DNA polymerase I.
This created a linear baculoviral DNA molecule that could be ligated with
any DNA fragment that had been digested with EcoRI to generate single-
stranded 50 -AATT-30 overhanging sequences and treated with Klenow and
dATP to convert those ends to 50 -AA-30 . The focused purpose in develop-
ing this approach was to facilitate the production of baculovirus-based
cDNA expression libraries. The other two-published examples of direct
cloning approaches both involved the introduction of homing endonucle-
ase sites into the baculovirus genome. One involved the introduction of a
unique I-Sce-I site into the AcMNPV genome to produce a recombinant
designated Ac-Omega (Ernst et al., 1994). Genomic DNA isolated from this
virus could be digested with I-Sce-I and then ligated with a DNA fragment
produced using the same enzyme. The other example, designated the
Homingbac system, involved the introduction of a unique I-Ceu-I site into
208 Donald L. Jarvis

the genomes of several different baculoviruses (Lihoradova et al., 2007).

Genomic DNA isolated from these viruses could be linearized with I-Ceu-I
and ligated with a DNA fragment containing compatible overhangs produced
using BstXI. At this time, none of the three direct cloning vectors described in
this paragraph have been commercialized, nor have they been widely used in
published studies. Thus, it is difficult to evaluate the relative success of these
direct cloning approaches or the general utility of these parental baculovirus
backbones for the production of baculovirus expression vectors.
In addition to solving the early technical problems associated with the
production of recombinant baculovirus vectors, the purpose of some paren-
tal baculovirus genome modifications was to enhance foreign protein pro-
duction by the recombinant vectors. The basic approach was to delete
baculovirus genes that were (1) known to be nonessential for baculovirus
replication in cultured insect cells, provisionally defined as ‘‘accessory’’
genes and (2) thought to interfere in some way with foreign protein
production and/or to degrade the foreign protein of interest. The first
baculoviral DNA of this type was a linearized viral DNA developed by
Bishop and his colleagues and commercialized as BacVectorÒ -2000 by
Novagen. In addition to polyhedrin, this viral DNA lacked five other
undisclosed baculoviral accessory genes. The impact of these deletions on
foreign protein production was and is unclear, as no studies comparing the
performance of a matched pair of recombinant baculoviruses produced
using viral DNAs with and without these deletions have been published.
Subsequently, two additional viral genes, one encoding a chitinase
(Hawtin et al., 1995) and the other a cathepsin-like protease (Slack et al.,
1995), were deleted from BacVectorÒ -2000 to produce an AcMNPV-
based parental viral DNA called BacVectorÒ -3000 (Novagen). Other
AcMNPV-based viral DNAs lacking the viral chitinase and cathepsin-like
protease genes include a commercial bacmid called BestBac (Expression
Systems) and a noncommercial modified bacmid, AcBacDCC (Kaba et al.,
2004). The bacmid used in the flashBACTM system also lacks a functional
viral chitinase gene. The impact of the viral chitinase and cathepsin-like
protease gene deletions in these different parental viral DNAs on foreign
gene expression by their recombinant baculoviral vector offspring is not
totally clear, but at least some information is available. One study showed
that there was less degradation of a foreign glycoprotein produced by
AcBacDCC, as compared to a matched baculovirus vector without the
viral chitinase and cathepsin-like protease deletions (Kaba et al., 2004).
This is consistent with the seemingly obvious expectation that deleting a
viral protease gene would have a generally positive impact on recombinant
protein and secretory pathway protein production. However, more work
needs to be done to determine if the deletion of this particular protease gene
broadly impacts foreign protein production in the baculovirus–insect cell
system. The potential impact of deleting the viral chitinase gene is less
Baculovirus–Insect Cell Expression Systems 209

obvious. It has been suggested that the viral chitinase gene product, which is
a resident ER protein (Thomas et al., 1998), interferes with the production
of secretory pathway proteins by contributing to the saturation of host
protein translocation machinery (Kaba et al., 2004). If this is true, then
deletion of the viral chitinase gene might be expected to increase secretory
pathway protein production levels by eliminating production of this prod-
uct. However, there are no published studies documenting the impact of
deleting the viral chitinase gene, alone, in an AcMNPV-based vector such
as flashBACTM, and the study documenting the reduced degradation of a
foreign glycoprotein produced by AcBacDCC is complicated by the fact
that this vector lacks both the viral chitinase and cathepsin-like protease
genes. On the other hand, one publication has documented the impact of
deleting the viral chitinase gene, cathepsin-like protease gene, or both, on
the production of a foreign protein by recombinant silkworm baculoviruses
(Bombyx mori nucleopolyhedrovirus; BmNPV; Lee et al., 2006). In this
system, recombinant BmNPV vectors encoding an insect cellulase pro-
duced higher levels of both the foreign protein and its enzymatic activity
in silkworms when either the chitinase or the cathepsin-like protease genes
were deleted. Furthermore, a vector lacking both of these viral genes
produced the highest levels of cellulase protein and enzyme activity. One
could speculate that these results indicate that deleting the viral chitinase
gene from AcMNPV-based vectors would have similar benefits. However,
this speculation would be weakened by the fact that silkworms, not insect
cell lines, were used as the hosts in this study.
DiamondBacTM is another example of a parental baculovirus DNA that has
a deletion in an accessory gene to potentially enhance recombinant protein
production in the baculovirus–insect cell system. As mentioned above, Dia-
mondBacTM is a commercial, prelinearized baculovirus DNA analogous to
BakPAK6, but this viral DNA also lacks a functional p10 gene, which has been
implicated in host cell lysis (Williams et al., 1989). Accordingly, the manufac-
turer’s commercial literature suggests that cells infected with a recombinant
baculovirus vector produced using this parental virus retain higher cell viabil-
ities throughout the course of infection, which contributes to higher recom-
binant protein production levels (Sigma-Aldrich, 2008). Another interesting
and potentially useful feature of DiamondBacTM is that the p10 gene in this
baculoviral genome has been replaced by a protein disulfide isomerase (PDI)
gene, which encodes a chaperone that drives disulfide bonding and contributes
to protein folding. Published evidence suggests that coexpressing PDI in the
baculovirus–insect cell system increases the solubility and secretion of recom-
binant IgG (Hsu et al., 1996). The commercial literature from Sigma-Aldrich
states that the p10 gene deletion and PDI gene insertion provide ‘‘up to a 10-
fold increase in overall protein production for many recombinant proteins’’
(Sigma-Aldrich, 2008). However, there are no published studies comparing
foreign protein production levels by matched pairs of recombinant baculovirus
210 Donald L. Jarvis

vectors with and without the p10 gene deletion and/or the PDI insertion.
Thus, in the final analysis, directed studies with appropriately matched pairs of
AcMNPV-based vectors will be needed to directly determine the general
impact of deleting nonessential viral genes, such as the viral chitinase, cathep-
sin-like protease, and p10 genes, on foreign protein production in the
baculovirus–insect cell system.
In addition to the DiamondBacTM parental baculovirus DNA, other
baculovirus vectors with gene insertions encoding heterologous protein pro-
cessing enzymes designed to enhance foreign protein production have been
described in the literature. One is a recombinant baculovirus that encodes a
polydnavirus vankyrin gene under the control of the p10 promoter (Fath-
Goodin et al., 2006). It has been found that the expression of certain vankyrin
gene products prolongs the viability of baculovirus-infected Sf 9 cells and this,
in turn, can increase the amounts of a foreign protein produced by coexpres-
sion under the control of the polyhedrin promoter. Other types of baculovirus
vectors that fall within this genre include those encoding heterologous glyco-
syltransferases ( Jarvis et al., 2001; Tomiya et al., 2003) or enzymes involved in
CMP-sialic acid biosynthesis (Hill et al., 2006) under the control of baculo-
virus ie1 promoters. The use of the immediate-early promoter in these vectors
allows the newly introduced processing functions to be expressed early in the
infection cycle, well before expression of the foreign gene encoding the
protein of interest. This allows for functional extension of the endogenous
insect cell protein N-glycosylation pathway and the establishment of a path-
way capable of producing glycoproteins with humanized carbohydrate side
chains prior to the onset of production of the glycoprotein of interest.
A transfer plasmid that can be used to simultaneously introduce a poly-
dnavirus vankyrin gene and a foreign gene encoding a protein of interest into
the baculovirus genome is commercially available from Paratechs. Similarly,
parental baculoviral DNAs encoding higher eukaryotic N-glycosylation
functions are likely to be commercially available for recombinant baculovirus
vector isolation in the near future. These tools will facilitate a more global
analysis of the capabilities of these types of vectors in increasing expression
levels, on the one hand, or producing foreign glycoproteins with humanized
carbohydrate side chains, on the other, as neither has been broadly assessed
with a wide array of different foreign protein/glycoprotein products.

8. The Other Half of the Baculovirus–Insect

Cell System
Considering the focus on baculovirus expression vectors to this point
in this chapter, it might be necessary to remind the reader that the
baculovirus–insect cell system is a binary system that consists of two essential
Baculovirus–Insect Cell Expression Systems 211

components. The first one is, of course, the baculovirus expression vector,
which is an insect virus whose obvious function is to deliver a foreign
gene(s) encoding a protein(s) of interest to the host. In addition, the
baculovirus expression vector typically serves another function, which is
to induce the production of the transcriptional complex needed to tran-
scribe the foreign gene of interest under the control of a late or very late
baculovirus promoter. The second component, which has been given
almost no attention so far, is the host, which is typically a lepidopteran
insect cell line, but alternatively can be a lepidopteran insect.
The Order Lepidoptera includes the moths and butterflies, which are the
hosts for many viruses in the Family Baculoviridae, including AcMNPV.
Grace described the first-established lepidopteran insect cell cultures in 1962
(Grace, 1962) and as of today, over 250 insect cell lines have been described
(Lynn, 2007). However, we can focus on just two lepidopteran insect cell
lines, which are the two most commonly used hosts for baculovirus expres-
sion vectors. One is Sf 9, a cell line described by Smith and Summers in
1987 (Summers and Smith, 1987) as a subclone of IPLB-SF-21, which is a
line that had been isolated from pupal ovarian tissue of Spodoptera frugiperda
(fall armyworm) in 1977 (Vaughn et al., 1977). Sf 9 cells can be obtained
from several different companies, including Invitrogen and Novagen, or the
American Type Culture Collection (ATCC). The other is High FiveTM, a
cell line originally isolated from adult ovarian tissue of Trichoplusia ni
(cabbage looper) and described as BTI Tn 5B-1 by Granados’ and
Wood’s groups in 1992 (Davis et al., 1992; Wickham et al., 1992) and
then commercialized by Invitrogen as High FiveTM.
One interesting and useful feature of Sf 9 and High FiveTM cells is that
they can grow in either adherent or suspension formats. Thus, it is conve-
nient to perform lab-scale protein production experiments by infecting a
few million Sf 9 or High FiveTM cells as adherent cultures in plates or flasks.
Adherent cultures of Sf 21 or Sf 9 cells also are routinely used to plaque
purify and quantify recombinant baculovirus expression vectors. On the
other hand, both Sf 9 and High FiveTM cells can be scaled up to varying
degrees in spinner flasks, shake flasks, or in airlift, stirred tank (Elias et al.,
2007), or Wave (Kadwell and Hardwicke, 2007) bioreactors to produce
larger amounts of recombinant proteins.
The conditions and methods used to culture insect cells are quite
different from those used to culture mammalian cells, which are usually
more familiar to most investigators. For example, the optimal temperature
for insect cell culture is 28  C, rather than 37  C. In addition, both Sf 9 and
High FiveTM cells are loosely adherent and neither EDTA nor trypsin is
required for their subculture. It is not necessary to have a CO2 incubator to
grow insect cells because insect cell culture media are buffered with phos-
phate, rather than carbonate. Sf 9 and High FiveTM cells both can be cul-
tured in either growth media supplemented with serum or in serum-free
212 Donald L. Jarvis

media and both types of growth media are now widely available. In fact,
since the publication of the first edition of this book in 1990, many different
insect cell culture media have become available from many different man-
ufacturers, including Expression Systems, Invitrogen (GIBCO), HyClone,
and Sigma-Aldrich, among others. However, it is important to note that
Sf 9 and High FiveTM cells tend to become adapted to a specific growth
medium. Thus, the abrupt transition of an Sf 9 or High FiveTM culture from
a growth medium supplemented with serum to a serum-free medium can
result in loss of the culture. A much higher rate of success can be achieved
by slowly weaning the cells out of serum containing and into serum-free
medium. For example, one can increase the proportion of serum-free
medium from 0 to 25, 50%, 75, and 100% over four consecutive passages.
Even with this relatively slow type of weaning process, these cells typically
undergo a temporary arrest or grow very slowly when first placed into the
100% serum-free medium. Thus, it is important to be patient, as the cells
will typically recover after a lag of a day or two and begin growing normally
again in the new medium. This type of weaning process is not only useful
for the transition from a serum containing to a serum-free medium, but is
also sometimes needed to move insect cell cultures from one specific type of
serum-free medium into another.

9. A New Generation of Insect Cell Hosts

for Baculovirus Expression Vectors
Technology for the genetic transformation of lepidopteran insect cells
was first described in 1990 ( Jarvis et al., 1990). At that time, the main reason
for developing this technology was to enable the creation of stably trans-
formed insect cell lines that would constitutively produce a recombinant
protein of interest in the absence of baculovirus infection. However, the
vectors and methods used to create these ‘‘producer’’ cell lines also set the
stage for the creation of transgenic insect cell lines with new traits that
would improve their ability to support recombinant baculovirus-mediated
foreign protein production. The first example of this type of host cell
improvement was directed toward an issue raised earlier in this chapter,
which is that the baculovirus–insect cell system has eukaryotic protein
processing capabilities, but these are not always identical to those of higher
eukaryotes. One of the best-recognized examples of this limitation is
the ability of both Sf 9 and High FiveTM cells to N-glycosylate newly
synthesized proteins, but their inability to process N-linked carbohydrate
side chains (N-glycans) to the same extent as mammalian cells (reviewed in
Harrison and Jarvis, 2006, 2007b; Shi and Jarvis, 2007). The first cell-based
approach used to begin to address this problem was to genetically transform
Baculovirus–Insect Cell Expression Systems 213

Sf 9 cells with a bovine b1,4-galactosyltransferase cDNA placed under the

control of the AcMNPV ie1 promoter (Hollister et al., 1998). The addition of
this cDNA, which encodes an N-glycan processing enzyme that is not
present in Sf 9 cells, was designed to extend the endogenous processing
pathway and permit the transgenic cell line to produce N-glycoproteins
with more extensively processed, partially ‘‘humanized’’ N-glycans. Indeed,
the resulting cell line, designated Sfb4GalT, encoded and constitutively
expressed this mammalian gene and, unlike the parental cell line, was able
to produce foreign glycoproteins with partially humanized N-glycans. How-
ever, this was only a first step, as further humanization of the Sf 9 cell protein
N-glycosylation pathway of Sf 9 cells, which involved knocking-in several
additional mammalian functions, was required. In general, this was accom-
plished by constructing ie1-driven mammalian genes encoding these addi-
tional functions and using the resulting constructs to produce additional
transgenic Sf 9 cell subclones (Aumiller et al., 2003; Hollister and Jarvis,
2001; Hollister et al., 2002; Seo et al., 2001). Each of these new insect cell
lines retained the ability to support baculovirus replication and baculovirus-
mediated foreign gene expression and encoded and constitutively expressed
the mammalian genes with which they had been transformed. Finally, each of
these transgenic insect cell lines also had more extensive N-glycan processing
capabilities than the parental Sf 9 cells and could produce recombinant
glycoproteins with increasingly humanized N-glycans. An analogous
transgenic subclone of High FiveTM cells was produced using ie1-
based transgenes encoding two different mammalian glycosyltransferases
(Breitbach and Jarvis, 2001). One Sf 9-based transgenic insect cell line
originally designated Sf SWT-1, which has an extended N-glycosylation
pathway and can produce humanized recombinant glycoproteins when
cultured in the presence of serum (Hollister et al., 2002), has been commer-
cialized by Invitrogen under the trade name ‘‘MIMICTM.’’ More advanced
transgenic insect cell lines with a larger repertoire of mammalian genes,
which are capable of producing recombinant glycoproteins with terminally
sialylated N-glycans when cultured in serum-free media (Aumiller et al.,
2003), are likely to be commercially available in the near future.
Another transgenic lepidopteran insect cell line that holds promise as an
improved host for baculovirus expression vectors is an Sf 9 cell derivative
transformed with a polydnavirus vankyrin gene under the control of an
immediate-early baculovirus promoter (Fath-Goodin et al., 2006). As dis-
cussed above, baculovirus-mediated expression of this polydnavirus vankyrin
gene prolonged the viability of baculovirus-infected Sf 9 cells and increased
the production of a foreign protein coexpressed under the control of the
polyhedrin promoter. By analogy, a transgenic Sf 9 cell derivative engineered
to constitutively express this vankryin gene had a longer life span after and
produced higher levels of yellow fluorescence protein when infected with
a recombinant baculovirus expression vector encoding this product
214 Donald L. Jarvis

(Fath-Goodin et al., 2006). Three different ‘‘vankyrin-enhanced’’ (VE) Sf 9

derivatives are now commercially available from ParaTechs.
The availability of MIMICTM and VE cells and the future availability of
other transgenic insect cell lines of these types will permit a more global
analysis of their abilities to produce foreign glycoproteins with humanized
carbohydrate side chains or to increase recombinant protein expression
levels, respectively. This is important because neither one of these two
different types of ‘‘improved’’ hosts for baculovirus expression vectors has
been broadly assessed using a wide variety of different foreign protein/
glycoprotein products. In fact, there is one published report of the
inability of MIMICTM (SfSWT-1) cells to sialylate one specific foreign
glycoprotein, equine lutenizing hormone/chorionic gondotropin
(Legardinier et al., 2005). These two equine glycopeptide hormones are
encoded by the same gene and, therefore, have identical amino acid
sequences. From a careful analysis of the N-glycosylation profiles of the
native products (Smith et al., 1993), one would expect the recombinant
hormone produced by MIMICTM (SfSWT-1) cells to have terminal alpha
2,3-linked sialic acid residues. Thus, the results of Legardinier and coworkers
most likely reflect the inability of MIMICTM (SfSWT-1) cells to sialylate
equine lutenizing hormone/chorionic gonadotropin in alpha 2,3-linked
fashion, which is a capability that had not been assessed before. In fact, we
recently found that MIMICTM (SfSWT-1) cells have no alpha 2,3-sialyl-
transferase activity, despite the fact that they encode and express a murine
alpha 2,3-sialyltransferase (data not shown). Thus, it is now clear that
MIMICTM (Sf SWT-1) cells do not have the universal capacity to sialylate
every foreign N-glycoprotein of interest and efforts are underway to produce
a new cell line that can provide both alpha 2,6- and alpha 2,3-sialylation.
However, even a new transgenic insect cell line with alpha 2,3-sialyltrans-
ferase activity might not have the capacity to sialylate every recombinant
glycoprotein of interest. Furthermore, it will not be surprising if this limita-
tion also applies to VE cells or any other type of transgenic insect cell line
engineered to serve as an improved host for baculovirus expression vectors.

10. Basic Baculovirus Protocols

Chapter 10 in the first edition of this book contained a significant
number of detailed protocols describing the construction of baculovirus
transfer plasmids, insect cell culture, the production, isolation and character-
ization of baculovirus expression vectors, and methods for the analysis of
recombinant protein production in the baculovirus–insect cell system
(Bradley, 1990), which are still relevant today. In addition, a large number
of other books and book chapters containing detailed protocols describing
these technical aspects of the baculovirus–insect cell system have been
Baculovirus–Insect Cell Expression Systems 215

published before and since that time (King and Possee, 1992; Murhammer,
2007; O’Reilly et al., 1992; Richardson, 1995; Summers and Smith, 1987).
Finally, each manufacturer of the new baculovirus transfer plasmids and
genome backbones described in this chapter, which were created to facilitate
the production and isolation of recombinant baculovirus expression vectors,
provides detailed maps, sequences, and stepwise protocols for the use of their
materials, which are freely available at their Web sites. Accordingly, I have
chosen to share a relatively small, but essential set of time-tested protocols
used in my lab, which were originally taken from the original baculovirus
users ‘‘manuals’’ by Summers and Smith (1987) and O’Reilly et al. (1992)
and, in some cases, modified to fit our needs. The reader is referred to the
sources given above for more direct and detailed information on any of the
specific approaches, transfer plasmids, and parental baculovirus vectors
described in this chapter, which need not be reiterated here.

10.1. Insect cell maintenance

We routinely maintain 50 mL Sf 9 cell cultures in 125 mL DeLong shake
flasks (Bellco) in either serum-containing or serum-free media. We use
TNM-FH medium supplemented with 10% (v/v) fetal bovine serum
(HyClone) and 0.1% pluronic F68 (Sigma-Aldrich) for the former and we
use ESF 921 (Expression Systems) for the latter. We do not heat-inactivate
the fetal bovine serum nor do we use cell culture antibiotics in our mainte-
nance cultures, as they can interfere with the creation of transgenic subclones
by our standard coselection approach (Harrison and Jarvis, 2007a; Jarvis and
Guarino, 1995). We split all of our maintenance cultures each Monday,
Wednesday, and Friday of the week, using seeding densities for Sf 9 cells of
0.5  106 cells/mL on Mondays and Wednesdays and 0.3  106 cells/mL on
Fridays. We measure cell densities using a standard trypan blue dye exclusion
procedure with a hemocytometer (Murhammer, 2007). Cell viabilities
should remain very high in shake flask cultures and Sf 9 cell doubling times
should be around 24 h. For larger numbers of cells, the culture can be scaled
up to 100 mL in 250 mL DeLong flasks, 200 mL in 500 mL DeLong flasks, or
800 mL in 2.8 L Fernbach flasks (Bellco). The seeding densities used for
transgenic insect cell lines are variable and typically higher than those used
for Sf 9 cells. In our recent experience, transgenic lines seeded at the same
densities used for Sf 9 cells tend to grow slowly and/or undergo a temporary
arrest before starting to grow at normal rates again.

10.2. Isolation of baculovirus genomic DNA

Unless you are in a position to constantly purchase prepurified viral DNA
from commercial sources, the isolation of adequately purified baculovirus
genomic DNA is a critical aspect of producing recombinant baculovirus
216 Donald L. Jarvis

expression vectors. We prepare viral DNA from budded virus progeny

isolated using a differential centrifugation protocol. Sf 9 cells are infected
with an appropriate parental baculovirus using a low input multiplicity (e.g.,
0.01 plaque-forming units/cell) to avoid the generation of defective inter-
fering particles (Kool et al., 1991). The cells are then cultured for 3–5 days in
TNM-FH medium supplemented with fetal bovine serum and pluronic
F68, as described above, and monitored under a phase-contrast microscope
for signs of cytopathology. Once clear signs of infection are observed, the
infected cells are aseptically transferred to disposable centrifuge tubes and
then centrifuged at  1000g for 15 min. The clear supernatants are trans-
ferred to ultracentrifuge tubes (e.g., 33 mL per Beckman SW28 tube) and
then carefully underlaid with a solution of 25% (w/v) sucrose in 5 mM NaCl
and 10 mM EDTA (e.g., 3 mL per Beckman SW28 tube). The tubes are
ultracentrifuged for 75 min at 4  C at  100,000g (e.g., 28,000 rpm for a
Beckman SW28 rotor) and the supernatant is carefully decanted away from
the resulting budded virus pellet. The pellet is then gently resuspended using
a blue tip with the narrow end cut off to produce a larger bore diameter and a
disruption buffer (0.1 mL for each mL of starting infected cell culture
medium) consisting of 10 mM Tris, pH 7.6, 10 mM EDTA, and 0.25%
(w/v) sodium dodecyl sulfate. Proteinase K (10 mg/mL in water) is then
added to a final concentration of 500 ug/mL and the preparation is incubated
at 37  C with occasional swirling until it clears, which typically requires 4 h
to overnight. If the solution does not clear, one should add more disruption
buffer and proteinase K. The solution is then gently extracted with phenol,
then phenol–chloroform and the viral DNA is ethanol precipitated in
the presence of 0.3 M sodium acetate. Finally, the DNA pellet is gently
resuspended in 10 uL TE for each mL of starting culture medium. The
concentration of the resuspended viral DNA can be estimated using spec-
trophotometry, but we find this is highly unreliable. Thus, we typically
refine the estimate by digesting 10 ug of the viral DNA with EcoRI or
HindIII, running the digests on an agarose gel against commercial DNA
size standards of known concentrations, and staining the gel with ethidium
bromide or another DNA stain. It should be noted that this protocol is
essentially the same as the one described in the Baculovirus Expression
Vectors manual by O’Reilly and coworkers (O’Reilly et al., 1992).

10.3. Baculovirus plaque assays

Plaque assays are a critical aspect of virology, in general, and are the single
best way to ensure that a recombinant baculovirus is initially isolated in a
clonal form. Two-independent steps that need to be taken in preparation
for baculovirus plaque assays include (1) preseeding the indicator cells
into culture plates and (2) making serial 10-fold dilutions of the virus
stock. We seed the cells first from a log-phase shake flask culture of Sf 9
Baculovirus–Insect Cell Expression Systems 217

cells in TNM-FH containing 10% fetal bovine serum and 0.1% pluronic
F68, as described above. The requisite number of cells is gently centri-
fuged out of the old growth medium, resuspended in fresh TNM-FH with
no additives, and then seeded into 6-well cell culture plates (Corning) at a
density of 0.75  106 cells/well. The cells are then allowed to settle and
attach to the plastic at 28  C for about an hour. The cells will adhere more
tightly to the plastic when seeded in the absence than when seeded in the
presence of serum. While the cells are attaching, the serial 10-fold dilutions
of the virus stock are performed, working under the assumption that a
typical baculovirus stock will have a plaque assay titer of  1  107 pfu/
mL. We use TNM-FH supplemented with fetal bovine serum and
pluronic F68 as the diluent, with 0.1 mL of virus and 9.9 mL dilution
blanks for 100-fold dilutions and 0.5 mL of virus and 4.5 mL dilution
blanks for 10-fold dilutions. The dilution blanks are set up and the
dilutions performed under aseptic conditions and each dilution is mixed
very well with a vortex mixer between each transfer. After the dilutions
have been performed and the cells have attached to the culture plates, the
cells are inoculated using two wells for each appropriately diluted virus.
The medium is removed from the wells and replaced with 2 mL/well of
the diluted virus. This is a critical step in the assay because the cells dry out
rather quickly and those along the edges will die if you move too slowly.
Thus, it is best to drain the media from only two wells at a time and
immediately inoculate those wells with one virus dilution in duplicate
before moving on to another pair of wells and a new dilution. We
typically plate the virus at dilutions of 10 5, 10 6, and 10 7 to try to
obtain reasonable numbers of plaques for counting. The virus is then
allowed to attach to the cells for  1 h at 28  C and the overlay medium
is prepared during the viral adsorption period. We prepare 125 mL bottles
containing 50 mL of 2% (w/v) SeaplaqueÒ low melting temperature
agarose (Cambrex) in water, then autoclave the bottles and allow the
agarose to solidify. These bottles of solid agarose then can be microwaved
to remelt the agarose, cooled to  60  C, and the liquid agarose can be
mixed with an equal volume of 2 Grace’s medium prewarmed to
 30  C (you will need 3 mL per well). If a blue–white screen is being
used to help identify putative recombinant viruses, the chromogenic
substrate should be added to the overlay at this time. The overlay is then
swirled to obtain an even mixture and cooled to about 40–42  C. Finally,
the infected cells are removed from the incubator, the viral inocula
are completely removed with pipettes, and 3 mL of overlay are very gentle
dribbled down the edge of the well. Again, draining two wells at a time
and overlaying those wells with agarose before moving on to another pair
of wells will prevent the cells from drying out and dying. After the overlay
hardens for  15 min, the plates are incubate in sealed plastic baggie,
upside down, for 7–10 days at 28  C. A dissecting microscope is used to
218 Donald L. Jarvis

visualize plaques for careful counting and the plaque counts are used to
calculate the titer of the original virus stock, taking the dilution factor and
plating volume into account.

REFERENCES
Aumiller, J. J., Hollister, J. R., and Jarvis, D. L. (2003). A transgenic lepidopteran insect cell
line engineered to produce CMP-sialic acid and sialoglycoproteins. Glycobiology 13,
497–507.
Azzouz, N., Kedees, M. H., Gerold, P., Becker, S., Dubremetz, J. F., Klenk, H. D., and
Eckert, V. R. T. S. (2000). An early step of glycosylphosphatidyl-inositol anchor
biosynthesis is abolished in lepidopteran insect cells following baculovirus infection.
Glycobiology 10, 177–183.
Berger, I., Fitzgerald, D. J., and Richmond, T. J. (2004). Baculovirus expression system for
heterologous multiprotein complexes. Nat. Biotechnol. 22, 1583–1587.
Bradley, M. K. (1990). Overexpression of proteins in eukaryotes. Methods Enzymol. 182,
112–132.
Breitbach, K., and Jarvis, D. L. (2001). Improved glycosylation of a foreign protein by
Tn-5B1–4 cells engineered to express mammalian glycosyltransferases. Biotechnol. Bioeng.
74, 230–239.
Chazenbalk, G. D., and Rapoport, B. (1995). Expression of the extracellular domain of the
thyrotropin receptor in the baculovirus system using a promoter active earlier than the
polyhedrin promoter Implications for the expression of functional highly glycosylated
proteins. J. Biol. Chem. 270, 1543–1549.
Davis, T. R., Trotter, K. M., Granados, R. R., and Wood, H. A. (1992). Baculovirus
expression of alkaline phosphatase as a reporter gene for evaluation of production,
glycosylation and secretion. Nat. Biotechnol. 10, 1148–1150.
Douris, V., Swevers, L., Labropoulou, V., Andronopoulou, E., Georgoussi, Z., and
Iatrou, K. (2006). Stably transformed insect cell lines: Tools for expression of secreted
and membrane-anchored proteins and high-throughput screening platforms for drug and
insecticide discovery. Adv. Virus Res. 68, 113–156.
Elias, C. B., Jardin, B., and Kamen, A. (2007). Recombinant protein production in large-
scale agitated bioreactors using the baculovirus expression vector system. Methods Mol.
Biol. 388, 225–246.
Ernst, W. J., Grabherr, R. M., and Katinger, H. W. (1994). Direct cloning into the
Autographa californica nuclear polyhedrosis virus for generation of recombinant
baculoviruses. Nucleic Acids Res. 22, 2855–2856.
Fath-Goodin, A., Kroemer, J., Martin, S., Reeves, K., and Webb, B. A. (2006). Polydna-
virus genes that enhance the baculovirus expression vector system. Adv. Virus Res. 68,
75–90.
Friesen, P. D. (1997). Regulation of baculovirus early gene expression. In ‘‘The Baculo-
viruses’’ (L. K. Miller, ed.), pp. 141–170. Plenum Press, New York.
Grace, T. D. C. (1962). Establishment of four strains of cells from insect tissues grown in vitro.
Nature 195, 788–789.
Harrison, R. L., and Jarvis, D. L. (2006). Protein N-glycosylation in the baculovirus-insect
cell expression system and efforts to engineer insect cells to produce ‘‘mammalianized’’
recombinant glycoproteins. Adv. Virus Res. 68, 159–191.
Harrison, R. L., and Jarvis, D. L. (2007a). Transforming lepidopteran insect cells for
continuous recombinant protein expression. Methods Mol. Biol. 388, 299–316.
Baculovirus–Insect Cell Expression Systems 219

Harrison, R. L., and Jarvis, D. L. (2007b). Transforming lepidopteran insect cells for
improved protein processing. Methods Mol. Biol. 388, 341–356.
Strausberg, R. L., and Strausberg, S. L. (2003). Overview of protein expression in Saccha-
romyces cerevisiae. Curr. Protoc. Protein Sci. Chapter 5, Unit 5.17.
Hawtin, R. E., Arnold, K., Ayres, M. D., Zanotto, P. M., Howard, S. C., Gooday, G. W.,
Chappell, L. H., Kitts, P. A., King, L. A., and Possee, R. D. (1995). Identification and
preliminary characterization of a chitinase gene in the Autographa californica nuclear
polyhedrosis virus genome. Virology 212, 673–685.
Hill, D. R., Aumiller, J. J., Shi, X., and Jarvis, D. L. (2006). Isolation and analysis of a
baculovirus vector that supports recombinant glycoprotein sialylation by SfSWT-1 cells
cultured in serum-free medium. Biotechnol. Bioeng. 95, 37–47.
Hill-Perkins, M. S., and Possee, R. D. (1990). A baculovirus expression vector derived from
the basic protein promoter of Autographa californica nuclear polyhedrosis virus. J. Gen.
Virol. 71, 971–976.
Hollister, J., and Jarvis, D. L. (2001). Engineering lepidopteran insect cells for sialoglyco-
protein production by genetic transformation with mammalian b1, 4-galactosyltransfer-
ase and alpha 2, 6-sialyltransferase genes. Glycobiology 11, 1–9.
Hollister, J. R., Grabenhorst, E., Nimtz, M., Conradt, H. O., and Jarvis, D. L. (2002).
Engineering the protein N-glycosylation pathway in insect cells for production of
biantennary, complex N-glycans. Biochemistry 41, 15093–15104.
Hollister, J. R., Shaper, J. H., and Jarvis, D. L. (1998). Stable expression of mammalian beta
1, 4-galactosyltransferase extends the N-glycosylation pathway in insect cells. Glycobiology
8, 473–480.
Hsu, T. A., Watson, S., Eiden, J. J., and Betenbaugh, M. J. (1996). Rescue of immunoglo-
bulins from insolubility is facilitated by PDI in the baculovirus expression system. Prot.
Exp. Purif. 7, 281–288.
Jarvis, D. L. (1997). Baculovirus expression vectors. In ‘‘The Baculoviruses’’ (L. K. Miller,
ed.), pp. 389–431. Plenum Press, New York.
Jarvis, D. L., Fleming, J. A., Kovacs, G. R., Summers, M. D., and Guarino, L. A. (1990). Use
of early baculovirus promoters for continuous expression and efficient processing of
foreign gene products in stably transformed lepidopteran cells. Nat. Biotechnol. 8,
950–955.
Jarvis, D. L., and Guarino, L. A. (1995). Continuous foreign gene expression in transformed
lepidopteran insect cells. In ‘‘Baculovirus Expression Protocols’’ (C. D. Richardson, ed.),
Vol. 39, pp. 187–202. Humana Press, Clifton, NJ.
Jarvis, D. L., Howe, D., and Aumiller, J. J. (2001). Novel baculovirus expression vectors that
provide sialylation of recombinant glycoproteins in lepidopteran insect cells. J. Virol. 75,
6223–6227.
Jarvis, D. L., and Summers, M. D. (1989). Glycosylation and secretion of human tissue
plasminogen activator in recombinant baculovirus-infected insect cells. Mol. Cell. Biol. 9,
214–223.
Jarvis, D. L., Weinkauf, C., and Guarino, L. A. (1996). Immediate early baculovirus vectors
for foreign gene expression in transformed or infected insect cells. Prot. Exp. Purif. 8,
191–203.
Kaba, S. A., Salcedo, A. M., Wafula, P. O., Vlak, J. M., and van Oers, M. M. (2004).
Development of a chitinase and v-cathepsin negative bacmid for improved integrity of
secreted recombinant proteins. J. Virol. Methods 122, 113–118.
Kadwell, S. H., and Hardwicke, P. I. (2007). Production of baculovirus-expressed recombi-
nant proteins in wave bioreactors. Methods Mol. Biol. 388, 247–266.
King, L. A., and Possee, R. D. (1992). The baculovirus system: A laboratory guide.
Chapman and Hall, London.
220 Donald L. Jarvis

Kitts, P. A., Ayres, M. D., and Possee, R. D. (1990). Linearization of baculovirus DNA
enhances the recovery of recombinant virus expression vectors. Nucleic Acids Res. 18,
5667–5672.
Kitts, P. A., and Possee, R. D. (1993). A method for producing recombinant baculovirus
expression vectors at high frequency. Biotechniques 14, 810–817.
Kool, M., Voncken, J. W., van Lier, F. L., Tramper, J., and Vlak, J. M. (1991). Detection
and analysis of Autographa californica nuclear polyhedrosis virus mutants with defective
interfering properties. Virology 183, 739–746.
Kost, T. A., Condreay, J. P., and Jarvis, D. L. (2005). Baculovirus as versatile vectors for
protein expression in insect and mammalian cells. Nat. Biotechnol. 23, 567–575.
Lee, K. S., Je, Y. H., Woo, S. D., Sohn, H. D., and Jin, B. R. (2006). Production of a
cellulase in silkworm larvae using a recombinant Bombyx mori nucleopolyhedrovirus
lacking the virus-encoded chitinase and cathepsin genes. Biotechnol. Lett. 28, 645–650.
Legardinier, S., Klett, D., Poirier, J. C., Combarnous, Y., and Cahoreau, C. (2005).
Mammalian-like nonsialyl complex-type N-glycosylation of equine gonadotropins in
Mimic insect cells. Glycobiology 15, 776–790.
Lihoradova, O. A., Ogay, I. D., Abdukarimov, A. A., Azimova Sh, S., Lynn, D. E., and
Slack, J. M. (2007). The Homingbac baculovirus cloning system: An alternative way to
introduce foreign DNA into baculovirus genomes. J. Virol. Methods 140, 59–65.
Lu, A., and Miller, L. K. (1996). Generation of recombinant baculoviruses by direct cloning.
Biotechniques 21, 63–68.
Lu, A., and Miller, L. K. (1997). Regulation of baculovirus late and very late gene
expression. In ‘‘The Baculoviruses’’ (L. K. Miller, ed.), pp. 193–216. Plenum Press,
New York.
Luckow, V. A., Lee, S. C., Barry, G. F., and Olins, P. O. (1993). Efficient generation of
infectious recombinant baculoviruses by site-specific transposon-mediated insertion of
foreign genes into a baculovirus genome propagated in Escherichia coli. J. Virol. 67,
4566–4579.
Lynn, D. E. (2007). Available lepidopteran insect cell lines. Methods Mol. Biol. 388, 117–138.
McLachlin, J. R., and Miller, L. K. (1994). Identification and characterization of vlf-1,
a baculovirus gene involved in very late gene expression. J. Virol. 68, 7746–7756.
Miller, L. K. (1997). The Baculoviruses. Plenum Press, New York.
Mistretta, T. A., and Guarino, L. A. (2005). Transcriptional activity of baculovirus very late
factor 1. J. Virol. 79, 1958–1960.
Murhammer, D. W. (2007). Baculovirus and Insect Cell Expression Protocols. Humana
Press, Totowa, NJ.
Murphy, C. I., Lennick, M., Lehar, S. M., Beltz, G. A., and Young, E. (1990). Temporal
expression of HIV-1 envelope proteins in baculovirus-infected insect cells: Implications
for glycosylation and CD4 binding. Gen. Anal. Tech. Appl. 7, 160–171.
Ooi, B. G., Rankin, C., and Miller, L. K. (1989). Downstream sequences augment
transcription from the essential initiation site of a baculovirus polyhedrin gene. J. Mol.
Biol. 210, 721–736.
O’Reilly, D. R., Miller, L. K., and Luckow, V. A. (1992). Baculovirus Expression Vectors.
W. H. Freeman and Company, New York.
Passarelli, A. L., and Guarino, L. A. (2007). Baculovirus late and very late gene regulation.
Curr. Drug Targets 8, 1103–1115.
Pennock, G. D., Shoemaker, C., and Miller, L. K. (1984). Strong and regulated expression
of Escherichia coli beta-galactosidase in insect cells with a baculovirus vector. Mol. Cell.
Biol. 4, 399–406.
Pijlman, G. P., van Schijndel, J. E., and Vlak, J. M. (2003). Spontaneous excision of BAC
vector sequences from bacmid-derived baculovirus expression vectors upon passage in
insect cells. J. Gen. Virol. 84, 2669–2678.
Baculovirus–Insect Cell Expression Systems 221

Possee, R. D., Hitchman, R. B., Richards, K. S., Mann, S. G., Siaterli, E., Nixon, C. P.,
Irving, H., Assenberg, R., Alderton, D., Owens, R. J., and King, L. A. (2008). Genera-
tion of baculovirus vectors for the high-throughput production of proteins in insect cells.
Biotechnol. Bioeng. 101, 1115–1122.
Possee, R. D., and King, L. A. (2007). Baculovirus transfer vectors. Meth. Mol. Biol. 388,
55–76.
Rankl, N. B., Rice, J. W., Gurganus, T. M., Barbee, J. L., and Burns, D. J. (1994). The
production of an active protein kinase C-delta in insect cells is greatly enhanced by the
use of the basic protein promoter. Prot. Exp. Purif. 5, 346–356.
Richardson, C. D. (1995). Baculovirus Expression Protocols. Humana Press, Totowa, NJ.
Seo, N. S., Hollister, J. R., and Jarvis, D. L. (2001). Mammalian glycosyltransferase expres-
sion allows sialoglycoprotein production by baculovirus-infected insect cells. Prot. Exp.
Purif. 22, 234–241.
Shi, X., and Jarvis, D. L. (2007). Protein N-glycosylation in the baculovirus-insect cell
system. Curr. Drug Targets 8, 1116–1125.
Sigma-Aldrich, D. L. (2008). D6192 DiamondBacTM Baculovirus DNA.
Slack, J. M., Kuzio, J., and Faulkner, P. (1995). Characterization of v-cath, a cathepsin L-like
proteinase expressed by the baculovirus Autographa californica multiple nuclear polyhedro-
sis virus. J. Gen. Virol. 76, 1091–1098.
Smith, P. L., Bousfield, G. R., Kumar, S., Fiete, D., and Baenziger, J. U. (1993). Equine
lutropin and chorionic gonadotropin bear oligosaccharides terminating with SO4–4-
GalNAc and Sia alpha 2, 3Gal, respectively. J. Biol. Chem. 268, 795–802.
Smith, G. E., Fraser, M. J., and Summers, M. D. (1983a). Molecular engineering of the
Autographa californica nuclear polyhedrosis virus genome: Deletion mutations within
the polyhedrin gene. J. Virol. 46, 584–593.
Smith, G. E., Summers, M. D., and Fraser, M. J. (1983b). Production of human beta
interferon in insect cells infected with a baculovirus expression vector. Mol. Cell. Biol.
3, 2156–2165.
Sridhar, P., Panda, A. K., Pal, R., Talwar, G. P., and Hasnain, S. E. (1993). Temporal nature
of the promoter and not relative strength determines the expression of an extensively
processed protein in a baculovirus system. FEBS Lett. 315, 282–286.
Summers, M. D. (2006). Milestones leading to the genetic engineering of baculoviruses as
expression vector systems and viral pesticides. Adv. Virus Res. 68, 3–73.
Summers, M. D., and Smith, G. E. (1987). A manual of methods for baculovirus vectors and
insect cell culture procedures. Tx. Agric. Exp. Stn. Bull. No. 1555.
Thomas, C. J., Brown, H. L., Hawes, C. R., Lee, B. Y., Min, M. K., King, L. A., and
Possee, R. D. (1998). Localization of a baculovirus-induced chitinase in the insect cell
endoplasmic reticulum. J. Virol. 72, 10207–10212.
Tomiya, N., Howe, D., Aumiller, J. J., Pathak, M., Park, J., Palter, K., Jarvis, D. L.,
Betenbaugh, M. J., and Lee, Y. C. (2003). Complex-type biantennary N-glycans of
recombinant human transferrin from Trichoplusia ni insect cells expressing mammalian
b1, 4-galactosyltransferase and b1, 2-N-acetylglucosaminyltransferase II. Glycobiology 13,
23–34.
Vaughn, J. L., Goodwin, R. H., Thompkins, G. J., and McCawley, P. (1977). The
establishment of two insect cell lines from the insect Spodoptera frugiperda (Lepidoptera:
Noctuidae). In Vitro 13, 213–217.
Vialard, J., Lalumiere, M., Vernet, T., Briedis, D., Alkhatib, G., Henning, D., Levin, D., and
Richardson, C. (1990). Synthesis of the membrane fusion and hemagglutinin proteins of
measles virus, using a novel baculovirus vector containing the beta-galactosidase gene.
J. Virol. 64, 37–50.
222 Donald L. Jarvis

Vialard, J. E., and Richardson, C. D. (1993). The 1, 629-nucleotide open reading frame
located downstream of the Autographa californica nuclear polyhedrosis virus polyhedrin
gene encodes a nucleocapsid-associated phosphoprotein. J. Virol. 67, 5859–5866.
Walhout, A. J., Temple, G. F., Brasch, M. A., Hartley, J. L., Lorson, M. A., van den
Heuvel, S., and Vidal, M. (2000). GATEWAY recombinational cloning: Application to
the cloning of large numbers of open reading frames or ORFeomes. Methods Enzymol.
328, 575–592.
Wickham, T. J., Davis, T., Granados, R. R., Shuler, M. L., and Wood, H. A. (1992).
Screening of insect cell lines for the production of recombinant proteins and infectious
virus in the baculovirus expression system. Biotechnol. Progr. 8, 391–396.
Williams, G. V., Rohel, D. Z., Kuzio, J., and Faulkner, P. (1989). A cytopathological
investigation of Autographa californica nuclear polyhedrosis virus p10 gene functions
using insetion/deletion mutants. J. Gen. Virol. 70, 187–202.
 C H A P T E R F I F T E E N

Recombinant Protein Production

by Transient Gene Transfer into
Mammalian Cells
Sabine Geisse* and Cornelia Fux†

Contents
1. Introduction 224
2. HEK293 and CHO Cell Lines Commonly Used in TGE Approaches 224
3. Expression Vectors for HEK293 and CHO Cells 226
4. Cultivation of HEK293 Cells and CHO Cell Lines in Suspension 228
5. Transfection Methods 228
5.1. Small-scale transfection by lipofection in six-well plates 229
5.2. Medium scale transfection in shake flasks with
polyethylenimine as transfer reagent 231
5.3. Large-scale transient transfection in the WaveTM Bioreactor
with polyethylenimine as transfer reagent 232
6. Conclusions 234
Acknowledgments 234
References 235

Abstract
The timely availability of recombinant proteins in sufficient quantity and of
validated quality is of utmost importance in driving drug discovery and the
development of low molecular weight compounds, as well as for biotherapeu-
tics. Transient gene expression (TGE) in mammalian cells has emerged as a
promising technology for protein generation over the past decade as TGE meets
all the prerequisites with respect to quantity and quality of the product as well
as cost-effectiveness and speed of the process. Optimized protocols have been
developed for both HEK293 and CHO cell lines which allow protein production
at any desired scale up to >100 l and in milligram to gram quantities. Along with
an overview on current scientific and technological knowledge, detailed proto-
cols for expression of recombinant proteins on small, medium, and large scale
are discussed in the following chapter.

* Novartis Institutes for BioMedical Research, Department of NBC/PPA, Basel, Switzerland

{
Novartis Pharma AG, Department of NBx/PSD: Scale up, Basel, Switzerland

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63015-9 All rights reserved.

223
224 Sabine Geisse and Cornelia Fux

1. Introduction
Recombinant proteins including antibodies are essential tools for drug
discovery in the early stage and used in a multitude of applications. While
for production of therapeutic proteins criteria such as growth behavior,
clonality, and stability, as well as the productivity of the producer cell line
are of major importance, recombinant protein production for research
purposes is mainly driven by the cost-effectiveness, simplicity, and speed
of the process in conjunction with adequate yields of the product.
Over the past decade efficient transient transfection protocols have
reportedly been developed, which meet and even exceed the above-men-
tioned prerequisites. Transient production of proteins nowadays is predom-
inantly done in HEK293-derived cell lines, yet strong efforts are ongoing to
develop similarly efficient protocols for CHO cells. This bears the advantage
of maintaining the same host cell background and thus product character-
istics throughout the entire process of therapeutic protein development in
R&D. Detailed summaries of current scientific knowledge with respect to
TGE can be found in some recently published reviews (Baldi et al., 2007;
Geisse, 2009; Geisse et al., 2005; Pham et al., 2006).

2. HEK293 and CHO Cell Lines Commonly Used

in TGE Approaches
Several descendants of the original HEK293 cell line established by
Graham and coworkers in 1977 (Graham et al., 1977; Shaw et al., 2002) are
currently in use in large-scale transient expression approaches, as summar-
ized in Table 15.1. While 293 Freestyle cells are a derivative of a 293 wild-
type strain selected for good growth performance and high expression levels
(Zhang et al., 2009), most of the other candidates used in TGE trials are
actually engineered cell lines expressing, for instance, the EBNA-1 gene
from Epstein–Barr virus (EBV). The EBNA-1 protein acts in conjunction
with the origin of replication from EBV, oriP, encoded on the expression
plasmids; upon transfection, the plasmids remain in the nucleus as extra-
chromosomal entities—episomes—and are segregated onto the daughter
cells once per cell cycle (Kishida et al., 2008; Lindner and Sugden, 2007).
In addition, the EBNA-1 gene contains a nuclear localization signal (NLS)
sequence as well as a transcriptional enhancer element which both foster
nuclear import and enhanced transcription rates (Dean et al., 1999, 2005;
Laengle-Rouault et al., 1998).
Similarly, HEK293 T cells feature an integrated SV40 large T antigen
copy, which upon interaction with the SV40 origin of replication (SV40ori)
Table 15.1 Overview of HEK293 cell lines and expression systems for transient expression

Used in conjunction
Expression system Features of cell line Cultivation mediuma with plasmid vectors References Comments

HEK.EBNA EBNA-1 Suspension: for pCEP4 Many, citations, for Most commonly used
(HE, 293- transformed example, ExCell (Invitrogen), example in Baldi cell line in literature,
EBNA) HEK293 cell 293 (SAFC pEAK8, pcDNA et al. (2007) and but has not been
line, originally Biosciences), 3.1, pTT (NRC Pham et al. marketed by
marketed by Freestyle 293 Canada) (2006) Invitrogen for
Invitrogen (Invitrogen) several years
HEK.EBNA EBNA-1 Adherent: DMEM Same as above ATCC/Stanford ATCC CRL-10852
(HE, 293- transformed þ 10% FCS þ University
EBNA) HEK293 cell line 400 mg/ml G418,
suspension not
done
293-SFE (293SF- Suspension adapted Suspension: pTT plasmid series Pham et al. (2005) System available upon
3F6, NRC) EBNA-1 Hybridoma license from NRC
transformed 293 serum-free
cell line medium þ 1%
BCS
HEK293 T SV40 T-antigen Adherent: DMEM pCMV/myc/ER Li et al. (2007) and ATCC CRL-11268,
transformed 293 þ 10% FCS, (Invitrogen) þ Pham et al. also available in
cell line suspension not derivatives (2005) many labs
routinely done
293 Freestyle HEK293 wild type Suspension: Freestyle pcDNA 3.1, pCMV Zhang et al. (2008) Cell line has been
(293-F) medium SPORT Manual on Web carefully selected for
(Invitrogen) site of growth in Freestyle
Invitrogen medium and for 293
fectin reagent
HKB-11 (Hybrid Fusion of 293 cell Suspension: Bayer pTAT/TAR vector Cho et al. (2001) System available upon
of Kidney and with B-Cell proprietary (Bayer) license from Bayer
B-cell) lymphoma cell medium Healthcare
line
a
According to our experience, all of these cell lines can be cultivated adherently in Dulbecco’s Modified Eagle medium (DMEM) or in a 1:2 mixture of DMEM/Ham’s F12 medium, supplemented
with 10% fetal calf serum.
226 Sabine Geisse and Cornelia Fux

supplied on the expression vector supports extrachromosomal maintenance

of plasmids. Furthermore, the SV40 enhancer element as part of the SV40ori
is capable of binding not only to SV40 large T antigen, but also to newly
synthesized cytosolic transcription factors such as AP-1 and NFkB, which
thus serve to ‘‘cargo’’ the plasmids into the nucleus (Dean et al., 1999, 2005;
Graessmann et al., 1989).
The HKB-11 cell line results from a fusion between a Burkitt
lymphoma-derived cell clone, 2B8 and 293 S cells (Cho et al., 2001, 2002),
facilitating single cell suspension growth while maintaining high transfection
rates and recombinant product yields.
All of the six HEK293-derived cell lines described in Table 15.1 were
extensively tested in parallel for growth behavior and ‘‘transfectability’’ in
our lab. For proteins, such as antigens, HEK293-6E, and HKB-11 cells gave
in most instances the best results, while for expression of antibodies, trans-
fection into HEK293 T cells frequently resulted in higher expression levels
(Geisse, 2009). In summary, it is certainly worthwhile to try parallel expres-
sion in several HEK293-derived cell lines in order to find the optimal
candidate for a given gene to be expressed.
CHO cell lines used for transient expression approaches are descendants
of the CHO K1 line or the dhfr-negative variants DG44 and DUK X B11
(Gottesman, 1987; Kao and Puck, 1968; Urlaub and Chasin, 1980; Urlaub
et al., 1983). Suspension-adapted CHO-S cells, a CHO K1 derivative, along
with a suitable serum-free medium and a lipofection reagent were recently
introduced as ‘‘Freestyle MaxTM system’’ by Invitrogen (Zhang et al., 2009).
Genetically engineered CHO sublines are as yet infrequently used, even
though a beneficial effect on protein expression exerted via the inclusion of
viral elements derived from bovine papillomavirus, polyoma virus, and EBV
on the expression vector could be demonstrated (Kivimäe et al., 2001;
Kunaparaju et al., 2005; Silla et al., 2005).

3. Expression Vectors for HEK293 and CHO

Cells
To take advantage of the multiple beneficial functions of the EBV-or
SV40-derived viral elements described above suitable expression vectors for
transient expression in HEK293 cells should harbor the oriP or SV40ori
sequences, respectively. The human cytomegalovirus-derived immediate
early promoter is most commonly chosen to drive transgene expression (Lee
et al., 2007; Makrides, 1999; Van Craenenbroeck et al., 2000), but also other
promoters of viral origin, such as the mouse CMV promoter or a mouse/
human hybrid promoter (Meissner et al., 2001; Xia et al., 2006) or of
nonviral origin (e.g., the human elongation factor 1a [EF1alpha] promoter
Transient Transfection, Mammalian Cells, HEK293, CHO 227

(Backliwal et al., 2008a; Pham et al., 2006)) have been described as well. In
brief, a striking advantage of a single promoter/cell line combination could
so far not be demonstrated. Moreover, genetic elements active at the
posttranscriptional level, such as an intron splice element or the Wood-
chuck hepatitis virus regulatory element (WPRE) impact significantly on
protein yields by enhancement of mRNA stability, increased nuclear export
rates of mRNA and increased translation rates (Backliwal et al., 2008a; Klein
et al., 2006; Le Hir et al., 2003; Matsumoto et al., 1998). More details on the
design of various expression plasmids and their performance can be found in
the public domain (Backliwal et al., 2008a; Durocher et al., 2002; Meissner
et al., 2001; Pham et al., 2006; Xia et al., 2006) (Fig. 15.1).
It should, however, be mentioned that despite all benefits to gene
expression by specifically tailored expression plasmids featuring additional
enhancing elements, a large sized expression vector impacts negatively the
stability and yield in plasmid preparation. We have observed a three- to
fivefold reduction in plasmid yields when the size of the empty backbone
vector was doubled or tripled. As quite large quantities of plasmid DNA
need to be prepared for large-scale transient expression approaches and the
size of genes/cDNAs to be expressed appears to be steadily increasing, this
point should be taken into consideration in the design of backbone expres-
sion vectors for TGE approaches.

Growth curve and productivity

5.00E + 06
4.50E + 06 90
Viability(%) / IgG (mg/L)

4.00E + 06 80
Viable cells/mL

3.50E + 06 70
3.00E + 06 60
2.50E + 06 50
2.00E + 06 40
1.50E + 06 30
1.00E + 06 20
5.00E + 05 10
0.00E + 00 0
0 50 100 150 200 250
Time (h)

Viable cells/mL Viability (%) IgG (mg/L)

Figure 15.1 Production of a human IgG2 antibody by PEI-mediated transient trans-

fection into HEK293 T cells cultivated in M11V3 medium (Novartis proprietary) in the
WaveTM bioreactor over a period of 213 h. Heavy and light chain genes were encoded
on the same plasmid backbone. Symbols denote viable cell count (diamonds), viability
in % (squares) and antibody titer (triangles) determined by Protein A HPLC.
228 Sabine Geisse and Cornelia Fux

4. Cultivation of HEK293 Cells and CHO Cell

Lines in Suspension
Transient transfections into these two most frequently used cell lines
are in most instances performed in suspension culture, even though for both
lines the option to use adherent cell cultures—either by cultivation in, for
example, classical DMEM-based media containing fetal calf serum, by
supplementation of serum-free media formulations with serum, or by cell-
surface attachment mediated by coating of tissue culture flasks with, for
example, poly-D-lysine prior to transfection—is retained.
CHO and HEK cell lines are readily adaptable to serum-free conditions
and a large variety of specially designed media formulations (serum-free,
protein-free, or chemically defined) are commercially available. Adaptation
to serum-free suspension conditions is done by gradually ‘‘weaning’’ the
cells from the serum by dilution; simultaneously, the cells are transferred
from stationary culture in tissue culture flasks to agitated cultivation in
Erlenmeyer shake flasks, spinner flasks, or roller bottles.
The suspension adapted cells should exhibit good growth behavior and
high cell viability. Thus, cell counting and viability determination, done
either manually using a Neubauer hemocytometer or by employing cell
counting devices such as the CedexTM (Innovatis AG, Bielefeld, FRG) or
the Vi-CellTM (Beckman Coulter, Krefeld, FRG), and dilution of cells
twice per week is mandatory. In particular, at the day of transfection the
cells should be in logarithmic growth phase; thus, a dilution of the culture
1 or 2 days prior to gene transfer is important. We keep our backup cultures
at cell densities between 3–4  105 cells/ml for HEK293 cells and at approx.
2  105 cells/ml for CHO cells in Erlenmeyer shake flasks of variable
size (Corning by Fisher Scientific, Wohlen, CH) in a shaker incubator
(Kuehner, Climo-Shaker ISF1-X, Kuehner AG, Birsfelden, CH) at
100 rpm/37  C and 5% CO2.

5. Transfection Methods
For lipofection a wealth of formulations with individual, mostly propri-
etary compositions are commercially available, and doubtlessly many of them
are highly efficacious in introducing the plasmid DNA into the cells. Their
drawback is the cost-of-goods—transient transfection beyond the scale of
several hundred milliliters is economically not feasible when using these
reagents. Large-scale transient transfection approaches call for transfection
reagents readily available in bulk quantities with little batch-to-batch varia-
bility to ensure consistency of the process. Calcium–phosphate-mediated
Transient Transfection, Mammalian Cells, HEK293, CHO 229

transfection as well as polyethylenimine (PEI)-based plasmid complexation

and uptake of DNA into the cells are the only two commonly applied methods
which meet these requirements, even though some other options have been
described, such as Chitosan and derivatives thereof, and 14DEA2 as transfer
reagents (Dang and Leong, 2006; Jiang and Sharfstein, 2008; Kusumoto et al.,
2006).
The details of calcium–phosphate-mediated transfection have been
extensively studied and described (Batard et al., 2001; Jordan and Wurm,
2004; Jordan et al., 1996, 1998; Meissner et al., 2001), and similarly, the
mechanisms of action of PEI-mediated gene transfer have attracted much
research efforts. An in-depth mechanistic overview on PEI-based transfec-
tion would exceed the scope of this chapter, but can be easily retrieved from
the literature (Bertschinger et al., 2006; Eliyahu et al., 2005; Lungwitz et al.,
2005; Payne, 2007; Vicennati et al., 2008).
In a nutshell, two individual aspects are essential to mention in this
context. For both methods, calcium–phosphate-mediated transfection and
PEI-mediated transfection, the compatibility of the cell cultivation medium
with the transfection reagent is of key importance. It has been shown that
the composition of some commonly used, commercially available culture
media may hamper or even preclude an efficient uptake of plasmid DNA
(Bertschinger et al., 2006; Schlaeger and Christensen, 1999; Sun et al., 2006),
necessitating either a medium exchange prior to transfection or a biphasic
expression strategy using two different media during the transfection and the
production phases.
Furthermore, an efficiency of 90–100% of transduced cells, as observed
with lipofection reagents, has not been achieved when using calcium–
phosphate or PEI as transfer reagents. At best, a transfection rate of 60% is
realistically obtained. However, for the sake of cost-effectiveness and scal-
ability of the process, a reduction in transfection efficiency and yields
appears to be acceptable. In a very recent publication by Chenuet and
coworkers (Chenuet et al., 2008) comparative analyses of calcium–
phosphate- and PEI-mediated transfection demonstrated a higher plasmid
uptake for PEI-mediated transfection into CHO DG44 cells; yet, for stable
recombinant cell line generation the uptake and integration of plasmid
copies resulting in high specific productivities appeared to be more favorable
when calcium–phosphate-mediated transfection was applied.

5.1. Small-scale transfection by lipofection in six-well plates

Prior to embarking on transfection approaches on large scale, it is desirable
to assess the integrity of the expression vector and the functionality of the
protein. To do so, possibly also in combination with multiconstruct screen-
ing, a rapid, small-scale approach using lipofection as reagent for gene
transfer is the method of choice. In the following, a protocol is described
230 Sabine Geisse and Cornelia Fux

which is applicable to HEK293 as well as CHO cells. The expression vector

we use is a composite plasmid derived from commercially available plasmids
and features elements such as the CMV promoter, an intron splice element,
and the oriP for extrachromosomal replication in HEK293–EBNA-1
positive cells. The inserted cDNA fragments are in most instances equipped
with a protein detection tag which can be used for analytical purposes as
well as for protein purification via affinity chromatography on Protein A
column, for example:
1. At the day of transfection harvest the appropriate amount of cells
(calculate 1  106 cells per well of a 6-well plate) and spin them
down at 216g (1000 rpm) for 5 min.
2. Resuspend the cell pellet in 2 ml of medium/well; transfer the cell
suspension to a 6-well plate and distribute the cells into the wells.
3. Place the plate into a cell culture incubator at 37  C/5% CO2.
4. For preparation of the lipoplexes, transfer 100 ml of OPTI-MEM
medium (Invitrogen) into a sterile polypropylene tube.
5. Add 2 mg of supercoiled recombinant plasmid DNA (ideally at 1 mg/ml
concentration) into this tube.
6. Add 7 ml of FuGENEÒ HD reagent (Roche Diagnostics, Indianapolis,
IN) to the same tube and tap the tube gently for a few seconds for
mixing.
7. Incubate the mixture for 15 min at room temperature.
8. Add the mix dropwise directly to one well of cells of the 6-well plate.
9. Incubate the 6-well plate at 37  C/5% CO2 in a tissue culture incubator
for 72 h.
10. Harvest the culture supernatant (in case of secreted proteins) and/or the
cells (for membrane-bound or intracellular expression) and perform the
desired analytical experiments (e.g., ELISA, Western Blot, HPLC
affinity chromatography, surface plasmon resonance, etc.).
Notes:
(i). We have achieved excellent results when using the FuGENEÒ HD
reagent for transfection, but other lipofection reagents from other
vendors, which we did not try, may work equally well. Care should
be taken, however, to optimize process parameters such as the DNA to
reagent ratio as well as the amount of cells in an appropriate volume of
medium.
(ii). Harvest of samples beyond 72 h of incubation/production is possible;
however, due to the relatively high cell density in a small volume
nutrient depletion, cell death and potentially proteolytic damage to the
protein may occur. Thus, if protein analysis following a prolonged
production phase is desired, the medium scale protocol described
below appears to be the better choice.
Transient Transfection, Mammalian Cells, HEK293, CHO 231

5.2. Medium scale transfection in shake flasks with

polyethylenimine as transfer reagent
1. At the day of transfection harvest 1.25  107 cells and pellet them by
centrifugation at 216g (1000 rpm) for 5 min.
2. Resuspend the cells in 9 ml of fresh culture medium and transfer the
suspension into a 125 ml disposable shake flask. A total medium replace-
ment at this stage enhances transfection efficiency.
3. Polyplex formation: Take two sterile 15 ml polypropylene tubes and add
into each 1.75 ml of fresh culture medium.
Note: The original protocol for PEI transfection by Boussif and co-
workers (Boussif et al., 1995) describes the polyplex formation in a
150 mM NaCl solution. This protocol has been applied frequently; how-
ever, more recently DNA complexation carried out in RPMI 1640
medium (Invitrogen) for HEK293 cells or ProCHO5 medium (Lonza,
Vervier, Belgium) for CHO cells has been shown to lead to successful
transfection as well (Backliwal et al., 2008c; Bertschinger et al., 2006). We
routinely use our cultivation medium M11V3 for complexing of the plas-
mid DNA without any adverse effects.
1. Into one of the tubes add 50 mg of polyethylenimine from a sterile stock
solution: 25 kDa linear PEI (Polysciences, Warrington PA) is dissolved
in water at a concentration of 1 mg/ml, and the pH is adjusted to 7.0.
Note: The PEI stock solution should be sterile, filtered, aliquotted, and
stored at  80  C until use. A transfection reagent based on linear PEI is also
commercially available under the trade name jetPEITM from Polyplus-
transfection, New York, NY.
2. Add 25 mg of supercoiled, recombinant plasmid DNA into the other
tube.
3. Tap both tubes a few times gently and incubate them for 15 min at room
temperature.
4. Pipet the diluted PEI-solution into the diluted DNA solution (not the
inverse!), mix gently and incubate the mix for 15 min at room tempera-
ture. The DNA:PEI ratio employed corresponds to 1:2 (mg:mg).
Notes:
(i). If the complexes mature in 150 mM NaCl solution, a maturation time
beyond 10 min does not lead to efficient release of DNA from the
particles. If, however, culture medium is used for maturation, the
duration of incubation does not seem to have any effect (Backliwal
et al., 2008b; Bertschinger et al., 2006).
(ii). The DNA:PEI ratio is of critical importance for the success of these
experiments. Commonly used are ratios of 1:2–1:3 (mg:mg), but for
232 Sabine Geisse and Cornelia Fux

CHO cells a ratio of 1:4–1:5 has been determined as optimal (Chenuet

et al., 2008). At high cell densities of 20  106 cells/ml, an a priori
complex formation does not even seem to be necessary—the plasmid
DNA can be added directly to the culture, followed by addition of the
PEI solution (Backliwal et al., 2008b).
5. Add the polyplexes formed in 3.5 ml to the 9 ml of suspension culture in
the shake flask and incubate the flask on an orbital shaker platform at
37  C/5% CO2/100 rpm. The cell density at transfection corresponds to
1  106 cells/ml.
6. After 4 h of incubation, add 12.5 ml of culture medium to the shake flask
and continue the incubation for 72 h (or more). The cell density at the
beginning of the production phase is then 5  105 cells/ml.
7. After 72 h (or at a later time-point; 5–6 days posttransfection appears to
be the maximum, otherwise severe nutrient limitations will occur)
harvest the cells and/or the cell supernatant for analyses of recombinant
protein expression.

5.3. Large-scale transient transfection in the WaveTM

Bioreactor with polyethylenimine as transfer reagent
The disposable WaveTM bioreactor has gained widespread acceptance in
recent years due its ease of maintenance and handling versus classical stirred
tank reactors. Most commonly used are working volumes of 10 and 20 l,
even though reactors of 100 l working volume are commercially available,
and a prototype of 300 l working volume is under developmental construc-
tion. Whether to use the WaveTM bioreactor for transfection cultures below
10 l is a question of economical feasibility—we prefer to use shake flask
cultures for smaller production runs (Fernbach flasks, Corning). Report-
edly, also other reactors or shaking devices (Muller et al., 2005; Pham et al.,
2006; Stettler et al., 2007) have been successfully used for transiently
transfected cultures.
The following protocol describes production of an antibody at a scale of
10 l using HEK293 T cells as host and M11V3 serum-free culture medium
(Novartis proprietary). Both chains of the antibody were cloned into the
same backbone expression vector.
1. A 20 l WaveTM bag (Sartorius Stedim Biotech, Göttingen, Germany) is
mounted onto a WaveTM platform (Wave-Bioreactor SPS50) and linked
to a DASGIP gas-mixing module (DASGIP, Juelich, Germany). Subse-
quently, the bag is inoculated with 4 l HEK293T cell culture at a cell
density of approximately 1.8  106 viable cells/ml.
2. The following process parameters and conditions are applied: gas flow:
20 l/h; gas mix consisting of 21–25% O2/0% CO2; temperature 37  C,
pH: 6.8–7.4, rocking speed: 10 rpm, rocking angle: 7 .
Transient Transfection, Mammalian Cells, HEK293, CHO 233

3. 10 mg plasmid DNA (1 mg/ml) are mixed with M11V3 medium to a

final volume of 500 ml and incubated for 10 min at room temperature.
Afterward, the diluted DNA is sterile filtered through a 0.22 mm GP
EXPRESS PLUS membrane (Millipore, Billerica, MA).
4. 30 ml PEI solution (1 mg/ml) are mixed with 500 ml M11V3 medium
and incubated for 10 min at room temperature. Afterwards, the diluted
PEI solution is sterile filtered through a 0.22 mm GP EXPRESS PLUS
membrane.
Note: The PEI stock solution should be sterile, filtered, aliquotted, and
stored at 80  C until use.
5. Next, the PEI-solution is added to the diluted DNA and the mix is
incubated for 15 min at room temperature for polyplex formation to
occur.
6. The DNA–PEI–M11V3 mix is then aseptically added to the cells in the
WaveTM bag to achieve a final volume of 5 l. Incubation is continued
for 5–6 h applying the following parameters: gas-flow: 25 l/h, gas mix:
25% O2/0% CO2, temperature: 37  C, pH: 6.8–7.4, rocking speed:
10 rpm, angle: 7 .
7. Subsequently, 5 l M11V3 medium supplemented with 100 ml RX1
combined feed (feed solution consisting of amino acids, glucose and
glutamine; custom-made by Irvine Scientific, Santa Ana, CA) are added
aseptically to the cells in the WaveTM bag. During the production phase
the following process parameters are applied: gas flow: 30–40 l/h, gas
mix: 30–40% O2/0% CO2, temperature: 37  C, pH: 6.8–7.4 (adjusted
with bicarbonate solution pH 9.2, Sigma Aldrich, Buchs, Switzerland),
air saturation: 60–100%, rocking speed: 14–18 rpm, angle: 7 .
8. The transfected cells are cultivated for 10 days in the WaveTM bioreac-
tor for antibody production.
Note: For production of recombinant proteins the production phase
until harvest is usually between 5 and 7 days, but may be even shorter, if
the desired protein is sensitive toward proteolytic degradation. Production
runs of antibody molecules may be extended to 10 days.
9. Each day a sample is taken and the cell density and viability is deter-
mined using a Vi-Cell cell counter device (Beckman Coulter). Nutri-
ent status, pH, and air saturation are measured using the Bioprofile 400
Analyzer (Labor-Systeme Flückiger, Switzerland). The IgG concentra-
tion is assessed by Protein A HPLC determination (Fig. 15.1).
10. After 10 days, the cells are aseptically harvested and the cell removal is
performed by cross-flow filtration (Fresenius Filter PlasmaFlux,
0.2 mm). Afterward, the cell free supernatant is concentrated 10-fold
by means of hollow fiber filtration (Hemoflow F 10 HPS, Fresenius,
Stans, Switzerland, 10 kDa cutoff ).
234 Sabine Geisse and Cornelia Fux

11. The concentrate is subjected to protein purification by Protein A

affinity chromatography and size exclusion chromatography.
A variety of modifications to these standard protocols described above
have been tried and published, some of which are listed in brief below:
a. Enhancement of expression rates by ‘‘temperature shift’’ to 30–32  C
during the production phase when using CHO cells (Backliwal et al.,
2008a; Galbraith et al., 2006).
b. Cell cycle arrest in G2/M by treatment with the microtubule depoly-
merizing agent nocodazole (Tait et al., 2004).
c. Treatment with inhibitors of histone deacetylase and DNA methyl
transferase (e.g., sodium butyrate, valproic acid) in CHO and HEK293
cells (Backliwal et al., 2008c).
d. Genetic engineering of cell lines to prevent apoptosis or to overcome
inhibitions exerted by the unfolded protein response (UPR) (Backliwal
et al., 2008c; Majors et al., 2007; Tigges and Fussenegger, 2006).

6. Conclusions
The generation of recombinant proteins by means of transient trans-
fection technologies has reached an impressive state in terms of method
development and yields during the past several years, as reflected in the
wealth of data published on the subject. As yet, it is probably premature to
envisage the replacement of the tedious process of cell line development for
production of biotherapeutics by large-scale transient approaches, even
though remarkable antibody titers of >1 g/l have been reported already
(Backliwal et al., 2008a). Our own observations based on >100 transient
expression trials indicate a huge heterogeneity in expression yields depend-
ing on the gene(s) expressed, ranging from 1 to >170 mg/l for proteins;
antibody titers amount to 20–40 mg/l on average with frequent outliers to
both sides. There is certainly a need for further technical advancement with
respect to construct design, cultivation conditions, and possibly cell line
engineering, but these efforts are more than justified by the overall impres-
sive success of the approach.

ACKNOWLEDGMENTS
We thank our collaborators Agnès Patoux, Thomas Cremer, Mirjam Buchs, Sibylle Bossart,
Stefan Dalcher, and Rainer Uhrhahn for the development of the protocols and the genera-
tion of the experimental data. Special thanks go to Prof. Bertram Opalka, Uniklinikum
Essen-Duesburg, FRG for critically reviewing the manuscript.
Transient Transfection, Mammalian Cells, HEK293, CHO 235

REFERENCES
Backliwal, G., Hildinger, M., Chenuet, S., Wulhfard, S., De Jesus, M., and Wurm, F. M.
(2008a). Rational vector design and multi-pathway modulation of HEK 293E cells yield
recombinant antibody titers exceeding 1 g/l by transient transfection under serum-free
conditions. Nucleic Acids Res. 36, e96.
Backliwal, G., Hildinger, M., Hasija, V., and Wurm, F. M. (2008b). High-density transfec-
tion with HEK-293 cells allows doubling of transient titers and removes need for a priori
DNA complex formation with PEI. Biotechnol. Bioeng. 99, 721–727.
Backliwal, G., Hildinger, M., Kuettel, I., Delegrange, F., Hacker, D. L., and Wurm, F. M.
(2008c). Valproic acid: A viable alternative to sodium butyrate for enhancing protein
expression in mammalian cell cultures. Biotechnol. Bioeng. 101, 182–189.
Baldi, L., Hacker, D. L., Adam, M., and Wurm, F. M. (2007). Recombinant protein
production by large-scale transient gene expression in mammalian cells: State of the art
and future perspectives. Biotechnol. Lett. 29, 677–684.
Batard, P., Jordan, M., and Wurm, F. (2001). Transfer of high copy number plasmid into
mammalian cells by calcium phosphate transfection. Gene 270, 61–68.
Bertschinger, M., Backliwal, G., Schertenleib, A., Jordan, M., Hacker, D. L., and
Wurm, F. M. (2006). Disassembly of polyethylenimine-DNA particles in vitro: Implica-
tions for polyethylenimine-mediated DNA delivery. J. Control Release 116, 96–104.
Boussif, O., Lezoualc’h, F., Zanta, M. A., Mergny, M. D., Scherman, D., Demeneix, B.,
and Behr, J. P. (1995). A versatile vector for gene and oligonucleotide transfer into cells in
culture and in vivo: Polyethylenimine. Proc. Natl. Acad. Sci. USA 92, 7297–7301.
Chenuet, S., Martinet, D., Besuchet-Schmutz, N., Wicht, M., Jaccard, N., Bon, A. C.,
Derouazi, M., Hacker, D. L., Beckmann, J. S., and Wurm, F. M. (2008). Calcium
phosphate transfection generates mammalian recombinant cell lines with higher specific
productivity than polyfection. Biotechnol. Bioeng. 101, 937–945.
Cho, M.-S., Yee, H., Brown, C., Jeang, K.-T., and Cahn, S. (2001). An oriP expression
vector containing the HIV Tat/TAR transactivation axis produces high levels of protein
expression in mammalian cells. Cytotechnology 37, 23–30.
Cho, M. S., Yee, H., and Chan, S. (2002). Establishment of a human somatic hybrid cell line
for recombinant protein production. J. Biomed. Sci. 9, 631–638.
Dang, J. M., and Leong, K. W. (2006). Natural polymers for gene delivery and tissue
engineering. Adv. Drug Deliv. Rev. 58, 487–499.
Dean, D., Dean, B., Muller, S., and Smith, L. (1999). Sequence requirements for plasmid
nuclear import. Exp. Cell Res. 253, 713–722.
Dean, D., Strong, D., and Zimmer, W. (2005). Nuclear entry of nonviral vectors. Gene Ther.
12, 881–890.
Durocher, Y., Perret, S., and Kamen, A. (2002). High-level and high-throughput recombi-
nant protein production by transient transfection of suspension-growing human 293-
EBNA1 cells. Nucleic Acids Res. 30, E9.
Eliyahu, H., Barenholz, Y., and Domb, A. J. (2005). Polymers for DNA delivery. Molecules
10, 34–64.
Galbraith, D. J., Tait, A. S., Racher, A. J., Birch, J. R., and James, D. C. (2006). Control of
culture environment for improved polyethylenimine-mediated transient production of
recombinant monoclonal antibodies by CHO cells. Biotechnol. Prog. 22, 753–762.
Geisse, S. (2009). Reflections on more than 10 years of TGE approaches. Protein Expr. Purif.
64, 99–107.
Geisse, S., Jordan, M., and Wurm, F. (2005). Large-scale transient expression of therapeutic
proteins in mammalian cells. In ‘‘Therapeutic Proteins’’, (C. Smales and D. James, eds.),
Vol. 308, pp. 87–98. Humana Press, Totowa, NJ.
236 Sabine Geisse and Cornelia Fux

Gottesman, M. M. (1987). Chinese hamster ovary cells. In ‘‘Methods in Enzymology’’,

(M. Gottesman, ed.), Vol. 151, pp. 3–8. Academic Press Inc, San Diego.
Graessmann, M., Menne, J., Liebler, M., Graeber, I., and Graessmann, A. (1989). Helper
activity for gene expression, a novel function of the SV40 enhancer. Nucleic Acids Res. 17,
6603–6612.
Graham, F. L., Smiley, J., Russell, W. C., and Nairn, R. (1977). Characteristics of a human
cell line transformed by DNA from human Adenovirus Type 5. J. Gen. Virol. 36, 59–72.
Jiang, Z., and Sharfstein, S. T. (2008). Sodium butyrate stimulates monoclonal antibody
over-expression in CHO cells by improving gene accessibility. Biotechnol. Bioeng. 100,
189–194.
Jordan, M., Koehne, C., and Wurm, F. (1998). Calcium-phosphate mediated DNA transfer
into HEK-293 cells in suspension: Control of physicochemical parameters allows trans-
fection in stirred media. Cytotechnology 26, 39–47.
Jordan, M., Schallhorn, A., and Wurm, F. M. (1996). Transfecting mammalian cells:
Optimization of critical parameters affecting calcium-phosphate precipitate formation.
Nucleic Acids Res. 24, 596–601.
Jordan, M., and Wurm, F. (2004). Transfection of adherent and suspended cells by calcium
phosphate. Methods 33, 136–143.
Kao, F.-T., and Puck, T. T. (1968). Genetics of somatic mammalian cells. VII. Induction
and isolation of nutritional mutants in Chinese hamster cells. Proc. Natl. Acad. Sci. 60,
1275–1281.
Kishida, T., Asada, H., Kubo, K., Sato, Y. T., Shin-Ya, M., Imanishi, J., Yoshikawa, K., and
Mazda, O. (2008). Pleiotrophic functions of Epstein-Barr virus nuclear antigen-1
(EBNA-1) and oriP differentially contribute to the efficacy of transfection/expression
of exogenous gene in mammalian cells. J. Biotechnol. 133, 201–207.
Kivimäe, S., Allikas, A., Kurg, R., and Ustav, M. (2001). Replication of a chimeric origin
containing elements from Epstein-Barr virus ori P and bovine papillomavirus minimal
origin. Virus Res. 75, 1–11.
Klein, R., Ruttkowski, B., Knapp, E., Salmons, B., Gunzburg, W. H., and Hohenadl, C.
(2006). WPRE-mediated enhancement of gene expression is promoter and cell line
specific. Gene 372, 153–161.
Kunaparaju, R., Liao, M., and Sunstrom, N.-A. (2005). Epi-CHO, an episomal expression
system for recombinant protein production in CHO cells. Biotechnol. Bioeng. 91,
670–677.
Kusumoto, K., Akao, T., Mizuki, E., and Nakamura, O. (2006). Gene transfer effects on
various cationic amphiphiles in CHO cells. Cytotechnology 51, 57–66.
Laengle-Rouault, F., Patzel, V., Benavente, A., Taillez, M., Silvestre, N., Bompard, A.,
Sczakiel, G., Jacobs, E., and Rittner, K. (1998). Up to 100-fold increase of apparent gene
expression in the presence of Epstein-Barr virus oriP sequences and EBNA1: Implications
of the nuclear import of plasmids. J. Virol. 72, 6181–6185.
Lee, J., Lau, J., Chong, G., and Chao, S. H. (2007). Cell-specific effects of human
cytomegalovirus unique region on recombinant protein expression. Biotechnol. Lett. 29,
1797–1802.
Le Hir, H., Nott, A., and Moore, M. J. (2003). How introns influence and enhance
eukaryotic gene expression. Trends Biochem. Sci. 28, 215–220.
Li, J., Menzel, C., Meier, D., Zhang, C., Dubel, S., and Jostock, T. (2007). A comparative
study of different vector designs for the mammalian expression of recombinant IgG
antibodies. J. Immunol. Methods 318, 113–124.
Lindner, S., and Sugden, B. (2007). The plasmid replicon of Epstein-Barr virus: Mechanistic
insights into efficient, licensed, extrachromosomal replication in human cells. Plasmid 58,
1–12.
Transient Transfection, Mammalian Cells, HEK293, CHO 237

Lungwitz, U., Breunig, M., Blunk, T., and Gopferich, A. (2005). Polyethylenimine-based
non-viral gene delivery systems. Eur. J. Pharm. Biopharm. 60, 247–266.
Majors, B. S., Betenbaugh, M. J., and Chiang, G. G. (2007). Links between metabolism and
apoptosis in mammalian cells: applications for anti-apoptosis engineering. Metab. Eng. 9,
317–326.
Makrides, S. C. (1999). Components of vectors for gene transfer and expression in mamma-
lian cells. Protein Expr. Purif. 17, 183–202.
Matsumoto, K., Wassarman, K. M., and Wolffe, A. P. (1998). Nuclear history of a pre-
mRNA determines the translational activity of cytoplasmic mRNA. EMBO J. 17,
2107–2121.
Meissner, P., Pick, H., Kulangara, A., Chatellard, P., Friedrich, K., and Wurm, F. M.
(2001). Transient gene expression: Recombinant protein production with suspension-
adapted HEK293-EBNA cells. Biotechnol. Bioeng. 75, 197–203.
Muller, N., Girard, P., Hacker, D. L., Jordan, M., and Wurm, F. M. (2005). Orbital shaker
technology for the cultivation of mammalian cells in suspension. Biotechnol. Bioeng. 89,
400–406.
Payne, C. K. (2007). Imaging gene delivery with fluorescence microscopy. Nanomed 2,
847–860.
Pham, P. L., Kamen, A., and Durocher, Y. (2006). Large-scale transfection of mammalian
cells for the fast production of recombinant protein. Mol. Biotechnol. 34, 225–237.
Pham, P. L., Perret, S., Cass, B., Carpentier, E., St-Laurent, G., Bisson, L., Kamen, A., and
Durocher, Y. (2005). Transient gene expression in HEK293 cells: Peptone addition
posttransfection improves recombinant protein synthesis. Biotechnol. Bioeng. 90, 332–344.
Schlaeger, E.-J., and Christensen, K. (1999). Transient gene expression in mammalian cells
grown in serum-free suspension culture. Cytotechnology 30, 71–83.
Shaw, G., Morse, S., Ararat, M., and Graham, F. L. (2002). Preferential transformation
of human neuronal cells by human adenoviruses and the origin of HEK 293 cells.
FASEB J. 16, 869–871.
Silla, T., Hääl, I., Geimanen, J., Janikson, K., Abroi, A., Ustav, E., and Ustav, M. (2005).
Episomal maintenance of plasmids with hybrid origins in mouse cells. J.Virol. 79,
15277–15288.
Stettler, M., Zhang, X., Hacker, D. L., de Jesus, M., and Wurm, F. M. (2007). Novel orbital
shake bioreactors for transient production of CHO derived IgGs. Biotechnol. Prog. 23,
1340–1346.
Sun, X., Goh, P. E., Wong, K. T., Mori, T., and Yap, M. G. (2006). Enhancement of
transient gene expression by fed-batch culture of HEK 293 EBNA1 cells in suspension.
Biotechnol. Lett. 28, 843–848.
Tait, A. S., Brown, C. J., Galbraith, D. J., Hines, M. J., Hoare, M., Birch, J. R., and
James, D. C. (2004). Transient production of recombinant proteins by Chinese hamster
ovary cells using polyethyleneimine/DNA complexes in combination with microtubule
disrupting anti-mitotic agents. Biotechnol. Bioeng. 88, 707–721.
Tigges, M., and Fussenegger, M. (2006). Xbp1-based engineering of secretory capacity
enhances the productivity of Chinese hamster ovary cells. Metab. Eng. 8, 264–272.
Urlaub, G., and Chasin, L. A. (1980). Isolation of Chinese hamster cell mutants deficient in
dihydrofolate reductase activity. Proc. Natl. Acad. Sci. 77, 4216–4220.
Urlaub, G., Käs, E., Carothers, A. M., and Chasin, L. A. (1983). Deletion of the diploid
dihydrofolate reductase locus from cultured mammalian cells. Cell 33, 405–412.
Van Craenenbroeck, K., Vanhoenacker, P., and Haegeman, G. (2000). Episomal vectors for
gene expression in mammalian cells. Eur. J. Biochem. 267, 5665–5678.
Vicennati, P., Giuliano, A., Ortaggi, G., and Masotti, A. (2008). Polyethylenimine in
medicinal chemistry. Curr. Med. Chem. 15, 2826–2839.
238 Sabine Geisse and Cornelia Fux

Xia, W., Bringmann, P., McClary, J., Jones, P. P., Manzana, W., Zhu, Y., Wang, S.,
Liu, Y., Harvey, S., Madlansacay, M. R., McLean, K., Rosser, M. P., et al. (2006). High
levels of protein expression using different mammalian CMV promoters in several cell
lines. Protein Expr. Purif. 45, 115–124.
Zhang, J., Liu, X., Bell, A., To, R., Baral, T. N., Azizi, A., Li, J., Cass, B., and Durocher, Y.
(2009). Transient expression and purification of chimeric heavy chain antibodies. Protein
Expr. Purif. 65, 77–82.
 C H A P T E R S I X T E E N

Tagging for Protein Expression

Arun Malhotra

Contents
1. Introduction 240
2. Some Considerations When Designing a Tagged Protein 241
2.1. Affinity and/or solubility? 241
2.2. Which tag(s) to use? 243
2.3. Tandem tags? 244
2.4. N- or C-terminal? 244
2.5. Cleavage sites to remove tags 245
3. Protein Affinity Tags 245
3.1. His-tag 245
3.2. GST tag 246
3.3. Other purification tags 247
4. Solubility Tags 249
4.1. MBP tag 249
4.2. Trx tag 250
4.3. NusA tag 250
4.4. Other solubility tags 250
5. Removal of Tags 251
6. Conclusions 253
Acknowledgment 254
References 254

Abstract
Tags are frequently used in the expression of recombinant proteins to improve
solubility and for affinity purification. A large number of tags have been devel-
oped for protein production and researchers face a profusion of choices when
designing expression constructs. Here, we survey common affinity and solubility
tags, and offer some guidance on their selection and use.

Department of Biochemistry and Molecular Biology, University of Miami Miller School of Medicine, Miami,
Florida, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63016-0 All rights reserved.

239
240 Arun Malhotra

1. Introduction
Most proteins are made using recombinant techniques in expression
systems. Additional residues or tags can be engineered on either the N- or
C-terminal end of the protein of interest during the cloning step. These
tags, which can range in size from just a few residues to full-length proteins
or domains, can be used to improve protein production or to confer new
properties that can be used for characterization and study of the target
protein. The term ‘‘fusion protein’’ is also often used instead of the term
‘‘tag,’’ and fusion sometimes refers to the simpler end-to-end joining of two
proteins while tags are typically shorter and include linker regions; here, we
will use these terms interchangeably.
The focus of this chapter is to survey some of the common tags that are used
for improving the production of proteins, and highlight the advantages and
pitfalls of their use. Given the wild profusion of tags, often in different
combinations and with different cleavage sites, this review is by no means
exhaustive and the reader is encouraged to look at constructs available from
many commercial sources as well as repository sources such as Addgene
(http://www.addgene.org) or ATCC (http://www.atcc.org). Detailed proto-
cols related to the use of many individual tags have been described previously
in Methods in Enzymology (e.g., Volumes 326 and 327), and in specialized
journals in this field such as Protein Expression and Purification and the newer
Microbial Cell Factories. Handbooks from suppliers of vectors for most expres-
sion systems (Amersham/GE Healthcare Inc., Clontech Inc., Invitrogen Inc.,
New England Biolabs Inc., Novagen/EMD Biosystems Inc., Roche Inc.,
Sigma-Aldrich Inc., and others) are another rich source of protocols.
Tags used to improve the production of recombinant proteins can be
roughly divided into purification and solubility tags. The former are used
along with affinity binding to allow rapid and efficient purification of
proteins, while the latter refer to tags that enhance the proper folding and
solubility of a protein. Tables 16.1 and 16.2 list some of the common
purification and solubility tags that are used for protein expression.
Table 16.3 summarizes common endoproteases used to remove tags and
recover the target protein of interest.
While such tags are quite useful, other tags can be fused to proteins for a
broad range of applications—labeling for imaging and localization studies,
protein detection and quantification, protein–protein interaction studies,
subcellular localization or transduction, and many others. It is important to
keep in mind some of these additional capabilities that can be engineered
into a protein as recombinant constructs are being designed, since multiple
tags can be added together in different combinations (Fig. 16.1).
Tagging for Protein Expression 241

Table 16.1 Common affinity tags

Tag Size Affinity matrix

His-tag 6–10 His Immobilized metal
ions—Ni, Co,
Cu, Zn
GST (glutathione- 211 aa Glutathione resin
S-transferase)
FLAG-tag 8 aa (DYKDDDDK) Anti-FLAG mAB
(22 aa for 3xFLAG)
Strep-II-tag 8 aa (WSHPQFEK) Strep-Tactin (modified
streptavidin)
Protein A 280 aa Immobilized IgG
(staphylococcal
Protein A)
MBP (maltose- 396 aa Cross-linked amylose
binding protein)
CBP (calmodulin- 26 aa Immobilized calmodulin
binding protein)
CBD (chitin-binding 51 aa Chitin
domain)
HaloTag
300 aa Chloroalkane

Table 16.2 Common solubility tags

Tag Size Protein

MBP 396 aa Maltose-binding protein
NusA 495 aa N-utilization substance
Trx 109 aa Thioredoxin
SUMO
100 aa Small ubiquitin modifier
GB1 56 aa IgG domain B1 of streptococcus Protein G
SET/ <20 aa Hydrophilic solubility enhancing peptide sequences
SEP
HaloTag
300 aa Mutated dehalogenase

2. Some Considerations When Designing

a Tagged Protein
2.1. Affinity and/or solubility?
Two challenges in the production of heterologous proteins in Escherichia coli,
the workhorse of protein expression systems, are poor or low expression,
and the misfolding of the expressed protein into insoluble aggregates called
242 Arun Malhotra

Table 16.3 Common endoproteases used to remove protein tags

Protease Cleavage site

Enterokinase DDDDK"
Factor Xa IEGR"
Thrombin LVPR"GS
PreScission (human LEVLFQ"GP
rhinovirus 3C protease)
TEV (tobacco etch virus ENLYFQ"G
protease)
TVMV (tobacco vein ETVRFQG"S
mottling virus protease)
SUMO protease (catalytic Recognizes SUMO tertiary structure and
core of Ulp1) cleaves at the C-terminal end of the
conserved Gly–Gly sequence in SUMO

A
TEV cleavage site

N Target protein C
His-tag

B
Thrombin cleavage site

N Target protein GST C

C
TEV cleavage site

N Target protein CBP Prot A C

TVMV cleavage site

Figure 16.1 Example schematics of single and tandem tagged proteins. (A) N-terminally
His-tagged protein; (B) C-terminally GST tagged protein; and (C) target protein with a
removable (using TVMV proteases) C-terminal tandem affinity tag (TAP tag).

inclusion bodies. Expression problems caused by weak promoters, poor

translation initiation, or the presence of rare codons, can be alleviated by
introducing corrective sequences in the gene when assembling overexpres-
sion plasmids and the use of E. coli strains supplemented with rare tRNAs.
Attaching a highly translated native gene as a fusion on the N-terminal end
of the heterologous target protein is another approach to improving yield,
Tagging for Protein Expression 243

and has also the added benefit of increasing the solubility of the target
protein; this is the basis of solubility tags (Table 16.2).
Affinity tags, on the other hand, are crucial during protein purification,
and allow the use of a variety of strategies to bind the target protein on an
affinity matrix (Table 16.1). Some protein tags can function in both affinity
and solubility roles—for example, the glutathione-S-transferase (GST) tag
improves solubility of some proteins, while solubility tags such as maltose-
binding protein (MBP) can be used for affinity purification of the target
protein.

2.2. Which tag(s) to use?

Protein affinity and solubility tags vary widely in their size, and thus in the
metabolic load that they impose on the host cell. For example, an MBP tag
(43 kDa) on a 100-residue target protein will require the expression of 5 mg
of fusion protein for every mg of the target protein produced. Affinity tags
also vary in the cost of purification, since different affinity media have
different expenses, both for the resin itself, and for operating costs (ease of
regeneration and reuse, elution agent, binding capacity, etc.). Lichty et al.
(2005) did a comparison of eight affinity tags looking at many different
aspects such as purity, yield, and cost, and many similar studies have been
carried out. Several recent reviews provide good overviews (Terpe, 2003),
and list advantages and disadvantages of affinity tags (Arnau et al., 2006b;
Waugh, 2005) and solubility tags (Esposito and Chatterjee, 2006).
Apart from the cost, the choice between different affinity tags often
depends on finding purification buffer conditions suitable for the target pro-
tein. For proteins susceptible to oxidation or proteolytic damage, the His-tag
may not be very suitable since immobilized metal affinity chromatographic
(IMAC) media cannot tolerate reducing agents or EDTA. Similarly, care has
to be taken when IMAC media is used with target proteins that are sensitive to
metal ions. Conversely, for target proteins requiring denaturing conditions or
refolding, the His-tag and IMAC purification is an excellent choice.
Expression levels can dictate the choice of tags in some cases—solubility
tags such as MBP, Trx, NusA, as well as GST have strong translational
initiation signals and can drive expression levels higher which is useful for
structural studies. On the other hand, when low expression levels are
desirable, such as when studying complexes or physiological interactions,
more stringent epitope tags or tandem tagging may be more appropriate.
A major source of comparative studies and innovation in protein tags is
coming from ongoing structural genomics efforts, since protein production
is a big roadblock for these projects (Pédelacq et al., 2002; Structural
Genomics Consortium et al., 2008; Yokoyama, 2003). The reader is encour-
aged to take advantage of the consensus ‘‘what to try first’’ strategies that are
emerging from these efforts (Structural Genomics Consortium et al., 2008).
244 Arun Malhotra

Many high-throughput studies have looked more specifically at the effec-

tiveness of individual solubility tags (e.g., Cabrita et al., 2006; Kataeva et al.,
2005), and reporter systems have been developed to monitor the solubility
and proper folding of proteins (Listwan et al., 2009; Liu et al., 2006; Waldo
et al., 1999). Directed evolution of proteins can also be used for improving
protein expression and solubility (Roodveldt et al., 2005; Waldo, 2003).
Use of tags for production of transmembrane proteins is another area of
active development (Cunningham and Deber, 2007).

2.3. Tandem tags?

Multiple tags can be attached on target proteins, allowing for improved
purification, expression, or tracking. The tandem affinity purification
(TAP) tag was originally developed to allow for two purification steps—
first on Protein A IgG beads, followed by cleavage of the Protein A tag and
subsequent purification on calmodulin coated beads (Fig. 16.1; Puig et al.,
2001). The use of tandem tags allows TAP-tagged proteins to be detected
and purified in native conditions even when expressed at very low levels.
TAP tagging has proved very useful for studying protein complexes (Bauer
and Kuster, 2003), and many other variations of the TAP tags have been
designed (Collins and Choudhary, 2008).
Solubility tags such as Trx or NusA can also be linked with affinity tags
such as His-tags for efficient purification of the fusion protein. Other tags
such as S-tags or FLAG-tags can be added to allow protein detection when
low levels of protein expression are expected and less specific antibodies
(such as those against His-tags) may not be adequate. Affinity tags can also be
attached at both ends of a target protein (Mueller et al., 2003).

2.4. N- or C-terminal?
Tags can be placed at either the N- or C-terminus of a target protein. One
advantage of placing a tag on the N-terminal end is that the construct can
take advantage of efficient translation initiation sites on the tag. Solubility
tags based on highly expressing proteins such as MBP, Trx, and NusA are
also more efficient at solubilizing target proteins when positioned at the
N-terminal end (Sachdev and Chirgwin, 1998), though recent high-
throughput studies have shown than the MBP tag is still quite effective
when positioned at the C-terminal end (Dyson et al., 2004). Another
advantage of placing a tag on the N-terminal site is that the tag can
be removed more cleanly, since most endoproteases cut at or near the
C-terminus of their recognition sites.
While placing a tag, care should be taken to preserve the positioning of
any signal sequences or modification sites. Sequences at termini of the fusion
protein should be examined for effects on the stability of the final construct,
Tagging for Protein Expression 245

especially at the N-terminal end, which should be inspected for the host
cell’s N-end rule degradation signals (Bachmair et al., 1986; Wang et al.,
2008). It is also useful to examine the sequence of the tagged protein for any
inadvertently created interaction or cleavage sites using motif databases such
as PROSITE (Hulo et al., 2008).

2.5. Cleavage sites to remove tags

Tags can interfere with the structure and function of the target protein, and
provision must be made to remove tags after the expression and purifica-
tions steps. Multiple cleavage sites can be engineered into the expression
construct to remove individual tags at different stages of purification
(Fig. 16.1). Table 16.3 lists some of the endoproteases used for tag removal,
and these are discussed in this chapter.

3. Protein Affinity Tags

3.1. His-tag
The His-tag (also called 6xHis-tag) is one of the simplest and most widely
used purification tags, with six or more consecutive histidine residues.
These residues readily coordinate with transition metal ions such as Ni2þ
or Co2þ immobilized on beads or a resin for purification. IMAC is the
preferred choice as a first step during the purification of His-tagged proteins,
though small batch reactions or spin columns with IMAC beads can be used
for expression tests or small-scale preparations. Metal ions are immobilized
using linkages such as Ni(II)-nitrilotriacetic acid (Ni-NTA) or Co2þ-
carboxymethyl-aspartate on resins and beads available from many commer-
cial sources. IMAC media typically has high binding capacities (5–40 mg of
His-tagged protein/ml of media), is relatively low cost, and can easily be
sanitized. For nickel binding media, the metal ion can often be stripped
(using buffers with EDTA) and recharged for multiple use cycles. Some
cobalt based resins (such as Talon, Clontech Inc.) use proprietary linkages
that are more durable and cannot be recharged; such resins can be reused
three and four times, but offer the advantage of being more specific for
polyhistidine tags and almost no metal leakage during protein elution.
IMAC can also be used under denaturing conditions, since the His-tag
does not need a specific protein conformation for metal binding; indeed,
binding to IMAC resins is stronger under denaturing conditions as the
His-tag becomes more exposed.
His-tags bind the immobilized metal via the histidine imidazole ring, and
tagged protein bound on IMAC media can be easily eluted using elution
buffers with imidazole (100–250 mM ) or low pH (4.5–6). Since some
246 Arun Malhotra

endogenous proteins also exhibit weak binding to IMAC media (Bolanos-

Garcia and Davies, 2006), a low level of imidazole (5–20 mM ) should be
included in the loading and wash buffers to minimize nonspecific binding.
While the mild elution conditions for IMAC is one of the positive aspects
of His-tags, care has to be taken to avoid EDTA (or EGTA) in any of the
buffers. Cell extracts loaded on IMAC columns should not contain any
EDTA, and only EDTA-free protease inhibitor cocktails should be used
for sample preparation. TRIS salts weakly chelate metal ions as well, and the
use of TRIS buffers should be minimized (50 mM or less). Most IMAC
media is also very sensitive to reducing agents such as DTT or DTE, and low
levels of b-mercaptoethanol (<10 mM ) should be used instead.
The small size of the His-tag minimizes interference with the folding and
structure of the target protein, and this tag can be positioned at either the
N- or C-terminal ends (Fig. 16.1A). While the tag can be removed by
introducing a protease cleavage site, there are many examples of proteins
that have been crystallized with intact His-tags and little or no impact on
the structure of the tagged protein (Carson et al., 2007). In some cases, the
His-tag can actually assist in crystal formation (e.g., Smits et al., 2008).
The His-tag can also be used with commercially available His-tag-specific
antibodies for protein detection.
There are several variations on the standard 6xHis sequence used in
His-tags (Terpe, 2003). These tags intersperse multiple histidines among
other residues to lower the charge or improve stability, such as the HAT-tag
and 6xHN tag which also bind better to the Co2þ-Talon resin (Clontech
Inc.), and the MAT-tag (Sigma-Aldrich Inc.).

3.2. GST tag

GST is an abundantly expressed 26 kDa eukaryotic protein, and GST cloned
from Schistosoma japonicum was shown to promote solubility and expression as
an N-terminal fusion (Smith and Johnson, 1988). When positioned at the
C-terminal end (Fig. 16.1B), GST is less efficient at improving protein solu-
bility but still functions well as an affinity tag. GST binds to resin immobilized
glutathione, and this property is used for affinity purification of GST tagged
proteins. After the fusion protein is bound to the resin, it can be eluted under
rather mild conditions using free reduced glutathione (10–40 mM ) at neutral
pH. Resins used for GST fusion purification, such as Glutathione-Sepharose
beads, are relatively cheap, have high binding capacity (5–10 mg of GST/ml of
resin), and can be regenerated and reused multiple times.
The GST tag has to be properly folded to bind glutathione, and thus the
fusion protein needs to be soluble and in nondenaturing conditions for
efficient purification. GST fusion proteins are often expressed at high levels,
and thus protein solubility has to be monitored carefully. For some proteins,
GST can act as a solubility tag (e.g., Kim and Lee, 2008). GST can be
Tagging for Protein Expression 247

detected by a colorimetric assay with the substrate 1-chloro-2,4 dinitroben-

zene (CDNB) (Habig et al., 1974); this assay can also be used to monitor
proper folding and accessibility of GST in the fusion protein when binding
to glutathione resin is suboptimal. Commercial anti-GST antibodies are also
available for detecting this tag.
The large size of the GST tag increases potential for degradation by
proteases, and GST fusion proteins need to be purified quickly to minimize
sample loss. Unlike for His-tagged proteins, EDTA can be used in buffers
during sample preparation to reduce proteolytic damage. Care should also
be taken to use reducing conditions since GST has four solvent exposed
cysteines that can be involved in oxidative aggregation (Kaplan et al., 1997).
GST forms a homodimer in solution, which makes it a poor choice for
tagging oligomeric proteins.
The kinetics of GST binding to glutathione and its elution are relatively
slow, and so GST fusion proteins need to be loaded and eluted from GST
columns at slow flow rates. The monitoring of protein during elution also
has to be done carefully since glutathione absorbs strongly at 280 nm.

3.3. Other purification tags

3.3.1. Epitope tags
A number of short amino acid (aa) sequences, recognized by commercially
available antibodies, can be used as tags for detection and purification of
proteins. Epitope tagging is widely used in molecular biology as a general
method for tracking recombinant proteins (Fritze and Anderson, 2000), and
has the advantage of high specificity and the use of short tags that minimizes
deleterious effects on the structure and function of the target protein. These
tags are mostly placed at the N- or C-termini, but can also be placed within a
target protein (in loops or between structural domains in a solvent exposed
region). Epitope tags are usually sequences absent in the host cell, making the
detection of the target protein straightforward. For purification, however,
epitope tag binding media (typically, monoclonal antibodies immobilized on
chromatographic resins) is expensive and less suitable for large-scale prepara-
tions than other affinity media. Short peptides corresponding to the tag, low
pH, or other approaches (e.g., chelation of calcium required for tag binding, or
use of salt and polyol) can be used to elute the target protein, but some of these
procedures tend to be more harsh than other affinity purification methods
(however, see Chapter 28 in this volume).
The FLAG-tag is a short, eight-residue (DYKDDDDK) hydrophilic
peptide tag that can be used for detection and purification of target proteins
(Einhauer and Jungbauer, 2001; Prickett et al., 1989). Apart from availabil-
ity of several high specificity anti-FLAG mABs, the FLAG-tag has the
advantage of incorporating the enterokinase cleavage site (DDDK) that
allows the complete tag to be cleanly removed after purification. Another
248 Arun Malhotra

variant of the FLAG-tag is the 3xFLAG-tag (Sigma-Aldrich, Inc.) that is

made up of three tandem repeats of FLAG-like sequences (Hernan et al.,
2000). Other commonly used epitope tags are the HA-tag and the c-Myc
tag (Fritze and Anderson, 2000).

3.3.2. The S-tag

S-tag, reviewed by Raines et al. (2000), takes advantage of the tight associa-
tion between the N-terminal S-peptide (residues 1–20) and the S-protein
(residues 21–124) in RNase S (Richards and Vithayarhil, 1959). The S-tag
system (Novagen Inc.) uses the N-terminal 15 residues of the S-peptide
(Potts et al., 1963), with S-protein immobilized on agarose beads for target
protein purification. S-tag vectors typically encode a site-specific protease
cleavage site, and elution of the target protein requires cleavage of the tag or
harsher denaturing conditions that disrupt the S-tag S-protein interaction.

3.3.3. STREP-II-tag
STREP-II-tag (WSHPQFEK) is a tag that takes advantage of the strong and
specific interaction between biotin and streptavidin (Schmidt and Skerra,
1994). This peptide tag binds in the biotin pocket of streptavidin;
Strep-Tactin, a recombinant form of streptavidin optimized to bind the
Strep-II-tag, is used in affinity media. Bound target protein can be eluted
with low levels (2.5 mM ) of desthiobiotin, a biotin analog that competes for
binding to Strep-Tactin in a reversible manner (Skerra and Schmidt, 2000).
The elution conditions are gentle, and buffers can include high levels (up to
50 mM ) of reducing agents such as DTT or b-mercaptoethanol, as well as
chelating agents such as EDTA, which makes this an excellent purification
technique for proteins that are very sensitive to oxidation. Strep-Tactin
chromatographic resins can regenerated and reused a few times. While there
is a low background of naturally biotinylated proteins in most cell extracts,
these usually do not interfere in the purification; if necessary, avidin can be
added to clear biotinylated host proteins (Schmidt and Skerra, 2007). There
has also been some work on using streptavidin/avidin in fusion tags (Sano and
Cantor, 2000), but they are difficult to use for affinity purification because the
very strong interaction between these proteins is difficult to disrupt.

3.3.4. CBP-tag
The calmodulin-binding peptide (CBP) is a short 26-residue sequence
derived from the C-terminus of skeletal muscle myosin light chain kinase
that binds specifically to calmodulin (reviewed in Terpe, 2003). Calmodulin
(CaM) immobilized on chromatographic media (such as CaM-Sepharose)
can be used for affinity purification of target proteins tagged with CBP.
Though calmodulin binds very strongly with CBP (nanomolar affinity), this
interaction is dependent on calcium and the bound protein can be eluted in
a single step using gentle buffers containing a calcium-chelating agent, such
Tagging for Protein Expression 249

as EGTA. Another application of the CBP-tag is for 32P isotopic labeling of

the fusion protein, since the tag includes a protein kinase A target sequence
(Vaillancourt et al., 2000). The CBP-tag cannot be used for expression in
eukaryotic systems, since many endogenous proteins in eukaryotes also
interact with calmodulin.

4. Solubility Tags
Production of well-folded, soluble proteins is a major bottleneck in
recombinant protein expression, especially when heterologous eukaryotic
proteins are expressed in bacterial cells. Several soluble proteins are used as
tags to improve folding of the target protein. These tags should be used in
conjunction with other approaches to improve protein folding such as
lowering temperature after protein induction or coexpression of chaperones
(Baneyx and Mujacic, 2004; de Marco et al., 2007; Sahdev et al., 2008). It is
also useful to screen multiple solubility tags (Peleg and Unger, 2008). For
recalcitrant proteins, protein purification using denaturing conditions and
refolding can be tried (Cabrita and Bottomley, 2004; Jungbauer and Kaar,
2007; Qoronfleh et al., 2007; see Chapter 17 in this volume).

4.1. MBP tag

MBP is a solubility enhancing tag (di Guan et al., 1988) that can also be used
for effective affinity purification, since it binds specifically to maltose or
amylose. MBP is a large 43 kDa secreted E. coli protein that can be expressed
at very high levels, and helps keep proteins fused at its C-terminal end
soluble (Kapust and Waugh, 1999). More recent surveys have shown than
the MBP tag is also effective when placed on the C-terminal end of target
proteins (Dyson et al., 2004). The large size of this tag puts a heavy
metabolic load on the host cell, but in high-throughput tests, MBP ranks
as one of the best tags for making soluble fusions (Dyson et al., 2004;
Kataeva et al., 2005). About a quarter of proteins expressed as MBP fusions,
however, remain insoluble or are prone to aggregation when the MBP tag is
removed. In our work on the expression of human calcitonin gene-related
peptide-receptor component protein (CGRP-RCP) in E. coli, for example,
we observed protein aggregation when the N-terminal MBP tag was
removed by enterokinase cleavage even after successful expression and
purification (Tolun et al., 2007). Each solubility tag also has different
effects, and several tags may need to be tried for recalcitrant proteins.
In the work on CGRP-RCP, a thioredoxin tag was not effective and the
fusion protein remained insoluble (Tolun et al., 2007). Nallamsetty and
Waugh (2006) have suggested that solubility tags such as MBP and NusA act
as passive partners in the folding of target proteins, and that solubility of
250 Arun Malhotra

aggregation-prone target proteins after removal of the tag appears to depend

on the characteristics of the target protein rather than the tag used.
Commercial vectors for tagging MBP are available for both cytoplasmic
and periplasmic expression of target proteins, with a variety of cleavage sites
(New England Biolabs Inc.). Cross-linked amylose resin is used to bind
MBP tagged proteins, and the bound fusion protein can be easily eluted by
adding 10 mM maltose to the wash buffers. This allows for an easy one-step
purification of MBP tagged proteins under very gentle conditions; how-
ever, amylose affinity purification cannot be carried out with reducing
agents or under denaturing conditions. Amylose resins are degraded some-
what by amylase activity in crude extracts, especially from cells grown in
rich LB media, but this can be minimized by including glucose in the media
(0.2%). Amylose resins can be regenerated and reused several times.

4.2. Trx tag

Thioredoxin (Trx) is a thermostable, 12-kDa intracellular E. coli protein
that is easily overexpressed and soluble even when overexpressed up to 40%
of the total cellular protein (LaVallie et al., 1993), and is very useful as a tag
in avoiding inclusion body formation in recombinant protein production
(LaVallie et al., 2000). Tests by Dyson et al. (2004) indicate that the Trx tag
is more effective when placed on the N-terminal end of the target protein.
Thioredoxin accumulates at cytoplasmic membrane adhesion sites
(Bayer et al., 1987), which allows Trx fusion proteins to be released by
simple osmotic shock or freeze/thaw treatments, providing a simple initial
purification step. Additional affinity tags such as His-tags are typically
attached to the Trx tag for further purification steps.

4.3. NusA tag

NusA is a large (495 aa) N-utilizing substance A transcription antitermina-
tion factor that was chosen as a potential tag as it ranked among the most
soluble E. coli proteins as a fusion for heterologous protein expression, based
on a statistical solubility model (Davis et al., 1999; De Marco et al., 2004). In
some large-scale screening tests (Busso et al., 2005; Kataeva et al., 2005),
NusA has performed as well as or better than MBP as a solubility tag, but the
results vary for individual proteins. NusA tags are used in conjunction with
other affinity tags such as His-tags (de Marco, 2006).

4.4. Other solubility tags

A number of smaller solubility enhancement tags (SETs) or solubility enhance-
ment peptide (SEP) tags have been developed, which use highly acidic
sequences to enhance solubility of some target proteins (Kato et al., 2007;
Tagging for Protein Expression 251

Zhang et al., 2004). Other small tags are the GB1-tag (56 residues) which is
based on the IgG binding B1 domain of the streptococcal Protein G (Cheng
and Patel, 2004; Zhou et al., 2001), as well as the IgG binding domain of
Protein A (ZZ domain, 116 residues; Inouye and Sahara, 2009; Rondahl et al.,
1992). These small tags are especially useful in protein preparations for NMR
studies (Kato et al., 2007), and some progress has been made in the creation of
NMR-invisible solubility tags using protein ligation methods (Durst et al.,
2008; Kobashigawa et al., 2009).
Small ubiquitin-like modifier (SUMO) protein (
11 kDa) has
been shown to significantly improve protein stability and solubility as an
N-terminal fusion (Marblestone et al., 2006). The tag can be removed after
purification using SUMO protease (catalytic domain of Ulp1) that recog-
nizes the SUMO structure (Lee et al., 2008; Panavas et al., 2009). The
SUMO-fusion system has also been adapted for use in insect cells and other
eukaryotic expression systems (Liu et al., 2008).
HaloTag is a recently created modular tagging system that uses a 34-kDa
modified haloalkane dehalogenase protein that can bind a variety of syn-
thetic ligands (HaloTag ligands; Promega Inc.). These ligands are comprised
of a constant reactive linker that binds covalently to the HaloTag, and a
variable reporter end that can impart a variety of useful properties to the
fusion protein. Thus, a single tag can be used for subcellular imaging within
live cells, cell labeling and sorting, affinity purification, or even immobili-
zation on solid supports (Los et al., 2008). Ohana et al. (2009) have used a
version of this tag (HaloTag7) for affinity purification, using the chloroalk-
ane linker attached to agarose beads. Since the HaloTag binds covalently to
this linker in a very specific nonreversible manner, even poorly expressed
proteins can be bound efficiently to the chloroalkane resin. The target
protein can then be eluted off the resin using tobacco etch virus (TEV)
protease that cuts at a cleavage site engineered between the HaloTag and the
target protein. Surprisingly, the monomeric compact HaloTag was seen to
dramatically improve fusion protein solubility when tested on a panel of
difficult-to-express recombinant human proteins being expressed in E. coli
(Ohana et al., 2009). In these tests, the HaloTag fared significantly better
than MBP, and appears to function as a bona fide solubility tag.

5. Removal of Tags
After a protein tag has been used for solubility enhancement or affinity
purification, it is often useful to remove it for biological and functional
studies since the tag can potentially interfere with the proper functioning of
the target protein. This is especially true for large tags such GST or MBP,
252 Arun Malhotra

though there are some examples where the fusion protein was more
amenable for crystallization (Smyth et al., 2003).
Most commercial expression vectors that are used to add tags on target
proteins also include cleavage sites with specific sequences that allow the tag
to be removed using recombinant endoproteases. After the initial affinity
purification step, the sample can be treated with the endoprotease to cleave
off the tag, which can subsequently be separated from the target protein by
passing the sample back on the affinity column and collecting the flow-
through. The recombinant endoprotease usually also comes with an affinity
tag, allowing for its easy removal after the cleavage reaction.
Some commonly used endoproteases are listed in Table 16.3. Enteroki-
nase and Factor Xa are useful for removal of N-terminal tags since they cut
at the C-terminal end of their recognition sequence, allowing for the
complete tag and recognition sequence to be removed. However, both
these enzymes are somewhat promiscuous and can cleave at secondary sites,
often at other basic residues. Similar secondary cleavage sites have also been
seen for thrombin, another protease used to remove tags ( Jenny et al., 2003;
Liew et al., 2005). One advantage with proteases such as thrombin, espe-
cially for large-scale protein production, is that it is cheaper and more
efficient than other more specific proteases.
The PreScission protease is a more specific protease with a longer and
stricter recognition sequence. This is the protease 3C from human rhinovi-
rus-14 (3Cpro) with a GST tag, which makes it especially useful for removing
GST tags. Another very specific and popular protease is the TEV protease
(Kapust et al., 2001). The TEV protease prefers a Gly after the cleavage site
(Table 16.3), but can tolerate other residues with only a modest decrease in
activity, which allows it to cleave N-terminal tags with no additional residues
remaining on the target protein in many cases (Kapust et al., 2002). The TEV
protease is easy to produce in-house in large quantities and is usually expressed
with a His-tag for convenient purification and removal after the cleavage
reaction (Tropea et al., 2009). Many forms of TEV proteases, with other
affinity tags or enhanced for higher activity and stability, are sold commercially.
A variant of the TEV protease is the tobacco vein mottling virus
(TVMV) protease that recognizes a different sequence (Nallamsetty et al.,
2004), and can be used to design separate cleavage sites between different
tags (Fig. 16.1C).
Exoproteases can also be used to remove N-terminal tags, such as the
TAGzyme system (Arnau et al., 2006a, available commercially from Qiagen
Inc.). This approach uses dipeptide aminopeptidase I (DAPase) to sequen-
tially chew up the N-terminal tag until a dipeptide ‘‘stop point’’ in the
sequence is reached. Additional variations of this system have been designed
to work with a variety of sequences (Arnau et al., 2006b, 2008).
Complete removal of C-terminal tags is more problematic, since most
endoproteases cut toward the C-terminal end of their recognition sequence.
Tagging for Protein Expression 253

If complete removal of a tag is necessary, more specialized cleavage sites can

be designed that take advantage of structure-based recognition (such as
with the SUMO protease; Malakhov et al., 2004) or an autocatalytic protein
self-splicing element (Inteins; Saleh and Perler, 2006). Both of these cleav-
age systems have been coupled with affinity purification tags—the SUMO-
fusion system (Butt et al., 2005; Lee et al., 2008), and intein-chitin-binding
domain (CBD; also commercialized as the IMPACT system by
New England Biolabs Inc.; Chong et al., 1997) or the intein-polyhydrox-
ybutyrate-binding (PHB) and similar protein purification systems (Gillies
et al., 2008).
In some cases, tags can also be cleaved chemically. Though chemical
cleavage uses cheap reagents and can be very efficient, these reactions
require harsh solvents and denaturing conditions, and are used typically
for preparation of small peptides. Cyanogen bromide (CNBr) cleaves at
methionines, and can be used when the fusion protein can be designed to
have a unique methionine between the tag and the target peptide (Döbeli
et al., 1998; Fairlie et al., 2002). Another chemical cleavage agent, hydrox-
ylamine, cleaves the peptide bond between Asn and Gly (Hu et al., 2008).
Practical considerations: While designing a specific cleavage site between a
tag and a target protein is relatively straightforward, efficient cleavage of the
tag is not always possible or easily predictable. Each construct has to be
experimentally tested both for cleavage efficiency, and for any secondary
cleavages that may occur when promiscuous proteases are used. Often, the
level of the protease and the duration of incubation have to be optimized.
Maximizing the efficiency of cleavage is especially important for oligomeric
proteins where tags have to be removed for each monomer for productive
yields of the target protein (Kenig et al., 2006). The cleavage sequence also
has to be sterically accessible to the protease and relatively unstructured;
poor cleavage can sometimes be alleviated by introducing a spacer or linker
of a few residues between the recognition site and the target protein, or by
using sequences around the cleavage site that are unlikely to form secondary
structures. For proteases that lack an affinity tag, on-column cleavage
(where the protease is injected onto the affinity column with the fusion
protein bound) is often more efficient than batch reactions.

6. Conclusions
A large variety of protein tags are available for facilitating the soluble
expression and purification of recombinant proteins. However, even with
this wide arsenal of tags, structural genomics protein production facilities are
getting soluble purified proteins with success rates of less than 50%
(Structural Genomics Consortium et al., 2008). While many good strategies
254 Arun Malhotra

have emerged from these large-scale studies, given the diversity in how
proteins fold and their distinct biochemical characteristics, there is no
common set of solubility or affinity tags that works for all proteins. Rather,
the choice of tags largely depends on the protein being expressed and the
task at hand. Herein, we have surveyed the most effective protein expres-
sion tags and issues related to their use.

ACKNOWLEDGMENT
The author is supported in part by a grant (R01-GM69972) from the National Institutes
of Health.

REFERENCES
Arnau, J., Lauritzen, C., and Pedersen, J. (2006a). Cloning strategy, production and purifi-
cation of proteins with exopeptidase-cleavable His-tags. Nat. Protoc. 1, 2326–2333.
Arnau, J., Lauritzen, C., Petersen, G. E., and Pedersen, J. (2006b). Current strategies for the
use of affinity tags and tag removal for the purification of recombinant proteins. Protein
Expr. Purif. 48, 1–13.
Arnau, J., Lauritzen, C., Petersen, G. E., and Pedersen, J. (2008). The use of TAGZyme for
the efficient removal of N-terminal His-tags. Methods Mol. Biol. 421, 229–243.
Bachmair, A., Finley, D., and Varshavsky, A. (1986). In vivo half-life of a protein is a function
of its amino-terminal residue. Science 234, 179–186.
Baneyx, F., and Mujacic, M. (2004). Recombinant protein folding and misfolding in
Escherichia coli. Nat. Biotechnol. 22, 1399–1408.
Bauer, A., and Kuster, B. (2003). Affinity purification-mass spectrometry. Powerful tools for
the characterization of protein complexes. Eur. J. Biochem. 270, 570–578.
Bayer, M. E., Bayer, M. H., Lunn, C. A., and Pigiet, V. (1987). Association of thioredoxin
with the inner membrane and adhesion sites in Escherichia coli. J. Bacteriol.
169, 2659–2666.
Bolanos-Garcia, V. M., and Davies, O. R. (2006). Structural analysis and classification of
native proteins from E. coli commonly co-purified by immobilised metal affinity chro-
matography. Biochim. Biophys. Acta 1760, 1304–1313.
Busso, D., Delagoutte-Busso, B., and Moras, D. (2005). Construction of a set Gateway-
based destination vectors for high-throughput cloning and expression screening in
Escherichia coli. Anal. Biochem. 343, 313–321.
Butt, T. R., Edavettal, S. C., Hall, J. P., and Mattern, M. R. (2005). SUMO fusion
technology for difficult-to-express proteins. Protein Expr. Purif. 43, 1–9.
Cabrita, L. D., and Bottomley, S. P. (2004). Protein expression and refolding—A practical
guide to getting the most out of inclusion bodies. Biotechnol. Annu. Rev. 10, 31–50.
Cabrita, L. D., Dai, W., and Bottomley, S. P. (2006). A family of E. coli expression vectors
for laboratory scale and high throughput soluble protein production. BMC Biotechnol.
6, 12.
Carson, M., Johnson, D. H., McDonald, H., Brouillette, C., and Delucas, L. J. (2007).
His-tag impact on structure. Acta Crystallogr. D Biol. Crystallogr. 63, 295–301.
Cheng, Y., and Patel, D. J. (2004). An efficient system for small protein expression and
refolding. Biochem. Biophys. Res. Commun. 317, 401–405.
Tagging for Protein Expression 255

Chong, S., Mersha, F. B., Comb, D. G., Scott, M. E., Landry, D., Vence, L. M.,
Perler, F. B., Benner, J., Kucera, R. B., Hirvonen, C. A., et al. (1997). Single-column
purification of free recombinant proteins using a self-cleavable affinity tag derived from a
protein splicing element. Gene 192, 271–281.
Collins, M. O., and Choudhary, J. S. (2008). Mapping multiprotein complexes by affinity
purification and mass spectrometry. Curr. Opin. Biotechnol. 19, 324–330.
Cunningham, F., and Deber, C. M. (2007). Optimizing synthesis and expression of trans-
membrane peptides and proteins. Methods 41, 370–380.
Davis, G. D., Elisee, C., Newham, D. M., and Harrison, R. G. (1999). New fusion protein
systems designed to give soluble expression in Escherichia coli. Biotechnol. Bioeng. 65, 382–388.
de Marco, A. (2006). Two-step metal affinity purification of double-tagged (NusA-His6)
fusion proteins. Nat. Protoc. 1, 1538–1543.
De Marco, V., Stier, G., Blandin, S., and de Marco, A. (2004). The solubility and stability of
recombinant proteins are increased by their fusion to NusA. Biochem. Biophys. Res.
Commun. 322, 766–771.
de Marco, A., Deuerling, E., Mogk, A., Tomoyasu, T., and Bukau, B. (2007). Chaperone-
based procedure to increase yields of soluble recombinant proteins produced in E. coli.
BMC Biotechnol. 7, 32.
di Guan, C., Li, P., Riggs, P. D., and Inouye, H. (1988). Vectors that facilitate the expression
and purification of foreign peptides in Escherichia coli by fusion to maltose-binding
protein. Gene 67, 21–30.
Döbeli, H., Andres, H., Breyer, N., Draeger, N., Sizmann, D., Zuber, M. T., Weinert, B.,
and Wipf, B. (1998). Recombinant fusion proteins for the industrial production of
disulfide bridge containing peptides: Purification, oxidation without concatemer forma-
tion, and selective cleavage. Protein Expr. Purif. 12, 404–414.
Durst, F. G., Ou, H. D., Löhr, F., Dötsch, V., and Straub, W. E. (2008). The better tag
remains unseen. J. Am. Chem. Soc. 130, 14932–14933.
Dyson, M. R., Shadbolt, S. P., Vincent, K. J., Perera, R. L., and McCafferty, J. (2004).
Production of soluble mammalian proteins in Escherichia coli: Identification of protein
features that correlate with successful expression. BMC Biotechnol. 4, 32.
Einhauer, A., and Jungbauer, A. (2001). The FLAG peptide, a versatile fusion tag for the
purification of recombinant proteins. J. Biochem. Biophys. Methods 49, 455–465.
Esposito, D., and Chatterjee, D. K. (2006). Enhancement of soluble protein expression
through the use of fusion tags. Curr. Opin. Biotechnol. 17, 353–358.
Fairlie, W. D., Uboldi, A. D., De Souza, D. P., Hemmings, G. J., Nicola, N. A., and
Baca, M. (2002). A fusion protein system for the recombinant production of short
disulfide-containing peptides. Protein Expr. Purif. 26, 171–178.
Fritze, C. E., and Anderson, T. R. (2000). Epitope tagging: General method for tracking
recombinant proteins. Methods Enzymol. 327, 3–16.
Gillies, A. R., Hsii, J. F., Oak, S., and Wood, D. W. (2008). Rapid cloning and purification
of proteins: Gateway vectors for protein purification by self-cleaving tags. Biotechnol.
Bioeng. 101, 229–240.
Habig, W. H., Pabst, M. J., and Jakoby, W. B. (1974). Glutathione S-transferases. The first
enzymatic step in mercapturic acid formation. J. Biol. Chem. 249, 7130–7139.
Hernan, R., Heuermann, K., and Brizzard, B. (2000). Multiple epitope tagging of expressed
proteins for enhanced detection. Biotechniques 28, 789–793.
Hu, J., Qin, H., Sharma, M., Cross, T. A., and Gao, F. P. (2008). Chemical cleavage of
fusion proteins for high-level production of transmembrane peptides and protein
domains containing conserved methionines. Biochim. Biophys. Acta 1778, 1060–1066.
Hulo, N., Bairoch, A., Bulliard, V., Cerutti, L., Cuche, B. A., de Castro, E., Lachaize, C.,
Langendijk-Genevaux, P. S., and Sigrist, C. J. (2008). The 20 years of PROSITE. Nucleic
Acids Res. 36, D245–D249.
256 Arun Malhotra

Inouye, S., and Sahara, Y. (2009). Expression and purification of the calcium binding
photoprotein mitrocomin using ZZ-domain as a soluble partner in E. coli cells. Protein
Expr. Purif. 66, 52–57.
Jenny, R. J., Mann, K. G., and Lundblad, R. L. (2003). A critical review of the methods for
cleavage of fusion proteins with thrombin and factor Xa. Protein Expr. Purif. 31, 1–11.
Jungbauer, A., and Kaar, W. (2007). Current status of technical protein refolding.
J. Biotechnol. 128, 587–596.
Kaplan, W., Hüsler, P., Klump, H., Erhardt, J., Sluis-Cremer, N., and Dirr, H. (1997).
Conformational stability of pGEX-expressed Schistosoma japonicum glutathione S-trans-
ferase: A detoxification enzyme and fusion-protein affinity tag. Protein Sci. 6, 399–406.
Kapust, R. B., and Waugh, D. S. (1999). Escherichia coli maltose-binding protein is uncom-
monly effective at promoting the solubility of polypeptides to which it is fused. Protein
Sci. 8, 1668–1674.
Kapust, R. B., Tözsér, J., Fox, J. D., Anderson, D. E., Cherry, S., Copeland, T. D., and
Waugh, D. S. (2001). Tobacco etch virus protease: Mechanism of autolysis and rational
design of stable mutants with wild-type catalytic proficiency. Protein Eng. 14, 993–1000.
Kapust, R. B., Tözsér, J., Copeland, T. D., and Waugh, D. S. (2002). The P10 specificity of
tobacco etch virus protease. Biochem. Biophys. Res. Commun. 294, 949–955.
Kataeva, I., Chang, J., Xu, H., Luan, C. H., Zhou, J., Uversky, V. N., Lin, D., Horanyi, P.,
Liu, Z. J., Ljungdahl, L. G., et al. (2005). Improving solubility of Shewanella oneidensis
MR-1 and Clostridium thermocellum JW-20 proteins expressed into Escherichia coli.
J. Proteome Res. 4, 1942–1951.
Kato, A., Maki, K., Ebina, T., Kuwajima, K., Soda, K., and Kuroda, Y. (2007). Mutational
analysis of protein solubility enhancement using short peptide tags. Biopolymers 85, 12–18.
Kenig, M., Peternel, S., Gaberc-Porekar, V., and Menart, V. (2006). Influence of the protein
oligomericity on final yield after affinity tag removal in purification of recombinant
proteins. J. Chromatogr. A 1101, 293–306.
Kim, S., and Lee, S. B. (2008). Soluble expression of archaeal proteins in Escherichia coli by
using fusion-partners. Protein Expr. Purif. 62, 116–119.
Kobashigawa, Y., Kumeta, H., Ogura, K., and Inagaki, F. (2009). Attachment of an NMR-
invisible solubility enhancement tag using a sortase-mediated protein ligation method.
J. Biomol. NMR 43, 145–150.
LaVallie, E. R., DiBlasio, E. A., Kovacic, S., Grant, K. L., Schendel, P. F., and McCoy, J. M.
(1993). A thioredoxin gene fusion expression system that circumvents inclusion body
formation in the E. coli cytoplasm. Biotechnology (NY) 11, 187–193.
LaVallie, E. R., Lu, Z., Diblasio-Smith, E. A., Collins-Racie, L. A., and McCoy, J. M.
(2000). Thioredoxin as a fusion partner for production of soluble recombinant proteins in
Escherichia coli. Methods Enzymol. 326, 322–340.
Lee, C. D., Sun, H. C., Hu, S. M., Chiu, C. F., Homhuan, A., Liang, S. M., Leng, C. H.,
and Wang, T. F. (2008). An improved SUMO fusion protein system for effective
production of native proteins. Protein Sci. 17, 1241–1248.
Lichty, J. J., Malecki, J. L., Agnew, H. D., Michelson-Horowitz, D. J., and Tan, S. (2005).
Comparison of affinity tags for protein purification. Protein Expr. Purif. 41, 98–105.
Liew, O. W., Ching Chong, J. P., Yandle, T. G., and Brennan, S. O. (2005). Preparation of
recombinant thioredoxin fused N-terminal proCNP: Analysis of enterokinase cleavage
products reveals new enterokinase cleavage sites. Protein Expr. Purif. 41, 332–340.
Listwan, P., Terwilliger, T. C., and Waldo, G. S. (2009). Automated, high-throughput
platform for protein solubility screening using a split-GFP system. J. Struct. Funct.
Genomics 10, 47–55.
Liu, J. W., Boucher, Y., Stokes, H. W., and Ollis, D. L. (2006). Improving protein
solubility: The use of the Escherichia coli dihydrofolate reductase gene as a fusion reporter.
Protein Expr. Purif. 47, 258–263.
Tagging for Protein Expression 257

Liu, L., Spurrier, J., Butt, T. R., and Strickler, J. E. (2008). Enhanced protein expression in
the baculovirus/insect cell system using engineered SUMO fusions. Protein Expr. Purif.
62, 21–28.
Los, G. V., Encell, L. P., McDougall, M. G., Hartzell, D. D., Karassina, N., Zimprich, C.,
Wood, M. G., Learish, R., Ohana, R. F., Urh, M., et al. (2008). HaloTag: A novel protein
labeling technology for cell imaging and protein analysis. ACS Chem. Biol. 3, 373–382.
Malakhov, M. P., Mattern, M. R., Malakhova, O. A., Drinker, M., Weeks, S. D., and
Butt, T. R. (2004). SUMO fusions and SUMO-specific protease for efficient expression
and purification of proteins. J. Struct. Funct. Genomics 5, 75–86.
Marblestone, J. G., Edavettal, S. C., Lim, Y., Lim, P., Zuo, X., and Butt, T. R. (2006).
Comparison of SUMO fusion technology with traditional gene fusion systems: Enhanced
expression and solubility with SUMO. Protein Sci. 15, 182–189.
Mueller, U., Büssow, K., Diehl, A., Bartl, F. J., Niesen, F. H., Nyarsik, L., and
Heinemann, U. (2003). Rapid purification and crystal structure analysis of a small protein
carrying two terminal affinity tags. J. Struct. Funct. Genomics 4, 217–225.
Nallamsetty, S., and Waugh, D. S. (2006). Solubility-enhancing proteins MBP and NusA
play a passive role in the folding of their fusion partners. Protein Expr. Purif. 45, 175–182.
Nallamsetty, S., Kapust, R. B., Tözsér, J., Cherry, S., Tropea, J. E., Copeland, T. D., and
Waugh, D. S. (2004). Efficient site-specific processing of fusion proteins by tobacco vein
mottling virus protease in vivo and in vitro. Protein Expr. Purif. 38, 108–115.
Ohana, R. F., Encell, L. P., Zhao, K., Simpson, D., Slater, M. R., Urh, M., and
Wood, K. V. (2009). HaloTag7: A genetically engineered tag that enhances bacterial
expression of soluble proteins and improves protein purification. Protein Expr. Purif.
68, 110–120.
Panavas, T., Sanders, C., and Butt, T. R. (2009). SUMO fusion technology for enhanced
protein production in prokaryotic and eukaryotic expression systems. Methods Mol. Biol.
497, 303–317.
Pédelacq, J. D., Piltch, E., Liong, E. C., Berendzen, J., Kim, C. Y., Rho, B. S., Park, M. S.,
Terwilliger, T. C., and Waldo, G. S. (2002). Engineering soluble proteins for structural
genomics. Nat. Biotechnol. 20, 927–932.
Peleg, Y., and Unger, T. (2008). Application of high-throughput methodologies to the
expression of recombinant proteins in E. coli. Methods Mol. Biol. 426, 197–208.
Potts, J. T., Young, D. M., and Anfinsen, C. B. (1963). Reconstitution of fully active RNase
S by carboxypeptidase-degraded RNase S-peptide. J. Biol. Chem. 238, 2593–2594.
Prickett, K. S., Amberg, D. C., and Hopp, T. P. (1989). A calcium-dependent antibody for
identification and purification of recombinant proteins. Biotechniques 7, 580–589.
Puig, O., Caspary, F., Rigaut, G., Rutz, B., Bouveret, E., Bragado-Nilsson, E., Wilm, M.,
and Séraphin, B. (2001). The tandem affinity purification (TAP) method: A general
procedure of protein complex purification. Methods 24, 218–229.
Qoronfleh, M. W., Hesterberg, L. K., and Seefeldt, M. B. (2007). Confronting high-
throughput protein refolding using high pressure and solution screens. Protein Expr.
Purif. 55, 209–224.
Raines, R. T., McCormick, M., Van Oosbree, T. R., and Mierendorf, R. C. (2000).
The S.Tag fusion system for protein purification. Methods Enzymol. 326, 362–376.
Richards, F. M., and Vithayarhil, P. J. (1959). The preparation of subtilisin-modified
ribonuclease and the separation of the peptide and protein components. J. Biol. Chem.
234, 1459–1465.
Rondahl, H., Nilsson, B., and Holmgren, E. (1992). Fusions to the 50 end of a gene
encoding a two-domain analogue of staphylococcal protein A. J. Biotechnol. 25, 269–287.
Roodveldt, C., Aharoni, A., and Tawfik, D. S. (2005). Directed evolution of proteins for
heterologous expression and stability. Curr. Opin. Struct. Biol. 15, 50–56.
258 Arun Malhotra

Sachdev, D., and Chirgwin, J. M. (1998). Order of fusions between bacterial and mamma-
lian proteins can determine solubility in Escherichia coli. Biochem. Biophys. Res. Commun.
244, 933–937.
Sahdev, S., Khattar, S. K., and Saini, K. S. (2008). Production of active eukaryotic proteins
through bacterial expression systems: A review of the existing biotechnology strategies.
Mol. Cell. Biochem. 307, 249–264.
Saleh, L., and Perler, F. B. (2006). Protein splicing in cis and in trans. Chem. Rec. 6, 183–193.
Sano, T., and Cantor, C. R. (2000). Streptavidin-containing chimeric proteins: Design and
production. Methods Enzymol. 326, 305–311.
Schmidt, T. G., and Skerra, A. (1994). One-step affinity purification of bacterially produced
proteins by means of the ‘‘Strep tag’’ and immobilized recombinant core streptavidin.
J. Chromatogr. A 676, 337–345.
Schmidt, T. G., and Skerra, A. (2007). The Strep-tag system for one-step purification and
high-affinity detection or capturing of proteins. Nat. Protoc. 2, 1528–1535.
Skerra, A., and Schmidt, T. G. (2000). Use of the Strep-tag and streptavidin for detection
and purification of recombinant proteins. Methods Enzymol. 326, 271–304.
Smith, D. B., and Johnson, K. S. (1988). Single-step purification of polypeptides expressed in
Escherichia coli as fusions with glutathione S-transferase. Gene 67, 31–40.
Smits, S. H., Mueller, A., Grieshaber, M. K., and Schmitt, L. (2008). Coenzyme- and
His-tag-induced crystallization of octopine dehydrogenase. Acta Crystallogr. Sect. F Struct.
Biol. Cryst. Commun. 64, 836–839.
Smyth, D. R., Mrozkiewicz, M. K., McGrath, W. J., Listwan, P., and Kobe, B. (2003).
Crystal structures of fusion proteins with large-affinity tags. Protein Sci. 12, 1313–1322.
Structural Genomics Consortium, China Structural Genomics Consortium, Northeast
Structural Genomics Consortium, Gräslund, S., Nordlund, P., Weigelt, J.,
Hallberg, B. M., Bray, J., Gileadi, O., Knapp, S., et al. (2008). Protein production and
purification. Nat. Methods 5, 135–146.
Terpe, K. (2003). Overview of tag protein fusions: From molecular and biochemical
fundamentals to commercial systems. Appl. Microbiol. Biotechnol. 60, 523–533.
Tolun, A. A., Dickerson, I. M., and Malhotra, A. (2007). Overexpression and purification of
human calcitonin gene-related peptide-receptor component protein in Escherichia coli.
Protein Expr. Purif. 52, 167–174.
Tropea, J. E., Cherry, S., and Waugh, D. S. (2009). Expression and purification of soluble
His(6)-tagged TEV protease. Methods Mol. Biol. 498, 297–307.
Vaillancourt, P., Zheng, C. F., Hoang, D. Q., and Breister, L. (2000). Affinity purification of
recombinant proteins fused to calmodulin or to calmodulin-binding peptides. Methods
Enzymol. 326, 340–362.
Waldo, G. S. (2003). Genetic screens and directed evolution for protein solubility. Curr.
Opin. Chem. Biol. 7, 33–38.
Waldo, G. S., Standish, B. M., Berendzen, J., and Terwilliger, T. C. (1999). Rapid protein-
folding assay using green fluorescent protein. Nat. Biotechnol. 17, 691–695.
Wang, K. H., Oakes, E. S., Sauer, R. T., and Baker, T. A. (2008). Tuning the strength of a
bacterial N-end rule degradation signal. J. Biol. Chem. 283, 24600–24607.
Waugh, D. S. (2005). Making the most of affinity tags. Trends Biotechnol. 23, 316–320.
Yokoyama, S. (2003). Protein expression systems for structural genomics and proteomics.
Curr. Opin. Chem. Biol. 7, 39–43.
Zhang, Y. B., Howitt, J., McCorkle, S., Lawrence, P., Springer, K., and Freimuth, P.
(2004). Protein aggregation during overexpression limited by peptide extensions with
large net negative charge. Protein Expr. Purif. 36, 207–216.
Zhou, P., Lugovskoy, A. A., and Wagner, G. (2001). A solubility-enhancement tag (SET)
for NMR studies of poorly behaving proteins. J. Biomol. NMR 20, 11–14.
 C H A P T E R S E V E N T E E N

Refolding Solubilized Inclusion

Body Proteins
Richard R. Burgess

Contents
1. Introduction 260
2. General Refolding Consideration 262
3. General Procedures 262
4. General Protocol 263
5. Comments on this General Procedure 264
5.1. Overexpression of recombinant proteins in E. coli 264
5.2. Washing IBs 265
5.3. Solubilizing IBs 265
5.4. Refolding 267
5.5. High-resolution ion-exchange chromatography 268
5.6. Reoxidation to form correct disulfide bonds 269
5.7. Characterization 270
6. Performing a Protein Refolding Test Screen 271
6.1. Systematic refolding screens 271
6.2. Variables in refolding 271
6.3. A practical, inexpensive rational approach 274
7. Other Refolding Procedures 275
8. Refolding Database: Refold 277
9. Strategies to Increase Proportion of Soluble Protein 277
10. Conclusion 279
References 279

Abstract
The vast majority of protein purification is now done with cloned, recombinant
proteins expressed in a suitable host. The predominant host is Escherichia coli.
Many, if not most, expressed proteins are found in an insoluble form called an
inclusion body (IB). Since the target protein is often relatively pure in a washed
IB, the challenge is not so much to purify the target, but rather to solubilize an
IB and refold the protein into its native structure, regaining full biological

McArdle Laboratory for Cancer Research, University of Wisconsin-Madison, Madison, Wisconsin, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63017-2 All rights reserved.

259
260 Richard R. Burgess

activity. While many of the operations of this process are quite general (expres-
sion, cell disruption, IB isolation and washing, and IB solubilization), the precise
conditions that give efficient refolding differ for each protein. This chapter
describes the main techniques and strategies for achieving successful refolding.

1. Introduction
When proteins are expressed at high levels in Escherichia coli and other
expression hosts, it is common to find that most of the protein is not soluble,
but in an insoluble form called an inclusion body (IB). While the exact
mechanism for IB formation is not fully understood and may vary with
different proteins and expression conditions, it is generally thought that
target proteins are being made faster than they can fold into the native
structure. If a protein is partially folded or misfolded, it will generally have
hydrophobic or ‘‘sticky’’ regions exposed that can interact with other
similar proteins and form aggregates. These aggregates form IBs that are
dense, refractile bodies on the order of 0.2–0.5 mm in diameter. Once in an
IB, the protein is usually protected from proteolytic attack and is the
predominant protein. IBs can vary from being mostly native protein,
relatively easily solubilized under mild condition to being misfolded, effec-
tively irreversibly insoluble material that requires high concentrations
of denaturants to be solubilized (see Bowden et al., 1991; Ventura and
Villaverde, 2006). The latter category is by far the most common and
thus, the major barrier to obtaining native, active material is to find condi-
tions under which the denatured protein can be refolded efficiently. This
refolding process and strategies to achieve it are the focus of this chapter.
Anfinsen, in the early 1960s (see Anfinsen, 1973), observed that small
protein molecules could fold spontaneously to the native state. This observa-
tion has led to a thermodynamic hypothesis: the native state has the lowest
free energy. While the thermodynamic hypothesis has become a paradigm of
protein folding, a simple and general model for describing how an unfolded
protein molecule finds its native state is still not available. One cannot yet use
computers to accurately predict protein structure from sequence.
Levinthal (1968) pointed out the paradox that an unfolded protein
molecule does not have the time needed to search all possible conforma-
tions, yet it reaches the single native state rapidly. Consider a (hypothetical)
small protein with 100 residues. Levinthal calculated that, if each residue can
assume only three different conformations (it is probably many more than
this), the total number of structures would be 3100, which is equal to
5
1047. If it takes only 10 13 s to convert one structure into another,
the total search time would be 5
1047 times10 13 s, which is equal to
5
1034 s or 1.6
1027 years! Clearly, it would take much too long for even
Refolding Proteins 261

a small protein to fold properly by randomly trying out all possible con-
formations. This paradox has shaped many studies on the theory of refolding
(e.g., see Clark, 2004; Karplus, 1997).
A great many experimental studies and theoretical analyses have been
carried out to better understand the folding process. The consensus is that
when a denatured protein is no longer in denaturing conditions, some
secondary structure (a-helix and b-sheet) forms, a hydrophobic collapse
occurs to form a ‘‘molten globule’’ state with some secondary structure, but
nonnative tertiary structures, the globular states go through multiple inter-
mediates, and finally reach the lowest energy state, the native conformation.
Solution conditions can have significant effects on the folding pathway of a
given protein. If you think of the refolding energy funnel as a deep, fog-filled
crater in the top of a volcanic mountain, then the rim is the denatured state
and the very bottom of the crater is the native state (the lowest energy state).
If you roll a ball down into the crater from one point on the rim of the crater,
then it may be able to roll all the way to the bottom. From another point on
the rim, the ball may become trapped in a local minimum and not reach the
bottom. In many ways the challenge of finding conditions for efficient
refolding is like trying to guess where on the rim to start rolling the ball.
The rate of protein refolding is rapid, with folding halftimes on the order
of seconds for proteins lacking prolines and on the order of minutes for
proteins with prolines (see Nall, 1994). The amino acid proline can consist
of two isomers, a cis and trans configuration, and the halftime of this
chemical isomerization is on the order of 1 min. In a native protein, each
proline is in a particular configuration, but in a denatured protein an
equilibrium mixture of the two isomers forms consisting of about 70%
trans and 30% cis isomer. Therefore, as a protein refolds, those prolines
that are in the incorrect conformation cannot form the native structure and
must wait for an isomerization event to occur. Thus, proline isomerization
becomes rate limiting for correct protein folding. The enzyme, protein
proline isomerase (PPI), can speed up this isomerization rate.
Even though recombinant proteins were first prepared by refolding
solubilized IBs in the late 1970s and early 1980s, and many improvements
have been made in refolding procedures and techniques, refolding of any
given protein still presents a significant challenge. During the last 10 years
the protein structure initiative (PSI), through its numerous Structural
Genomics Centers, has cloned and expressed over 110,000 proteins (as of
December 2008), but were only able to purify about 29,000 (about 26%).
On the order of 50% of Eubacterial and Archaeal proteins and only about
10% of Eukaryal proteins could be expressed as soluble proteins in E. coli
(Graslund et al., 2008). The proteins that were insoluble were generally
abandoned rather than trying to find effective refolding conditions. I believe
that many of these insoluble proteins could have been salvaged or rescued
by finding suitable refolding conditions as described below.
262 Richard R. Burgess

Many excellent reviews report approaches to protein refolding (Baneyx

and Mujacic, 2004; Cabrita and Bottomley, 2004; Clark, 2001; Jungbauer and
Kaar, 2006; Middelberg, 2002; Panda, 2003; Quronfleh et al., 2007; Singh and
Panda, 2005; Swietnicki, 2006; Thatcher and Hitchcock, 1994; Tsumoto et al.,
2003; Vallejo and Rinas, 2004). I will try to present the kind of advice below that
I would give to colleagues embarking on a project that involves protein refolding.

2. General Refolding Consideration

The main problem observed in many refolding attempts is that upon
dilution or dialysis of the solubilized IB to decrease the concentration of the
denaturant to a nondenaturing level, you get major precipitation, especially if
the protein concentration is high during refolding. In a way this is likely to be
similar to the process in the cell that leads to formation of IBs in the first place.
You want to go from denatured to partially folded intermediate to native
structure, and bias the reaction away from aggregation of the sticky folding
intermediate and precipitation. Aggregation is dependent on collision of sticky
folding intermediates. At low protein concentrations the probability of colli-
sion is diminished. The best general strategy in refolding is to refold at the lowest
protein concentration that is feasible. In screening refolding conditions, one
searches for conditions where little aggregation or turbidity is observed. But
even when there is no obvious precipitation (the dilute protein in refolding
solution is not turbid), there can be significant ‘‘soluble multimer.’’ Soluble
multimers are generally from 2- to 20-mers that are soluble and not large
enough to cause much light scattering (turbidity). Usually these soluble multi-
mers are not active and tend to bind tightly to ion exchange resins.

3. General Procedures
Given that it is very often the case that protein expressed in E. coli and
other hosts is found in IBs, the following steps are almost always needed to
obtain active, native protein:
1. Cell growth and protein overexpression
2. Cell lysis, isolating and washing IBs
3. Solubilizing IBs
4. Identifying suitable refolding conditions
5. Protein refolding
6. Reoxidation to form disulfide bonds where necessary
7. High-resolution ion-exchange chromatography
8. Characterization of final material to determine if refolding has been
successful
Refolding Proteins 263

4. General Protocol
Below is a typical procedure that works well for many proteins. This
protocol is based on and adapted from the protocol first developed
by Nguyen et al. (1993) and used in the Cold Spring Harbor Protein
Purification and Characterization Course section on Purification of Insolu-
ble Recombinant Proteins (Burgess and Knuth, 1996). Other similar
procedures are likely to give similar results, but this is the procedure used
almost exclusively in my laboratory. The critical steps will be discussed in
the following steps.
1. E. coli BL21(DE3)pLysS containing the target gene cloned into a pET
expression vector (Studier et al., 1990) is grown in 1 L of LB in a 2-L
flask at 37  C with shaking until the A600 nm reaches about 0.6.
2. IPTG is added to 0.5–1 mM to induce expression of the T7 RNA
polymerase that then transcribes the target gene.
3. Three to four hours after induction, the cells are harvested by centrifu-
gation at 15,000 rpm for 15 min, the cells are resuspended in a small
volume of the culture supernatant, transferred into a preweighed,
40-ml Oak Ridge tube, and centrifuged to pellet the cells. The wet
weight of the cell pellet is noted and the cells are stored frozen
at  80  C until use. It is common to get 1.5–2.0 g wet weight
E. coli cells per liter using this procedure.
4. Cells are thawed, resuspended in 30 ml of a lysis buffer (50 mM
Tris–HCl, pH 7.9, 0.1 mM EDTA, 5% glycerol, 0.1 mM DTT,
0.1 M NaCl), and then sonicated at 60% power for 3–4 intervals of
20 s each with about 1 min on ice to cool between each interval.
5. Purified Triton X-100 (from a 10%, w/v stock) is added to 1% (w/v) to
dissolve membranes and solubilize membrane proteins. The lysate is
incubated on ice for 10 min and then centrifuged at 15,000 rpm for
15 min to pellet the IB and remove the soluble supernatant.
6. The IB is resuspended in 30 ml of lysis buffer with 1% Triton X-100,
incubated on ice for 10 min, and then centrifuged for 15,000 rpm for
15 min.
7. The drained pellet is resuspended in 30 ml of lysis buffer without Triton
X-100 to remove the Triton X-100 and recentrifuged as above. The pellet
is called the washed IB fraction and is usually over 90% pure.
8. The washed IB pellet is resuspended in a suitable denaturant, incubated,
to allow denaturation and solubilization, and then centrifuged at
15,000 rpm for 15 min to remove any residual insoluble material. We
routinely use the lysis buffer above (lacking the NaCl if diluting from
GuHCl) containing either 6 M guanidine hydrochloride (GuHCl), 8 M
urea, or 0.3% Sarkosyl (n-lauroyl sarcosinate) for IB solubilization.
264 Richard R. Burgess

9. More recently, we have carried out a relatively simple refolding test

(see below) to identify a suitable refolding buffer.
10. Adjust the protein concentration in the denatured sample to about
1 mg/ml.
11. Dilute the denatured protein about 15–60 fold to dilute the denaturant
concentration to a point where the protein can refold. We usually either
flash dilute or slowly drip dilute the denatured IB into refolding buffer in a
beaker with vigorous stirring to cause very rapid mixing. The dilution
process is performed at room temperature and after mixing is complete,
we let the solution stand for 1–2 h to allow the refolding process to
complete and to allow any aggregated material to form and flocculate.
12. The refolded material is filtered through a low protein binding,
0.22 mm membrane filter (e.g., Stericup-GV 0.22 mm, 500 ml,
Millipore #SCGVU05RE) to remove any particulate material (this
can be preceded by a centrifugation step if there is significant precipi-
tated material that would otherwise immediately clog the filter).
13. Pump the filtered solution onto a 10–15-ml ion-exchange column as
fast as the pressure constraints of the column and system will allow
(hopefully, at least 10 ml/min). Monitor the absorbance at 280, 260,
and 320 nm, if possible. The absorbance at 320 nm is a measure of light
scattering and for some proteins can indicate a peak that is composed of
multimers.
14. Wash the column with 5–10 column volumes of Buffer A (50 mM
Tris–HCl, pH 7.9, 5% glycerol, 0.1 mM EDTA, 0.1 mM DTT) plus
0.1 M NaCl and then elute with a 10-column volume, linear salt
gradient in the same buffer from 0.1 to 1.0 M NaCl at a flow rate of
5 ml/min, collecting 3–4 ml fractions.
15. Analyze the A280 nm peaks by SDS–PAGE to determine purity of the
various fractions. If an enzyme assay is available, assay fractions to
determine those with the highest specific activity.
16. Pooled peak fractions are dialyzed against lysis buffer containing 50%
glycerol and stored at  20 or  80  C.
17. Characterize the pooled peak material, if possible, by specific activity
compared to a standard sample of the protein purified by a more
conventional procedure that does not involve refolding.

5. Comments on this General Procedure

5.1. Overexpression of recombinant proteins in E. coli
The issue of achieving a high level of expression is a topic somewhat
separate from the refolding issue and will not be discussed further here
(see Chapter 12 in this volume). Many systems will allow expression levels
Refolding Proteins 265

of the target protein of 10–40% of the total cell protein (Makrides, 1996;
Murby et al., 1996; Sorensen and Mortensen, 2005; Studier et al., 1990)
(see Section 9). Usually one carries out a test to see at what temperature to
grow the cells, how much inducer to use, and how long after induction you
should harvest the cells. After induction of the high-level expression of
the target protein, the cells are usually centrifuged, and frozen at  80  C
until use. There are anecdotal reports that IBs from cells stored for weeks in
the frozen state are harder to refold, but in my experience, we have gotten
excellent refolding results from cells stored frozen for months. Unfortunately,
to my knowledge, no systematic study has addressed this point.

5.2. Washing IBs

We originally used 2% sodium deoxycholate to wash IBs (Burgess and
Knuth, 1996; Nguyen et al., 1993), but later found that 1% Triton X-100
worked better and was much easier to work with. It is highly recommended
that one use what I call ‘‘gourmet’’ Triton X-100 (Pierce, Surfact-Amps
X-100, 10% (w/v) solution) that has been purified (to remove peroxides
sometimes found in old, yellowish bottles of Triton X-100) and stored
under nitrogen in sealed ampoules. It is possible to test various salts and
detergents to see which are particularly effective at washing out protein
impurities from an IB, but not solubilizing a given target protein.
Sometimes washes with 1–2 M urea can be used. In general, it is desirable
to remove as much membrane material, DNA, and other proteins as
possible from an IB before solubilization.

5.3. Solubilizing IBs

As mentioned earlier, some IBs are nearly native and can be solubilized by
mild conditions such as nonionic detergent or even 0.5 M NaCl (Vera et al.,
2006). In most cases, this will not work and much stronger denaturing
conditions must be used.
The solubilizing agent/denaturant is prepared in a buffer, much like the
lysis buffer above, to control pH, chelate heavy metals, and maintain a
reducing environment. Usually one resuspends the IB with the solubilizing
agent, lets it incubate for 30–60 min, and then centrifuges out any insoluble
material. Since the material is already at least partially denatured in the IB,
this solubilization can be performed at room temperature and in some cases it is
even necessary to heat the solution to achieve full solubilization. For example,
when native green fluorescent protein (GFP) is adjusted to 6 M GuHCl
it retains its fluorescence at room temperature, but at 75  C it denatures
and loses it fluorescence within 1 min (R. Burgess and N. Thompson, unpub-
lished results). A detailed discussion of solubilization is found in Marston and
Hartley (1990).
266 Richard R. Burgess

Some comments on the use of several solubilizing agents follow:

1. GuHCl. This is perhaps the most commonly used solubilizing agent,
usually used at 6 M in a compatible buffer. Most proteins will rapidly
denature in this strong chaotropic agent, but in many cases it may help to
incubate at higher temperatures to achieve complete denaturation. Once
solubilization occurs, it is often possible to dilute to 3 M GuHCl without
a problem, since the transition from denatured to native state often
occurs in the 1–2 M GuHCl range (Pace, 1986). One usually dilutes
to about 0.1 M GuHCl so that the salt concentration is low enough to
allow the refolded target protein to bind to a suitable ion-exchange
column (see below).
2. Urea. 8 M urea is a common solubilizing agent, and is especially useful if
one wants to further purify the target protein in the denatured state
(Knuth and Burgess, 1987). The 8 M urea is generally not as effective as
6 M GuHCl in solubilizing protein and causing complete denaturation.
One must be aware of the possibility of carbamylation of proteins by the
cyanate always present in urea solutions (see Chapter 44 in this volume
for advice on how to minimize cyanate in urea).
3. Sarkosyl or SDS. Sarkosyl has been found to be a quite effective solubiliz-
ing agent that allows refolding at higher protein concentrations
(see Burgess, 1996). Sarkosyl is a strong anionic detergent, much like
SDS, but it seems to bind to proteins more weakly and dissociate more
easily than SDS. Usually an IB pellet can be solubilized in 0.3% Sarkosyl,
but you must remember that you cannot solubilize more than about an
equal weight of protein per weight of Sarkosyl. For example, if you have
a large washed IB that contains 150 mg of target protein, it will not be
solubilized completely by 20 ml of 0.3% Sarkosyl (60 mg of Sarkosyl).
You will have to add at least 50 ml of 0.3% or 30 ml of 0.5% Sarkosyl.
We have noticed that if there is much Triton X-100 present in an IB, it
will take more Sarkosyl to solubilize the target protein. This is almost
certainly due to the fact that Triton X-100 forms large micelles and absorbs
some of the Sarkosyl to form mixed micelles that are not effective at
solubilization.
If one dilutes 0.3% Sarkosyl-solubilized target protein to about 0.01%
Sarkosyl, most of the Sarkosyl will dissociate from the protein and the
protein will refold. The residual detergent seems to bind to the hydrophobic
regions (the ‘‘sticky’’ sites) on partially refolded protein and prevent aggre-
gation. It is effectively acting as a chemical chaperone. If your target protein
is able to bind to a cation-exchange column such as a POROS HS, then the
diluted solution can be filtered and pumped on a 10-ml column, washed
with 10–20 column volumes of 0.1 M NaCl in buffer, and eluted with a salt
gradient (see Section 5.5). The target protein will bind, but the free Sarkosyl
will flow-through and any residual Sarkosyl bound to the protein will
Refolding Proteins 267

dissociate during the long wash. Experimental analysis of the eluted target
protein indicates essentially no (<1 molecule detergent per 10 molecules of
protein) residual Sarkosyl in the protein (R. Burgess, unpublished). Do not
use MonoS for this purpose since Sarkosyl binds to the column and can
damage the column and contaminate your eluted protein.
SDS can also be used in a somewhat similar fashion, but it is important to
use SDS that is free from higher alkane lengths such as C14, C16, etc.
Refolding experiments using GFP denatured in SDS work quite well
(R. Burgess N. Thompson, and R. Chumanov, unpublished), so SDS
should be considered a viable solubilizing agent.
The successful use of the cationic detergent cetyltrimethylammonium
chloride (CTAC) in solubilization and refolding has also been reported
(Puri et al., 1992). Even distilled water at very low salt has been reported
to achieve solubilization (Song, 2009).

5.4. Refolding
Assuming you have carried out refolding trials as described below, then one
basically dilutes out the denaturant by diluting the solubilized protein into a
suitable refolding buffer. However, there are three ways to dilute during
refolding.
1. Reverse dilution: Add refolding buffer to denatured protein with mixing
between each addition. This method has been successful in several cases
(e.g., Gribskov and Burgess, 1983). However, on reflection, it is clear
that this method of dilution results in the highest concentration of
protein at the critical time when the protein is starting to refold. For
example, if the solubilized protein concentration is 1 mg/ml in 6 M
GuHCl, then if one adds 2–5 volumes of refolding buffer (dilutes to
1–2 M GuHCl), one is likely to go through the refolding transition
between protein concentrations of 330–160 mg/ml.
2. Flash dilution: Add denatured protein to refolding buffer quickly. For
example, add 10 ml of solubilized protein in 6 M GuHCl at one time
with rapid mixing to 590 ml of a suitable refolding buffer to achieve a
60-fold dilution. The protein concentration will be 16 mg/ml during the
refolding, much lower than with reverse dilution and less likely to result
in aggregation.
3. Drip dilution: Add denatured protein to refolding buffer very slowly,
drop-by-drop, over period of 1 h. In theory, this should be the optimal
method, since protein is always refolding at minimal protein concentra-
tion (see Singh and Panda, 2005; Vallejo and Rinas, 2004). For example,
if you add 0.1 ml of solubilized protein as above into 590 ml refolding
buffer, then the protein concentration will only be 0.16 mg/ml. As you
add small incremental amounts of the sample, the protein will always be
268 Richard R. Burgess

low and when the last has been added, the protein concentration will be
16 mg/ml. However, if the denatured protein is added over a period of
1 h, and significant protein refolding to native state is achieved with a
refolding half-life of 1–3 min, then the concentration of partially
refolded sticky protein will always be quite low and the native protein
will not be available to aggregate. While adding solubilized protein over
1 h period is tedious when added drop-by-drop, one can set up a
peristaltic pump at a very slow flow rate to deliver the denatured protein
over 1 h period. This is more reproducible and causes considerably less
graduate student burnout.
Let the protein sit after dilution for 30–60 min and then filter as
described in the generic protocol. If this is not done, then as the protein
solution is loaded on the column (see below) very small particulate material
is likely to clog up the column, causing the pressure to rise, either aborting
the run or requiring one to decrease the flow rate to such low levels that it
takes many hours to load the column.

5.5. High-resolution ion-exchange chromatography

After refolding and filtering, a final, high-resolution ion-exchange column
step accomplishes five important jobs:
1. It concentrates protein (e.g., from 600 to 4–8 ml).
2. It removes denaturant (in the flow-through).
3. It removes minor impurities (in flow-through or binding weaker or stronger
than the target protein to the column). If a high-resolution anion-
exchange column is used, like POROS HQ or MonoQ, any DNA
contamination will bind tightly and elute at 0.6–0.9 M NaCl. If the
presence of E. coli lipopolysaccharide (LPS) in the final product is
particularly undesirable, one can wash the column with isopropanol to
greatly reduce LPS before beginning the salt gradient elution.
4. It selects for a homogeneous, active monomer. If one has a sample of the native
protein, purified by nonrefolding methods, then one can see at what salt
it elutes from this same ion-exchange column. If a major peak is eluted at
the same salt from the refolded material, then one has some confidence
that the material is properly refolded because misfolded molecules are
likely (but not necessarily) to elute slightly differently if the column is
capable of high resolution.
5. It removes soluble protein multimers. Soluble multimers are very often
found in refolded material (see Section 5.7). If you have chosen the
refolding solution well, they may be minimal; if not, they may domi-
nate. These soluble N-mers have N times as much charge as the
monomer. They tend to bind tighter to the ion-exchange column
and elute later due to their higher charge. Removal of multimers is
Refolding Proteins 269

an extremely important step if the protein is being used for crystalliza-

tion for structure determination, since heterodisperse material is much
less likely to crystallize.
In my experience, in most cases the protein eluting in the first
major peak from the column will be high quality, pure material with full
biological activity.

5.6. Reoxidation to form correct disulfide bonds

If your protein has no cysteines this is not an issue. However, most
proteins contain cysteines, and many have disulfide bonds as important
parts of their three-dimensional structure. Since the cytoplasm of E. coli is
very reducing, most internal proteins are in the reduced state. If you add a
reducing agent like 0.1–1 mM DTT to the lysis buffer, you can usually
keep the protein reduced and prevent unwanted or incorrect disulfide
bonds from forming during the early steps above. However, once you
refold your target protein, you must at some stage allow the reoxidation of
the protein to form disulfide bonds if they can form. Proteins that have
cysteines, but that do not normally contain disulfide bonds have cysteines
that are not positioned in the precise geometric configuration needed to
form a disulfide bond (see Anthony et al., 2002). Therefore, even though
there are often a very large number of possible incorrect disulfide bonds,
they do not form readily. The best strategy is to have a redox buffer
present during protein refolding. Such a buffer (see below) has a mixture
of reducing and oxidizing agents that allow disulfide shuffling to occur (see
Gilbert, 1994). Disulfide shuffling consists of allowing disulfide bonds to
be formed and be reduced repeatedly. If an incorrect bond forms in a
misfolded protein and cannot be reduced, the protein is frozen into an
incorrect conformation and will not reach the native state. If that bond is
reduced, the protein can continue to move between intermediate folding
states until it reaches the correct structure. If the correct bond now forms,
it stabilizes the final native protein. Even if it is occasionally reduced, the
protein is in a stable state and will reform the correct bond upon reoxida-
tion. For this reason, it is often successful merely to have reducing agent in
the solubilized protein but not in the refolding buffer. After refolding, the
protein is allowed to undergo slow air oxidation (exposed to air for a day
or two) and often achieves the correct disulfide bonded structure. For
more on reoxidation see Vallejo and Rinas (2004) and Kersteen and
Raines (2003).
Some typical methods for forming disulfide bonds are
1. Air oxidation: expose to air, without reducing agent for several days.
2. Redox buffer: for example, reduced/oxidized glutathione ¼ 10/1 (3 mM
GSH/0.3 mM GSSG). Various redox pairs are used including reduced
270 Richard R. Burgess

and oxidized cysteine, dithiothreitol, and glutathione. The molar ratio of

reduced to oxidized form is sometimes varied to achieve optimal disul-
fide reoxidation.
3. Protein disulfide isomerase (PDI): this is an enzyme that catalyzes disulfide
shuffling (see review by Kersteen and Raines, 2003). One can also use
small molecule PDI mimics (see Woycechowsky et al., 1999).

5.7. Characterization
One of the most common problems I see when reviewing papers involving
refolding is that one has no idea if the material at the end of the procedure is
correctly refolded, monodisperse, or has full biological activity. It is com-
mon to see an enzyme or biological assay, but often without any standard.
You see phrases like ‘‘since the final product has activity it is shown that it is
correctly folded and fully active.’’ Only a comparison with a known, fully
active standard will allow an estimation of the percent of the final product
that is active. Without this comparison you know there is activity, but do
not know if 100%, 10%, 1%, 0.1%, or 0.01% of the refolded material is
active. Such a determination of specific activity of the final product is
essential in a refolding procedure, if possible.
Another important characterization is to determine the size of the
refolded material. Is it monomeric (monodisperse) or does it contain
large amounts of soluble multimer (heterodisperse)? Crystallographers
routinely use dynamic light scattering (DSL) to determine if purified
protein, refolded or not, is monodisperse. Another method, developed
by Mark Knuth at GNF in La Jolla, CA is to analyze the final product by
analytical size exclusion chromatography (ANSEC). We have found this
to be very useful since very small amounts of protein can be analyzed.
Typically, we run 20–50 mg of protein on a 12-ml Shodex KW-802.5
column at 0.5–1.0 ml/min in a buffer containing 0.25 M NaCl, moni-
toring absorbance at 280 and 215 nm. When the column is calibrated
with suitable MW markers, one can quickly determine if most of the
protein is in a single peak (monodisperse) of the size expected for the
monomer (this is good), or if much of the protein elutes earlier indicating
that it is much larger (heterodisperse) due to formation of multimers (this
is bad). Use of this procedure is an important part of a thorough set of
refolding test trials (see below).
One cannot use circular dichroism (CD) or Western blot analysis to
determine activity or monodispersity. There are plenty of examples of
proteins whose CD spectra are very similar or identical to native standard
material and which react just fine in a Western blot analysis that are not
monodisperse or active, respectively.
Refolding Proteins 271

6. Performing a Protein Refolding Test Screen

6.1. Systematic refolding screens
In the general procedure and discussion earlier, it was assumed that you knew
a suitable refolding solution for your protein. This, however, is the most
essential and the most difficult part in designing an efficient refolding proto-
col. In early refolding papers, usually a single refolding solution was chosen
and it either worked or did not work. The ones that worked were published
and the unsuccessful attempts were often abandoned. As a result, the early
successes were usually proteins that refolded readily under many different
conditions. In recent years, it has become increasingly clear that many
proteins can only be refolded under very specific conditions. The challenge
has become how to screen the very many possible conditions to find one that
is effective in promoting refolding. An important advance was the use of
fractional factorial approaches to systematically determine the effects of many
different variables (Armstrong et al., 1999; Cowieson et al., 2006; Quronfleh
et al., 2007; Trésaugues et al., 2004; Vincentelli et al., 2004; Willis et al., 2005).
Several commercial products have been developed to aid the researcher in
identifying suitable refolding conditions. Much more information on these
refolding screens and related protocols are available at the company web sites.
AthenaES QuickFoldTM (15-solution kit); (http://www.athenaes.com/
QuickFoldProteinRefolding Kit).
EMD/Novagen’s iFOLD1TM, iFOLD2TM, and iFOLD3TM (96-solution
kit); (http://www.novagen.com).
Pierce Biotechnology’s ProMatrixTM (96-solution kit components);
(http://www.fishersci.com).
It is quite likely that this list will grow and that the screens will evolve as
more experience is gained in the use of this type of screen and new refolding
aids are identified. It should be noted that nothing prevents a researcher
who cannot afford to buy these rather expensive kits from designing a set of
test solutions of their own and customizing the refolding screen to fit their
special needs. The key concept here is the systematic, parallel screening of multiple
refolding conditions.

6.2. Variables in refolding

There are many solution variables (such as pH, temperature, salt concentra-
tion, redox environment, and the presence of divalent ions) and additives
that have been reported to improve the efficiency of refolding of solubilized
proteins from IBs. Such variables and additives have been taken into
272 Richard R. Burgess

account in designing the commercial refolding screens mentioned above

and serve as examples for those wishing to design their own refolding
screen. These variables are discussed below. Other reviews on protein
refolding go into greater detail on some of these (Armstrong et al., 1999;
Cowieson et al., 2006; Middelberg, 2002; Quronfleh et al., 2007; Schein,
1991; Singh and Panda, 2005; Trésaugues et al., 2004; Vallejo and Rinas,
2004; Vincentelli et al., 2004; Willis et al., 2005).
pH: Most refolding is done in the range of pH 5–9 and, in our experi-
ence, most proteins refold best at about pH 8–8.5. In general it is a good
idea to use a pH that is at least one pH unit away from the isoelectric point
or pI (the pH at which the protein has zero net charge and is most prone to
precipitation).
Temperature: So far no obvious trend has developed as to any generally
optimal temperature to use during refolding. Most people carry out the
refolding process at near room temperature. This is a low enough to prevent
thermal damage to the protein and high enough to increase the thermal
motion of the molecules that is likely important in melting out transient
misfolded conformations and reaching the native state. One could argue
that higher temperatures will strengthen the hydrophobic interactions that
likely lead to aggregation, but then again it should also strengthen the
hydrophobic interactions needed to bury hydrophobic residues in the
interior of a native protein.
A very interesting paper by Xie and Wetlaufer (1996) reported the
refolding of carbonic anhydrase II at 4 mg/ml in 1 M GuHCl and 50 mM
Tris sulfate, pH 7.5 at different temperatures. It appears that refolding at low
temperatures (4–12  C) for 120 min followed by a ‘‘temperature-leap’’ to
36  C for 30 min gave excellent recovery of activity (>90%). They
argue that hydrophobic interaction is diminished at the low temperature,
minimizing aggregation and allowing slow conversion to an intermediate
that is not active, or aggregation prone. Upon temperature-leap from 4 to
36  C, this intermediate converts to a highly active, native form.
Salt concentration: Some salt is probably desirable to cause ‘‘salting in’’ to
increase solubility of the native protein (see Chapter 20 in this volume).
Often 50–100 mM salt is used. In the case of a 60-fold dilution of 6 M
GuHCl, the final concentration of GuHCl is 0.1 M. Additional salt is often
avoided to keep the salt low enough to allow protein binding to an
appropriate ion-exchange column.
Redox agents: The use of redox buffers to aid in correct disulfide bond
formation is discussed in the section on reoxidation above. In some of the
commercial kits there are solutions with no reducing agent, with reducing
agent and redox buffers with several ratios of reducing to oxidizing agents.
Divalent cations: This variable has not been thoroughly explored, but it is
clear that native proteins often have divalent cations such as Mg2þ or Zn2þ
as part of their structure. Obviously, anything that will stabilize the native
Refolding Proteins 273

structure will bias refolding toward native and away from aggregated
protein.
Arginine and other amino acids: There is an extensive literature on the use
of arginine to promote refolding of solubilized protein and how it functions
to diminish aggregation (Arakawa et al., 2007; Baynes et al., 2005; Das et al.,
2007; Dong et al., 2004; Reddy et al., 2005). It appears that arginine
decreases aggregation by slowing the rate of protein–protein interactions.
Das et al. argue that this is because arginine forms supramolecular assemblies
in solution. One of its drawbacks is that it is usually used and is most
effective at concentrations of 0.5–1.0 M. At these concentrations, it inter-
feres with subsequent chromatography on Ni2þ-chelate affinity columns
and will prevent the binding of most protein to an ion-exchange column
without further dilution or dialysis. The natural osmoprotectant proline has
also been found to be effective in some cases at increasing solubility both
in vitro and in vivo (Ignatova and Gierasch, 2006).
Glycerol and sugars (sucrose, mannitol, sorbitol, and trehalose): We have found
glycerol to be an excellent refolding additive in many cases, usually used in
the 5–30% range. One extreme method by Shimamoto et al. (1998) involves
solubilization of IB protein to 2–10 mg/ml by 6 M GuHCl, addition of
glycerol to 50% (v/v), dialysis into 75% (v/v) glycerol to remove denaturant,
and then rapid dilution to lower glycerol to 5–10%. Several sugars in the
0.5–1.5 M range have been seen to promote successful refolding (Bowden
and Georgiou, 1988).
Other chemical additives: A large variety of papers have reported the use of
materials such as polyethylene glycol (PEG) (Cleland et al., 1992), cyclo-
dextrin (Rozema and Gellman, 1996), and various nondetergent sulfobe-
taines (NDSBs) (Expert-Bezancon et al., 2003) to increase refolding for
certain proteins. The effects of certain ionic liquids (i.e., organic salts with
melting points below 100  C) such as N 0 -alkyl-N-methylimidazolium
chlorides as refolding additives have been reported (Lange et al., 2005).
The ethyl and propyl derivatives, at concentrations of 0.5–1.0 M, showed
effects in increasing activity and refolding yields comparable to L-arginine.
Detergents: In theory, detergents should be helpful in preventing aggre-
gation during refolding. At low concentrations they bind weakly to exposed
hydrophobic or sticky regions and mask them, preventing aggregation.
As their concentration decreases they dissociate and allow reformation of
native structure. This is thought to be why Sarkosyl works as well as it does.
At high concentration it is a denaturant, but at low concentration it acts as
an artificial chaperone and promotes refolding without aggregation.
Chaotropic agents (denaturants at high concentrations): Since aggregation is
due to interactions among partially folded intermediates, it has sometimes
been useful to have a nondenaturing amount of a chaotrope in the refolding
solution to dissociate aggregates, but not the more stable, native structures.
1 M urea and 0.5 M GuHCl have been used.
274 Richard R. Burgess

Target protein-specific additives: The same argument given above for the
potential benefit of certain divalent cations can be used to suggest that the
presence of appropriate substrates, or cofactors, or essential heme groups, for
example, would stabilize native structure and improve refolding yield.

6.3. A practical, inexpensive rational approach

With all the refolding screens, you still have to obtain some sort of a readout
that indicates which condition works best. The best situation is when your
target enzyme is a known enzyme and can be easily assayed. You merely
make your dilutions from solubilized protein into the various refolding
solutions, wait some time for refolding (often several hours), and then
assay a portion of the dilute solution for enzyme activity.
It is even easier if you want to examine the effects of conditions on
refolding of a fluorescent protein such as GFP since the fluorescence is
restored when the GFP refolds correctly and can very conveniently be
measured during and after the refolding with an appropriate fluorescence
plate reader. This makes an excellent exercise for teaching protein refolding
and is used at the Protein Course at Cold Spring Harbor (R. Burgess,
N. Thompson, R. Chumanov, unpublished results).
However, most proteins being studied these days are either not enzymes
or have a very difficult biological assay. How do you know which refolding
condition worked best? The following protocol has been used with great
success (R. Chumanov and R. Burgess, unpublished results). This proce-
dure is quite general and takes advantage of the turbidity of solutions where
protein aggregation has occurred (see Trésaugues et al., 2004; Vincentelli
et al., 2004).
1. Prepare washed IBs solubilized in several different chaotropic agents as
listed above and centrifuged to remove insoluble material. Protein
concentration should be 3–5 mg/ml.
2. Prepare a set of test solutions based on variables and additives that you
think might be effective.
3. Pipette 10 ml of solubilized protein into the appropriate number of wells
of a 96-well plate microtiter dish (Linbro, flat bottomed, polystyrene).
4. Carry out a 20-fold flash dilution by quickly adding 190 ml of the test
refolding solution and mixing the solution up and down several times in
the pipette at room temperature.
5. Wait 15–60 min and then read the absorbance of the plate at 320 nm.
The solution is not absorbing the light, but rather the protein aggregates
scatter light and decrease the amount of transmitted light measured. This
works well because light scattering increases as you go to lower wave-
lengths, proteins do not absorb at 320 nm, and 320 nm is about a low as
you can go in wavelength before absorbance of the plate gets too great.
Refolding Proteins 275

At 320 nm, the absorbance of the empty well due to the plastic is about
0.2 and can be subtracted from the apparent absorbance readings due to
solution turbidity to give a corrected turbidity.
6. The results of the absorbance readings with the plate absorbance
subtracted can be graphed versus the solution number and one can see
which solutions give the lowest turbidity. Corrected turbidity values
range from 0.0 to about 0.4. We often find several conditions that give
low- or zero-corrected turbidity. One often gets somewhat different
results with protein denatured using different solubilizing agents such as
6 M GuHCl, 8 M urea, or 0.3% Sarkosyl.
7. Choose the best 2–3 conditions that are compatible with loading directly
(after filtration) onto an analytical size exclusion column (ANSEC, see
Section 5.7). To be really useful for large-scale refolding, it must also be
compatible with loading onto an appropriate ion-exchange column at a
salt concentration of about 0.1 m NaCl. The solution (about 200 ml) can
be removed from the corresponding well, filtered through a syringe filter
or centrifuged in a microfuge tube and loaded onto a 12-ml ANSEC
column. A refolding condition that gives low turbidity and mostly monomeric
material on the ANSEC column is the condition that is likely to be useful in the
large-scale refolding of the target protein.

7. Other Refolding Procedures

Although the main focus of this chapter has been on refolding of
solubilized IB protein by dilution into a suitable refolding solution, there
are several other significant approaches to protein refolding that need to be
described briefly.
1. On-column refolding. This is a large and active area of research and
application. The basic principle is that denatured proteins when bound
to a column resin or residing within the pores of a gel filtration resin are
tethered or sequestered and if they experience a gradual decrease in the
concentration of denaturant, they will refold, but be less likely to
aggregate. This idea was first proposed by Tom Creighton in the late
1980s and is now widely used in various column modes such as ion
exchange, affinity, metal chelate affinity, hydrophobic interaction, and
gel filtration chromatography. For example, one can load an IB protein
solubilized with 8 M urea onto an ion-exchange column, wash away
some impurities with 8 M urea, wash with a gradually decreasing urea
gradient, and elute with a linear salt gradient. Another approach is to
preload a gel filtration column with a reverse gradient of denaturant. The
top of the column is 8 M urea, the gradient decreases linearly to 0 M urea
276 Richard R. Burgess

in the middle of the column, and the bottom half of the column is 0 M
urea. One then loads the denatured protein and denaturant is applied at a
slow flow rate to the column. The protein will move down the column
faster than the buffer (because it is partially or completely excluded from
the bead) and move gradually into lower and lower concentrations of
denaturant until it has refolded and elutes from the column. The diffi-
culty here is to decide on the steepness and volume of the reverse
gradient. The various protocols and applications are summarized nicely
in several recent reviews and articles ( Jungbauer and Kaar, 2006;
Jungbauer et al., 2004; Oganesyan et al., 2005; Swietnicki, 2006;
Veldkamp et al., 2007). While this on-column refolding method seems
ideal, it is often the case that as the denaturant concentration is decreased,
the protein precipitates on the column and is lost. This is usually due to
the tendency of many researchers to overload their column or to load
from one end that becomes saturated with protein. In these cases, the fact
that the protein is tethered to the resin does not help because the bound
proteins are close enough together on the column that they can still
aggregate during refolding. One general procedure that sometimes helps
is to bind the protein in denaturant to the column in batch (obviously
not applicable to size exclusion chromatography) at well below saturat-
ing amounts, then load the resin into a column, and finally carry out the
subsequent washing and elution steps in column.
2. Use of folding catalysts such as protein and chemical chaperones, PPI, PDI.
Another nice approach is to covalently attach various enzymes and cha-
perones to a gel filtration column resin and then refold in column. The
enzymes that have been immobilized, either individually or perhaps better
in combinations are: PPI (catalyze proline isomerization), PDI (promote
disulfide shuffling), DsbA and DsbC to catalyze disulfide formation, and
the chaperones GroEL/GroES to aid in refolding (Baneyx and Mujacic,
2004; Jungbauer et al., 2004; Paul et al., 2007; Swietnicki, 2006).
3. High-pressure refolding. A number of papers, summarized by Quronfleh
et al. (2007) report the successful solubilization of IBs by exposure to
high hydrostatic pressure (1.5–2.5 kbar). It is claimed that exposure of
IBs to high pressure within a certain range will disaggregate the aggre-
gated protein but not denature it. The process of pressurizing, holding at
high pressure, and slowly decreasing the pressure in a variety of buffers
has been reported to result in high recovery of soluble and active
enzymes. While this method has been commercialized by Barofold
(Boulder, CO) using a specialized pressure cells (PreEMPTM high-
pressure chamber), the equipment is not widely available. In some
cases, while the IB is solubilized, the protein is largely in the form of
soluble multimers. Like the solubilization with denaturants and refolding
discussed for most of this chapter, one must find the right conditions for
optimal recovery of active protein by screening multiple conditions.
Refolding Proteins 277

4. Alkaline pH shift to solubilize and refold. Singh and Panda (2005) argue that
proteins in IBs are largely folded and can more effectively be refolded if
not subjected to high concentrations of GuHCl or urea. They state that
IB protein (recombinant human growth hormone) was solubilized effi-
ciently at pH 12.5 in 2 M urea and a concentration of 2 mg/ml, but
retained significant secondary structure. The solubilized protein was
then diluted into 2 M urea, pH 8 with a 40% recovery of activity.

8. Refolding Database: Refold

A database of protein refolding information is available at http://
refold.med.monash.edu.au.
This database, developed by S. Bottomley and his colleagues at Monash
University, contains over 1000 proteins that have been successfully refolded
and presents protocols and statistics on the frequency of use of various
refolding techniques, disruption methods, fusion proteins, and preparation
prior to refolding (Chow et al., 2006). For example, almost 85% of the
examples utilize some form of dilution or dialysis to remove denaturant,
while about 12% utilize on-column refolding.

9. Strategies to Increase Proportion of

Soluble Protein
The standard method for determining the proportion of soluble and
insoluble material is to prepare a cell lysate and then centrifuge to sediment
insoluble material (unbroken cells, cell debris, and IBs). The total (crude
lysate), soluble (supernatant), and insoluble (pellet) fractions are analyzed by
SDS–PAGE and the intensity of the target protein band in the three
fractions is compared by stain intensity or by Western blot analysis. If the
vast majority of the target protein is in the soluble fraction, then you say that
the protein is soluble. If the majority is in the pellet, then you say the protein
is insoluble. Often target protein is in both soluble and insoluble fractions.
While this is often an accurate estimate of the proportion of the protein that
is soluble, I have seen two cases where the results can be misleading. The
first is when cell breakage is incomplete and the SDS gel pattern of the pellet
looks much like that of the lysate. This suggests that more effective lysis
conditions be used (see Chapter 18 in this volume). Sometimes one sees
significant amounts of target protein in the soluble fraction due to soluble
multimers or to ineffective pelleting of IBs. If sonication is too vigorous,
sometimes the IBs are broken up into smaller particles that may not
be completely sedimented. This can be remedied by longer, harder
278 Richard R. Burgess

centrifugations. Also sedimentation of IBs can be incomplete if the viscosity

of the lysate is high due to the presence of too much large DNA. Treatment
of the lysate with a nuclease such as Benzonase can decrease this viscosity.
Even though I believe it is possible to find effective refolding conditions
for most expressed proteins, many researchers strongly prefer to decrease IB
formation and increase the proportion of soluble product if possible.
In many expression systems so much protein is produced that even if only
20% of the protein is soluble there can be quite significant amounts of
protein available for purification by more standard methods. Some of the
strategies for increasing the amount of soluble product are listed below
(see Schein, 1989; Chapters 11 and 12 in this volume).
1. Induce overproduction at lower temperatures, for example, 20  C. It was observed
in the late 1980s (Schein, 1989) that one could increase the proportion of
soluble material by inducing the target protein overexpression in E. coli
grown below the normal growth temperature of 37  C. Temperatures of
20–25  C are commonly used, although cell growth is quite slow at the
lower temperature. The basic rationale seems to be that at the lower
temperature, transcription and translation are slower, so that proteins
have more time to refold into native structures and the internal concentra-
tion of the partially folded, sticky intermediate is lower, resulting in less
aggregation, and IB formation (see Vera et al., 2006).
2. Add 0.4 M sucrose to the culture medium. This approach has been reported
by Bowden and Georgiou (1988). This phenomenon may be due to
the high osmolyte concentration inducing the osmotic shock response,
which increases the internal level of glutamate, proline, and trehalose,
thus providing conditions more favorable for refolding.
3. Coexpression of molecular chaperones. Since molecular chaperones such
as DnaK/J and GroEL/ES are known to suppress misfolding and aggre-
gation in the cell, one can increase the levels of these proteins
by introducing a separate plasmid that contains inducible genes for
these chaperones.
4. Subject cells to brief heat shock at 42  C, then shift to 20  C, and induce. This is an
elegant approach that avoids the need to coexpress chaperones as above, but
rather takes advantage of the natural heat shock response where DnaK/J and
GroEL/ES levels increase. A reasonable procedure is to grow cells at 35  C
to A600 nm of about 0.5, shift the cells rapidly to 42  C for 15–20 min to
induce the heat shock response, decrease the temperature to 20–25  C, and
then induce expression of target protein for 4–8 h.
5. Coexpression of several subunits of a complex. It has been observed that if two
proteins (say protein X and Y) that form a stable heterodimer are
expressed in E. coli individually they are mostly insoluble, but if they
are coexpressed they refold, form their normal dimer (XY) and are
soluble. Vectors for such coexpression are marketed by EMD/Novagen
as the Duet vectors.
Refolding Proteins 279

6. Fuse to easily refolded protein, for example, a ‘‘solubility tag’’ such as NusA,
Trx, MBP, GST, SUMO, and HaloTag (see Chapter 16 in this volume).
7. Form disulfide bonds in the E. coli cytoplasm. The environment in the E. coli
cytoplasm is reducing, preventing most disulfide bond formation.
Mutations in the glutathione reductase (gor gene) and thioredoxin reduc-
tase (trxB gene) enhance formation of disulfide bonds in the E. coli cyto-
plasm (Bessette et al., 1999). Strains AD494 (DE3) and BL21trxB(DE3) are
trxB deficient. Novagen Origami strains are trxB and gor deficient.

10. Conclusion
Recombinant proteins overexpressed in E. coli often form insoluble
IBs in the cell. The IBs are easy to purify and solubilize, but finding suitable
conditions for efficient refolding of the solubilized protein is sometimes
difficult. The best general strategy seems to be to screen many different
refolding conditions in parallel to find ones that give the highest recovery of
enzyme activity or the lowest turbidity due to aggregation and that show the
least amount of soluble multimers. Many additives may prove useful in
refolding, but the surest way to prevent aggregation and precipitation upon
refolding is to refold at low protein concentration.

REFERENCES
Anfinsen, C. B. (1973). Principles that govern the folding of protein chains. Science
181, 223–230.
Anthony, L. C., Dombkowski, A. A., and Burgess, R. R. (2002). Using disulfide bond
engineering to study conformational changes in the b0 260–309 region of E. coli RNA
polymerase b0 during s70 binding. J. Bacteriol. 184, 2634–2641.
Arakawa, T., Ejima, D., Tsumoto, K., Obeyama, N., Tanaka, Y., Kita, Y., and
Timasheff, S. N. (2007). Suppression of protein interactions by arginine: A proposed
mechanism of the arginine effects. Biophys. Chem. 127, 1–8.
Armstrong, N., De Lencastre, A., and Gouaux, E. (1999). A new protein folding screen:
Application to the ligand binding domains of a glutamate and kainite receptor and to
lysozyme and carbonic anhydrase. Protein Sci. 8, 1475–1483.
Baneyx, F., and Mujacic, M. (2004). Review. Recombinant protein folding and misfolding
in E. coli. Nat. Biotechnol. 22, 1399–1408.
Baynes, B. M., Wang, D. I. C., and Trout, B. L. (2005). Role of arginine in the stabilization
of proteins against aggregation. Biochemistry 44, 4919–4925.
Bessette, P. H., Åslund, F., Beckwith, J., and Georgiou, G. (1999). Efficient folding of
proteins with multiple disulfide bonds in the E. coli cytoplasm. Proc. Natl. Acad. Sci. USA
96, 13703–13708.
Bowden, G. A., and Georgiou, G. (1988). The effect of sugars on b-lactamase aggregation in
E. coli. Biotech. Prog. 4, 97–101.
Bowden, G. A., Paredes, A. M., and Georgiou, G. (1991). Structure and morphology of
protein inclusion bodies in E. coli. Biotechnology (NY) 9, 725–730.
280 Richard R. Burgess

Burgess, R. R. (1996). Purification of overproduced E. coli RNA polymerase sigma factors

by solubilizing inclusion bodies and refolding from Sarkosyl. Meth. Enzymol.
273, 145–149.
Burgess, R. R., and Knuth, M. (1996). Purification of a recombinant protein overproduced
in E. coli. In ‘‘Strategies for Protein Purification and Characterization: A Laboratory
Manual’’, (D. Marshak, J. Kadonaga, R. Burgess, M. Knuth, W. Brennan Jr., and
S.-H. Lin, eds.), pp. 205–274. Cold Spring Harbor Press, New York.
Cabrita, L. D., and Bottomley, S. P. (2004). Protein expression and refolding—A practical
guide to getting the most out of inclusion bodies. Biotechnol. Annu. Rev. 10, 31–54.
Chow, M. K., Amin, A. A., Fulton, K. F., Whisstock, J. C., Buckle, A. M., and
Bottomley, S. P. (2006). REFOLD: An analytical database of protein refolding methods.
Protein Expr. Purif. 46, 166–171.
Clark, E. D. (2001). Protein refolding for industrial processes. Curr. Opin. Biotechnol.
12, 202–207.
Clark, P. L. (2004). Protein folding in the cell: Reshaping the folding funnel. Trends Biochem.
Sci. 29, 527–534.
Cleland, J. L., Builder, S. E., Swartz, J. R., Winkler, M., Chang, J. Y., and Wang, D. I.
(1992). Polyethylene glycol enhanced protein refolding. Biotechnology 10, 1013–1019.
Cowieson, N. P., Wensely, B., Listwan, P., Hume, D. A., Kobe, B., and Martin, J. L.
(2006). An automatable screen for the rapid identification of proteins amenable to
refolding. Proteomics 6, 1750–1757.
Das, U., Hariprasad, G., Ethayathulla, A. S., Manral, P., Das, T. K., Pasha, S., Mann, A.,
Ganguli, M., Verma, A. K., Bhat, R., Chandrayan, S. K., Ahmed, S., Sharma, S.,
Kaur, P., Singh, T. P., and Srinivasan, A. (2007). Inhibition of protein aggregation:
Supramolecular assemblies of arginine hold the key. PLoS ONE 2, e1176.
Dong, X.-Y., Huang, Y., and Sun, Y. (2004). Refolding kinetics of denatured-reduced
lysozyme in the presence of folding aids. J. Biotechnol. 114, 135–142.
Expert-Bezancon, N., Rabilloud, T., Vuillard, L., and Goldberg, M. E. (2003). Physico-
chemical features of non-detergent sulfobetaines active as protein folding helpers.
Biophys. Chem. 100, 469–479.
Gilbert, H. F. (1994). The formation of native disulfide bonds. In ‘‘Mechanisms of Protein
Folding’’, (R. H. Pain, ed.), pp. 104–136. IRL Press, Oxford.
Graslund, S., et al. (2008). Protein production and purification. Nat. Methods 5, 135–146.
Gribskov, M., and Burgess, R. R. (1983). Overexpression and purification of the sigma70
subunit of E. coli RNA polymerase. Gene 26, 109–118.
Ignatova, Z., and Gierasch, L. M. (2006). Inhibition of protein aggregation in vitro and
in vivo by a natural osmoprotectant. Proc. Natl. Acad. Sci. USA 103, 13357–13361.
Jungbauer, A., and Kaar, W. (2006). Review. Current status of technical protein refolding.
J. Biotechnol. 128, 587–596.
Jungbauer, A., Kaar, W., and Schlegl, R. (2004). Folding and refolding of proteins in
chromatographic beds. Curr. Opin. Biotechnol. 15, 487–494.
Karplus, M. (1997). The Levinthal paradox: Yesterday and today. Fold. Des. 2, S69–S75.
Kersteen, E. A., and Raines, R. T. (2003). Minireview. Catalysis of protein folding by
protein disulfide isomerase and small-molecule mimics. Antioxid Redox Signal.
5, 413–424.
Knuth, M. W., and Burgess, R. R. (1987). Purification in the denatured state. In ‘‘Protein
Purification: Micro to Macro’’, (R. R. Burgess, ed.), pp. 297–305. Alan R. Liss,
New York.
Lange, C., Patil, G., and Rudolph, R. (2005). Ionic liquids as refolding additives: N 0 -alkyl
and N 0 -(o-hydroxyalkyl) N-imidazolium chlorides. Protein Sci. 14, 2693–2701.
Levinthal, C. (1968). Are there pathways for protein folding? J. Chim. Phys. 65, 44–45.
Refolding Proteins 281

Makrides, S. C. (1996). Strategies for achieving high-level expression of genes in E. coli.

Microbiol. Rev. 60, 512–538.
Marston, F. A. O., and Hartley, D. L. (1990). Solubilization of protein aggregates. Meth.
Enzymol. 182, 264–276.
Middelberg, A. P. J. (2002). Preparative protein refolding. Trend Biotechnol. 20, 437–443.
Murby, M., Uhlen, M., and Stahl, S. (1996). Review. Upstream strategies to minimize
proteolytic degradation upon recombinant production in E. coli. Protein Expr. Purif.
7, 129–136.
Nall, B. (1994). Proline isomerization as a rate limiting step. In ‘‘Mechanisms of Protein
Folding’’, (R. H. Pain, ed.), pp. 80–103. IRL Press, Oxford.
Nguyen, L., Jensen, D. B., and Burgess, R. R. (1993). Overproduction and purification of
sigma32, the E. coli heat-shock transcription factor. Protein Expr. Purif. 4, 425–433.
Oganesyan, N., Kim, S.-H., and Kim, R. (2005). On-column protein refolding for crystal-
lization. J. Struct. Funct. Genomics 6, 177–182.
Pace, C. N. (1986). Determination and analysis of urea and guanidine hydrochloride
denaturation curves. Meth. Enzymol. 131, 266–280.
Panda, A. K. (2003). Bioprocessing of therapeutic proteins from the inclusion bodies of
E. coli. Adv. Biochem. Eng. Biotechnol. 85, 43–93.
Paul, S., Punam, S., and Chaudhuri, T. K. (2007). Chaperone-assisted refolding of E. coli
maltodextrin glucosidase. FEBS J. 274, 6000–6010.
Puri, N. K., Crivelli, E., Cardamone, M., Fiddes, R., Bertolini, J., Ninham, B., and
Brandon, M. R. (1992). Solubilization of growth hormone and other recombinant proteins
from E. coli inclusion bodies by using a cationic surfactant. Biochem. J. 285, 871–879.
Quronfleh, M. W., Hesterberg, L. K., and Seefeldt, M. B. (2007). Review. Confronting
high-throughput protein refolding using high pressure and solution screens. Protein Expr.
Purif. 55, 209–224.
Reddy, R. C. K., Lilie, H., Rudolph, R., and Lange, C. (2005). L-Arginine increases the
solubility of unfolded species of hen egg white lysozyme. Protein Sci. 14, 929–935.
Rozema, D., and Gellman, S. H. (1996). Artificial chaperone-assisted refolding of carbonic
anhydrase B. J. Biol. Chem. 271, 3478–3487.
Schein, C. H. (1989). Production of soluble recombinant proteins in bacteria. Biotechnology
7, 1141–1147.
Schein, C. H. (1991). Optimizing protein refolding to the native state in bacteria. Curr.
Opin. Biotechnol. 2, 746–750.
Shimamoto, N., Kasciukovich, T., Nagai, H., and Hayward, R. S. (1998). Efficient solubi-
lization of proteins overproduced as inclusion bodies by use of an extreme concentration
of glycerol. J. Tech. Tips Online 23, t01576.
Singh, S. M., and Panda, A. K. (2005). Review. Solubilization and refolding of bacterial
inclusion body protein. J. Biosci. Bioeng. 99, 303–310.
Song, J. (2009). Insight into insoluble proteins with pure water. FEBS Lett. 583, 953–959.
Sorensen, H. P., and Mortensen, K. K. (2005). Review. Advanced genetic strategies for
recombinant protein expression in E. coli. J. Biotechnol. 115, 113–128.
Studier, F. W., Rosenberg, A. H., Dunn, J. J., and Dubendorff, J. W. (1990). Use of T7
RNA polymerase to direct expression of cloned genes. Meth. Enzymol. 185, 60–89.
Swietnicki, W. (2006). Folding aggregated proteins into functionally active forms. Curr.
Opin. Biotechnol. 17, 367–372.
Thatcher, D. R., and Hitchcock, A. (1994). Protein folding in biotechnology.
In ‘‘Mechanisms of Protein Folding’’, (R. H. Pain, ed.), pp. 229–261. IRL Press, Oxford.
Trésaugues, L., Collinet, B., Minard, P., Henkes, G., Aufrère, R., Blondeau, K., Liger, D.,
Zhou, C.-Z., Janin, J., and van Tilbeurgh, H. (2004). Refolding strategies from inclusion
bodies in a structural genomics project. J. Struct. Funct. Genomics 5, 195–204.
282 Richard R. Burgess

Tsumoto, K., Ejima, D., Kumagai, I., and Arakawa, T. (2003). Practical considerations in
refolding proteins from inclusion bodies. Protein Expr. Purif. 28, 1–8.
Vallejo, L. F., and Rinas, U. (2004). Review. Strategies for the recovery of active proteins
through refolding of bacterial inclusion body proteins. Microb. Cell Fact. 3, 11–22.
Veldkamp, C. T., Person, F. C., Hayes, P. L., Mattmiller, J. E., Haugner, J. C. III, de la
Cruz, N., and Volkman, B. F. (2007). On-column refolding of recombinant chemokines
for NMR studies and biological assays. Protein Expr. Purif. 52, 202–209.
Ventura, S., and Villaverde, A. (2006). Protein quality in bacterial inclusion bodies. Trends
Biotechnol. 24, 179–185.
Vera, A., Gozalez-Montalban, N., Aris, A., and Villaverde, A. (2006). The conformation
quality of insoluble recombinant proteins is enhanced at low growth temperatures.
Biotechnol. Bioeng. 96, 1101–1106.
Vincentelli, R., Canaan, S., Campanacci, V., Valencia, S., Maurin, D., Frassinetti, F., Scap-
pucini-Calvo, L., Bourne, Y., Cambillau, C., and Bignon, C. (2004). High-throughput
automated refolding screening of inclusion bodies. Protein Sci. 13, 2782–2792.
Willis, M. S., Hogan, J. K., Prabhakar, P., Liu, X., Tsai, K., Wei, Y., and Fox, T. (2005).
Investigation of protein refolding using a fractional factorial screen: A study of reagent
effects and interactions. Protein Sci. 14, 1818–1826.
Woycechowsky, K. J., Wittrup, K. D., and Raines, R. T. (1999). A small-molecule catalyst
of protein refolding in vitro and in vivo. Chem. Biol. 6, 871–879.
Xie, Y., and Wetlaufer, D. B. (1996). Control of aggregation in protein refolding: The
temperature-leap tactic. Protein Sci. 5, 517–523.
 C H A P T E R E I G H T E E N

Advances in Preparation of Biological

Extracts for Protein Purification
Anthony C. Grabski

Contents
1.Introduction 286
2.Chemical and Enzymatic Cell Disruption 286
3.Mechanical Cell Disruption 290
4.Concluding Remarks 293
5.Procedures, Reagents, and Tips for Cell Disruption 293
5.1. Buffer composition 293
5.2. Cell disruption buffers 296
5.3. Microscale protocols for E. coli cell lysis 296
5.4. Gram-scale mechanical disruption of E. coli 299
References 301

Abstract
There are a variety of reliable methods for cellular disintegration and extraction
of proteins ranging from enzymatic digestion and osmotic shock to ultrasonica-
tion, and pressure disruption. Each method has inherent advantages and
disadvantages. Generally vigorous mechanical treatments reduce extract vis-
cosity but can result in the inactivation of labile proteins by heat or oxidation,
while gentle treatments may not release the target protein from the cells, and
resulting extracts are extremely viscous. Depending on the cell type selected as
the source for target protein expression, cellular extracts contain large amounts
of nucleic acid, ribosomal material, lipids, dispersed cell wall polysaccharide,
carbohydrates, chitin, small molecules, and thousands of unwanted proteins.
Isolation and recovery of a single protein from this complex mixture of macro-
molecules presents considerable challenges. The first and possibly most impor-
tant of these challenges is generation of a cellular extract that can be efficiently
manipulated in downstream processes without inactivation or degradation of
labile protein targets. Cell disruption techniques must rapidly and efficiently
lyse cells to extract proteins with minimal proteolysis or oxidation while reduc-
ing extract viscosity caused by cell debris and genomic DNA contamination.

Department of Research and Development, Semba Biosciences, Inc., Madison, Wisconsin, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63018-4 All rights reserved.

285
286 Anthony C. Grabski

Advanced bioprocessing equipment and reagents have been developed over

the past twenty years to complement established disruption procedures and
accomplish these tasks with even greater success. This chapter will summarize
these advances and describe detailed protocols for some of the most popular
methods for protein extraction.

1. Introduction
Early protein chemists worked primarily with small extracellular
proteins that were stable, plentiful, and easily isolated without cell disrup-
tion. Intracellular proteins expressed at physiological levels might constitute
<0.001% of the total cellular protein and were difficult to extract
and recover without proteolysis or loss of enzymatic activity. Modern
recombinant protein production through genetic engineering of Escherichia
coli can result in expression of the target protein at over 40% of the cellular
total. Recombinant yeasts such as Saccharomyces cerevisiae, Pichia pastoris, and
Kluvyeromyces lactis are capable of expressing complex proteins requiring
posttranslational modifications (Spencer and Spencer, 1997). Recovery of
the intracellular recombinant proteins from these hosts requires effective cell
disruption and extraction techniques. Well established protocols have been
reviewed and detailed procedures described for laboratory and process scale
disruption of various cell types by both mechanical and nonmechanical
methods (Andrews and Asenjo, 1987; Cull and McHenry, 1990; Dignam,
1990; Engler, 1985; Gegenheimer, 1990; Harrison, 1991; Hopkins, 1991;
Jazwinski, 1990; Middelberg, 1995). Radical changes to standard disruption
equipment and procedures had not occurred up to 1995 (Foster, 1995).
However, the evolution of structural and functional proteomics demanded
new reagents and automated methods to streamline the process steps
converting gene sequences to purified proteins (Grabski et al., 2002).
Proteomics and structural genomics efforts require cell extraction meth-
ods that allow screening of multiple host vector combinations for hundreds
of proteins in parallel (Stevens et al., 2001). The best combinations for
correctly folded high-level expression identified in these multiparallel
screens are scaled up in order to produce milligram quantities of purified
protein for functional and structural analysis (Dieckman et al., 2002; Lesley,
2001; Stevens and Wilson, 2001; Zhu et al., 2001). These high-throughput
(HT) protein expression and purification strategies have fostered develop-
ment and optimization of reagents and instrumentation that allow efficient
cell lysis and protein extraction at both micro- and macroscales.
Preparation of Biological Extracts 287

2. Chemical and Enzymatic Cell Disruption

Micro-scale cell disruption is often accomplished through chemical and
enzymatic methods or combinations of the two. Lysis methods such as
sonication and French press are not easily applied to cell pellets harvested
from 5 ml of culture or less, and excess heat and oxidation are common
problems. A significant advancement for simplification of microscale cell lysis
and has been the development of detergent-based reagents such as B-PerÒ
(Chu et al., 1998) and BugBusterÒ (Grabski et al., 1999). These reagents do
not require expensive equipment, are very fast and easy to use, and are the
most practical and effective means to HT cell extract preparation. Highly
active enzymes and enzyme mixtures have also been developed to improve
lysis efficiencies and decrease extract viscosity by completely digesting geno-
mic DNA. Used alone and in combinations these commercially available lysis
reagents and enzymes (Table 18.1) efficiently disrupt cells and extract pro-
teins from bacteria, yeasts, plants, insect cells, and higher eukaryotes. Deter-
gent-based methods for extraction and enrichment of proteins and subcelluar
organelles from eukaryotic cells and tissues are covered extensively in chapter
19 by Michelsen and von Hagen of this volume.
Another advancement in chemical cell disruption tailored for HT
robotic processing of samples, is the development of reagents for extraction
of recombinant proteins from E. coli without cell harvest from the culture
media (Grabski et al., 2001, 2003; Stevens and Kobs, 2004). Traditional
protein purification methods require an initial cell harvest step, which
concentrates cell mass and removes spent media components (Burgess,
1987; Deutscher, 1990; Scopes, 1994). This step, as well as any mechanical
lysis step is difficult to automate and scale down for parallel extraction
and purification of small amounts of proteins simultaneously. Concentrated,
detergent-based reagents (Table 18.1) such as PopCultureTM, FastBreakTM,
and B-PERÒ Direct, combined with high-activity lysozyme and nuclease
overcome these bioprocessing hurdles to automation. The advantages of
these reagents for HT extraction and purification are eliminating the need
for multiple centrifugation steps to separate cells from culture media and
clarify the crude extract, and eliminating the need for mechanical disruption
(Nguyen et al., 2004). These innovations allow direct affinity adsorption of
target proteins from the total culture extract, and the entire cell growth,
extraction, and purification can be done in a single tube or well.
The active ingredients in many of the bacterial lysis reagents are
nonionic or zwitterionic detergents functioning to disrupt cell membrane
and cell wall structures weakening the cells for rupture by osmotic shock,
Table 18.1 Commercial cell disruption reagents and enzymes

Vendor Reagents Cell types Vendor web site

Ò TM
EMD Chemicals- BugBuster , YeastBuster , Bacteria, yeast, insect, www.emdbiosciences.
Novagen CytoBusterTM mammalian com
PopCultureÒ
rLysozymeTM, BenzonaseÒ , LysonaseTM
M-Pek, S-Pek
Epicentre EasyLyseTM, OmniCleaveTM, ReadyLyseTM Bacteria www.epicentre.com
Promega FastBreakTM Bacteria www.promega.com
Roche CompleteTM Lysis (B, M, Y) Bacteria, mammalian, yeast www.roche-applied-
science.com
Semba Recombinant Lysozyme, BenzonaseÒ , Bacteria www.sembabio.com
Biosciences, LiquisonicTM
Inc.
Sigma Aldrich CelLyticTM (B, IB, M, MEM, MT, Bacteria, mammalian, plant, www.sigmaaldrich.
NuClearTM, P, PN, and Y) yeast com
Thermo Fisher- POP-PERSÒ (B-, I-, M-, NE-,P-, and Bacteria, insect, mammalian, www.thermofisher.
Pierce Y-PerÒ ) plant, yeast com
Preparation of Biological Extracts 289

freeze-thaw, or enzymatic attack by phage or chicken egg white lysozyme.

Gram positive bacteria are easily disrupted by lysozyme treatment alone,
with a few exceptions (Cull and McHenry, 1990). Gram negative bacteria
are difficult-to-disrupt with lysozyme alone because their outer lipid bilayer
must be permeabilized first to expose the peptidoglycan cell wall to lyso-
zyme digestion. Tris buffer plus EDTA liberates about 50% of the poly-
anionic lipopolysaccharide in the bilayer (Lieve, 1974), but EDTA
interferes with the most popular downstream purification step for recombi-
nant proteins, immobilized metal affinity chromatography (IMAC). The
detergent lysis reagents are extremely effective for outer membrane per-
meabilization without IMAC interference. However, detergent-based
reagent cell disruption has several disadvantages. Solubilities for some pro-
teins extracted with detergents may be enhanced by detergent–protein
interactions, and protein solubilities will change depending on the
detergent–protein ratio. The solubility of a particular protein extracted by
these reagents may not correlate with the protein’s solubility as extracted by
scale-up mechanical methods. Detergents and detergent impurities can inter-
fere with downstream processing and structural characterization (Swiderek
et al., 1997), and fractionation methods that depend on hydrophobic interac-
tion should be avoided (Marshak et al., 1996). Despite these drawbacks,
detergent-based lysis reagents are widely used in modern proteomics. The
exact concentrations and types of detergents employed in these lysis reagents
are proprietary and the cellular architecture to be disrupted dictates the
chemical properties and concentration of detergent required for the task.
Several references from this chapter (Neugebauer, 1990; Chu and Mallia,
2001; Eshaghi et al., 2005; Kashino, 2003) on applications of detergents for
protein extraction and solubilization provide a wealth of information allowing
home brewed extraction reagents to be formulated with reasonable success.
Yeasts are more difficult-to-disrupt than bacteria. Their thick complex
cell wall can compose up to 25% of the cells’ dry weight. Typical compo-
nents are glucans, cellulose, mannoproteins, and chitin interconnected by
covalent, disulfide, hydrogen, and hydrophobic bonds. Yeast cells should be
harvested in late log or early stationary phase, for more efficient disruption.
Yeasts cultured longer into stationary phase develop thicker cell walls and
multiple budding scars or aggregates in the case of budding yeast (Catley,
1988). Protease inhibitors should be included in lysis buffers or added to
commercial lysis reagents and protease deficient strains used for expression.
Yeast lysis with reagents such as Y-PerÒ and YeastBusterTM (Table 18.1)
and lysis by controlled manipulation of genes involved in cell wall biogene-
sis have provided new tools to complement previously developed mechan-
ical and enzymatic methods (Drott et al., 2002).
Enzymatic treatment of microbial cells for lysis and viscosity reduction
has several distinct advantages over mechanical and chemical methods.
The lytic enzymes are highly specific for targeted cell wall components,
290 Anthony C. Grabski

they are gentle and do not generate shear, high temperature or oxidative
damage, and they are simple to use requiring no specialized equipment for
processing. Enzyme treatments of cells and extracts are also combined with
mechanical disruption methods to increase the selectivity of product release,
to increase the rate and yield of extraction, to minimize product damage,
and to reduce viscosity for downstream processes (Andrews and Asenjo,
1987; Grabski et al., 1999). Improved bioprocessing enzymes and enzyme
mixtures (Table 18.1) include high-specific activity phage lysozymes such
as rLysozymeTM, Ready-LyseTM, and Recombinant Lysozyme, nonspecific
nucleases like BenzonaseÒ and OmniCleaveTM, and the lysozyme-nuclease
cocktails LiquisonicTM, LysonaseTM, and EasyLyseTM. Recombinant Lyso-
zyme, Ready-LyseTM and rLysozymeTM phage lysozymes have specific
activities over 200-fold greater than that of chicken egg white lysozyme.
The nuclease of Serratia marcesens, available as a highly purified recombinant
enzyme BenzonaseÒ , is one of the most promiscuous nucleases known. The
enzyme cleaves all forms of DNA and RNA (Meiss et al., 1995; Nestle and
Roberts, 1969). Compared to bovine pancreatic DNase I, Benzonase
attacks substrates more evenly making it an excellent choice for reducing
extract viscosity caused by nucleic acids. ZymolaseÒ and Lyticase glucanases
are enzymes useful for yeast protoplasting or yeast cell lysis.
Potential problems with enzymatic treatment limit its use especially for
industrial cell disruption. These problems include the added lytic enzyme
and impurities in the enzyme preparation that complicate downstream
purification, degradation of the recovered product during lysis, tempera-
ture, and pH optima for the lytic enzyme may be incompatible with the
target protein, and the expense and limited availability of most lytic enzymes
prohibits their use at an industrial scale (Hopkins, 1991).

3. Mechanical Cell Disruption

Sonication and high-pressure disruption have been effectively applied
for disruption of microorganisms, plants, and animal cells. These methods
have been used for decades and the equipment and mechanisms involved
described in detail (Harrison, 1991; Hopkins, 1991; Middelberg, 1995).
Sonication is based on the shear forces created by high frequency ultrasonic
vibrations generated by resonation (15–25 kHz) of a tuned probe or
horn. The sonic pressure waves created cause the collapse of formed
microbubbles and their implosion generates shock waves with sufficient
energy to disrupt cell walls, and reduce viscosity by shearing nucleic acids.
High-pressure homogenizers and pressure extruders operate by forcing a
pressurized cell suspension through a narrow orifice valve. The valve may
be a simple restricting needle valve as in the French Press or a more complex
Preparation of Biological Extracts 291

design with a combined valve seat and impact ring found in the APV
Manton-Gaulin homogenizer. The mechanism for disruption is a combined
pressure drop and shear at the nozzle in pressure extrusion. Numerous
mechanisms for cell disruption with pressure homogenizers have been
proposed including turbulence, cavitation, viscous shear, and impingement
(Kleinig and Middelberg, 1998; Pandolfe, 1999). Universal acceptance for a
single mechanism has not been reached. However, cavitation and impinge-
ment have been reported as the major forces responsible for disruption
(Shirgaonkar et al., 1998; SPX Corp., 2009).
Cell disruption using glass beads to grind the cells in suspension, also
known as bead milling or bead homogenization, is a frequently used
procedure at both laboratory and production scale. Bead milling can be
accomplished with laboratory equipment as simple as a magnetic stirrer,
vortex mixer, or blender and with commercial equipment specialized for
the process in the form of high speed mills, agitators, and mixers. The cell
disintegration is dependent on cell concentration, bead size and composi-
tion, ratio of beads to suspension, processing duration, and the forces
applied (Ramanan et al., 2008). The method is effective with difficult-
to-disrupt cells like yeasts, spores, and microalgae. It is used for bacteria,
plant and animal cells and is a preferred method for large-scale disruption of
fungi (Hopkins, 1991). The mechanism and models of the bead milling
process have been reviewed (Harrison, 1991; Middelberg, 1995), but basi-
cally the cells are crushed, ground, and torn apart by abrasive contact and
shear forces created between the beads, cells, and reaction chamber itself.
The specialized equipment necessary for mechanical cell disruption
techniques has undergone functional innovation and new instruments
have been developed. The majority of mechanical methods are not easily
applied to cell pellets harvested from 5 ml of culture or less, and excess heat
and oxidation are common problems with mechanical disruption. How-
ever, 96-well sonicator heads and microplate horns are available (Misonix,
Inc. Farmingdale, NY; www.misonix.com) and for cases where traces of
detergent in the purified protein preparation might interfere with biochem-
ical characterization or crystallography, this HT physical method is a viable
option. The SonicMan high-throughput sonication system (MatriCal, Inc.
Spokane, WA; www.matrical.com) is a stand alone or integrated unit for
96-, 384-, and 1536-well plate sonications. This touch screen controlled
microplate sonicator uses disposable gasketed pin lids to prevent well-
to-well cross-contamination and has a plate shuttle allowing direct integra-
tion into robotics platforms. Another new sonicator available from BioSpec
Products is the cordless handheld Sonozap Ultrasonic Homogenizer. The
1/8 in. diameter auto-tuned probe is ideal for small samples of 0.3–5.0 ml.
Pressure Biosciences, Inc. (www.pressurebiosciences.com) has developed
the BarocyclerTM, bench top, and PCT Shredder, hand held, sample
preparation systems. These instruments are capable of rapid, high-pressure
292 Anthony C. Grabski

(up to 35 kpsi) cycling using specialized PULSETM tubes. The 1.2–1.5 ml

tubes contain a ram and pored lysis disk that facilitate the hydrostatic
pressure induced lysis of plant, animal, insect, or microbial cells (Garrett
et al., 2002). The FastPrep (MP Biomedicals, Irvine, CA; www.mpbio.
com), Geno/Grinder (SPEX Certiprep, Inc., Metuchen, NJ; www.
spexcsp.com) MagNA Lyser (Roche Diagnostics, Penzberg, Germany;
www.roche.com), Mikro-Dismembrator (Sartorius Stedim Biotech,
Aubagne, France; www.sartorius-stedim.com), Mini-Bead Beater (BioSpec
Products, Bartlesville, OK; http://www.biospec.com), and Retsch Mixer
Mill (Retsch GmbH, Haan, Germany; www.retsch.com) are all shaking-
type bead mills that process multiple milliliter size samples. These devices
are not truly HT but are capable of high-yield disruption results from
recalcitrant cells in 1–5 min.
The Microfluidics Microfluidizer (Newton, MA; www.
microfluidicscorp.com) differs from other high-pressure homogenizers.
The instrument pressurizes and accelerates split streams of the cell suspen-
sion by gas driven pumping of the liquid through fixed-geometry micro-
channels. The two high-velocity streams directly impinge on each other
creating very high shear forces and pressure drop as the suspension exits
the device. Microfluidizer models range from the M-110P capable of
processing small sample volumes of 25 ml and ideal for laboratory use to
the M-700 biopharmaceutical process models with up to 900 l/h through-
put, complete process control monitoring, CIP, and validatable under 21
CFR to cGMP. Constant Systems Ltd. (Low March Daventry, Northants,
England; www.constantsystems.com) has designed hydraulically operated
disruption instruments that function similar to the original French press by
exerting the disruptive forces of pressure extrusion but under controlled,
contained and repeatable conditions. The Constant Systems instruments
are available in single shot bench (1–20 ml) to continuous process scale
(405–565 ml/min.) models. Useful features include touch screen moni-
toring and control, cooling jacket, and CIP/SIP processes. Avestin,
Inc. (www.avestin.com) offers high-pressure EmulsiFlex homogenizers
capable of cell disruption from 1.0–1000 l/h. The homogenizers generate
disrupting pressure from 500 to 30,000 psi using either gas or electrically
driven piston pumps. They can be equipped with heat exchangers for
cooling cell extracts and are SIP sterilizable for GMP manufacturing. The
BioNeb cell disruption systems of Glas-Col, LLC (Terre Haute, IN;
www.glascol.com) employ nebulizing gas pressures from 10 to 250 psi
to break open cells without heat generation. Laminar flow in the nebuli-
zation channel creates cell disrupting shear forces. The magnitude of
the force depends on the pressure applied, type of gas (argon < nitrogen
< helium) and viscosity of the liquid. The smaller the nebulized droplets,
the higher the viscosity, and the greater the gas pressure applied, the
greater the shear force created (Surzycki et al., 1996).
Preparation of Biological Extracts 293

4. Concluding Remarks
The cell disruption methods, reagents, and instrumentation described
in this chapter are useful and effective if appropriately modified for different
cell types and scales, but the question of which method is superior for each
application has certainly been asked. Studies have been conducted compar-
ing methods of microbial disruption and protein quantitation in resulting
extracts (Benov and Al-Ibraheem, 2002; De Mey et al., 2008; Guerlava
et al., 1998; Ho et al., 2008). The mechanical methods of high-pressure
homogenization and bead milling are favorite methods at large scale because
of their effectiveness and ability to rapidly handle various process volumes at
low cost. Sonication, chemical reagents including detergents, enzymatic
treatments, freeze-thaw, and enzymatic plus chemical or physical method
combinations are also very effective and are used more frequently at labora-
tory and especially microscale. The uniqueness of proteins and differences in
host cell structure force a high degree of empiricism in selection and
optimization of cell disruption and extract preparation techniques.
However, the tools and methods for success in biological extract prepara-
tion are available and have kept pace with the demands of modern structural
and functional proteomics.

5. Procedures, Reagents, and Tips for Cell

Disruption
The following strategies and guidelines serve as starting points to
produce quality cellular extracts suitable for most downstream purification
and analysis processes. Although the focus is on disrupting E. coli at micro-
and macrolaboratory scale, many of the techniques can be applied to
disrupting and isolating intracellular proteins from other sources and at
larger scales. Comprehensive references for protein purification provide
additional information on some of the guidelines for protein extract prepa-
ration outlined here as well as on many important aspects of protein
purification and characterization (Burgess, 1987; Deutscher, 1990; Harris
and Angal, 1989; Hopkins, 1991; Marshak et al., 1996; Scopes, 1994).

5.1. Buffer composition

The components and volume of cell disruption buffers are critical not only
for efficient disruption but will also affect subsequent purification steps and
the target protein’s stability and recovery after it has been released from the
cells. Each protein extracted is an individual and ideally would have an
294 Anthony C. Grabski

extraction and purification buffer tailored specifically for its own biochemi-
cal requirements and intended direction through the purification pipeline.
In most cases a more generic extraction buffer can be used with good results
if a few basic criteria are met. These criteria are pH, ionic strength, additives
to prevent degradation and improve stability, and buffer to cell paste ratio.
A very good reference for maintenance of active enzymes through buffer
composition and other means is Scopes (1994; ‘‘Protein Purification’’).
Technical information on pH and buffers including tables for preparation
of pH 1–13 buffers, buffer properties, and the influence of salt, temperature,
and dilution values can be found by Dawson et al. (1986).
The pH selected for the extraction buffer should be at least one pH unit
above or below the protein’s isoelectric point. This pH difference from the
pI will prevent isoelectric precipitation by maintaining a positive or nega-
tive charge on the protein and also facilitate ion-exchange purification as a
purification step. The buffer ionic strength should be 20–50 mM with a pKa
within 0.5 unit of the desired pH in order to maintain buffering capacity
with minimal conductivity increase. The ionic strength inside the cytoplasm
of a typical cell is 150–200 mM with very high concentrations of charged
biomolecules for ionic protein interaction. The lysis buffer should contain at
least 50–100 mM NaCl. Increased ionic strength of the extraction buffer
will reduce these ionic interactions and precipitating losses from charged
particulates that would adsorb protein and be removed by centrifugation or
filtration steps. Finally, buffer components to prevent degradation and
improve stability should be added as necessary. These include protease
inhibitors (Table 18.2), reducing agents such as dithiothreitol (DTT), tris
(2-carboxyethyl)phosphine (TCEP), or tris(hydroxypropyl)phosphine
(THP), divalent cations, cofactors, and kosmotropes like glycerol, sorbitol,
or trehalose. The water soluble odorless phosphines TCEP and THP are
more stable and effective for the maintenance of reduced protein disulfide
bonds than the more commonly used thiol reductants 2-mercaptoethanol
and DTT (Cline et al., 2004; Getz et al., 1999; Han and Han, 1994).
Nonionic or zwitterionic detergents may be added to increase solubility
of hydrophobic proteins. Reducing agents, protease inhibitors, and deter-
gents may interfere with some purification and detection methods and
assays. The potential interference of these buffer components must be
considered when selecting the chemicals and concentrations employed.
A very versatile buffer (Buffer B) for bacterial cell disruption is 50 mM
Tris/HCl or sodium phosphate pH 7.5–8.0, 50 mM NaCl, 5% glycerol,
0.5 mM EDTA, and 0.5 mM DTT. Buffer formulations recommended for
yeast disruption (Buffer Y) and insect cell disruption (Buffer I) are given
below.
The volume of buffer used for resuspension of cell pellets must be at least
three times the volume of the original pellet for effective disruption and
good recovery of the liquid fraction after removal of insoluble cell
Table 18.2 Protease inhibitors

Inhibitor Proteases inhibited Stock solution Effective concentration

Aprotinin Serine 10 mg/ml PBS 0.6–2.0 mg/ml
Benzamidine Serine 50 mg/ml water 0.5–4.0 mM
E-64 (L- transepoxysuccinyl-leucylamido- Cysteine 5.0 mg/ml water 1.0–10 mM
[4-guanidino]butane)
EDTA Metallo 1.9 g/10 ml water 1.0–10 mM
(ethylenediaminetetraacetic acid)
Leupeptin Serine/cysteine 50 mg/11 ml water 10–100 mM
PMSF (phenylmethylsulfonyl fluoride) Serine 120 mg/ml water 0.1–1.0 mM
AEBSF (4-[2-aminoethyl]
benzenesulfonyl fluoride hydrochloride)
AEBSF is less toxic
than PMSF
Pepstatin A Aspartic 5 mg/7.3 ml (1.0 mM) 1.0–10 mM
methanol or DMSO
Complete tabletsa Broad spectrum Tablets 1X of multiple inhibitors
Inhibitor cocktails I–VIIb Broad spectrum 100X 1X of multiple inhibitors
Inhibitor mixes Broad spectrum 100X 1X of multiple inhibitors
B, FY, G, HP, M, Pc
a
www.roche.com
b
www.emdchemicals.com
c
www.serva.de
296 Anthony C. Grabski

debris and precipitated material by centrifugation or filtration. Since

approximately 50% of the volume of the insoluble residue will be trapped
liquid, at least three volumes are required for greater than 85% liquid
recovery. Proteins are more stable at high concentration, but highly con-
centrated extracts are difficult to process and aggregation can occur.
Although a 3:1 ratio of disruption buffer to cell paste may be used to
produce a more concentrated extract, 5–10 volumes of buffer are preferred
and will yield more soluble protein and a less viscous extract.

5.2. Cell disruption buffers

Buffer B: 50 mM Tris/HCl (pH 8.0), 50 mM NaCl, 5% glycerol, 0.5 mM
EDTA, 0.5 mM DTT or 1.0 mM THP, 0.5 mM benzamidine, 1.0 mM
AEBSF.
Buffer I: 50 mM sodium phosphate (pH 7.5), 150 mM NaCl, 5%
glycerol, 0.03% Brij 35, 0.5 mM EDTA, 0.5 mM DTT or 1.0 mM THP,
1.0% Triton X-100, 0.5 mM benzamidine, 1.0 mM pepstatin A, 1.0 mM
AEBSF. Alternatively, to avoid nuclear lysis in insect cells, substitute 1.0%
NP-40 for Triton X-100.
Buffer Y: 50 mM Tris/HCl (pH 8.0), 150 mM LiCl, 10% ethylene
glycol, 0.5 mM EDTA, 2.0 mM THP, 0.1% Triton X-100, 0.5 mM
benzamidine, 1.0 mM E-64, 1.0 mM pepstatin A, 1.0 mM AEBSF.
Note: Protease inhibitors may be omitted from or supplemented to
(Table 18.2) these buffers depending on the sensitivity of the protein target
to proteolysis, and the degree or class of inhibition required. The ionic
strength of buffers I and Y may need to be reduced or appropriate dilution
of the extract performed if the initial protein fractionation step is
ion-exchange chromatography.

5.3. Microscale protocols for E. coli cell lysis

5.3.1. Detergent-based reagent lysis
Lysis reagent recipe for 10 ml solution: 10 ml B-Per or BugBuster, 20 ml
Lysonase Bioprocessing Reagent. Additives such as EDTA, protease inhi-
bitors, 5–10% glycerol, and reducing agents may be included in the lysis
reagent as required to limit proteolysis and improve target protein solubility
and stability. Downstream purification and analysis requirements must be
considered when selecting lysis buffer components and their concentrations.
1. Culture and express the target proteins using 1.0 ml medium in
2 ml  96 or 5 ml medium in 10 ml 24-deep well plates with air
permeable sealing film or BugStopperTM cap mats.
2. Pellet the cells by centrifugation.
3. Remove and discard the spent medium by aspiration.
Preparation of Biological Extracts 297

4. Resuspend the cell pellets in 100–200 ml of lysis reagent by pipetting.

5. Mix the reaction on a shaking platform 10 min.
6. Remove 10 ml of the mixture for total cell extract analysis.
7. Centrifuge the mixture 5 min and remove a 10-ml sample of the
supernatant for analysis of soluble proteins. The remaining clarified
supernatant is ready for other fractionation or analysis procedures.
8. The amount of target protein present in the insoluble fraction may be
estimated indirectly by differential comparison of the total cell extract in
step 6 and the soluble protein in step 7. The samples are typically
compared by electrophoresis, ELISA, or enzymatic assay. Direct analysis
of the insoluble fraction can be done electrophoretically by aspiration of
the soluble supernatant, solubilization of the pelleted fraction in sodium
dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) sample
buffer, and SDS-PAGE.
Note: Procedures and tips for preparation of protein samples for SDS-PAGE
are available in Grabski and Burgess (2001). The article provides details for
SDS-PAGE sample buffer recipes, protein sample preparation, and gel
loading recommendations, and gives alternatives for analysis of difficult
samples.

5.3.2. Detergent-based reagent whole culture in-media cell lysis

Lysis reagent recipe for 10 ml solution: 10 ml PopCulture, FastBreak, or
B-Per Direct, 200 ml Lysonase Bioprocessing Reagent. The following
protocol can be performed on multiple plates using robotic liquid handling
platforms, on single plates by multichannel pipette, and on individual
cultures. Multiwell filtration plates are available from several manufacturers
for separation of affinity resin from culture extract and processing solutions.
Separation of magnetic particles is accomplished using specialized magnetic
pin plates or platforms.
1. Culture and express the target proteins using 1.0 ml medium in
2 ml  96 or 5.0 ml medium in 10 ml x 24-deep well plates with air
permeable sealing film or BugStopperTM cap mats.
2. Add 0.1 culture volume lysis reagent.
3. Pipette to mix the reaction and continue to mix on a shaking platform
for 10 min.
4. Remove 50 ml of the mixture for total cell extract analysis. The sample
can be analyzed by electrophoresis, ELISA or enzymatic assay. Direct
SDS-PAGE analysis and Coomassie staining requires high-expression
levels and maximum sample loading or sample concentration by precip-
itation to detect dilute protein. Screening soluble expression levels is
easily performed at this step using specialized filter plates (Pall, Millipore,
3 M ) capable of separating protein aggregates and inclusion bodies from
soluble proteins.
298 Anthony C. Grabski

5. Add equilibrated affinity capture resin or magnetic affinity particles

directly to the crude cell lysate. Typically 50–100 ml/ml (50% particle
slurry in capture buffer) of the original culture volume is added to the
lysate depending on resin capacity for the recombinant protein and
target protein expression level.
6. Wash the capture resin with 10–30 resin volumes of wash buffer to
remove contaminants.
7. Elute the target protein with three resin volumes elute buffer.

5.3.3. Freeze-thaw plus enzymatic lysis

Lysis reagent recipe for 10 ml solution: 10 ml lysis buffer at 50–100 mM
concentration (glycine, sodium acetate, Tris–HCl, sodium phosphate, or
HEPES depending on pH desired), 50–300 mM NaCl, and 30 ml Lysonase
Bioprocessing Reagent. Additives such as EDTA, detergents, protease
inhibitors, 5–10% glycerol, and reducing agents may be included in the
lysis buffer as required to limit proteolysis and improve target protein
solubility and stability. Downstream purification and analysis requirements
must be considered when selecting lysis buffer components and their
concentrations.
Freeze-thaw is dependent on the density of the suspension, number of
freeze-thaw cycles, and rate of freezing, but has been shown to be less
effective for protein release from E. coli than bead vortexing, French press,
or sonication (Benov and Al-Ibraheem, 2002). However, when combined
with enzymatic lysis by lysozyme it is an effective and gentle method of
protein release for labile proteins and does not require specialized equip-
ment. The slow freeze-thaw cycle ruptures the cell membrane exposing the
cell wall to enzymatic digestion.
1. Culture and express the target proteins using 1.0 ml medium in
2 ml  96 or 5 ml medium in 10 ml  24-deep well plates with air
permeable sealing film or BugStopperTM cap mats.
2. Pellet the cells by centrifugation.
3. Remove and discard the spent medium by aspiration.
4. Freeze the cell pellets completely at 20  C.
5. Resuspend the cell pellets in 100–200 ml of lysis reagent by pipetting.
6. Mix the reaction on a shaking platform 10 min at room temperature.
7. Freeze the suspension again at 20  C.
8. Thaw at room temperature, mix, and remove 10 ml of the mixture for
total cell extract analysis.
9. Centrifuge the mixture 5 min and remove a 10-ml sample of the
supernatant for analysis of soluble proteins. The remaining clarified
supernatant is ready for other fractionation or analysis procedures.
10. The amount of target protein present in the insoluble fraction may be
estimated indirectly by differential comparison of the total cell extract
Preparation of Biological Extracts 299

in step 8 and the soluble protein in step 9. The samples are typically
compared by electrophoresis, ELISA or enzymatic assay. Direct analysis
of the insoluble fraction can be done electrophoretically by aspiration
of the soluble supernatant, solubilization of the pelleted fraction in
SDS-PAGE sample buffer, and SDS-PAGE.

5.4. Gram-scale mechanical disruption of E. coli

5.4.1. Sonication
The sonication procedure below is convenient for cell pellets of 2–50 g.
1. Pellet the cells by centrifugation in tared centrifuge bottles or tubes
at 9000g for 15 min.
2. Decant and discard the spent medium and drain residual medium by
inverting for a few minutes on paper towels. Record the cell pellet
weights.
3. Freeze the cell pellets completely at 20  C and process the cells within
1 week for best results. Longer term storage of cell paste should be at
80  C to minimize proteolytic degradation. Fresh unfrozen cells may
be used, but enzymatic pretreatment with lysozyme will be less effective
because the cell membrane is intact.
4. Completely resuspend the cell pellets in 7–10 ml/g 4  C lysis buffer
using a hand held homogenizer at low speed. Avoid vigorous mixing
and foaming to prevent oxidation.
5. Optional: add chicken egg (0.2 mg/ml) or recombinant lysozyme
(60 KU or 0.2 ml/ml) and nuclease such as DNase (10 mg/ml) or
BenzonaseÒ (1.0 ml/ml).
6. Sonicate the suspension in a glass beaker on ice using 60–90 s bursts.
Avoid excessive heating and aeration by adjustment of instrument set-
tings, probe immersion depth, and duration of bursts. Caution: Do not
allow the sonicator tip to contact the glass beaker during operation or the
glass may break. For 100 ml cell suspension, four bursts at a setting of
eight and 70% duty cycle using a Branson Model S-450 sonifier (Branson
Sonic Power Co., Danbury, CT; www.bransonultrasonics.com) and
1/2-in. diameter horn is usually sufficient. Cool the extract with stirring
on ice for several minutes between each burst. If desired, save a 50-ml
sample of the lysate at this point (prior to centrifugation) for analysis of
total cellular protein.
7. Centrifuge the lysate at 15,000g for 15 min at 4  C. Carefully transfer
the supernatant to another container without disturbing the pellet. The
soluble fraction may be further clarified by filtration with a low protein
binding filter depending on the tolerance for particulates in the next
fractionation step. If using syringe filters, it is convenient to attach a
0.8-mM filter on top and a 0.45-mM filter on the bottom in tandem.
300 Anthony C. Grabski

The larger pore size filter in-line first prevents rapid fouling of the
smaller pore size filter. This final clarified supernatant represents the
soluble cell protein fraction and is ready for chromatography or other
downstream fractionation or analysis procedures.

5.4.2. High-pressure homogenization

The batch, high-pressure homogenization procedure below is convenient
for cell pellets of 50–500 g with an APV Gaulin homogenizer (APV Fluid
Handling, Delevan, WI; www.apv.com), Microfluidizer, or Constant Sys-
tems hydraulic disrupter. Cell disruption of greater than 500 g of cells should
be performed in multiple batches or using continuous disruption equipment
described above.
1. Follow steps 3 and 4 in the sonication procedure above to resuspend
the fresh or frozen cell paste, but resuspend the cell paste using a Waring
or similar blender (Waring Products, Torrington, CT; www.
waringproducts.com) at slow speed. Do not blend excessively and
keep the extract cold at all times. The resuspension volume should be
less than 10 ml/g depending on the amount of buffer removed to rinse
the disruption instrument after processing. Remove and set aside
50–100 ml of the lysis buffer from the calculated resuspension volume
prior to blending. This buffer will be used for rinsing the instrument after
sample processing. The rinse solution contains residual extract from the
processing equipment’s dead volume.
2. If the Gaulin press is used, two process passes at 10,000 psi gives excellent
disruption of E. coli. Additional process passes or continuous recycling
may be necessary for difficult-to-disrupt cells such as yeast. Follow the
manufacturer’s operating instructions if the Microfludizer or hydraulic
disrupter is used. Significant heating of the sample will occur during
processing especially with smaller volumes. Chill the lysate to 5–10  C
between passes by immersion of the collection vessel in a dry ice ethanol
bath. Minimize freezing of the lysate to the sides and bottom of the
collection vessel by continuous stirring and periodic removal of the
vessel from the cooling bath.
3. Centrifuge the lysate at 15,000g for 30 min at 4  C. Carefully decant
and filter the supernatant using Miracloth (EMD Biosciences-
Calbiochem, Gibbstown, NJ; www.emdbiosciences.com) into another
container without disturbing the pellet. The soluble fraction may be
further clarified by filtration with a smaller pore size low protein binding
filter depending on the tolerance for particulates in the next fractionation
step. This final clarified supernatant represents the soluble cell protein
fraction and is ready for chromatography or other downstream fraction-
ation or analysis procedures.
Preparation of Biological Extracts 301

REFERENCES
Andrews, B. A., and Asenjo, J. A. (1987). Enzymatic lysis and disruption of microbial cells.
TIBTECH 5, 273–277.
Benov, L., and Al-Ibraheem, J. (2002). Disrupting Escherichia coli: A comparison of methods.
J. Biochem. Mol. Biol. 35, 428–431.
Burgess, R. R. (1987). In ‘‘Protein Purification: Micro to Macro’’, (R. R. Burgess, ed.),
Alan R. Liss, New York.
Catley, B. J. (1988). Isolation and analysis of cell walls. In ‘‘Yeast: A Practical Approach’’,
(I. Campbell and J. H. Duffus, eds.), pp. 163–183. IRL Press, Washington.
Chu, R., and Mallia, K. (2001). Method for recovery of proteins prepared by recombinant
DNA procedures. U. S. Patent 6,174,704.
Chu, R., Mallia, K., Brennan, T., and Klenk, D. (1998). Recombinant protein extraction
from E. coli using B-PERÒ Bacterial Protein Extraction Reagent. Previews 2, 12–13.
Cline, D. J., Redding, S. E., Brohawn, S. G., Psathas, J. N., Schneider, J. P., and Thorpe, C.
(2004). New water-soluble phosphines as reductants of peptide and protein disulfide
bonds: Reactivity and membrane permeability. Biochemistry 43, 15195–15203.
Cull, M., and McHenry, C. S. (1990). Preparation of extracts from prokaryotes. In ‘‘Guide
to Protein Purification’’, (M. P. Deutscher, ed.), Meth. Enzymol. 182, pp. 147–153.
Dawson, R. M. C., Elliott, D. C., Elliott, W. H., and Jones, K. M. (1986). pH, buffers,
and physiological media. In ‘‘Data for Biochemical Research’’, pp. 417–448. Oxford
University Press, Inc., New York.
De Mey, M., Lequeux, G. J., Maertens, J., De Muynck, C. I., Soetaert, W. K., and
Vandamme, E. J. (2008). Comparison of protein quantification and extraction methods
suitable for E. coli cultures. Biologicals 36, 198–202.
Guide to protein purification. Deutscher, M. P. (ed.) (1990). Meth. Enzymol.
182, pp. 147–203.
Dieckman, L., Gu, M., Stols, L., Donnelly, M. I., and Collart, F. R. (2002). High-
throughput methods for gene cloning and expression. Protein Expr. Purif. 25, 1–7.
Dignam, J. D. (1990). Preparation of extracts from higher eukaryotes. In ‘‘Guide to Protein
Purification’’, (M. P. Deutscher, ed.), Meth. Enzymol. 182, pp. 194–203.
1
Drott, D., Bahairi, S., and Grabski, A. (2002). Yeast Buster protein extraction reagent for
fast, efficient extraction of proteins from yeast. inNovations 15, 14–16.
Engler, C. R. (1985). Disruption of microbial cells. In ‘‘Comprehensive Biotechnology’’,
(C. L. Cooney and A. E. Humphrey, eds.), Vol. 2, pp. 305–324. Pergamon Press,
Toronto.
Eshaghi, S., Hedren, M., Abdel Nasser, M. I., Hammarberg, T., Thornell, A., and
Norlund, P. (2005). An efficient strategy for high-throughput expression screening of
recombinant integral membrane proteins. Protein Sci. 14, 676–683.
Foster, D. (1995). Optimizing recombinant protein recovery through improvements in cell-
disruption technologies. Curr. Opin. Biotechnol. 6, 523–526.
Garrett, P. E., Tao, F., Lawrence, N., Ji, J., Schumacher, R. T., and Manak, M. M. (2002).
Tired of the same old grind in the new genomics and proteomics era? TARGETS
5, 156–162.
Gegenheimer, P. (1990). Preparation of extracts from plants. In ‘‘Guide to Protein Purifica-
tion’’, (M. P. Deutscher, ed.), Meth. Enzymol. 182, pp. 174–193.
Getz, E. B., Xiao, M., Chakrabarty, T., Cooke, R., and Selvin, P. R. (1999). A comparison
between the sulfhydryl reductants tris(2-carboxyethyl) phosphine and dithiothreitol for
use in protein biochemistry. Anal. Biochem. 273, 73–80.

1
References available online: www.emdbiosciences.com.
302 Anthony C. Grabski

1
Grabski, A., and Burgess, R. R. (2001). Preparation of samples for SDS-polyacrylamide gel
electrophoresis: Procedures and tips. inNovations 13, 10–12.
1
Grabski, A., McCormick, M., and Mierendorf, R. (1999). BugBusterTM and BenzonaseÒ :
The clear solutions to simple efficient extraction of E. coli proteins. inNovations 10, 17–19.
1
Grabski, A., Drott, D., Handley, M., Mehler, M., and Novy, R. (2001). Extraction and
purification of proteins from E. coli without harvesting cells. inNovations 13, 1–4.
1
Grabski, A., Mehler, M., Drott, D., and Van Dinther, J. (2002). Automated purification of
recombinant proteins in a 96-well format with RoboPopTM Kits and robotic sample
processing. inNovations 14, 2–5.
1
Grabski, A., Mehler, M., and Luedke, R. (2003). Automated methods for solubility
screening and recombinant protein purification. Amer. Biotechnol. Lab. 35, 34–38.
Guerlava, P., Izac, V., and Tholozan, J.-L. (1998). Comparison of different methods of cell
lysis and protein measurements in Clostridium perfringens: Application to the cell volume
determination. Curr. Microbiol. 36, 131–135.
Han, J. C., and Han, G. Y. (1994). A procedure for quantitative determination of tris
(2-carboxyethyl)phosphine, an odorless reducing agent more stable and effective than
dithiothreitol. Anal. Biochem. 220, 5–10.
Harris, E. L. V., and Angal, S. (eds.) (1989). In ‘‘Protein Purification Methods, a Practical
Approach’’, IRL Press, Oxford.
Harrison, S. T. L. (1991). Bacterial cell disruption: A key unit operation in the recovery of
intracellular products. Biotechnol. Adv. 9, 217–240.
Ho, C. W., Tan, W. S., Yap, W. B., Ling, T. C., and Tey, B. T. (2008). Comparative
evaluation of different cell disruption methods for the release of recombinant hepatitis B
core antigen from E. coli. Biotechnol. Bioproc. Eng. 13, 577–583.
Hopkins, T. R. (1991). Physical and chemical cell disruption for the recovery of intracellular
proteins. In ‘‘Purification and Analysis of Recombinant Proteins’’, (R. Seetharam and
S. K. Sharma, eds.), pp. 57–83. Marcel Dekker, Inc., New York.
Jazwinski, S. M. (1990). Preparation of extracts from yeast. In ‘‘Guide to Protein Purifica-
tion’’, (M. P. Deutscher, ed.), Meth. Enzymol. 182, pp. 154–174.
Kashino, Y. (2003). Separation methods in the analysis of protein membrane complexes.
J. Chromatgr. B 797, 191–216.
Kleinig, A. R., and Middelberg, A. P. J. (1998). On the mechanism of microbial cell
disruption in high-pressure homogenization. Chem. Eng. Sci. 53, 891–898.
Lesley, S. A. (2001). High-throughput proteomics: Protein expression and purification in
the postgenomic world. Protein Expr. Purif. 22, 159–164.
Lieve, L. (1974). The barrier function of the gram-negative envelope. Ann. N.Y. Acad. Sci.
235, 109–129.
Marshak, D. R., Kadonaga, J. T., Burgess, R. R., Knuth, M. W., Brennan, W. A., and
Lin, S.-H. (1996). Solubilization and purification of the rat liver insulin receptor.
In ‘‘Strategies for Protein Purification and Characterization: A Laboratory Manual’’,
pp. 275–336. Cold Spring Harbor Laboratory Press, New York.
Meiss, G., Friedhoff, P., Hahn, M., Gimadutdinow, O., and Pingoud, A. (1995). Sequence
preferences in cleavage of dsDNA and ssDNA by extracellular Serratia marcescens endo-
nuclease. Biochemistry 34, 11979–11988.
Middelberg, A. P. J. (1995). Process-scale disruption of microorganisms. Biotechnol. Adv.
13, 491–551.
Nestle, M., and Roberts, W. K. (1969). An extracellular nuclease from Serratia marcescens.
J. Biol. Chem. 244, 5219–5225.
Neugebauer, J. M. (1990). Detergents: An overview. In ‘‘Guide to Protein Purification’’,
(M. P. Deutscher, ed.), Meth. Enzymol. 182, pp. 239–253.
Nguyen, H., Martinez, B., Oganesyan, N., and Kim, R. (2004). An automated small-scale
protein expression and purification screening provides beneficial information for protein
production. J. Struct. Funct. Genom. 5, 23–27.
Preparation of Biological Extracts 303

Pandolfe, W. D. (1999). Homogenizers. In ‘‘Wiley encyclopedia of food science and

technology’’, (F. J. Francis, ed.), pp. 1289–1294. John Wiley and Sons, Inc., New York.
Ramanan, R. N., Ling, T. C., and Ariff, A. B. (2008). The performance of a glass bead
shaking technique for the disruption of E. coli cells. Biotechnol. Bioproc. Eng. 13, 613–623.
Scopes, R. K. (1994). Protein Purification: Principles and Practice. 3rd edn. Springer Verlag,
New York.
Shirgaonkar, I. Z., Lothe, R. R., and Pandit, A. B. (1998). Comments on the mechanism of
microbial cell disruption in high-pressure and high-speed devices. Biotechnol. Prog.
14, 657–660.
Spencer, J. F. T., and Spencer, D. M. (1997). Yeasts and the life of man: Part II: Genetics and
molecular biology of industrial yeasts and processes. In ‘‘Yeasts in Natural and Artificial
Habitats’’, ( J. F. T. Spencer and D. M. Spencer, eds.), pp. 243–263. Springer-Verlag,
Berlin.
SPX Corp. (2009). In ‘‘Cell Disruption by Homogenization’’, pp. 1–18; Pub. #3006-01-06-
2008-US.
Stevens, J., and Kobs, G. (2004). Taking the spin out of cell lysis. Promega Notes. 86, 23–24.
Stevens, R. C., and Wilson, I. A. (2001). Industrializing structural biology. Science
293, 519–520.
Stevens, R. C., Yokoyama, S., and Wilson, I. A. (2001). Global efforts in structural
genomics. Science 294, 89–92.
Surzycki, S., Togasaki, R. K., and Kityama, M. (1996). Process and apparatus for fragment-
ing biomaterials. U. S. Patent 5(506), 100.
Swiderek, K. M., Alpert, A. J., Heckendorf, A., Nugent, K., and Patterson, S. D. (1997).
Structural analysis of proteins and peptides in the presence of detergents: Tricks of the
trade. ABRF News Article Methodol. 8, 17–25.
Zhu, H., Bilgin, M., Bangham, R., Hall, D., Casamayor, A., Bertone, P., Lan, N.,
Jansen, R., Bidlingmaier, S., Houfek, T., Mitchell, T., Miller, P., et al. (2001). Global
analysis of protein activities using proteome chips. Science 293, 2101–2105.
 C H A P T E R N I N E T E E N

Isolation of Subcellular Organelles

and Structures
Uwe Michelsen and Jörg von Hagen

Contents
1. Introduction 306
2. Extraction and Prefractionation of Subproteomes 308
2.1. Density velocity centrifugation 308
2.2. Density gradient centrifugation 311
2.3. Mitochondria enrichment 311
2.4. Endomembrane enrichment (LOPIT) 314
2.5. Differential detergent fractionation 316
2.6. Lipid raft enrichment (detergent-resistant fractionation) 320
2.7. Nuclei and histone enrichment 322
2.8. Purification of core histones 324
2.9. Nuclear extract enrichment 325
2.10. Immunoaffinity reagents for organelle enrichment 326
References 327

Abstract
One of the major challenges in functional proteomics is the separation of
complex protein mixtures to allow detection of low abundance proteins and
provide for reliable quantitative and qualitative analysis of proteins impacted by
environmental parameters. Prerequisites for the success of such analyses are
standardized and reproducible operating procedures for sample preparation
prior to protein separation. Due to the complexity of total proteomes, especially
of eukaryotic proteomes, and the divergence of protein properties, it is often
beneficial to prepare standardized partial proteomes of a given organism to
maximize the coverage of the proteome and to increase the chance to visualize
low abundance proteins and make them accessible for subsequent analysis.
In this chapter we will describe with detailed recipes procedures for the
enrichment and isolation of the currently most investigated organelles and
subcellular compartments in mammalian cells using classical centrifugation
techniques to more sophisticated immunoaffinity-based procedures.

Merck KGaA, Darmstadt, Germany

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63019-6 All rights reserved.

305
306 Uwe Michelsen and Jörg von Hagen

Abbreviations

2D Two-dimensional
DDF Differential detergent fractionation
DRF Detergent-resistant fractions
DRM Detergent-resistant membranes
EMSA electro mobility shift assay
g Gravity unit
GPCR G-Protein coupled receptor
HM Homogenization medium
IEF Isoelectric focussing
iTRAQ Isotope tags for relative and absolute quantification
LC Liquid chromatography
LOPIT Localization of organelle proteins by isotope tagging
M Molar
b-MCD b-Methylcyclodextrin
MS Mass spectrometry
MudPIT Multidimensional protein identification technology
PBS Phosphate buffered saline
PMSF Phenylmethylsulfonyl fluoride
RAM Restricted access material
RP Reversed phase
SCX Strong cation exchange
SDS Sodium dodecyl sulfate

1. Introduction
In proteomics research, one essential step among enrichment
techniques is subcellular fractionation, which is of special importance for
analysis of intracellular organelles and multiprotein complexes. To reduce
sample complexity, subcellular fractionation is a flexible and adjustable
approach and is most efficiently combined with high-resolution 2D gel/
mass spectrometry analysis as well as with gel-independent techniques.
Isolating distinct subcellular compartments by fractionation has been
among the standard strategies established in biochemistry-oriented labora-
tories for decades. Determination of marker enzyme activities was to assay
the efficiency of the subcellular fractionation and a major goal was the
identification of individual new proteins. At that time, the power of protein
identification techniques was very limited, the characterization of the
Isolation of Subcellular Organelles and Structures 307

protein subsets specific to subcellular compartments was time-consuming,

of limited sensitivity, or even impossible.
By increasing their degree of complexity, organisms acquire a broader
repertoire of options to meet environmental challenges. Increased
complexity is realized in two ways: by cell differentiation and by
compartmentalization within cells. Subsets of cells of a given organism
serve different purpose and own distinct properties, for example, neurons,
germ cells, or epithelial cells. In addition certain cell functions are compart-
mentalized such as storage of genetic material, degradation of proteins, or
the generation of the cell fuel.
The identification of proteins at the subcellular level is therefore pivotal
for understanding of cellular function. There are proteins that are associated
with subcellular structures only in certain physiological states, but localized
elsewhere in the cell in other states. Protein translocation between different
compartments, cycling of proteins between the cell surface and intracellular
pools or shuttling between nucleoplasm and cytoplasm is very common in
mammalian cells. To be able to monitor the subcellular distribution of gene
products, the classic proteome analysis approach must be refocused on a
subcellular level. Such an approach fits very well to the way in which cells
are organized.
Finally one should note that many current proteome analysis projects are
aimed at the comparative analysis of tissue samples. Tissue samples are more
complex than samples from cultured cells as any tissue contains many
different cell types and contains structural material like connective tissue
that may not be the target of the analysis. This fact implies an even higher
challenge to sample preparation procedures to avoid artifacts. An overview
of the proteomics workflow for all types of samples is presented in Fig. 19.1
highlighting the steps for the enrichment of organelles and subproteomes.
Based on the scheme shown in Fig. 19.1, the following chapter is
subdivided into three sections summarizing the most important and promi-
nent sample preparation techniques used for enrichment and fractionation.
It is important to mention at this point, that based on the various
downstream separation techniques no general workflow is established. All
the diverse analysis techniques require a specific sample preparation to assure
the compatibility of the remaining sample with the selected technique.
More detailed information how to specifically prepare samples and various
sample types is found in ‘‘Proteomics Sample Preparation’’ (von Hagen,
2008). The described procedures are enrichment-based techniques and do
not claim to exclusively purify organelle proteomes. Moreover, the outlined
procedures are a starting point and need to be optimized depending on
species, tissue, and cell type.
In contrast, techniques based on chromatographic principles like
multidimensional protein identification technology (MudPit) (Washburn
et al., 2001), PF2D (Billecke et al., 2006), or FFE (Weber et al., 2004) are
308 Uwe Michelsen and Jörg von Hagen

Body fluids Tissues/cells Plants Yeast/fungi Microorganisms

Sample preparation
Remove high Extraction and Organelle Specific Specific
abundance proteins pre-fractionation isolation enrichment labeling

Separation
1D/2D Liquid Thin layer Free flow
gel electrophoresis chromatography chromatography electrophoresis

Characterization/functional validation
Mass Protein visualization
Crystallography NMR
spectrometry and quantitation

Enzyme Protein-protein Protein

assays interaction arrays

Figure 19.1 Schematic representation of the proteomic workflow.

commonly used to reduce proteome complexicity and do rely on preen-

riched organelles. These separation techniques are briefly described and
protocols can be found elsewhere.

2. Extraction and Prefractionation

of Subproteomes
2.1. Density velocity centrifugation
2.1.1. Introduction
Centrifugation is a common sample preparation technique that accom-
plishes separation based on the density and size differences in a mixture of
components. In the absence of Brownian motion and thermal mixing, a
centrifuge is often not required. If samples are diverse with regard to density
or shape and given enough time, sample components eventually will settle
at the bottom of a tube at 1g. The process can be sped up using a centrifuge.
The density velocity centrifugation takes advantage of the difference in the
terminal velocities of different particles as determined by Stoke’s law:

2R2 ðrS  rÞa

Vt ¼
ð9mÞ
where Vt is the terminal velocity of the biomolecule, m is the viscosity of the
medium, R is the radius of the particle, rs is the density of the particle, a is
the centrifugal acceleration, and r is the density of the medium.
Isolation of Subcellular Organelles and Structures 309

The terminal velocity is a function of the particle radius and density.

Thus, on the average, bigger and heavier particles migrate through the
medium faster and settle at the bottom of a centrifuge tube in a shorter
time. Because a typical mixture of cell homogenate contains organelles of
varying sizes and densities, as well as shapes, they can be separated according
to the sedimentation speed. The decrease in the sedimentation time in a
centrifuge over gravitational settling is mainly due to the considerable
increase in the variable a in Stoke’s equation (Table 19.1).
Each fraction obtained through differential centrifugation contains quite
a few different types of organelles which have similar sedimentation velo-
cities. Because sedimentation velocity is a combination of both the size and
the density, the fraction can be further separated based on density alone
irrespective of the sizes. This can be accomplished by a process known as
density gradient centrifugation (Fig. 19.2).
Procedure:
1. Centrifuge the cellular homogenate for 2 min at 100g to remove unbro-
ken cells.
2. Centrifuge the supernatant at 600g for 10 min to sediment nuclei.
3. Centrifuge the supernatant at 15,000g for 5 min to deposit mitochon-
dria, chloroplasts, lysosomes, and peroxisomes.
4. Ultracentrifuge at 100,000g for 60 min for sedimentation of the plasma
membrane, fragments of the endoplasmic reticulum, and large
polyribosomes.
5. The cytosol remains unpelleted after centrifugation at 300,000g for 2 h.

Table 19.1 Density and dimensions of cellular organelles

Organelle/compartment Diameter (mm) Density (g/cm3)

Nuclei 3–12 1.4
Mitochondria 0.5–2 1.15
Golgi apparatus – 1.08
Lysosomes 0.5–0.8 1.2
Peroxisomes 0.5–0.8 1.25
Plasma membranes 0.05–3 1.15
smooth endoplasmic reticulum 0.05–3 1.16
Nucleic acid 0.03 1.7
Ribosomes 0.03 1.6
soluble proteins 0.001–0.002 1.3
310 Uwe Michelsen and Jörg von Hagen

Sample
Mechanically disrupted/precleared

600 x g/10 min

Su
pe
rn
at
15,000 x g/5 min an
Nuclei t

Mitochondria, lysosomes
peroxisomes and 100,000 x g/60 min
chloroplasts

Plasma membrane, microsomal

Pe 300,000 x g/120 min
lle fraction, endoplasmic reticulum
t large polyribosomes
Cytosolic fraction
Ribosomal subunits
small polyribosomes

Figure 19.2 Cell fractionation by differential velocity centrifugation of a precleared

mechanical disrupted cell homogenate.

2.1.2. Further reduction in proteome complexity

2.1.2.1. Free-flow electrophoresis Free-flow electrophoresis (FFE) is a
highly versatile technique for the separation of a wide variety of charged or
chargeable analytes such as low-molecular-weight organic compounds,
peptides, proteins, protein complexes, membranes, organelles, and whole
cells. A detailed protocol is outlined in Current Protocols in Proteins Sciences
(Weber et al., 2004). This technology is used mainly for the purification of
prepurified organelle extracts as obtained by the procedure described above
and thus represents solely an additional separation technology.

2.1.2.2. Multidimensional protein identification technology MudPIT is

based on a gel-free two-dimensional high-performance liquid chromatog-
raphy (HPLC) approach initially developed by John Yates III lab
(Washburn et al., 2001). A complex protein mixture is denatured and
digested with proteases and separated in a capillary packed with a strong
cation exchange (SCX) resin back-to-back with a reversed phase (RP)
material. The column is interfaced with a quaternary HPLC system coupled
directly to a tandem mass spectrometer.
Moreover other orthogonal chromatographic resins are described to
achieve efficient reduction in proteome complexity like restricted access
materials (RAM) (Willemsen et al., 2004) or the PF2 D system (Billecke
et al., 2006) using IEF columns. This technology can be used to reduce
organelle proteome complexity obtained from prepurified organelles from
density velocity centrifugation or density gradient centrifugation described
in Section 2.2.
Isolation of Subcellular Organelles and Structures 311

2.2. Density gradient centrifugation

2.2.1. Introduction
A density gradient is simply established by placing sugar crystals at the bottom
of a test tube (Fisher and Cline, 1963). The sugar dissolves in the solution and
diffuses slowly toward the top which is a time-consuming process and not
very reproducible. A more practical approach is to place layer after layer of
sucrose solutions of decreasing concentrations in a test tube, resulting in the
heaviest layer at the bottom and the lightest layer on top. Finally, the cell
fraction is placed on top of the sucrose layers. If the density of a given particle
is higher than that of the surrounding solution it will sink until the density of
the surrounding solution is exactly the same as that of the particle. Using a
centrifuge will accelerate this process. Unlike the differential centrifugation
technique described above, centrifugation time does not matter, as long as the
whole system is allowed to come to quasiequilibrium.
This chapter will give a more detailed view on the classical density
gradient technique with regard to the isolation of mitochondria which are
currently in the focus of apoptosis research. For the enrichment of other
organelles please refer to the previous chapter in Methods in Enzymology
(Storrie and Madden, 1990). This section gives detailed protocols for the
isolation of lysosomes, postnuclear supernatants, mitochondria, nuclei as
well as assays to determine marker enzyme activity and organelle intactness
(Fig. 19.3).

2.3. Mitochondria enrichment

Proteomic analyses of particular mitochondrial proteins and their complexes
have provided important information about mitochondrial function
(Edelman et al., 1966). Percoll gradients are generally used in protocols for
the purification of mitochondrial fractions after tissue homogenization and
differential centrifugation of a cell homogenate.
One drawback is the removal of Percoll which is necessary before
further analyses. To circumvent difficulties in Percoll removal from the
isolated mitochondrial fraction, some workers used sucrose in the gradient
buffer as described in detail below. One should note that this again may
cause problems because disaccharides are able to enter the mitochondrial
matrix through membrane porin channels and thus might interfere with the
physiological interpretation of results obtained.
The described sucrose gradient separation procedure is a protein
subfractionation method optimized for mitochondria. It can be used to
reduce sample complexity and resolves a sample into at least 10 fractions.
The sucrose gradient is designed for an initial sample volume of up to
0.5 mL at 5 mg/mL protein. The sample should be solubilized in a nonionic
detergent. It has been determined that at protein concentration of 5 mg/mL
312 Uwe Michelsen and Jörg von Hagen

Fixed pipet tip

Sample 0.5 ml
15% 1.0 ml
20% 1.0 ml
25% 1.0 ml
27.5% 0.75 ml
30% 0.5 ml
35% 0.5 ml

Figure 19.3 Sucrose gradient setup: 15–35% sucrose density gradient. The gradient is
prepared by layering less dense sucrose solutions upon one another; therefore, the first
solution applied is the 35% sucrose solution. Firstly, a Beckman polyallomer tube is held
upright in a tube stand. Next a tip is placed held steady by a clamp stand and the end of
the tip is allowed to make contact with the inside wall of the tube as shown above.

mitochondria are completely solubilized by 20 mM n-dodecyl-b-D-

maltopyranoside (1% w/v lauryl maltoside). By this treatment membranes
are disrupted while the previously membrane embedded multisubunit com-
plexes remain intact, a step necessary for the procedure described below:
Material and equipment:
 Sucrose solutions 15, 20, 25, 27.5, 30, and 35% (w/v).
 Protease inhibitor cocktail.
 Phosphate buffered saline solution (PBS): 1.4 mM KH2PO4, 8 mM
Na2HPO4, 140 mM NaCl, 2.7 mM KCl, pH 7.3.
 200 mM n-dodecyl-b-D-maltopyranoside (10% w/v lauryl maltoside).
 Swing-out compatible ultracentrifuge and rotor.
 Polyallomer centrifuge tubes.

Homogenize and lyse cells as described above. Preferentially mechanical

disruption methods (Goldberg, 2008) should be used to avoid addition of
detergents which might result in leakage of the mitochondria and other
cellular compartments as well.
General procedure:
 Wash cells twice in PBS.
 Resuspend cells in PBS (1 mL/109 cells) including phosphatase and
protease inhibitors. Shear the cells with >20 strokes using a Potter–
Elvehjem or Dounce homogenizer (Check by viability stain).
Isolation of Subcellular Organelles and Structures 313

 Alternatively use five and more freeze thaw cycles, five or more soni-
cations (5 pulses of 5 s), French press, high pressure homogenization
(2 cycles of 500–1000 bar) or a bead mill according to the protocol
provided by the various suppliers. For the following application preclear
the lysate by a centrifugation step of 5 min at 500g at 4  C.
Procedure:
1. A mitochondrial membrane suspension at max. 5 mg/mL protein in PBS
is adjusted with lauryl maltoside to a final concentration of 1%.
2. Mix well and incubate on ice for 30 min.
3. Centrifuge at 70,000g for 30 min.
4. The supernatant is collected and the pellet discarded.
5. Add a protease inhibitor cocktail and keep the sample on ice until
centrifugation is performed.
6. Once the sucrose gradient is poured, discrete layers of sucrose should be
visible. Having applied the sample to the top of the gradient the tube
should be handled carefully.
7. The polyallomer tubes should be centrifuged in a swinging bucket type
rotor at 130,000g for 16 h at 4  C with the lowest acceleration and
deceleration profile.
8. Carefully remove 500 mL fractions from top to the bottom and analyze
the fractions. The fractions can now be stored at  80  C.

2.3.1. Other density gradient formulations

For larger amounts of proteins, multiple gradients can be prepared or
scaling-up to the gradients is also possible. For example, 40 mg of mito-
chondria can be resolved by a sucrose gradient of 40 mL total volume.
1. Combine as follows 55 (5 mL), 50 (5 mL), 35 (10 mL), 30 (10 mL), 25%
(2 mL) sucrose solution.
2. Add on top 8 mL mitochondria supernatant (solubilized in 1% lauryl
maltoside centrifuged 30 min at 72,000g).
3. Centrifuge at 115,000g for 17 h at 4  C with the lowest acceleration and
deceleration profile.
Besides the classical sucrose density medium, different media from
various suppliers are also used routinely to isolate subcellular structures
and organelles. The most prominent materials to mention are Ficoll
(Kurokawa et al., 1965), Percoll (Pertoft et al., 1977), Nycodenz
(Rickwood et al., 1982), and Optiprep (van Veldhoven et al., 1996). For
detailed protocols, please refer to the corresponding manuals provided by
the different suppliers.
314 Uwe Michelsen and Jörg von Hagen

2.4. Endomembrane enrichment (LOPIT)

2.4.1. Introduction
Identifying the localization of a novel protein is an important step toward
assigning function and will also improve our understanding of cell
compartments. To allow confident protein localization organelle prepara-
tions must be free of contaminants. For some organelles, such as mitochon-
dria, it is relatively easy to produce a largely homogeneous preparation of
the organelle. However, Golgi apparatus and endoplasmic reticulum of the
endomembrane system are difficult to separate from one another. Further-
more, proteins associated with this class of organelle traffic through the
endomembrane system to their final destination or cycle between the
compartments. So it is a tough job to discriminate between genuine
organelle residents and contaminants.
Several techniques have emerged which enable assignment of proteins to
organelles. Here we discuss the sample preparation necessary for the use
with one of these methods—localization of organelle proteins by isotope
tagging (LOPIT).
The LOPIT workflow involves partial separation of organelles using
equilibrium density gradient centrifugation. The relative abundance of
proteins amongst the fractions of the gradient is compared with the profiles
of known organelle markers. The relative abundance is measured by a stable
isotope tagging procedure and tandem mass spectrometry (MS/MS)
whereby primary amine groups present on tryptic peptides are derivatized
with iTRAQ reagents (Ross et al., 2004). The LOPIT technology is
applicable to the study of protein localization in many sample types
including cultured cells, tissues, and whole organisms.
The procedure starts with the collection of crude membrane extracts and
partial separation of organelles through a dense medium. The degree of
organelle separation is monitored using one-dimensional SDS–PAGE and
western blotting employing specific antibodies. Alternatively, enzyme assays
measuring the activity characteristic for a particular organelle can be
utilized. This has been described in detail by Storrie and Madden (1990)
in Methods in Enzymology (Table 19.2).
The gradient fractions enriched in specific organelles are selected for
quantitative labeling by tagging tryptic peptides with the iTRAQ reagents.
The labeled mix of peptides is separated by SCX chromatography, and
further separated by reverse phase chromatography coupled to tandem mass
spectrometer (LC-MS/MS). Finally peptides are identified and quantified
using fragmentation data generated by MS/MS and analyzed with statistical
approaches. The protocol described here is based on the work of Sherrier
et al., (1999) and can be optimized for membranes from various sources
(Ford et al., 1994; Graham et al., 1994). A typical preparation would involve
the homogenization of 60 g of tissue in an equal volume of homogenization
Isolation of Subcellular Organelles and Structures 315

Table 19.2 Enzyme and substrate markers for different cellular organelles

Organelle Enzyme or substance

Nuclei DNA, RNA, Parp
Mitochondria Cytochrome c oxidase, succinate dehydrogenase
Plasma membrane Alkaline phosphodiesterase I, GPCRs
Lysosome b-Galactosidase, b-hexosaminidase,
acid-phosphatase
Endosome Peroxidase
Peroxisome Catalase
Golgi apparatus a-Mannosidase II, galactosyl transferase
Smooth endoplasmic Glucose-6-phosphatase
reticulum
Rough endoplasmic RNA
reticulum
Cytosol Lactate dehydrogenase
Cytoskeleton Vimentin, keratin

medium using a Polytron homogenizer for two 7 s pulses. The amount of

medium used and the time of homogenization are crucial to minimize
organelle disruption during this stage of the preparation. The amount of
starting material should yield at a minimum 100 mg of membrane protein in
each gradient fraction which will be labeled with iTRAQ tags. The first part
of the protocol describes membrane fractionation using equilibrium density
gradient centrifugation. Utilizing a 16% iodixanol gradient enables the
separation of the Golgi apparatus, endoplasmic reticulum, vacuoles, plasma
membrane, and mitochondria/plastids
After harvesting the fractions in step 8 it is important to check the degree
of membrane separation by performing western blot analysis of gradient
fractions with antibodies specific for different organelle marker proteins.
Selected fractions can then be used in the following wash steps. The high
pH of the carbonate buffer disrupts electrostatic interactions between
peripheral proteins and the membrane. In addition, closed vesicles are
transformed into membrane sheets and soluble proteins are released. The
generated pellets can be used directly for the iTRAQ labeling reaction
which is described in Sadowski et al. (2006) and Shadforth et al. (2005).
Material and equipment:
 Homogenization buffer: 250 mM sucrose. 10 mM HEPES-NaOH (pH 7.4),
1 mM EDTA (pH 8.0), 1 mM DTT.
 Optiprep stock solution: 60% solution of iodixanol in water.
316 Uwe Michelsen and Jörg von Hagen

 System buffer: 60 mM HEPES-NaOH (pH 7.4), 6 mM EDTA (pH 8.0),

6 mM DTT.
 Iodixanol solution: Prepared by diluting five parts of Optiprep stock solu-
tion with one part of system buffer.
Note: All solution must be prepared freshly before use and kept at 4  C.
Procedure:
1. Preclear the homogenate from intact cells, nuclei, and cell wall frag-
ments by centrifuging for 5 min at 2200g at 4  C.
2. Decant supernatants into clean centrifuge tubes and recentrifuge as
above.
3. Aliquot the supernatants into a total of six centrifuge tubes and underlay
each supernatant with 6 mL of 18% iodixanol cushion solution.
4. Centrifuge at 100,000g at 4  C for 2 h using an ultracentrifuge.
5. Collect the crude membranes from the interphases (typically 6–8 mL)
6. Dilute membranes with iodixanol gradient solution and adjust to 16%
iodixanol using iodixanol solution.
7. Decant the adjusted solution into two polycarbonate tubes and centri-
fuge at 350,000g at 4  C for 3 h in an ultracentrifuge with the slowest
deceleration rate selected.
8. Harvest twenty 0.5 mL fractions from the top of each gradient into
polycarbonate tubes.
9. Add 800 mL of 162.5 mM Na2CO3 to 0.5 mL of each selected density
fraction to give a final concentration of 100 mM of Na2CO3.
10. Incubate for 30 min on ice and centrifuge for 25 min at 100,000g at
4  C using an ultracentrifuge and discard the supernatants.
11. Wash the pellets with 1 mL of deionized water at 4  C, centrifuge for
10 min like above and discard the supernatants.

2.5. Differential detergent fractionation

2.5.1. Introduction
To solubilize proteins, especially proteins with extended hydrophobic
domains the uses of amphipathic molecules like detergents are essential.
Typical detergents contain both polar and hydrophobic groups. They
contain a polar head at the end of a long hydrophobic chain or tail. In
contrast to purely polar or nonpolar molecules, detergent exhibit unique
properties in water. Their polar groups form hydrogen bonds with water
molecules, while the hydrocarbon chains aggregate due to hydrophobic
interactions. These properties allow detergents to be soluble in water. In
aqueous solutions, they form organized spherical structures called micelles.
Because of their amphipathic nature, detergents are able to solubilize
hydrophobic compounds in water. The following procedure describes the
Isolation of Subcellular Organelles and Structures 317

Digitonin H
O

H O
HO H
H H OH
Xyl-Glc-Gal
Glc-Gal H

In
cr
Triton-X 100 ea
O H sin
O g
n ex
tra
cti
on
eff
ica
cy
Tween 20
HO O O OH
O O
w x
O OH
O y
O
O
O O
z
O

OH
Deoxycholate O
OH

H
HO

Figure 19.4 Commonly used detergents with increased extraction efficacy from
Digitonin to sodium Deoxycholate.

use of detergents with increased extraction efficacy properties which result

in the enrichment of four distinct subproteom fractions (Ramsby et al.,
1994) (Fig. 19.4).
A typical sequential extraction method based on detergents enables
simple fractionation of proteins in their native state according to their
subcellular localization, yielding four subproteomes enriched in
(Fig. 19.4A) cytosolic; (B) membrane and membrane organelle-localized;
(C) soluble and DNA-associated nuclear, and (D) cytoskeletal proteins.
Due to the complexity of total proteomes and the divergence of protein
properties, it is beneficial to prepare standardized partial proteomes of a
given organism to make them accessible for subsequent analysis. Such
analyses often require that the proteins of interest remain in a nondenatured
state and their subcellular localization can be determined. Changes in the
topology of proteins, that is, spatial rearrangements to another subcellular
compartment, are important cellular events and crucial for biological pro-
cesses such as signal transduction or apoptosis. This cannot be achieved by
selective purification of cellular organelles. Furthermore, the homogeniza-
tion techniques employed, require usually relatively large amounts of start-
ing material, and are generally more efficient with tissues than with tissue
culture cells (Abdolzade-Bavil et al., 2004).
The protocol described enables simple fractionation of the total prote-
ome into four distinct subcellular fractions. Various related protocols in kit
318 Uwe Michelsen and Jörg von Hagen

Sample
Adherent or suspension cells

Digitonin extraction

Pe
lle
Cytosolic fraction Triton-X 100 extraction

t
Membrane fraction
(organelles) Tween 20/deoxycholate extraction

Nuclear fraction
Su

SDS extraction
pe
rn
at
an
t

Cytoskelatal fraction

Figure 19.5 DDF procedure: Nonultracentrifugation-based organelle enrichment

technique.

format have recently become commercially available by various suppliers

(e.g., ProteoExtract kits from Calbiochem, San Diego, USA) (Fig. 19.5).
Reagents and equipment:
 100 mM EDTA: dissolve 3.36 g EDTA in 100 mL water. Store at room
temperature.
 100 mM phenylmethylsulfonyl fluoride (PMSF): dissolve 174 mg PMSF
in 10 mL isopropanol.
 Piperazine-N,N 0 -bis(2-ethanesulfonic acid) (PIPES) stock buffer (10):
Filter through 0.45 mm sterile filter. Store dark at 4  C.
 Cytosolic protein extraction buffer (0.015% digitonin, pH 6.8 at 4  C):
dissolve by heating 18.75 mg digitonin (Calbiochem) in 4 mL 10
stock buffer in a small flask with a stir bar. Add 1 mL PMSF. Combine
with remaining reagents: 6 mL 10 stock buffer and 5 mL EDTA. Add
water to 100 mL.
 Membrane protein extraction buffer (0.5% Triton X-100, pH 7.4 at 4  C):
combine 10 mL 10 buffer, 0.1 mL PMSF, 3 mL EDTA, and 5 mL
freshly prepared 10% Triton X-100. Add water to 100 mL.
Isolation of Subcellular Organelles and Structures 319

 Nuclear protein extraction buffer (1% Tween/0.5% DOC, pH 7.4 at 4  C):

separately dissolve 0.5 g DOC in 2.5 mL 10 buffer and 1 mL Tween-40
in 2.5 mL 10 buffer (heat to dissolve if necessary). Combine and add
5 mL 10 buffer and 1 mL PMSF. Add water to 100 mL.
 Cytoskeleton solubilization buffer (5% sodium dodecyl sulfate, 10 mM
sodium phosphate, pH 7.4): for nonreducing buffer, solubilize 0.5 g
SDS in 5 mL 20 mM sodium phosphate buffer, pH 7.4. Add water to
10 mL. For denaturing buffer, add 1 mL b-mercaptoethanol. Adjust
water appropriately.
Procedure:
1. Add ice-cold cytosolic protein extraction buffer to washed cell pellets
(5 volumes/g wet weight, gently resuspend by swirling) or monolayers
(1 mL/T25 flask).
2. Incubate cells on ice with gentle agitation (platform mixer) until
95–100% of cells are permeabilized (5–10 min) as assessed by trypan
blue exclusion.
3. For suspension-cultured cells, centrifuge the extraction mixture at 500g
and remove supernatant.
4. For cell monolayers, tilt the culture flask and remove extract (cytosolic
protein-enriched fraction) with a pipet and store at  70  C.
5. Carefully resuspend insoluble pellet in ice-cold membrane protein extrac-
tion buffer in a volume equivalent to that used for cytoplasma protein
extraction (5 volumes relative to starting wet weight) to obtain a
homogeneous suspension.
6. For monolayer cultures, add 1 mL membrane protein extraction buffer per
T25 flask equivalent (approx. 5  106 cells).
7. Incubate on ice with gentle agitation for 30 min.
8. Remove the extract (membrane and organellar protein-enriched fraction) by
centrifuging the suspensions for 10 min at 5000g or decanting monolayers.
9. Aliquot, and store at  70  C.
10. Resuspend the insoluble pellets from suspension cultures in nuclear
protein extraction buffer at one-half the volume used for the membrane
protein extraction.
11. Remove the extract (nuclear protein-enriched fraction) by pelleting the
detergent-resistant residue 10 min at 6000g at 4  C.
12. Extract cell monolayers with 0.5–1 mL nuclear protein extraction buffer per
T25 flask equivalent.
13. Aliquot, and store at  70  C.
14. The detergent-resistant pellet is washed in ice-cold PBS (pH 7.4,
1.2 mM PMSF) by resuspension (Teflon/glass homogenizer) and cen-
trifugation 20 min at 4  C and 12,000g to mechanically shear DNA or
use 1000 units Benzonase (EMD/Novagen) to enzymatically degrade
nucleic acids.
320 Uwe Michelsen and Jörg von Hagen

15. Pellets from suspension cultures are washed once with  20  C of 90%
acetone, lyophilized, and weights determined in tared centrifuge tubes.
Samples are stored at  70  C.
16. Monolayers are rinsed in situ with PBS, and the detergent-resistant
residue (Cytoskeleton) is suspended directly into 0.5–1 mL nondenaturing
cytoskeleton solubilization buffer without b-mercaptoethanol. Store at
 70  C.

2.6. Lipid raft enrichment (detergent-resistant fractionation)

2.6.1. Introduction
A lipid raft is a cholesterol-enriched microdomain in cell membranes. Many
functions have been attributed to rafts, from cholesterol transport, endocy-
tosis, and signal transduction. The analysis of membrane signaling com-
plexes in the natural environment requires a tailor made procedure
depending on the complex of interest. Based on the various sample types
no general protocol can be outlined. The protocol given is a good starting
point. For further optimization, Table 19.3 gives a brief summary of mild
detergents commonly used to isolate membrane proteins and lipid rafts.
One drawback of the use of mild detergents is that classes of pharmaco-
logical relevant proteins which are anchored in the membrane by various
hydrophobic transmembrane domains are resistant to efficient extraction.
Besides the usage of sodium carbonate and OptiPrep also a magnetic-bead
immunoisolation approach might be used to isolate subpopulations of rafts
enriched for different markers such as caveolin-1, flotillin (Shah and Sehgal,
2007). Recently, another nondetergent-based extraction procedure claims
to efficiently extract membrane proteins with several passing transmem-
brane domains (7TMs or GPCRs; Nakamura et al., 2008) and this opens an
in-depth analysis of GPCR transmembrane signaling complexes in a two-
step procedure using a mild detergent to enrich lipid rafts first and after
enrichment extract the entire components of such complexes using the
novel chemistry approach as mentioned above or by diluting the small
molecule based chemistry and use it for both steps to enrich and extract
lipid rafts and the membrane signaling complexes.
Material and equipment:
 TNE-buffer (25 mM Tris–HCl pH 7.5, 50 mM NaCl, 5 mM EDTA)
containing 1%.
 Triton X-100.
 Protease and phosphatase inhibitors (Calbiochem).
 Nycodenz.
 Swing-out compatible ultracentrifuge and rotor.
 Polyallomer centrifuge tubes.
Table 19.3 Commonly used detergents for the solubilization of lipid rafts, membrane protein complexes and transmembrane proteins

cmc cmc Percentage for

Name Abbr. Typea MW (mM)b (% w/v)c solubilization (%)
n-Dodecyl-b-D-maltopyranoside DDM N 510.6 0.17 0.0087 1
n-Undecyl-b-D-maltopyranoside UDM N 496.6 0.59 0.029 1
n-Decyl-b-D-maltopyranoside DM N 482.6 1.8 0.087 1
Cymal 7 Cymal 7 N 522.5 0.19 0.0099 1
Cymal 6 Cymal 6 N 508.5 0.56 0.028 2
n-Octyl-b-D-glucopyranoside nOG N 292.4 18 0.53 2
N,N-Dimethyldodecylamine N-oxide LDAO Z 229.4 1–2 0.023 1
3[(3-Cholamidopropyl) CHAPS Z 614.9 8 0.49 1
dimethylammonio]propane
sulfonic acid
Triton X-100 TX-100 N 647 0.23 0.015 1 (v/v)
a
The type of detergents: N ¼ Nonionic, Z ¼ Zwitterionic.
b
The cmc is the critical micelle concentration and depends on temperature and solution conditions. For solubilizing and purification of membrane proteins one has to
work always above the cmc.
c
Percentage (w/v) of detergent as used in the detergent screen.
322 Uwe Michelsen and Jörg von Hagen

Procedure:
1. Plate 2.0–3.0  107 culture cells (adherent or suspension) in culture
flasks.
2. Wash the cells twice with prewarmed PBS.
3. Where required treat the cells either with b-MCD (10 mM) for 45 min
and/or with cholesterol for 1 h in the serum free medium at 37  C.
4. After treatment or 72 h of infection stimulate the cells where required
5. Lyse the cells in cold 750 mL of TNE-buffer containing 1% Triton
X-100 supplemented with protease and phosphatase inhibitors for
30 min on ice.
6. Mix the lysate with 1500 mL of 70% Nycodenz, dissolved in TNE-
buffer.
7. Load this mix in the bottom of 4 mL of polyallomer ultracentrifuge
tube.
8. Overlay this bottom loaded mix with 25, 21.5, 18, 15, and 8% Nyco-
denz, prepared in TNE-buffer.
9. Spin the tube at 200,000g for 4 h at 4  C using a swing-out rotor.
10. Fractions each of 350 mL from top to the bottom of the tube were
analyzed by western blotting.
11. Probe the blot to confirm caveolin-1 positive (fractions 3–6) or
detergent-resistant membrane fractions.

2.7. Nuclei and histone enrichment

2.7.1. Introduction
Understanding mechanisms of regulation of transcription by signal trans-
duction events and other processes that occur on DNA in vitro analysis of
chromatin is often important. The basic unit of chromatin is the nucleo-
some core, containing two copies each of the core histones H2A, H2B, H3,
and H4 to form a histone octamer that wraps 145 bp of DNA in a left-
handed superhelix. In vivo, chromatin is associated with linker histones (such
as H1), which facilitate the ordered packing of nucleosomes. The particle
with the linker histone is properly termed the nucleosome (or chromato-
some), while the linker histone-free particle is the nucleosome core. The
procedure described below is divided in two parts: preparation of washed
nuclei and isolation of histones.

2.7.2. Preparation of a washed nuclear pellet

The isolation of nucleosomes and free histones requires clean nuclei as
starting material. In this protocol, nuclei are released from cultured cells
by homogenization and nonionic detergent NP-40. Several washes with a
buffer containing detergent are then used to remove membranes and yield a
pellet containing relatively clean nuclei. Subsequent extraction with sodium
Isolation of Subcellular Organelles and Structures 323

chloride removes most loosely bound proteins from chromatin. However,

the NP-40 washes result in extracts less active than nuclear extracts prepared
by other means. If both nuclear extract and chromatin are desired, the
nuclear extract should be prepared as described below and the nuclear pellet
from this procedure can be used to prepare chromatin (Schnitzler, 2001).
Materials and equipment:
 2 M NaCl.
 Nuclear Pellet Buffer2 (NPB)/0.6 M KCl/10% (v/v) glycerol.
 Dounce homogenizer with type B pestle.
 Nuclear Pellet Buffer (NPB).
 20 mM HEPES, pH 7.5.
 3 mM MgCl2.
 0.2 mM EGTA.
 Store up to several weeks at 4  C. Immediately before use add:
 3 mM 2-mercaptoethanol.
 0.5 mM PMSF.
 1 mM pepstatin A.
 1 mM leupeptin.
 Leupeptin.
 Prepare a 1 mM aqueous stock solution and store up to 1 year at
 20  C.
 Lysis buffer (LB).
 20 mM HEPES, pH 7.5.
 0.25 M sucrose.
 3 mM MgCl2.
 0.5% (v/v) Nonidet P-40 (NP-40).
 Store up to several weeks at 4  C. Immediately before use add:
 3 mM 2-mercaptoethanol.
 0.5 mM PMSF.
 1 mM pepstatin A.
 1 mM leupeptin.

Note: Unless otherwise indicated, keep all solutions and materials on ice
or at 4  C.
Procedure:
1. 3  109 cells (HeLa) were washed twice in 1 L PBS.
2. Resuspend pellet in 40 mL lysis buffer.
3. Lyse the cells with  15 strokes of a type B pestle Dounce homoge-
nizer. Monitor lysis by light microscopy.
4. Centrifuge 15 min at 3000g, 4  C.
5. Repeat resuspension and pelleting twice with lysis buffer and once with
NPB.
324 Uwe Michelsen and Jörg von Hagen

6. Resuspend nuclei in 2 pellet vol NPB and measure the total volume of
the suspension.
7. While gently stirring, add dropwise 1 total vol of NPB2.
8. Continue gentle stirring for 10 min at 4  C.
9. Pellet nuclei 30 min at 17,500g, 4  C.
10. Nuclear pellets can be frozen in liquid nitrogen or dry ice and kept for
more than a year at  80  C.

2.8. Purification of core histones

2.8.1. Introduction
The four core histones can be readily purified by binding chromatin frag-
ments to hydroxylapatite resin. Hydroxylapatite binds DNA very strongly
(Simon and Felsenfeld, 1979). Histone H1 is eluted with 0.6 M NaCl and
core histones are eluted from the resin-bound DNA at 2.5 M NaCl
(Zagariya et al., 1993).
Materials and equipment:
 Nuclear pellet (see above).
 HP buffer (see below) with and without 2.5 M NaCl .
 BioGel HTP powder (Bio-Rad), with adsorption capacity 0.6 mg DNA
per g dry powder.
 2  15-cm column and accessories.
 Centriprep-10 concentrators (Amicon; optional).
 Additional reagents and equipment to determine protein concentration
and SDS–PAGE.
 Hydroxylapatite (HP) buffer.
50 mM sodium phosphate, pH 6.8, 0.6 M NaCl.
Store up to several weeks at 4  C.
Immediately before use add: 1 mM 2-mercaptoethanol, 0.5 mM PMSF.
Procedure:
1. Resuspend  2 mL nuclear pellet containing  6 mg DNA in 25 mL HP
buffer.
2. Stir gently for 10 min at 4  C. To avoid proteolysis add 1 mM pepstatin A
and 1mM leupeptin to HP buffer.
3. Add 10 g dry BioGel HTP powder (hydroxyl apatite).
4. Add HP buffer to prepare 1:1 slurry. Pour this into a 2  15-cm column
and collect the eluent.
5. Wash resin with 300 mL HP buffer (30 mL/h). The eluent from step 3
and the first column volume of wash contain H1.
6. Elute the core histones with HP buffer containing 2.5 M NaCl, collect-
ing 8-mL fractions.
7. Analyze the purity of the core histones by SDS–PAGE.
Isolation of Subcellular Organelles and Structures 325

8. If necessary, concentrate the core histones up to 10 mg/mL using

Centriprep-10 concentrators.
9. Divide into aliquots, freeze on dry ice, and store up to 4 years
at  80  C.

2.9. Nuclear extract enrichment

2.9.1. Introduction
The interaction of proteins with DNA is central to the control of many
cellular processes including DNA replication, recombination and repair,
transcription, and viral assembly. One technique to studying gene regula-
tion and determining protein–DNA interactions is the electrophoretic
mobility shift assay (EMSA) also referred as gel retardation analysis.
Material and equipment:
 Nuclear Buffer 1 (NB1): 25 mM HEPES; pH 7.9, 5 mM KCl, 0.5 mM
MgCl2, 1 mM DTT, 100 mM PMSF and ad. 100 mL.
 Nuclear Buffer 2 (NB2): 100 mL Buffer 1 þ 1 mL NP-40 (1% IGEPAL).
 Nuclear Buffer 3 (NB3): 1:1 mixture of buffer 1 þ 2.
 Nuclear Buffer 4 (NB4): 25 mM HEPES, pH 7.9; 350 mM NaCl.
 leupeptin, PMSF, aprotinin, pepstatin, and DTT (Calbiochem).
 IGEPAL CA-630 (Sigma).
Note: All components (buffers, protease inhibitors, etc) must be kept on
ice during the procedure
Procedure:
The following protocol is optimized for about 107 cells.
1. Discard media and add 1 mL of ice-cold PBS.
2. Scrape cells into PBS, transfer to eppendorf tube and spin at 10,000g for
5 min.
3. Discard supernatant and resuspend pellet in 200 mL of NB1 (scale-up or
down proportionally).
4. Add 200 mL of NB2 and rotate at 4  C for 15 min.
5. Spin at 2500g for 1 min.
6. Transfer supernatant to new eppendorf tube (cytoplasmic protein-enriched
fraction).
7. Add 100 mL of NB3 and gently mix. Transfer the wash to the cytoplas-
mic protein preparation from step 6. A short spin may be necessary if the
pellet was disturbed.
8. Add 500 mL of NB4 and rotate for 1 h at 4  C.
9. Spin at max speed for 10 min. Supernatant after this spin is the nuclear
protein preparation.
326 Uwe Michelsen and Jörg von Hagen

2.10. Immunoaffinity reagents for organelle enrichment

2.10.1. Introduction
Cell fractions often contain more than one type of organelle even after
differential and equilibrium density gradient centrifugation. The usage of
antibodies is described since 1975 (Thompson and Miller, 1975) for diverse
organelles and depends heavily on the quality of the antibody used. Besides
antibodies also the labeling of organelles like plasma membranes with biotin
and the isolation via streptavidin is a common procedure (Zeheb and Orr,
1986). Preenriched density gradient fractions can be further purified by
immunological techniques, using monoclonal antibodies for various organ-
elle-specific membrane proteins. One example is the purification of a partic-
ular class of cellular vesicles whose outer surface is coated with the protein
clathrin. An antibody to clathrin, bound to protein A or G (depending on
antibody subtype and species) coupled to a solid phase like agarose, polymeric
resins or preferably to magnetic particles which allows for an orthogonal
separation, not only gravity, which might interfere with large organelles in
crude extracts. The target specific antibody selectively binds to the vesicles in
a crude preparation of membranes, and the whole antibody complex can then
be isolated by low-speed centrifugation or magnetic separation.
All cells contain a dozen or more different types of small membrane-
limited vesicles of about the same size (50–100 nm in diameter) and density.
Because of their similar size and density, these vesicles are difficult to separate
from one another by centrifugation techniques. Immunological techniques
are particularly useful for purifying specific classes of such vesicles.
Material and equipment:
 Protein A/G coupled solid phase in HEPES, pH 7.0, containing 1% BSA,
100 mM sodium chloride, and 0.01% sodium azide.
 Monoclonal or polyclonal AB.
 Wash buffer I (WB1).
 20 mM HEPES, pH 7.0
 1% BSA
 Wash buffer II (WB2):
 20 mM HEPES, pH 7.0
 100 mM sodium chloride
 Elution buffer (EB):
 0.1% SDS, 50 mM Tris, pH 6.5
 100 mM DTT and 10% glycerol

Procedure:
Prepare a precleared cell lysate as described above:
1. Incubate the lysate, antibody (according to the suppliers information),
Protein A/G and solid phase (30 mL 1:1 slurry solution) overnight at
4  C with gentle agitation.
Isolation of Subcellular Organelles and Structures 327

Coated vesicle/organelle

Protein A or G
Antibody

ic/ ria
er acte
y m b
ol r
(p c) o
d
a e t i
Be agn
m

Figure 19.6 Scheme of affinity organelle and vesicle isolation.

2. Centrifuge at 500g for 5 min at 4  C.

3. Wash the pellet three times with 500 mL Wash buffer I.
4. Wash the pellet twice with 500 mL Wash buffer II.
5. Elute the coated vesicles/organelles with 30 mL elution buffer or under
mild extraction conditions if required.
6. Centrifuge at 5000g for 1 min. The supernatant contains the enriched
clathrin coated vesicles (Fig. 19.6).

REFERENCES
Abdolzade-Bavil, A., Hayes, S., Goretzki, L., Kröger, M., Anders, J., and Hendriks, R.
(2004). Convenient and versatile subcellular extraction procedure, that facilitates classical
protein expression profiling and functional protein analysis. Proteomics 4, 1397–1405.
Billecke, C., Malik, I., Movsisyan, A., Sulghani, S., Sharif, A., Mikkelsen, T., Farrell, N. P.,
and Bögler, O. (2006). Analysis of glioma cell platinum response by metacomparison of
two-dimensional chromatographic proteome profiles. Mol. Cell Proteomics 5, 35–42.
Edelman, M., Epstein, H. T., and Schiff, J. A. (1966). Isolation and characterization of DNA
from the mitochondrial fraction of Euglena. J. Mol. Biol. 17, 463–469.
Fisher, W. D., and Cline, G. B. (1963). A density gradient for the isolation of metabolically
active thymus nuclei. Biochim. Biophys. Acta 68, 640–642.
Ford, T., Graham, J., and Rickwood, D. (1994). Iodixanol: A nonionic iso-osmotic
centrifugation medium for the formation of self-generated gradients. Anal. Biochem.
220, 360–36612.
Goldberg, S. (2008). Mechanical/physical methods of cell disruption and tissue homogeni-
zation. Methods Mol. Biol. 424, 3–22.
Graham, J., Ford, T., and Rickwood, D. (1994). The preparation of subcellular organelles
from mouse liver in self-generated gradients of iodixanol. Anal. Biochem. 220, 367–373.
Kurokawa, M., Kato, M., and Sakamoto, T. (1965). Distribution of sodium-plus-potassium-
stimulated adenosine-triphosphatase activity in isolated nerve-ending particles. Biochem.
J. 97, 833–844.
328 Uwe Michelsen and Jörg von Hagen

Nakamura, E., Kozaki, K., Tsuda, H., Suzuki, E., Pimkhaokham, A., Yamamoto, G.,
Irie, T., Tachikawa, T., Amagasa, T., Inazawa, J., and Imoto, I. (2008). Frequent
silencing of a putative tumor suppressor gene melatonin receptor 1 A (MTNR1A) in
oral squamous-cell carcinoma. Cancer Sci. 99, 1390–1400.
Pertoft, H., Rubin, K., Kjellén, L., Laurent, T. C., and Klingeborn, B. (1977). The viability
of cells grown or centrifuged in a new density gradient medium. Percoll. Exp. Cell Res.
110, 449–457.
Ramsby, M. L., Makowski, G. S., and Khairallah, E. A. (1994). Differential detergent
fractionation of isolated hepatocytes: Biochemical, immunochemical and two-dimensional
gel electrophoresis characterization of cytoskeletal and noncytoskeletal compartments.
Electrophoresis 15, 265–277.
Rickwood, D., Ford, T., and Graham, J. (1982). Nycodenz: A new nonionic iodinated
gradient medium. Anal Biochem. 123, 23–31.
Ross, P. L., Huang, Y. L. N., Marchese, J. N., Williamson, B., Parker, K., Hattan, S.,
Khainovski, N., Pillai, S., Dey, S., Daniels, S., et al. (2004). Multiplexed protein
quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents.
Mol. Cell. Proteomics 3, 1154–1169.
Sadowski, P. G., Dunkley, T. P., Shadforth, I. P., Dupree, P., Bessant, C., Griffin, J. L., and
Lilley, K. S. (2006). Quantitative proteomic approach to study subcellular localization of
membrane proteins. Nat. Protoc. 1, 1778–1789.
Schnitzler, G. R. (2001). Isolation of histones and nucleosome cores from mammalian cells.
Curr. Protoc. Mol. Biol. Chapter 21, Unit 21.5.
Shadforth, I., Dunkley, T., Lilley, K., and Bessant, C. (2005). i-Tracker: For quantitative
proteomic using iTRAQ. BMC Genomics 6, 145.
Shah, M. B., and Sehgal, P. B. (2007). Nondetergent isolation of rafts. Methods Mol. Biol.
398, 21–28.
Sherrier, D. J., Prime, T. A., and Dupree, P. (1999). Glycosylphosphatidylinositol-anchored
cell-surface proteins from Arabidopsis. Electrophoresis 20, 2027–2035.
Simon, R. H., and Felsenfeld, G. (1979). A new procedure for purifying histone pairs H2A þ
H2B and H3 þ H4 from chromatin using hydroxylapatite. Nucleic Acids Res. 6, 689–696.
Storrie, B., and Madden, E. A. (1990). Isolation of subcellular organelles. Methods Enzymol.
182, 203–225.
Thompson, E. B., and Miller, J. V. (1975). Enrichment of polysomes synthesizing a specific
protein by use of affinity chromatography. Methods Enzymol. 40, 266–273.
Van Veldhoven, P. P., Baumgart, E., and Mannaerts, G. P. (1996). Iodixanol (Optiprep), an
improved density gradient medium for the iso-osmotic isolation of rat liver peroxisomes.
Anal. Biochem. 237, 17–23.
von Hagen, J. (2008). Proteomics Sample Preparation. New York: Wiley VCH, New York.
Washburn, M. P., Wolters, D., and Yates, J. R. (2001). Large-scale analysis of the yeast
proteome by multidimensional protein identification technology. Nat. Biotechnol. 19,
242–247.
Weber, P. J., Weber, G., and Eckerskorn, C. (2004). Isolation of organelles and prefractio-
nation of protein extracts using free-flow electrophoresis. Curr. Protoc. Protein Sci.
Chapter 22, Unit 22.5.
Willemsen, O., Machtejevas, E., and Unger, K. K. (2004). Enrichment of proteinaceous
materials on a strong cation-exchange diol silica restricted access material: protein-
protein displacement and interaction effects. J. Chromatogr. A. 1025, 209–216.
Zagariya, A., Khrapunov, S., and Zacharias, W. (1993). Rapid method for the fractionation
of nuclear proteins and their complexes by batch elution from hydroxyapatite. J. Chro-
matogr. 648, 275–278.
Zeheb, R., and Orr, G. A. (1986). Use of avidin-iminobiotin complexes for purifying
plasma membrane proteins. Methods Enzymol. 122, 87–94.
 C H A P T E R T W E N T Y

Protein Precipitation Techniques

Richard R. Burgess

Contents
1. Introduction 332
2. Ammonium Sulfate Precipitation 332
2.1. Principles 332
2.2. Basic procedure 334
2.3. Doing an ammonium sulfate precipitation test 336
2.4. Comments/problems/solutions 337
3. Polyethyleneimine Precipitation 337
3.1. Principles 337
3.2. Different modes of use of PEI 338
3.3. Basic procedure for Strategy C 338
3.4. Doing an PEI precipitation test 340
3.5. An example of using PEI to precipitate a basic protein
bound to DNA 340
4. Other Methods 341
4.1. Ethanol and acetone precipitation 341
4.2. Isoelectric precipitation 341
4.3. Thermal precipitation 341
4.4. Polyethylene glycol (nonionic polymer) precipitation 341
5. General Procedures When Fractionating Proteins by Precipitation 341
References 342

Abstract
After cell lysis, the most often used second step in a protein purification
procedure is some sort of a rapid, bulk precipitation step. This is commonly
accomplished by altering the solvent conditions and taking advantage of the
changes in solubility of your protein of interest relative to those of many of
the other proteins and macromolecules in a cell extract. This chapter will focus
on the two most widely used precipitation methods: (1) ammonium sulfate
precipitation and (2) polyethyleneimine (PEI) precipitation. These two methods
work through entirely different principles, but each can achieve significant
enrichment of target protein if optimized and applied carefully.

McArdle Laboratory for Cancer Research, University of Wisconsin–Madison, Madison, Wisconsin, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63020-2 All rights reserved.

331
332 Richard R. Burgess

1. Introduction
For both laboratory scale and larger scale protein fractionation, there is
a need for a quick, bulk precipitation to remove much of cellular protein
and other components. It is especially important to remove proteases as
early in the procedure as possible to avoid protein degradation. This precip-
itation must be rapid, gentle, scalable, and relatively inexpensive. In addi-
tion to the fractionation, it can also achieve a significant concentration of
the enriched protein. While many different precipitation methods have
been used over the last hundred years, ammonium sulfate (AS) has remained
the most widely used and polyethyleneimine (PEI) has increased in popu-
larity, especially for acidic proteins. These two methods will be discussed in
detail, followed briefly by several other precipitation methods and some
general advice on handling precipitates and obtaining maximal purification
from your precipitation step. An extensive and very useful general overview
of various types of protein precipitation procedures can be found in Scopes
(1994). An excellent review of AS and organic solvent (ethanol and
acetone) precipitation can be found in Englard and Seifter (1990).

2. Ammonium Sulfate Precipitation

2.1. Principles
While several salts can be used as precipitants, AS has several properties that
make it the most useful. It is very stabilizing to protein structure, very soluble,
relatively inexpensive, pure material is readily available, and the density of a
saturated solution (4.1 M ) at 25
C ( r ¼ 1.235 g/cm3) is not as high as
another salting-out agent, potassium phosphate (3 M, r ¼ 1.33 g/cm3).
Figure 20.1 shows a typical protein solubility curve where the log of the
protein solubility is plotted as a function of AS concentration. The main
features of this curve are a region at low salt where the solubility increases
(called ‘‘salting in’’), and then a region where the log solubility decreases
linearly with increasing AS concentration (called ‘‘salting out’’). The latter
part of the curve can be described by the equation log10S ¼ b  Ks(G/2)
where S in the solubility of the protein in mg/ml of solvent, G/2 is the ionic
strength, and b and Ks are constants characteristic of the protein in question.
Ks is a measure of the slope of the line and b is the log of the solubility if the
salting-out curve is extrapolated to zero ionic strength. In general, most
proteins have similar Ks values but vary considerably in their b value.
Suppose that the curve in Fig. 20.1 is valid for your protein and that the
concentration of your protein in a cell extract is 1 mg/ml. The upper
Precipitation Techniques 333

“Salting in” “Salting out”

Slope = Ks
1
Log10S (mg/ml)

0
1 mg/ml

−1
0.1 mg/ml

−2
0.01 mg/ml

−3

26 32 38

0 10 20 30 40 50
Ammonium sulfate (% saturation)

Figure 20.1 Ammonium sulfate solubility curve for a hypothetical protein. This
represents the log solubility of a hypothetical protein as a function of percent saturation
of ammonium sulfate. The ‘‘salting-out’’ line follows the relationship log S ¼ b 
Ks(G/2) as described in the text, where G/2 is the ionic strength, which here is given
as ammonium sulfate percent saturation.

horizontal dotted line intercepts the solubility line at log S ¼ 0

(S ¼ 1 mg/ml) and at an AS percent saturation of 26%. This means that if
you add AS to 26% saturation, all of your protein would be soluble. Now if
you increased the AS to 32% saturation (the middle horizontal dotted line),
the log S would be  1 (S ¼ 0.1 mg/ml) so 90% of your protein would
become insoluble and precipitate out. For this extract, an excellent strategy
would be to make a 26–32% saturated AS cut: add AS to 26%, spin out
insoluble material, and then make the supernatant 32% saturated and collect
what precipitates, which would contain 90% of your protein. You would
remove those proteins and cell components that precipitate at 26%
saturation and those that fail to precipitate at 32% saturation.
It is instructive to consider what would happen if you diluted the extract
10-fold with buffer. Now the initial concentration of your protein in the
334 Richard R. Burgess

extract is 0.1 mg/ml or log S ¼  1. You can add AS to 32% saturation and
your protein will not precipitate. To achieve 90% precipitation of your
protein, you would have to increase the AS to about 38% saturation
(bottom horizontal dotted line) or carry out a 32–38% saturated AS cut.
You would end up having to use more than 10 times as much AS with the
diluted extract to obtain your protein. This illustrates how important it is to
specify the concentration of your extract.
You do not usually have a curve like that shown in Fig. 20.1 for your
protein of interest so you have to determine the appropriate AS concentra-
tions experimentally as described below.

2.2. Basic procedure

While there are numerous variations on AS precipitation, the most
common ones are to add solid AS to a protein extract to give a certain
percent saturation. Adding an amount of solid AS based on Table 20.1 is
convenient, reproducible, and practical.
1. Generally one determines a lower percent saturation at which the
protein of interest just does not precipitate and a higher percent satura-
tion that gives > 90% precipitation as described in the section below.
2. Add solid AS to reach the lower value. Take care to add the AS slowly
with rapid stirring so that the local concentration does not ‘‘overshoot’’
the target value. Some people carefully grind the solid AS with a mortar
and pestle to a fine powder that dissolves rapidly. Once the AS is
completely dissolved, allow the precipitation to continue for about
30 min. This is a compromise between waiting several hours as precipi-
tation slowly approaches equilibrium and the desire to move along with
the purification and not to introduce long delays in the procedure.
Generally, one carries out all operations in an ice bucket or cold room.
3. Centrifuge at about 10,000  g for about 10 min in a precooled rotor to
pellet the material that is insoluble.
4. Carefully pour off the supernatant and determine its volume. Determine
the grams of AS from Table 20.1 to go from the lower desired percent
saturation to the final higher percent saturation. Again add the AS slowly
with rapid mixing to avoid high local concentrations and let the solution
sit for 30 min to allow precipitation to occur.
5. Centrifuge as above. Let the pellet drain for about 1 min to remove as
much as possible of the supernatant. If you have carried out the test
precipitation carefully, the pellet will contain 90% or more of your target
protein. This protein can be dissolved in an appropriate buffer and after
either dialysis, desalting, or dilution used in the next step of the
purification.
Table 20.1 Final concentration of ammonium sulfate: Percentage saturation at 0
Ca

Initial concentration of Percentage saturation at 0
C

ammonium sulfate
(percentage saturation 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
at 0
C) Solid ammonium sulfate (g) to be added to 1 l of solution

0 106 134 164 194 226 258 291 326 361 398 436 476 516 559 603 650 697
5 79 108 137 166 197 229 262 296 331 368 405 444 484 526 570 615 662
10 53 81 109 139 169 200 233 266 301 337 374 412 452 493 536 581 627
15 26 54 82 111 141 172 204 237 271 306 343 381 420 460 503 547 592
20 0 27 55 83 113 143 175 207 241 276 312 349 387 427 469 512 557
25 0 27 56 84 115 146 179 211 245 280 317 355 395 436 478 522
30 0 28 56 86 117 148 181 214 249 285 323 362 402 445 488
35 0 28 57 87 118 151 184 218 254 291 329 369 410 453
40 0 29 58 89 120 153 187 222 258 296 335 376 418
45 0 29 59 90 123 156 190 226 263 302 342 383
50 0 30 60 92 125 159 194 230 268 308 348
55 0 30 61 93 127 161 197 235 273 313
60 0 31 62 95 129 164 201 239 279
65 0 31 63 97 132 168 205 244
70 0 32 65 99 134 171 209
75 0 32 66 101 137 174
80 0 33 67 103 139
85 0 34 68 105
90 0 34 70
95 0 35
100 0
a
Reprinted from Englard and Seifter (1990), which was adapted from Dawson et al. (1969).
336 Richard R. Burgess

2.3. Doing an ammonium sulfate precipitation test

Generally one can precipitate 90% of a given protein with a 10% increase in
AS saturation so one should restrict the range of the ‘‘AS cut’’ to no more
than 10% (the proteins that are just soluble at 30% saturation but precipitate
at 40% saturation are referred to as the 30–40% AS cut).
Figure 20.2 illustrates a method to determine the optimal AS precipita-
tion conditions using only two centrifugation steps. Basically you place a
volume of cell extract, for example, 10 ml in each of five tubes. You add
with mixing amounts of solid AS to give 20%, 30%, 40%, 50%, and 60%
saturation based on Table 20.1, let sit 30 min to allow precipitation, and
then centrifuge to pellet the insoluble material. The pellets represent the
20%, 30%, 40%, 50%, and 60% saturated AS pellets. The volumes of the
corresponding supernatants are determined and again solid AS is added to
raise each to a 10% higher level of saturation. Again you mix, allow 30 min
to precipitate, and then spin. The five pellets are the 20–30%, 30–40%,
40–50%, 50–60%, and 60–70% AS cuts. Each of these is dissolved in buffer
and assayed for enzyme activity and total protein and perhaps subjected to
SDS gel analysis. Most of the activity should be in one of the cuts, but if, for
example, half is in 30–40% cut and half is in the 40–50% cut, then perhaps a
35–45% cut would be optimal. While this test may seem onerous, it is really

10 ml

AS 1.06 g 1.64 g 2.26 g 2.91 g 3.61 g

% saturation 20% 30% 40% 50% 60%

Mix, incubate 30 min, 0 ⬚C, centrifuge

10 ml
supernatant
AS 0.55 g 0.56 g 0.58 g 0.60 g 0.62 g
% saturation 30% 40% 50% 60% 70%

Mix, incubate 30 min, 0 ⬚C, centrifuge

Pellet 20–30% 30–40% 40–50% 50–60% 60–70%

(% cut)
Analyze

Figure 20.2 How to do an ammonium sulfate precipitation test. This test is carried out
as described in the text and is self-explanatory.
Precipitation Techniques 337

quite an efficient way to determine the optimal conditions that will result in
higher enrichment in this important step.

2.4. Comments/problems/solutions
1. Pellet is not solid. If the AS pellet is not firm after centrifugation it will be
difficult to cleanly pour off the supernatant. One solution is simply to
centrifuge 50% longer such that the precipitate that sedimented to the
bottom of the tube has more time to compact. Another reason for a loose
pellet is the presence of DNA that increases viscosity and slows sedimen-
tation rates. If viscosity is a problem, it can be reduced by sonicating longer
to break cells and shear DNA into shorter pieces. Another approach is to
treat with the recombinant nuclease Benzonase (EMD/Novagen).
2. Pellet floats in high concentrations of AS. Since the density of very high
concentrations of AS approach that of protein aggregates, the AS pre-
cipitate might float rather than sediment to the bottom of the tube
during centrifugation. This can be a problem especially when your
protein contains lipid or if there are nonionic detergents around that
bind to the protein and decrease its density.
3. Published protocols are often hard to follow. Many published proce-
dures fail to indicate the AS saturation convention (saturated at 0, 20, or
25
C) or the protein concentration of the extract. As cautioned above,
the amount of AS needed to achieve a given precipitation is dependent
on the protein concentration.
4. You must interrupt an AS precipitation procedure. If you must stop the
procedure, leave the protein as an AS precipitate. Proteins are remarkably
stable in AS, either as a suspension of precipitated protein or as a pellet.

3. Polyethyleneimine Precipitation
3.1. Principles
PEI, whose trade name is Polymin P, is a basic cationic polymer made by
BASF in large quantities for use in the textile and paper industry. PEI is a
product of polymerization of ethyleneimine to yield a basic polymer with
the structure: CH3CH2N–(–CH2CH2–NH–)n–CH2CH2NH2. Typically,
n equals 700–2000 to give a molecular weight range of 30,000–90,000 Da.
Since the pKa value of the imino group is 10–11, PEI is a positively charged
molecule in solutions of neutral pH. The use of PEI in protein fractionation
originated at Boehringer Mannheim and was published by Zillig et al.
(1970). More extensive examples of its application in protein purification
338 Richard R. Burgess

and several reviews have been published (Burgess, 1991; Burgess and
Jendrisak, 1975; Jendrisak, 1987; Jendrisak and Burgess, 1975).
PEI can be thought of as similar to soluble DEAE cellulose. It binds to
negatively charged macromolecules such as nucleic acid and acidic proteins
and forms a network of PEI and bound acidic molecules that rapidly
precipitates. The binding is stoichiometric. A heavy precipitate rapidly
forms that can be pelleted by centrifugation for 5 min at 5000 rpm. Whether
an acidic protein binds to PEI depends on the salt concentration. At low salt
(0.1 M NaCl), a mildly acidic protein will bind and precipitate, but at
intermediate salt (0.4 M NaCl), it will elute from the polymer and become
soluble. A highly acidic protein will bind at low salt, not be solubilized at
intermediate salt, and will be eluted at high salt (0.9 M NaCl). It should be
noted that when a protein is eluted from a PEI pellet both the protein and
the PEI become soluble. Thus it is necessary to remove the PEI from the
protein before returning to low salt (see below).

3.2. Different modes of use of PEI

There are three different strategies for the use of PEI precipitation.
Strategy A: Precipitate with PEI at high salt (1 M NaCl). This precipitates the
nucleic acids and leaves almost all protein in the supernatant.
Strategy B ( for neutral or basic proteins): Precipitate with PEI at 0.1 M NaCl to
remove nucleic acids and acidic proteins. This leaves protein of interest in
the supernatant.
Strategy C ( for acidic proteins such as E. coli RNA polymerase): This protocol
will be presented in detail below and is based on Burgess and Jendrisak
(1975) as refined by Burgess and Knuth (1996).

3.3. Basic procedure for Strategy C

1. Prepare a 10% (v/v) [5% (w/v)] PEI stock solution. PEI comes as viscous
liquid that is 50% (w/v). (We use PEI from MP Biochemicals,
Mw ¼ 50,000–100,000, but one can also use material from other sources,
for example, Sigma and Aldrich, Mw ¼ 750,000). Ten milliliter is diluted
to 70 ml with ddH2O and concentrated HCl (3.8–4.0 ml) is added until
the pH reaches 7.9. The volume is made up to a final volume of 100 ml
with ddH2O. This stock solution is stable at cold room or room
temperature indefinitely. It should be noted that some companies sell
PEI solutions that have been diluted one to one with ddH2O to reduce
viscosity and aid in dispensing. This material is only 25% (w/v).
2. Break E. coli cells (about 3 g wet weight cell pellet) by sonication in
30 ml buffer containing 50 mM Tris–HCl, pH 7.9, 5% glycerol, 0.1 mM
Precipitation Techniques 339

EDTA, 0.1 mM DTT, and 0.15 M NaCl. Centrifuge out cell debris at
15,000 rpm for 15 min. All operations are carried out on ice.
3. Based on a PEI precipitation test for your system (see below) add 10%
(v/v) PEI, pH 7.9 to a final concentration of, for example, 0.3% (v/v)
and mix well for 5 min to allow formation of a dense white precipitate.
4. Centrifuge at 5000 rpm for 5 min. Note: Do not centrifuge too hard or the
pellet will be harder to resuspend. Decant the 0.3% PEI supernatant and save
for later analysis. Let the pellet drain for 1–2 min to get rid of as much of
the supernatant as possible.
5. Thoroughly resuspend the 0.3% PEI pellet in 30 ml of the above buffer
containing 0.4 M NaCl. If available, the Tissue Tearor homogenizer
(BioSpec Products, Inc. Cat # 985370-07) works very well to finely
resuspend the pellet and effectively washes out proteins physically
trapped in the pellet, and elutes out mildly acidic protein that are weakly
bound to the PEI in the pellet. Let sit 5 min and then centrifuge at
5000 rpm for 5 min and decant the 0.4 M NaCl wash.
6. Resuspend thoroughly the 0.4 M NaCl pellet in 30 ml of buffer above
but containing 0.9 M NaCl. This elutes more acidic proteins (like E. coli
RNA polymerase), but leaves the nucleic acids in the pellet (it takes
about 1.6 M NaCl to elute the nucleic acids). Let sit about 5 min, mix,
and centrifuge at 15,000 rpm for 10 min.
7. To the 0.9 M NaCl eluate, add solid AS to 60% saturation (add
3.61 g per 10 ml of eluate). Mix well and let precipitation occur for
at least 30 min. Centrifuge at 15,000 rpm for 10 min. Let pellet drain
for 5 min. The pellet contains the AS precipitated protein, but almost
all of the PEI remains in the supernatant. While traces of PEI are
trapped in the pellet, they usually do not interfere with subsequent
operations. If it is necessary to more completely remove these PEI
traces, one can resuspend the pellet in buffer containing 60%
saturated AS and recentrifuge.
This procedure typically gives a sixfold purification of RNA polymerase
from other proteins, greater than 90% recovery, and a nearly complete
removal of nucleic acids in 1–2 h.

3.3.1. Additional comments

1. It is important to emphasize that it is necessary to remove PEI from the
0.9 M NaCl eluate. If you merely dilute to low salt or dialyze to low salt,
the proteins will again bind PEI and reprecipitate.
2. We have found that the PEI precipitation can occur even in the presence
of 1% Triton X-100.
3. In contrast to AS precipitation, when you dilute an extract 10-fold with
buffer, you use the same total amount of PEI (e.g., if you found that
0.3% PEI gave good precipitation of your protein with a normal extract,
340 Richard R. Burgess

you would only have to add 0.03% PEI to achieve the same precipitation
with the 10-fold diluted extract, but of course there would be 10 times
the volume of extract). This reflects the fact that PEI binds tightly to the
acidic components and essentially titrates them.

3.4. Doing an PEI precipitation test

Since one cannot predict how much PEI to add to precipitate a given acidic
protein nor how much salt it will take to elute the protein from the PEI
pellet, one is advised to carry out simple PEI precipitation and elution tests
(Burgess and Jendrisak, 1975; Burgess and Knuth, 1996). Basically this
involves taking about six 200-ml samples into six microfuge tubes and
adding 10% (v/v) PEI to final concentrations of 0%, 0.1%, 0.2%, 0.3%,
0.4%, and 0.5% (v/v). Mix well and microfuge for 1 min at high speed.
Analyze the supernatants by enzyme activity or by SDS polyacrylamide gel
electrophoresis and, if necessary, by Western blot to determine the mini-
mum amount of PEI needed to precipitate all of the target protein. Let us
say that it takes 0.3%. Now prepare a set of six microfuge tubes which each
contain a small 0.3% PEI pellet prepared as above and add 200 ml of buffer
containing 0, 0.2, 0.4, 0.6, 0.8, and 1.0 M NaCl. Resuspend well. Let sit for
15 min, microcentrifuge, and analyze supernatant for your protein as above.
The highest salt concentration that does not elute any of your protein is used
as a wash and the salt concentration where all of your protein is eluted is
used as the eluent.

3.5. An example of using PEI to precipitate a basic protein

bound to DNA
Recently (Duellman and Burgess, 2008), in trying to purify the very basic
protein Epstein–Barr virus nuclear antigen 1 (EBNA1) expressed in E. coli,
we carried out a PEI precipitation test at 0.1 M NaCl to see if we could
precipitate the nucleic acid and acidic proteins and leave the basic EBNA1
in the supernatant (this is Strategy B). To our surprise, the EBNA1 pre-
cipitated as we added some PEI (0.15%), but then failed to precipitate when
we added more PEI (0.4%). It appears that the EBNA1 was bound to DNA
and precipitated out with the DNA. However, at higher PEI concentrations
the PEI preferentially bound the DNA and displaced the EBNA1. We
found that precipitating with 0.15% PEI, washing with 0.3 M NaCl, and
then eluting out of the PEI pellet with 0.8 M NaCl gave both a good
enrichment for EBNA1 and rapid removal of nucleic acids.
Precipitation Techniques 341

4. Other Methods
This chapter has focused on AS and PEI precipitation. Other methods
for protein precipitation mentioned very briefly below have been well
described in numerous publications (see, Englard and Seifter, 1990;
Ingham, 1990; Scopes, 1994).

4.1. Ethanol and acetone precipitation

Precipitation with organic solvents, such as ethanol and acetone, has been in
use for well over a hundred years, but is probably best known for its use in
fractionating human serum in the classic work of Cohen and Edsall. Care
must be taken to carry out precipitations at very cold temperatures to avoid
protein denaturation.

4.2. Isoelectric precipitation

Proteins are less soluble at their isoelectric point where they have zero net
charge and can most easily approach each other with minimal charge
repulsion. Since proteins are also less soluble at very low ionic strength,
isoelectric precipitation is usually done at very low or no salt.

4.3. Thermal precipitation

In this method, cell extracts are heated to a temperature at which many
proteins denature and precipitate, where the protein of interest is more
stable and stays soluble. This approach is particularly useful in purifying
enzymes from thermophiles expressed in E. coli where the extract is heated
to a high enough temperature, often to denature and precipitate almost all
E. coli protein, leaving the heat stable enzyme in solution.

4.4. Polyethylene glycol (nonionic polymer) precipitation

This subject has been reviewed extensively by Ingham (1990).

5. General Procedures When Fractionating

Proteins by Precipitation
1. Thorough resuspension of precipitated protein pellets is important dur-
ing washing or elution. While a pellet may seem quite solid, there is a
very significant amount of supernatant trapped in a pellet and adhering
342 Richard R. Burgess

to the walls of the centrifuge tube. As mentioned earlier, it is wise to let

the pellet drain well to remove as much of the supernatant as possible.
If the pellet is large compared to the total amount of supernatant then it is
recommended that one resuspend the pellet in 10 volumes of an appro-
priate buffer to remove supernatant proteins trapped in the pellet. For
example, with a 40% saturated AS precipitate, one can wash by resus-
pending the pellet in 40% saturated AS and recentrifuging. For washing
pellets of PEI precipitated material, washing is very useful and the use of
a homogenizer like a Tissue Tearor is recommended to break up the
precipitate into a very fine suspension. If resuspension is not thorough,
then material that you want to wash out is not efficiently removed and
the fractionation is less effective.
2. Try to avoid frothing when mixing. Whipping air into a protein solution
can promote oxidation of proteins and also cause protein denaturation at
the air–water interface.

REFERENCES
Burgess, R. R. (1991). The use of polyethyleneimine in the purification of DNA binding
proteins. Meth. Enzymol. 208, 3–10.
Burgess, R. R., and Jendrisak, J. J. (1975). A procedure for the rapid, large-scale purification
of E. coli DNA-dependent RNA polymerase involving Polymin P precipitation and
DNA-cellulose chromatography. Biochemistry 14, 4634–4638.
Burgess, R. R., and Knuth, M. W. (1996). Purification of a recombinant protein over-
produced in E. coli. In ‘‘Strategies for Protein Purification and Characterization:
A Laboratory Manual’’, (D. Marshak, J. Kadonaga, R. Burgess, M. Knuth,
W. Brennan, Jr., and S.-H. Lin, eds.), pp. 219–274. Cold Spring Harbor Press, Cold
Spring Harbor, NY.
Dawson, R. M. C., Elliot, D. C., Elliot, W. H., and Jones, K. M. (1969). In ‘‘Data for
Biochemical Research.’’ 2nd edn., p. 616. Oxford University Press, Oxford.
Duellman, S. J., and Burgess, R. R. (2008). Large-scale Epstein-Barr virus EBNA1 protein
purification. Protein Expr. Purif. 63, 128–133.
Englard, S., and Seifter, S. (1990). Precipitation techniques. Meth. Enzymol. 182, 287–300.
Ingham, K. C. (1990). Precipitation of proteins with polyethylene glycol. Meth. Enzymol.
182, 301–306.
Jendrisak, J. J. (1987). The use of PEI in protein purification. In ‘‘Protein Purification: Micro
to Macro’’, (R. Burgess, ed.), pp. 75–97. A.R. Liss, New York.
Jendrisak, J. J., and Burgess, R. R. (1975). A new method for the large-scale purification of
wheat germ DNA-dependent RNA polymerase II. Biochemistry 14, 4639–4645.
Scopes, R. K. (1994). Protein Purification, Principles and Practice. 3rd edn. pp. 7–101.
Springer-Verlag, New York.
Zillig, W., Zechel, K., and Halbwachs, H. J. (1970). A new method of large scale preparation
of highly purified DNA-dependent RNA-polymerase from E. coli. Hoppe Seylers Z.
Physiol. Chem. 351, 221–224.
 C H A P T E R T W E N T Y- O N E

Affi-Gel Blue for Nucleic Acid Removal

and Early Enrichment of Nucleotide
Binding Proteins
Murray P. Deutscher

Contents
1. A Representative Protocol 344
Reference 345

Abstract
Passage of an extract or supernatant fraction through a column of Affi-Gel Blue
and batchwise elution can be a rapid and effective early procedure for removal
of nucleic acid, concentration of the sample and purification of nucleotide
binding proteins.

Important points to be considered for the early steps of a protein

purification procedure are: (1) how to deal with the large volume and
large amounts of material; (2) how to remove nucleic acids; and (3) how
to enrich, at least to a small degree, the protein of interest, being mindful as
well of points 1 and 2.
In the initial steps of a protein purification, it is usual to be working with
large volumes of fairly concentrated extract. This can be quite cumbersome,
and as a consequence, one generally will use a bulk procedure such as
precipitation to partially fractionate the extract and also to reduce the
volume of material as the precipitate can be redissolved in a smaller volume.
Common precipitants for these purposes include ammonium sulfate,
ethanol or acetone, polyethylene glycol, and polyethyleneimine (see
Chapter 20).
The nucleic acids present in crude extracts can cause a variety of
problems and need to be removed early in the purification procedure.
High-molecular-weight DNA can lead to elevated viscosity making samples
difficult to work with. DNA and RNA also can interfere with assays,

Department of Biochemistry and Molecular Biology, University of Miami School of Medicine, Miami,
Florida, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63021-4 All rights reserved.

343
344 Murray P. Deutscher

particularly with enzymes that act on nucleic acids. As a consequence, assays

of extracts may dramatically underestimate enzyme activity. Most impor-
tantly, nucleic acids may hamper subsequent purification steps. In as much
as DNA and RNA are polyanions, they can compete with chromatographic
resins for protein binding. For all of these reasons, early removal of nucleic
acids is beneficial, and various protocols have been used for this purpose.
Among them are treatment with nucleases, precipitation with certain
reagents or binding to anion exchange resins.
An ideal early purification step would be one that could simultaneously
concentrate the sample, remove nucleic acids and afford some enrichment
of the desired protein as well. We have found that the resin, Affi-Gel Blue,
is particularly useful in this regard. Over the years, Affi-Gel Blue has been
used as an affinity chromatography resin in the purification of hundreds of
proteins, primarily later in a purification procedure (see Bio-Rad Bulletin
1107 for references, available at biorad.com). In those situations, the par-
tially purified protein of interest is bound to the resin and eluted with a salt
gradient or with ligand specific to the protein. For the purposes of the
discussion here, we focus on the use of Affi-Gel Blue very early in a
purification protocol, generally in a bulk fashion, to accomplish the goals
stated above.
Affi-Gel Blue is a cross-linked agarose bead to which the dye Cibacron
Blue F3GA has been covalently attached. Conjugates of this dye to other
supports also are commercially available, but have not been examined with
regard to the procedure described here. The Cibacron Blue dye has ionic,
hydrophobic, and aromatic character, and consequently, displays affinity for
many proteins. One large class of enzymes for which binding to the dye
appears particularly strong are those that have nucleotide cofactors or which
act on nucleotide-containing substrates, including DNA and RNA. Thus,
the procedure to be described has proven to be extremely useful for
enzymes of this group (e.g., Deutscher and Marlor, 1985).

1. A Representative Protocol
The sample to be loaded may be a low-speed (S10) or high-speed
(S100) supernatant fraction. The latter is preferable as there will be no
interference by ribosomes or microsomes. The size of the column to be
used will be determined by the amount of sample to be loaded. A useful rule
of thumb is to use 1 ml of packed gel (50–100 mesh) per 100 mg of protein
in the sample. According to the manufacturer (Bio-Rad, Bulletin 1107),
1 ml of packed Affi-Gel Blue binds 11 mg of albumin. Generally, only
5–10% of total protein binds to the columns under the conditions described,
so the suggested ratio of protein to gel should be sufficient. However, this
should be determined for each extract used by specific assay of the flow-
Early Purification with Affi-Gel Blue 345

through material to ensure that the protein of interest is completely retained

on the column. If not, the ratio should be adjusted accordingly. A column
with a large diameter is preferable to allow rapid flow rates.
We have found that a column buffer of 10 mM Tris–Cl, pH 7.5,
containing 0.1 M KCl and appropriate stabilizing agents to be a good
starting point. The sample should also be adjusted to 0.1 M KCl. As
noted, under these conditions, less than 10% of the total protein is retained
by the column. In addition, nucleic acid is not bound and is washed through
the column. After the sample is added, the column continues to be washed
with the same solution until the A280 of the eluate has ceased decreasing.
The protein of interest and most of the bound proteins can be eluted
batchwise in the same buffer solution containing 1 M KCl. Usually two
column volumes will be sufficient to elute most of the protein, although
some trailing may be observed. This general procedure will suffice for most
situations, and will result in removal of most nucleic acid, concentration of
the sample and a 5- to 10-fold purification of the protein of interest.
However, this protocol can be modified to increase the amount of
purification. One can use an intermediate elution step with a KCl concen-
tration that removes as much extraneous protein as possible, but not the
protein of interest. The desired protein can then be eluted with the lowest
concentration of KCl necessary, leaving other extraneous proteins still
bound. This enhanced protocol will require some trial and error elutions,
but 20-fold purification of the desired protein is often attainable. In some
cases, the protein of interest may be bound extremely tightly and requires
stronger elution conditions. We have found 1 M KBr to be a good choice,
but the eluate should be dialyzed rapidly because Br can lead to denatur-
ation of some proteins. Alternatively, increasing the KCl concentration to
2 M may also be successful.
The Affi-Gel Blue column can be reused, but the column should be
regenerated with several column volumes of 2 M guanidine HCl or 1.5 M
sodium thiocyanate followed by equilibration with the starting buffer.
Columns should be stored in the cold with 0.1% sodium azide. Under
these conditions, columns and resin are stable for years.

REFERENCE
Deutscher, M. P., and Marlor, C. W. (1985). Purification and characterization of Escherichia
coli RNase T. J. Biol. Chem. 260, 7067–7071.
 C H A P T E R T W E N T Y- T W O

Ion-Exchange Chromatography
Alois Jungbauer and Rainer Hahn

Contents
1. Introduction 349
2. Principle 351
3. Stationary Phases 353
4. Binding Conditions 355
5. Elution Conditions 361
6. Operation of Ion-Exchange Columns 363
7. Example: Separation of Complex Protein Mixture 366
7.1. Pretreatment of milk 366
7.2. Chromatography conditions 366
8. Example: High-Resolution Separation with a Monolithic Column 367
8.1. Mn peroxidase production 367
8.2. Chromatography conditions 367
References 370

Abstract
Ion-exchange chromatography is the most popular chromatographic method for
separation of proteins. It is a versatile and generic tool and is suited for
discovery of proteins, high-resolution purification, and industrial production
of proteins. Separation conditions are within physiological range of salt and
pH and in the most cases a native protein can be obtained. In this chapter, the
guidance will be provided for binding and elution conditions and selection of
stationary phases.

1. Introduction
Proteins, polynucleotides, and other biomacromolecules expose
charged moieties at the surface and thus they can interact with ion exchan-
gers. Ion-exchange chromatography is a versatile and generic tool for
protein and plasmid separation. It is frequently used for analytical and

Department of Biotechnology, University of Natural Resources and Applied Life Sciences, Vienna, Austria

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63022-6 All rights reserved.

349
350 Alois Jungbauer and Rainer Hahn

preparative purposes. From the early application to the current state of the
art of ion-exchange chromatography, the method substantially improved
over the years. In the early days of protein separation using ion exchangers,
only extremely low flow rates were applicable due to the soft nature of the
chromatography medium. Often the columns have been operated under
gravity flow and linear flow velocities of only 1–2 cm/h were obtained.
This led to purification cycles lasting several days. As a consequence, protein
purification had to be performed in the cold room to maintain the biological
activity and to prevent microbial growth. Currently, ion-exchange chro-
matography is operated at linear flow velocities of up to 500 cm/h. This is
two orders of magnitude faster than two decades ago. Whenever the
stability of a protein solution allows, ion-exchange chromatography is
performed at room temperature. A substantial step toward improvement
and acceleration of the purification cycles was achieved by introduction of
FPLC in the early 1980s ( Jungbauer, 1993). Since then, ion-exchangers
have been constantly improved ( Jungbauer, 2005). The latest jump in
development was introduction of monoliths ( Jungbauer and Hahn, 2008).
With this chromatography media it is possible to separate proteins in less
than 5 min ( Jungbauer and Hahn, 2004).
The popularity of the methods is based on the high resolution that can be
achieved with ion-exchange chromatography. A dilute protein or polynu-
cleotide solution can be rapidly concentrated and simultaneously purified.
Usually simple salt buffers are sufficient and concentrations are used in a
concentration range, where the so-called salting-in effect on proteins is
observed. This is the range where protein becomes more soluble with
increasing salt concentration. In contrast, this is not always possible with
hydrophobic interaction chromatography. With hydrophobic interaction
chromatography often salt concentrations are required which are in the
salting-out regime.
One drawback is frequently associated with ion-exchange chromatogra-
phy—this is incompatibility with mass spectrometry—especially in case of
ionization mode. The ionization of proteins and peptides is severely per-
turbed by ions.
We can define four general cases, where ion-exchange chromatography
is applied for protein and biomolecular purification and separation.
1. De novo purification: Here only limited information on the molecular
structure of the compounds is available. Often only a biological activity
of a protein solution or a cell extract is known. The task is to identify the
compounds responsible for the biological activity. In this case, intuition
is requested. Only limited rules are available and by observation of how
the biological activity behave on an ion-exchange column will guide to
selection of buffers and materials. Here, the work is really exploratory
and often several methods are interconnected to identify a protein.
Ion-Exchange Chromatography 351

The most difficult situation is the presence of protein complex. Often

the complexes are destroyed during purification and the biological
activity is lost or if not lost the biophysical properties of a complex is
different than a single protein.
2. Purification of a protein with known pI, size, and primary structure: Here, it is
possible to develop purification strategies based on rational rules. The
most popular one is that a protein binds at pH above pI to anion
exchangers and below pI to cation exchangers. From primary sequence,
the pI can be easily calculated. This design rule works as long as the
protein did not undergo posttranslationial modification such as glycosyl-
ation, phosphorylation, etc. Then, the surface charge is substantially
altered and the rule becomes obsolete. It has also been observed that
certain proteins bind also at pH identical to pI. This has been explained
that the surface charge of a protein is not homogenously distributed and
the binding occurs in an oriented manner. Of course, if the protein is
part of a multiprotein complex, then all predictions are useless (see
Chapter 3 in this volume).
3. Preparative and industrial separation of proteins: In this case, a purification
protocol already exists and the focus is on productivity, throughput or
scale-up. In such a scenario, attention is paid on maximizing/optimizing
dynamic binding capacity, geometry of the column, and the residence
time. Often gradient elution is replaced by step gradients, since in an
industrial environment sophisticated gradient mixing is not available.
Here the pressure drop is also an important parameter.
4. High resolution separation of protein variants and/or isoforms: Proteins
undergo posttranslational modifications, chemical modifications such as
deamidation or they are artificially modified such as PEGylation. Pro-
teins with the same primary structure have different surface structure and
thus different retention behavior on ion exchangers. This can be
exploited for separation into the individual isoforms or variants. Proteins
differing in a single charge can be separated by ion exchangers. This
high-resolution purification became very important since the different
isoforms may also differ in biological activity. Ion-exchange chromatog-
raphy is also often used for separation of intact from truncated forms.

2. Principle
The simplest explanation of ion-exchange chromatography of pro-
teins is based on the attraction of oppositely charged molecules. The protein
carries surface charges depending on pI and pH of the environment. The
ion exchanger is composed of a base matrix usually in the form of porous
beads to provide enough surface area for adsorption. On this base matrix,
352 Alois Jungbauer and Rainer Hahn

a charged ligand, either positively or negatively charged, is immobilized. To

improve capacity, instead of a small charged ligand, a charged polymer
has also been grafted onto the matrix. Irrespectively of the molecular
size of the charged ligand, we distinguish between cation and anion exchan-
gers. Cation exchangers are negatively charged and anion exchangers are
positively charged. Above its pI, the protein is negatively charged and binds
to an anion exchanger, below its pI, it is positively charged and binds
to cation exchangers. Earlier we have already discussed the limitations
of this simple rule. The ion exchanger itself behaves like an acid or base
and the disproportionation of the charges depends on the pH. Strong ion
exchangers behave like a strong acid or base and do not change the charge
within a wide range of pH, whereas weak ion exchangers do. This property
can also be exploited to gain selectivity or by applying pH gradients for
elution.
The binding of a charged species to an ion exchanger can be described
by the mass action law which has been postulated by Boardman and
Partridge (1955). Fred Regnier and his group have extended this work
(Kopaciewicz et al., 1983). The retention of a protein and the salt concen-
tration has been used to determine the number of binding sites interacting
with the surface. The binding of a protein is achieved at low and elution at
high salt concentration (Fig. 22.1). The elution window for isocratic elution
is very low, thus linear salt gradients are very popular.
K is the distribution coefficient of the protein between mobile and
stationary phase. K is inversely proportional to salt concentration and
power of interacting sites (z); Eq. (22.1):


log K ¼ log Ke qz0  z logðNaþ Þ; ð22:1Þ
where Ke is the affinity of the protein to the stationary phase and q0 is
the ion-exchange capacity. Steve Cramer and his group have further refined
the relationship and have also considered the shielded charges at the ion-
exchange surface by the adsorbed protein (Brooks and Cramer, 1992).
In respect to how many binding sites are involved in binding to a surface,
a different behavior of oligo-, polynucleotides and proteins have been
observed (Yamamoto et al., 2007). Especially for small oligonucleotides,
the number of binding sites correlated with the number of moieties.
Exceeding a certain size the number of binding sites does not increase
linearly and levels off. Plasmids bind with many binding sites thus it is
possible to bind them even at salt concentration of up to 800 mM. Proteins
usually bind with only a few charges.
In proteins, the number of binding sites compared to the charged
moieties on the surface is rather small. Only up to 10 binding sites have
been observed (Fig. 22.2).
Ion-Exchange Chromatography 353

A
100
Ke = 0.4, z = 5
Ke = 0.2, z = 3
80
Distribution coefficient, K
Ke = 0.2, z = 5

60

40

20

0
0 100 200 300 400
Salt concentration in mobile phase (Na+), mM

B
100

Ke = 0.4, z = 5
Ke = 0.2, z = 3
Distribution coefficient, K

Ke = 0.2, z = 5

10

0.1
100 200 300
Salt concentration in mobile phase (Na+), mM

Figure 22.1 Relationship between protein retention and salt concentration in ion-
exchange chromatography: (A) linear plot and (B) log–log relationship.

3. Stationary Phases
The evolution of ion-exchange media paralleled the development of
stationary phases, which took place over the last several decades. The
driving force for development of new media has been the need for better
354 Alois Jungbauer and Rainer Hahn

The number of charges

5⬘-ATAT-3⬘
0 20 40 60 80 100 5⬘-GCGC-3⬘
60 5⬘-GAGC-3⬘
5⬘-AATATT -3⬘
5⬘-AATGTT -3
The number of binding sites, B

5⬘-6A-3⬘
50 5⬘-6T-3⬘
5⬘-AATATATT -3⬘
5⬘-AATGTATT -3⬘
40 5⬘-GGCACGCC -3⬘
5⬘-GGCGCGCC -3⬘
5⬘-9A-3⬘
30 5⬘-9T-3⬘
5⬘-12A-3⬘
5⬘-12T-3⬘
20 5⬘-20A-3⬘
5⬘-20T-3⬘
5⬘-50A-3⬘
10 5⬘-50T-3⬘
5⬘-9A-3⬘ + 5⬘-9T-3⬘
b -lactoglobulin 5⬘-12A-3⬘ + 5⬘-12T-3⬘
5⬘-20A-3⬘ + 5⬘-20T-3⬘
0 5⬘-50A-3⬘ + 5⬘-50T-3⬘
10,000 20,000 30,000
24mer(P)
Molecular weight, MW 24mer(C)
24mer(P) + 24mer(C)

Figure 22.2 Relationship between the number of binding sites and the molecular mass
or number of charges of polyA and polyT DNAs. As comparison, the number of
binding sites of b-lactoglobulin is shown. Figure from Yamamoto et al. (2007).

mechanical stability, reduced tendency to unspecific adsorption, higher

binding capacity, and accelerated mass transfer. From an engineering
point of view, these features are to some extent incompatible with one
another (Mueller, 2005). A strategy for the development of new materials
must balance contradictive physical characteristics on the chromatogra-
pher’s wish list: Large particles allow a low-pressure drop but slow mass
transfer kinetics due to long diffusional pathways through the porous
network of the bead. Large pores enable fast mass transfer but reduce the
available surface area and thus the equilibrium capacity. Mechanical stability
also suffers from oversized pores. The chemical composition of the station-
ary phase (natural or synthetic polymers, inorganic materials) places intrinsic
limits for the preparation of porous spherical materials. Traditionally, sta-
tionary phases common in chromatography of biomolecules are typically
composed of a bead-shaped matrix comprising liquid filled pores. A variety
of materials has been used in the design of chromatography matrices ( Janson
and Ryden, 1998). Among these the most common are polysaccharides
(cellulose, dextran, and agarose), synthetic organic polymers (polyacryl-
amide, polymethacrylate, polystyrene), and inorganic materials (silica,
hydroxyapatite). To produce a mechanically stable and functional matrix,
the materials are chemically cross-linked and provided with a functional
ligand. The physical and chemical conditions of the raw materials present in
the production (solvents, concentration of base materials and cross-linkers,
temperature, etc.) determine the properties of the stationary phase. Particle
sizes of such media range from 2 mm for analytical purposes up to about
200 mm for low-pressure preparative applications. Pore sizes are in the range
Ion-Exchange Chromatography 355

of 10–100 nm. In many cases, chromatographic media exhibit a typical size

distribution of both particles and pores. The mean surface area for media
used for preparative applications is in the range of 5–100 m2/ml of gel.
A major improvement to the early, single material-based porous media has
been the invention of composite media which combine the properties of
two different base materials. Different types of preparations are available
(Mueller, 2005). One example is to fill polymers in pores of a rigid matrix
with subsequent cross-linking. These types of media are distinguished by a
solid (or surface-type) diffusion mechanism of proteins though the hydrogel
which is formed inside the pore space. Another, more frequently used
approach is the attachment of polymers to the surface by grafting techni-
ques. Extremely high binding capacities and accelerated mass transfer have
been shown for such media although the exact mechanism(s) for this
behavior is not clear. These media also exhibit a certain salt tolerance and
thus direct application of fermentation broths or feed stocks with moderate
salt concentrations is possible. Other developments worth mentioning
include the introduction of perfusion media (Afeyan et al., 1990) and
monoliths (Tennikova et al., 1990). In perfusion chromatography, the
pressure drop at high velocity causes a convective fluid flow in large,
permeable pores, while capacity is provided by smaller pores with high
surface areas. Monoliths are polymerized as single blocks, where the trans-
port of fluid is solely based on convective flow through the channels.
The entire mobile phase is driven through the whole volume of the matrix.
Components to be separated are conveyed to the active groups located on
the surface of flow-through pores, which form an interconnected network.
A selection of cation exchangers for the separation of proteins is given in
Table 22.1. The choice is much larger, but would exceed the scope of this
manuscript. Typical types with different transport mechanism have been
selected to cover the spectrum of available material.
Stationary phases for analytical applications are not discussed here.
Equivalent to the cation exchangers, one can also find anion exchangers
with the same base matrix but different ligands. Several suppliers also offer
the same base matrix but different particle diameter and/or ligands.

4. Binding Conditions
According to the general principles of ion-exchange equilibria, it is
clear that loading of proteins should be preferably at very low salt concen-
tration (Fig. 22.1). Depending of the ligand density of the ion exchanger,
the salt concentration must be adjusted. Usually a concentration of
10–100 mM buffer concentration is recommended. This corresponds to a
conductivity of 1–4 mS/cm. Regular feedstock from bacterial culture or
356

Table 22.1 Selected cation exchangers suited for preparative protein separation

Name of Mean particle Mass transport Typical application and

stationary phase Supplier Base matrix diameter (mm) mechanism features
SP Sepharose GE Healthcare Cross-linked agarose 90 Pore diffusion Preparative protein
fast flow capture
Capto S GE Healthcare Highly cross-linked 90 Unknown Preparative protein
agarose with flexible mechanisms capture; salt tolerant,
dextran surface (accelerated very high capacity
extender transport)
UNOsphere Bio-Rad Polyacrylamide 80 Parallel diffusion Preparative protein
rapid S interpenetrating (pore and capture; very high
network surface capacity
diffusion)
S-HyperD M Pall Ceramic shell filled 80 Surface diffusion Preparative protein
with capture; very high
polyacryalamide soft capacity
gel
CIM SO3 BIA Separations Polymethacrylate 1–2 Convective mass Ultrafast analytical
monolith transport separations; capture of
large biomolecules
Fractogel EMD Merck Polymethacrylate with 65 Unknown Preparative protein
SO3 M polyacrylamide mechanisms capture, salt tolerant,
surface extender (accelerated very high capacity
transport)
 POROS HS Applied Polystyrene– 50 Pore diffusion, Preparative protein
Biosystems divinylbenzene convective purification, improved
with through pores mass transport resolution
at high flow
rates
Source 30 S GE Healthcare Polystyrene– 30 Pore diffusion Preparative protein
divinylbenzene purification, improved
monobeads with resolution
hydrophilic coating
Toyopearl Tosoh Polymethacrylate 65 Pore diffusion Preparative protein
SP-650 M Bioscience capture
GigaCap S Tosoh Polymethacrylate with 65 Unknown Preparative protein
Bioscience flexible polymeric mechanisms capture; salt tolerant;
surface extender (accelerated very high capacity
transport)
357
358 Alois Jungbauer and Rainer Hahn

cell culture has a conductivity ranging from 15 to 40 mS/cm. Thus the

protein solution has to be either diluted or desalted. Dilution is a simple
method, but only suitable for small scale. The more appropriate methods are
dialysis, diafiltration, or desalting by size exclusion chromatography. For
laboratory scale quantities size exclusion chromatography is recommended.
This can be effected either by simple columns in centrifuge tube, prepacked,
or self-packed columns. The maximal loading volume is 1/3 of the total
column volume. Limiting factor is the viscosity of the protein solution. Due
to viscous fingering, band spreading increases and the load must be reduced.
Ion exchangers with grafted surfaces can be loaded with a feedstock of
higher conductivity (Necina et al., 1998). Sometimes it is possible to load a
clarified feedstock directly. This is definitely only possible with feedstock of
low conductivity. The grafted layer binds a lot of ions and the equilibrium is
shifted to conditions still allowing protein adsorption. A compromise must
be found between binding capacity and efforts of preconditioning the
feedstock. As trade off of a simple process by direct loading, a lower capacity
must be taken into account (Fig. 22.3). With modern ion changers a lower
capacity can be often accepted because they can be run at higher velocity.
The selection of the buffering species depends on the isoelectric point of
the protein. Above the isoelectric point, the protein is negatively charged
and binds to anion exchangers, below positively charged and binds to cation
exchangers. The actual selection depends on the impurity profile and the
stability of the protein. A protein with a high isoelectric point is preferably
purified with a cation exchanger. Such proteins are definitely easier to
purify. In Fig. 22.4, a histogram of isoelectric points of proteins is shown.
It is clear that slightly acidic proteins are more difficult to separate, since the
chance of impurities present in the feedstock with similar pI is very high.
The second rule for binding is that a buffer should be selected with a
pKa value close to the selected pH, in extreme maximum one pH unit
below or above.
The third rule is that the buffering species such as acetate or Tris should
not bind to the ion exchanger. So do not use, for example, acetate with
anion exchangers and Tris with cation exchangers. Binding of the buffering
species causes unstable pH and nonreproducible conditions.
The fourth rule is that the protein is exposed to a different pH when it is
close to the ion-exchanger surface. In cation exchangers, the pH can be up
to one pH unit lower due to binding of hydronium ions, whereas in anion
exchangers the pH can be one unit higher due to binding of hydroxyl ions.
This has to be obeyed especially when working with labile proteins.
When the pI and other information on the protein are not available,
then the binding conditions must be searched on small scale, either in a
batch or in a column mode. For the batch mode, the ion exchangers are
filled in small test tubes or in microtiter plates. The adsorbent is equilibrated
with the appropriate buffer of varying pH or/and salt concentration and
Ion-Exchange Chromatography 359

30

25

20
q = [mg/ml]

15

10

8
6 0.0
0.1
C

4
=

0.2
[m

0.3
g/

2
[M ]
m

0.4 I=
l]

0 0.5
0.6

Figure 22.3 Relationship among protein binding capacity, protein concentration, and
salt concentration in a protein solution; I is the ionic strength of the loading buffer, C is
the protein concentration in the mobile phase in equilibrium, and q is the concentration
of protein in the stationary phase.

then the protein solution is added. The supernatant is tested for the protein
and by addition of salt the protein is eluted from the column and it is again
tested in the supernatant. The advantage of microtiter plates are the multiple
experiments which can be run in parallel. Separation of liquid and solid
phase can be easily achieved in a centrifuge. Currently, chromatography
manufacturers also provide ready-to-use microtiter plates.
Several manufacturers have also designed small-scale, prepacked col-
umns, which can be operated by a syringe or a laboratory robot. With
these tiny columns it is easy to optimize binding conditions without wasting
precious proteins. Optimization of ion exchangers is very material consum-
ing, since the capacity is extremely high. These small columns such as
MediaScout MiniColumns from Atoll (Weingarten, Germany) are of
5 mm internal diameter and up to 10 mm bed height (Fig. 22.5). They
contain each up to 0.2 ml of the chromatography adsorbent and are
prepacked to defined density, corrected for each individual resin material,
in order to replicate conditions in larger columns.
360 Alois Jungbauer and Rainer Hahn

12

9
Relative abundance (%)

3 4 5 6 7 8 9 10 11
pl

Figure 22.4 Distribution of isoelectric points of proteins according to G. Righetti 13.

Figure 22.5 Robotic applications of prepacked minicolumns for operation with a

positive displacement pipette and columns situated at frame equivalent to microtiter
plates, reproduced by kind permission of Atoll (Weinheim, Germany).
Ion-Exchange Chromatography 361

In several steps, the different solutions are applied. Flow is controlled by

a positive displacement pipette. The applied volume varies from ml to ml
range depending on the type of pipette used.

5. Elution Conditions
Elution of a bound protein can be effected by four modes:
1. Linear salt gradient
2. Step salt gradient
3. pH gradient
4. Displacement development.
Sometime combination between pH gradient and salt is used, but this is
difficult to optimize and robust conditions often are not achieved. Linear
salt and pH gradients are preferred for high-resolution separation, while step
salt gradients are preferred for concentration of proteins. Displacement
chromatography is in principle suited for separation of related species, but
due to the more complicated optimization, is less used in general.
The retention window of proteins is very small; at low salt concentration
proteins are tightly bound and then with increasing salt concentration not
retained at all; compare (Fig. 22.1) the relationship between k0 and salt
concentration in the mobile phase. So in ion-exchange chromatography
linear gradients and even more complex segmented gradients are used.
The resolution Rs
tA  tB
Rs ¼ 2 ð22:2Þ
WA þ WB

with the retention time of two different compounds (tA, tB) and the peak
width at the base (WA, WB) depends on the slope of the gradient b. Where
b is defined as
CM  CM
0
b¼ ð22:3Þ
tG
with C 0M and CM as the start and final salt concentration of the buffer, and
tG the gradient time. The retention of a compound depends in a complex
manner on the normalized gradient slope (g). A detailed discussion would
exceed the scope of this chapter. A detailed analysis can be found in the
book published by Yamamoto et al. (1988):

bL bV0
g¼ ¼ ; ð22:4Þ
v Q
362 Alois Jungbauer and Rainer Hahn

L is the length of the column and v is the interstitial velocity, V0 the void
volume, and Q is the flow rate.
Resolution is increased with the normalized gradients (g). From
Eq. (22.4) one can infer three possibilities to increase resolution.
1. Increasing the gradient length
2. Increasing the column length/volume
3. Decreasing the interstitial velocity/flow rate.
All three parameters can be varied at once. Equation (22.4) also helps to
transfer a successful elution condition to a larger column. The same resolu-
tion is obtained when g is kept constant.
The resolutions in these cases are 5.00 and 4.97 min, respectively, with a
difference ion scale by a factor of 7 (Fig. 22.6).
As a consequence of the interrelation between protein retention and g, it
has been observed that the peak salt concentration at shallower gradients is
lower compared with steeper gradients. So the protein does not always elute
at the same salt concentration. Resolution depends on g in a reciprocal
manner.
rffiffiffiffiffiffiffiffiffiffiffiffiffi
1 L
Rs / ð22:5Þ
g HETP

0.014
CV = 7.99 ml
0.012 CV = 55.61 ml

0.010
UV 280 nm

0.008

0.006

0.004

0.002

0.000
0 500 1000 1500 2000 2500 3000
Time [s]

Figure 22.6 Example of a a-lactoglobulin isoform separation on Source30Q (GE

Healthcare, Uppsala, Sweden). The flow rate, gradient length, and load are scaled to
the column volume thus consequently g was kept constant. The flow rate is 30 CV/h.
Column dimensions are: 1  10 and 2  17.7 cm. Reproduced by kind permission of
Jorgen Mollerup.
Ion-Exchange Chromatography 363

With a shallower gradient (low g) resolution increases. Resolution can

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
also be influenced by the particlepsize.
ffiffiffiffiffi This is reflected by the term
L=HETP which is equivalent to N , with N the number of theoretical
plates, for protein chromatography being a mass transfer limited process.
In this case for well-packed columns, HETP is proportional to the velocity
and the square of the particle diameter. So with reduction of the particle
diameter the resolution can be substantially improved. Once nonporous
material is used then HETP is only proportional to the diameter not to the
square of the diameter of the particle. Thus, materials for ultrahigh
resolution have particle diameters below 5 mm. They are manufactured as
nonporous particles.
Peak volume also increases with the length of the gradient. A step
gradient can be considered as a very steep linear gradient. In practice, a
step never can be generated. So in a step gradient the protein elutes at lower
salt concentration compared to a linear gradient.

6. Operation of Ion-Exchange Columns

Here some general rules are given on how to operate an ion-exchange
column. Several dedicated steps are required to ensure a stable operation
with an ion exchanger. These steps are
1. Loading with salt
2. Equilibration
3. Loading of protein
4. Washout of unbound material
5. Elution
6. Regeneration
7. (Sanitization)
General conditions are listed in Table 22.2.
The loading with salt is crucial since the charged ligands must be entirely
saturated with counter ions. This is usually done with a buffer of a molarity
between 10 and 100 mM supplemented with 1 M NaCl. Less frequently
KCl is used. Bivalent ions should be avoided. A least one column volume
of buffer should be prepared in order to saturate the column. Further
optimization needs an on- or off-line conductivity probe.
Equilibration is performed with the buffer which is suited for protein
binding. pH, buffer, salt, and molarity depend on the individual conditions.
Very rough guidance for selection of the appropriate buffers is given in the
section on binding conditions. For this purpose, prepare at least 10 column
volumes. The large volume for equilibration is required to get a stable pH.
Since the different ions are migrating differently through the column, the
364 Alois Jungbauer and Rainer Hahn

Table 22.2 Overview of operation conditions for ion-exchange chromatography of

proteins

Step Molarity of buffer Buffer volume

Loading with salt 1 M and higher At least 1 CV
Equilibration 10–100 mM Up to 10 CV
Loading of Similar to equilibration Depends on dynamic
protein buffer binding capacity
Washout of Equilibration buffer with At least 1 CV
unbound eventually supplements
material
Elution
Linear gradient 0–500 mM salt 10 CV
Step gradient Between 150 and 500 mM 1 CV
salt
Linear pH Below 50 mM
gradient
Induced pH 10 mM and lower
gradient
Regeneration High salt concentration 1 M 1 CV
Sanitization At least 1 h residence
time

counter ion is more retained than the species with the same charge. Thus a
wave with different ions is traveling through the column. To maintain
electroneutrality, the pH changes. A detailed explanation of this phenome-
non can be found at Pabst and Carta (2007). Especially for weak ion-
exchange ligands, a larger volume is required to reach the actual pH.
Loading of the protein solution should be performed at the same pH and
conductivity as the equilibration buffer. Often this is not possible; especially
the composition of the ions in a crude feedstock and the equilibration buffer
may be completely different. Again as a consequence, a pH transition will be
observed during loading. When the protein solution is desalted by size
exclusion chromatography, dialysis, or diafiltration, the same ion composi-
tion can be obtained leading to most robust loading conditions.
After loading the protein solution, the unbound material is washed out.
A first try is always the equilibration buffer. To obtain a higher purity in the
eluate, the wash procedure may consist of several steps; for a tightly bound
protein of interest, washing with equilibration buffer with salt added is often
used. A lot of protocols exist and it seems very protein specific which
compounds are added to a washing buffers. These may be detergents,
urea, sugars, ethylene glycol, etc.
Elution can be effected by a linear or step salt gradient or by displace-
ment development. It is best to start with a linear salt gradient. The gradient
Ion-Exchange Chromatography 365

volume should be about 10 column volumes. The two buffers for the
preparation of the gradient are in the simplest case the equilibration buffer
(Buffer A) and the buffer used for loading with counterions (Buffer B).
In most cases, the second half of the gradient is not required, since most
proteins elute from a conventional ion exchanger in the range between
150 and 500 nM salt. So gradients to maximal 50% B are often sufficient.
The step gradient can also be generated from Buffers A and B. It should be
taken into consideration that an ideal step never can be achieved in a
column. First, due to retention of salt and the dead volume in HPLC
often denoted as dwell time and secondly, due to the washing-out char-
acteristics of a buffer mixer only a sigmoidal-shaped curve will be obtained
(Kaltenbrunner and Jungbauer, 1997). According to the selection of the salt
concentration in the step elution, the protein will migrate with the salt front or
travel behind the salt front. Yamamoto et al. (1988) named this behavior type I
or II elution. When a concentrated eluate is washed then salt concentration
must be selected for type I elution. This is a salt concentration somewhat higher
than the corresponding peak salt concentration in linear gradient elution. For
the type I elution less than one column volume buffer is required.
Elution by pH gradient is more difficult to optimize. So the best start is
to find a condition according to the pI of the protein where the protein does
not bind to the ion exchanger. This is below the pI for an anion exchanger
and above the pI for a cation exchanger. So pH has to be decreased or raised.
Keep in mind that the ion exchanger behaves like an acid or a base and the
pH gradient produced follows a titration curve. It is extremely difficult to
produce a linear pH gradient. This can be done by reacting boronate buffers
with diols such as mannitol (Kaltenbrunner et al., 1993). The fourth possi-
bility is the elution with induced pH gradients. When working with buffer
of low molarity in the range of 10 mM, the application of salt induces a pH
gradient due to the differential migration of the different salt species. Such
an induced pH gradient can be used for high resolution and additional for
focusing of the protein peak (Pabst et al., 2008a,b). In this case also, several
column volumes of buffer are required.
For regeneration of the column, the optimal condition is to use the same
buffer as for loading with salt. This is only possible when a relatively clean
feedstock is processed. Then it is possible to combine two steps into one.
Especially when lipids, endotoxins, and other sticky compounds are loaded
onto an ion-exchange column, then the regeneration with a caustic agent,
preferably up to 1 M NaOH, is recommended. The maximum concentra-
tion of NaOH depends on the resistance of the stationary phase. One
column volume is sufficient but for very tough impurities the column
should be soaked in NaOH. So after pumping one column volume over
the column, flow is stopped and NaOH is left several hours in the column.
The NaOH treatment also has some sanitization effects. It will definitely
reduce the bacterial count in a column, but will not sterilize a column.
366 Alois Jungbauer and Rainer Hahn

It is extremely difficult to sterilize a chromatography column ( Jungbauer

and Lettner, 1994; Jungbauer et al., 1994). So many dead-end holes are in a
packed column and the sanitization agent can only diffuse into these holes.
Oxidizing agents are often not compatible with the stationary phase and the
equipment. Smaller columns can be autoclaved and the connected to the
system under sterile conditions. Usually it is not necessary to sterilize a
chromatography column.
Storage of columns should be done after careful cleaning with regenera-
tion puffers, sanitization when necessary, and under conditions where the
column is fully saturated with counter ions. After saturation, the column can
be washed with buffer and stored, for instance in 20% ethanol. For labora-
tory use, it is also recommended to add bacteriocidal compounds to the
storage solution; NaN3 is often used.

7. Example: Separation of Complex

Protein Mixture
At typical example for ion-exchange chromatography of a complex
protein mixture is the separation of proteins from sweet whey. This has
some industrial relevance but may also serve as an inexpensive test system
for evaluation of ion exchangers. We have actually used this protein mixture
for comparison of materials of different manufactures of ion-exchange
adsorbents (Hahn et al., 1998).

7.1. Pretreatment of milk

Milk was centrifuged at 4420g at room temperature for 30 min for
delipidation. The pH of the skimmed milk was adjusted to 4.7 by slow
addition of 5 M HCl. After casein precipitation, the solution was stirred for
further 30 min to complete precipitation. Casein was removed by centrifu-
gation at 17,700g and 4  C for 30 min. The obtained whey was diluted with
distilled water until a conductivity of 2.7 mS/cm was obtained. The pH was
readjusted to 4.7 since pH shifted during dilution. Prior to chromatography,
the whey was filtered through a 0.45-mm Millipak-60 filter (Millipore,
Bedford, MA, USA).

7.2. Chromatography conditions

The cation exchange resins were packed into HR 10 and HR 16 columns
(GE Healthcare, Uppsala, Sweden). Column dimensions were 30/10 and
35/16 mm height  I.D., respectively. The buffers were prepared from
citric acid and the pH was adjusted with 1.0 N NaOH. The cation
Ion-Exchange Chromatography 367

exchangers were regenerated with 1 M NaCl in 20 mM citric acid and

equilibrated with 20 mM citric acid, pH 4.7, 2.7 mS/cm. After loading
and washing with equilibration buffer, the bound proteins were eluted with
sequential gradients with increasing NaCl concentration.
The chromatograms are shown in Fig. 22.7. The flow through and the
eluates were collected and then further analyzed by analytical size exclusion
chromatography and SDS–PAGE.
The composition of the flow through at beginning and ending of the
loading step is shown in Fig. 22.8.
Although cation exchangers are used, the selectivity is different. It is
difficult to predict the selectivity for the various materials.

8. Example: High-Resolution Separation

with a Monolithic Column
An impressive example for high-resolution separation using mono-
lithic columns is the separation of manganese peroxidase (MP) using an
anion-exchanger monolithic column (Podgornik and Podgornik, 2004).
The separation is completed in less than 3 min.

8.1. Mn peroxidase production

Phanerochaete chrysosporium MZKI B-223 (ATCC 24725) was grown in a
nitrogen-limited medium in agitated 500 ml Erlenmeyer flasks containing
100 ml of the growth medium at 32  C. The growth medium contained
1 mM Mn(II) favoring MnP production only. Two hundred and fifty
milliliters of P. chrysosporium growth medium was harvested after 4 days of
cultivation, when the highest MnP activities were recorded. The growth
medium filtrate was frozen overnight and rethawed to remove mucilagi-
nous polysaccharides by centrifugation.

8.2. Chromatography conditions

MnP isoenzymes were isolated using CIM disk monolithic column contain-
ing 0.34 ml CIM QA (quaternary amine) disk (BIA Separations, Ljubljana,
Slovenia). Experiments were performed on a gradient HPLC system built
with two pumps, an injection valve with a 100-ml stainless steel sample loop,
a variable wavelength monitor with a 10-mm optical path set to 280 or
409 nm and with a 10-ml volume flow cell, connected by means of
0.25 mm I.D. PEEK capillary tubes. Isoenzymes were separated with a
linear gradient of sodium acetate of different concentrations (from 10 mM
to 1 M ) and pH values (from 4 to 6), at a flow rate of 4 ml/min (Fig. 22.9).
 A
2.0 80
#3 #4
S-hyperD-F
1.5 # 5 60
#2

OD 280 nm

mS/cm
#1
1.0 40

0.5 20

0.0 0
0 50 100 150 700 750 800 850
B ml
2.0 80
S-sepharose FF #4
#3
1.5 60
#2
OD 280 nm

mS/cm
#1
1.0 40

0.5 20

0.0 00
0 50 100 150 200 650 700 750 800 850
C ml
2.0 80
#3
Fractogel EMD (S)
#4
1.5 #2 60
OD 280 nm

#1
mS/cm

1.0 40

0.5 20

0.0 0
0 50 100 150 200 650 700 750 800 850
D ml
2.0 80
Macro-prep high S #4
#2 # 5 60
1.5 #3
OD 280 nm

mS/cm

#1
1.0 40

0.5 20

0.0 0
0 50 100 150 500 550 600 650
ml

Figure 22.7 Purification of bovine whey proteins by cation-exchange chromatogra-

phy. Clarified whey was applied to four different cation-exchange columns
(3.531.6 cm), equilibrated with 20 mM citric acid, pH 4.7. The flow rate was
3.3 ml/min (200 cm/h). Elution of bound material was carried out with sequential
NaCl gradients. Unbound material and eluted fractions were characterized by analytical
size-exclusion chromatography (Fig. 22.8). (A) S-HyperD-F, (B) S-Sepharose FF,
(C) Fractogel EMD SO S, and (D) Macro-Prep High-S Support (, theoretical salt
gradient; —, UV absorbance at 280 nm), according to Hahn et al. (1998).
Ion-Exchange Chromatography 369

A
400 400
#1 #2
300 la la 300
AU 214 nm

AU 214 nm
200 lgG 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
B t (min) t (min)
400 400
#1 #2 la
300 300
AU 214 nm

AU 214 nm
la lg
200 200
lgG
100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
t (min) t (min)
C
400 400
#1 la #2 la
300 300
AU 214 nm

AU 214 nm
lg
200 200
lgG

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
t (min) t (min)
D
400 400
lg
#1 #2 la
300 300
AU 214 nm

AU 214 nm

200 lgG 200

100 100

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
t (min) t (min)

Figure 22.8 Analytical size-exclusion chromatography on Superdex 200 of unbound

material (flow through) when bovine whey was loaded onto cation-exchange sorbents.
Fractions of the flow through, as indicated in Fig. 22.1, were analyzed. (A) S-HyperD-F
1 and 2, (B) S-Sepharose FF 1 and 2, (C) Fractogel EMD SO S 1 and 2, and
(D) Macro-Prep High-S Support 1 and 2; lg, lactoglobulin; la, lactalbumin.
370 Alois Jungbauer and Rainer Hahn

160 6.5
MnP1 409 nm
140 280 nm
pH gradient
Relative absorbance (409 nm)

6
120

100

pH value (−)
5.5
80
MnP2 5
60

40
MnP4 4.5
3
20 5

0 4
0 0.5 1 1.5 2
Time (min)

Figure 22.9 Separation of Mn-peroxidase variants by combination of pH and concen-

tration gradient. Stationary phase: CIM QA disk monolithic column; mobile phase:
buffer A: 10 mM sodium acetate, pH 6; buffer B: 20 mM sodium acetate, pH 4;
gradient: 0–100% of buffer B in 100 s; sample: 100 ml of desalted protein solution;
flow rate: 4 ml/min; detection: UV at 280 nm, Vis at 409 nm, and on-line pH meter
[pH gradient?].

This very fast method allows separation of isoforms and two other forms
in less than 2 min. The method is suited for preparative and analytical
applications.

REFERENCES
Afeyan, N. B., Gordon, N. F., Mazsaroff, I., Varady, L., Fulton, S. P., Yang, Y. B., and
Regnier, F. E. (1990). Flow-through particles for the high-performance liquid chro-
matographic separation of biomolecules: Perfusion chromatography. J. Chromatogr. 519,
1–29.
Boardman, N. K., and Partridge, S. M. (1955). Separation of neutral proteins on ion-
exchange resins. Biochem. J. 59, 543–552.
Brooks, C. A., and Cramer, S. M. (1992). Steric mass-action ion exchange: Displacement
profiles and induced salt gradients. AIChE J. 38, 1969–1978.
Hahn, R., Schulz, P. M., Schaupp, C., and Jungbauer, A. (1998). Bovine whey fractionation
based on cation-exchange chromatography. J. Chromatogr. A 795, 277–287.
Janson, J.-C., and Ryden, L. (1998). Protein Purification: Principles, High Resolution
Methods, and Applications 2nd edn. Wiley-VCH, New York.
Jungbauer, A. (1993). Preparative chromatography of biomolecules. J. Chromatogr. 639,
3–16.
Jungbauer, A. (2005). Chromatographic media for bioseparation. J. Chromatogr. A 1065,
3–12.
Ion-Exchange Chromatography 371

Jungbauer, A., and Hahn, R. (2004). Monoliths for fast bioseparation and bioconversion and
their applications in biotechnology. J. Sep.. Sci. 27, 767–778.
Jungbauer, A., and Hahn, R. (2008). Polymethacrylate monoliths for preparative and
industrial separation of biomolecular assemblies. J. Chromatogr. A 1184, 62–79.
Jungbauer, A., and Lettner, H. (1994). Chemical disinfection of chromatographic resins,
Part 1: Preliminary studies and microbial kinetics. BioPharm 7, 46–56.
Jungbauer, A., Lettner, H., Gurrier, L., and Boschetti, E. (1994). Chemical sanitization in
process chromatography, Part 2: In situ treatment of packed columns and long-term
stability of resins. BioPharm 7, 37–42.
Kaltenbrunner, O., and Jungbauer, A. (1997). Simple model for buffer blending in ion-
exchange chromatography. J. Chromatogr. A 769, 37–48.
Kaltenbrunner, O., Tauer, C., Brunner, J., and Jungbauer, A. (1993). Isoprotein analysis by
ion exchange chromatography using a linear pH gradient combined with a salt gradient.
J. Chromatogr. 639, 41–49.
Kopaciewicz, W., Rounds, M. A., Fausnaugh, J., and Regnier, F. E. (1983). Retention
model for high-performance ion-exchange chromatography. J. Chromatogr. A 266, 3–21.
Mueller, E. (2005). Properties and characterization of high capacity resins for biochromato-
graphy. Chem. Eng. Technol. 28, 1295–1305.
Necina, R., Amatschek, K., and Jungbauer, A. (1998). Capture of human monoclonal
antibodies from cell culture supernatant by ion exchange media exhibiting high charge
density. Biotechnol. Bioeng. 60, 689–698.
Pabst, T. M., and Carta, G. (2007). pH transitions in cation exchange chromatographic
columns containing weak acid groups. J. Chromatogr. A 1142, 19–31.
Pabst, T. M., Carta, G., Ramasubramanyan, N., Hunter, A. K., Mensah, P., and
Gustafson, M. E. (2008a). Separation of protein charge variants with induced pH
gradients using anion exchange chromatographic columns. Biotechnol. Prog. 24,
1096–1106.
Pabst, T. M., Antos, D., Carta, G., Ramasubramanyan, N., and Hunter, A. K. (2008b).
Protein separations with induced pH gradients using cation-exchange chromatographic
columns containing weak acid groups. J. Chromatogr. A 1181, 83–94.
Podgornik, H., and Podgornik, A. (2004). Separation of manganese peroxidase isoenzymes
on strong anion-exchange monolithic column using pH-salt gradient. J. Chromatogr.
B Analyt. Technol. Biomed. Life Sci. 799, 343–347.
Tennikova, T. B., Belenkii, B. G., and Svec, F. (1990). High-performance membrane
chromatography. A novel method of protein separation. J. Liq. Chromatogr. 13, 63–70.
Yamamoto, S., Nakanishi, K., and Matsuno, R. (1988). Ion-exchange chromatography of
proteins Marcel Dekker, New York.
Yamamoto, S., Nakamura, M., Tarmann, C., and Jungbauer, A. (2007). Retention studies of
DNA on anion-exchange monolith chromatography: Binding site and elution behavior.
J. Chromatogr. A 1144, 155–169.
 C H A P T E R T W E N T Y- T H R E E

Gel Filtration1
Earle Stellwagen

Contents
1. Principle 373
2. Practice 374
2.1. Matrices 376
2.2. Sample preparation 380
2.3. Chromatographic solvents 380
2.4. Preliminary screening 380
2.5. Chromatography using conventional matrix 382
2.6. Scaling upward 383
2.7. Trouble shooting 384
2.8. Further information 385

Among the chromatographic techniques employed for protein purifi-

cation, gel filtration is unique in that fractionation is based on the relative
size of protein molecules. In contrast to conventional filtration, none of the
proteins is retained by a gel-filtration column. This feature is at once both
the strength and weakness of gel filtration; a strength because the function of
fragile proteins is not damaged by binding to a chromatographic support,
and a weakness because the absence of such binding limits the resolution of
the chromatography.

1. Principle
Gel filtration is performed using porous beads as the chromatographic
support. A column constructed from such beads will have two measurable
liquid volumes, the external volume, consisting of the liquid between the
beads, and the internal volume, consisting of the liquid within the pores of
the beads. Large molecules will equilibrate only with the external volume
while small molecules will equilibrate with both the external and internal
volumes. A mixture of proteins is applied in a discrete volume or zone at the
top of a gel-filtration column and allowed to percolate through the column.
Department of Biochemistry, University of lowa, lowa City, lowa
1
Reprinted from Methods in Enzymology, Volume 182 (Academic Press, 1990)

Methods in Enzymology, Volume 463 Copyright # 1990 by Academic Press

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63023-8 All rights reserved.

373
374 Earle Stellwagen

The large protein molecules are excluded from the internal volume and
therefore emerge first from the column while the smaller protein molecules,
which can access the internal volume, emerge later.
The dimensions important to gel filtration are the diameter of the pores
that access the internal volume and the hydrodynamic diameter of the protein
molecules. The latter is defined as the diameter of the spherical volume
created by a protein as it rapidly tumbles in solution. Proteins whose
hydrodynamic diameter is small relative to the average pore diameter of
the beads will access all of the internal volume and are described as being
included in the gel matrix. Proteins whose hydrodynamic diameter is compa-
rable to the average pore diameter will access some but not all of the internal
volume and are described as being fractionally excluded. Proteins whose
hydrodynamic diameter is large relative to the average pore diameter will be
unable to access the internal volume and are described as being excluded.
This conceptualization has led to the gradual renaming of gel filtration as
size-exclusion chromatography. The order of elution of a mixture of
proteins from a size-exclusion column will then be the inverse of their
hydrodynamic diameters. If all the proteins in a mixture are known, or can
be assumed to have the same shape, then the order of elution will be the
inverse of their molecular weights. This discussion will treat protein dimen-
sions in terms of molecular weight since common usage assumes that
protein mixtures contain only globular proteins. However, the reader
should bear in mind that hydrodynamic volume is the operative protein
dimension and that an asymmetrical protein will appear to elute with an
abnormally high-molecular weight compared with globular proteins of
similar molecular weight.

2. Practice
An elution profile obtained by size-exclusion chromatography is illu-
strated in Fig. 23.1A. Zero elution volume is defined as the entry of the
sample into the chromatographic support. The elution volume for the
excluded component is designated V0 for the void volume, which repre-
sents the volume external to the beads. The elution volume for the included
component is designated Vt for the total volume, which represents the sum
of the external volume and the internal volume within the beads. Elution
volumes intermediate between these values are designated Ve. A partition
coefficient, designated Kav, relating these values is given in Eq. (1):
Ve  V0
Kav ¼ ð1Þ
Vt  V0
Gel Filtration 375

A Volume (ml)
0 20 40 60 80

Vt
Ve
V0

Al

1 2 3

B
1.0
0.8
0.6
Kav

0.4
0.2
0.0

1 10 100 1000
Molecular weight (10−3)

Figure 23.1 Chromatographic performance of a size-exclusion matrix. Part (A)

illustrates a relatively simple elution profile. The ordinate represents concentration
expressed as spectral absorbance at some fixed wavelength, l, and the abscissa repre-
sents effluent volume subsequent to the application of the sample into the column.
If the effluent flow rate is constant then the abscissa could be expressed in time.
Component 1 is excluded from the matrix and its elution position is denoted as
V0. Component 2 is partially excluded and its elution position is denoted as Ve.
Component 3 is included and its elution position is denoted as Vt. The assignment of
a component to an elution position is established by application of each component
individually to the column. Part (B) illustrates the sigmoidal dependence of the partition
coefficient Kav as defined in Eq. (1) on the logarithm of the molecular weight of a series
of components having the same shape.

A semilogarithmic plot of the dependence of the partition coefficient on

molecular weight is illustrated in Fig. 23.1B. The separation of proteins
based on molecular weight will be greatest in the central linear region of this
sigmoidal relationship, spanning Kav values between 0.2 and 0.8. This span
is described as the fractionation range of a size-exclusion matrix.
The steeper the slope of the sigmoidal relationship in the fractionation
range the greater the resolving power of a matrix. Accordingly, the best
separation among proteins having similar molecular weights will be
achieved using a matrix with a narrow fractionation range.
Fewer than 10 proteins can be resolved from one another in the
effluent from any size-exclusion column. This relatively low resolution
376 Earle Stellwagen

occurs because none of the proteins is retained by the column during

chromatography and because nonideal flow occurs around the beads.
Accordingly, prospects for a significant enhancement in purification
(-fold) by size-exclusion chromatography are most promising if the desired
protein has a molecular weight either considerably larger or smaller than
that of the majority of proteins in a mixture. Since this will generally not
be the case, an investigator can anticipate only a modest enhancement in
purification (-fold). Accordingly, it is wise to perform size-exclusion
chromatography relatively late in a purification procedure when the num-
bers of other proteins are small and when the preceding step has fractio-
nated the protein mixture on the basis of a completely different property.
For example, pooled fractions obtained from ion-exchange chromatogra-
phy will likely contain a mixture of proteins, each having about the same
net charge but a range of molecular weights.

2.1. Matrices
The properties of some conventional and high-performance size-exclusion
matrices are given in Tables 23.1–23.4. It should be noted that suppliers use
a variety of terms and abbreviations to index these products in their catalogs,
including gel-filtration chromatography (GFC), gel-permeation chroma-
tography (GPC), and size-exclusion chromatography (SEC).
The conventional matrices are distinguished by their relative economy
and slow flow rates. These matrices are available in bulk, requiring an
investigator to pour columns of any desired dimensions to accommodate
the volume of the sample to be chromatographed. The flow rates normally
used for chromatography are obtained by multiplying the linear flow rate
listed in Table 23.2 by the cross-sectional area of the column in cm2 to yield
the flow rate in mm/h. A column can be packed with a flow rate approxi-
mately five times that used during chromatography.
The high-performance matrices are distinguished by their convenience,
rapid flow rates, and expense. These matrices are usually purchased as
poured columns which are attached to an existent high-performance chro-
matograph available to the investigator. The smaller analytical columns,
about 8  300 mm, are normally loaded with not more than a few milli-
grams of protein and operated at a flow rate of about 1 ml/min. The larger
preparative columns generally contain beads having a diameter of 30 mm.
The approximately 20  300 mm columns can be loaded with between
10 and 100 mg of protein and can be operated at a flow rate of about
5 ml/min while the very large columns can be loaded with up to 2 g of
protein and be operated at a flow rate of up to 30 ml/min.
Table 23.1 Matrix parameters

Stability
Name Supplier Chemistrya Form supplied pH Temperature ( C) Bead diameter (mm)
Conventional
BioGel A Bio-Rad AG Suspension 4–13 1–30 40–300b
BioGel P Bio-Rad PA Powder 2–10 1–80 40–30b
Sephacryl HR Pharmacia DX Suspension 2–13 1–100 25–75
Sephadex G Pharmacia DX/PA Powder 2–10 1–100 20–300b
Sepharose Pharmacia AG Suspension 4–10 1–40 45–200b
Ultrogel A IBF AG Suspension 3–10 2–36 60–140
Ultrogel AcA IBF AG/PA Suspension 3–10 2–36 60–140
High performance
Protein Pak Waters S Packed column 2–8 1–90 10
Shodex Showa Denko S Packed column 3–7.5 10–45 9
Superose Pharmacia AG Packed column 1–14 4–40 10–13
SynChropak SynChrom S Packed column 2–7 1–60 5–10
TSK-SW Toyo-Soda S Packed column 3–7.5 1–45 10–13
Zorbax DuPont S Packed column 3–8.5 1–100 4–6
a
The following symbols are used to denote the chemical nature of the matrix: AG, cross-linked agarose; PA, cross-linked polyacrylamide; DX, cross-linked dextran;
DX/DA, copolymer of allyl dextran and bisacrylamide; AG/PA, mixture of agarose and polyacrylamide; S, bonded silica.
b
Individual matrices have narrower ranges.
Table 23.2 Powdered matrix parameters

Hydration time (h)

Name Code Fractionation rangea (kDa) Swollen volume (ml/g) 20  C 90  C Linear flowa,b (cm/h)
BioGel P-60 3–60 14 4 1 5
P-100 5–100 15 4 1 5
P-200 30–200 29 4 1 4
P-300 60–400 36 4 1 3
Sephadex G-50 2–30 10 3 1 5
G-100 4–150 18 72 5 5
G-150 5–300 25 72 5 3
G-200 5–600 35 72 5 2
a
The values listed are for beads of a medium mesh size.
b
The linear flow indicated is appropriate for moderately high-resolution chromatography. The volume flow in mm/h is obtained by multiplying the linear flow by the
cross-sectional area of a column in cm2.
Gel Filtration 379

Table 23.3 Suspended matrix parameters

Name Code Fractionation range (kDa) Linear flowa (cm/h)

BioGel A-0.5m <10–500 18
A-1.5m 10–1500 18
A-5m 10–5000 18
Sephacryl S-200 HR 5–600 15
S-300 HR 10–1500 15
S-400 HR 20–8000 15
Sepharose CL-6B 10–4000 18
Ultrogel A6 25–2400 5
A4 55–9000 4
AcA 54 5–70 4.5
AcA 44 10–130 4.5
AcA 34 20–350 4
AcA 22 100–1200 2.5
a
The linear flow indicated is appropriate for moderately high-resolution chromatography. The volume
flow in mm/h is obtained by multiplying the linear flow by the column cross-sectional area in cm2.

Table 23.4 Packed column matrix parameters

Pore Diameter  length Fractionation

Name Code diameter (Å) (mm) range (kDa)
Protein Pak60 50 7.8  300 1–20
125 125 2–80
300 300 10–500
Shodex WS 802.5 150 8  300a 4–150
WS 803 300 10–700
WS804 500 10–2000
Superose 12 – 10  300 1–300
6 – 5–5000
SynChropak 60 60 7.8  300b 5–30
100 100 5–130
300 300 15–800
500 500 30–2000
TSK G2000SW 125 7.5  300c 5–60
G3000SW 250 1–300
G4000SW 500 5–1000
Zorbax GF-250 150 9.4  250 4–500
GF-450 300 5–900
a
Also 8  500 and 20  300 (mm).
b
Also 21.5  250 (mm).
c
Also 21.5  600 and 55  600 (mm).
380 Earle Stellwagen

2.2. Sample preparation

The sample should not have a protein concentration in excess of about
50 mg/ml and should be clarified by centrifugation, if necessary, in order to
prevent particulate matter from slowing the flow rate of the column. The
solvent for the protein sample is of little consequence since the protein will
advance ahead of the application solvent during chromatography.

2.3. Chromatographic solvents

The solvents used to flow through the column have wide latitude, subject
only to the pH and temperature constraints listed in Table 23.1. However,
the ionic strength of the chromatographic solvent should be at least 0.2 M to
minimize the binding of proteins to the matrix by electrostatic or by van der
Waals interactions. Most proteins are inherently stable at room temperature
and require only low temperatures in order to reduce the rate of peptide
hydrolysis catalyzed by any proteolytic enzymes present in the protein
sample. However, proteolysis becomes an increasing problem during puri-
fication as the desired protein becomes the more abundant substrate for the
proteases. In some cases, rather expensive proteolytic inhibitors or effectors
need be present in the chromatographic solvent in order to maintain the
function of the desired protein. Some economy can be realized by
equilibration with only one column volume of the solvent containing the
expensive component(s) prior to application of the sample, since the sample
advances into the column solvent during chromatography. The solvent
following the sample application need not contain the expensive
component(s).
Columns poured in glass cylinders should be equilibrated with a simple
solvent, such as 0.1 M NaCl containing about 0.02% sodium azide, to
prevent the growth of microorganisms. Methanol is the preferred storage
solvent for columns poured in stainless steel cylinders in order to avoid the
corrosion accelerated by the continued presence of salt solutions.

2.4. Preliminary screening

In order to optimize the purification (-fold) achieved by size-exclusion
chromatography, it is necessary to use a matrix which will best resolve the
desired protein from the remaining proteins. Accordingly, a preliminary
screening is useful to estimate the molecular weight of the desired protein
and the molecular weights over which the remaining proteins are distributed.
The elements needed for screening in addition to a protein sample include a
size-exclusion column, a fraction collector, an assay for total protein, an assay
for the desired protein, and a molecular weight calibration mixture. The assay
for total protein can either be ultraviolet absorbance or a colorimetric
Gel Filtration 381

procedure (see Chapter 6 by Stoscheck, Vol. 182). A sufficient concentration

of the sample must be applied to the column so that the function of the desired
protein can be measured with confidence in the eluate fractions. It should be
anticipated that the concentration of the desired protein will be diluted at
least an order of magnitude by the chromatography.
Molecular weight calibration mixtures, often termed gel-filtration stan-
dards, can be purchased from several suppliers, including Bio-Rad Labora-
tories (Richmond, CA), Pharmacia LKB Biotechnology (Piscataway, NJ),
and Sigma Chemical (St. Louis, MO). These calibration mixtures contain
several identified proteins of known molecular weight as well as compo-
nents to establish V0 and Vt. Alternatively, an investigator can customize a
calibration mixture using purified components. Blue dextran and DNA
restriction fragments are frequently used to determine V0. It is important
not to use a small aromatic or heterocyclic compound to determine Vt since
such molecules are particularly prone to reversible adsorption by
size-exclusion chromatographic matrices.
If a high-performance size-exclusion analytical column and chromato-
graph is available, the screening is both rapid and simple. The column used
for screening should have a broad fractionation range. A guard column
should be placed in front of the analytical column to retain any particulate
material which has escaped notice. A protein sample containing a minimal
volume appropriate for analysis of the desired protein in the column effluent
should be injected. The effluent should be monitored for protein concen-
tration using an absorbance flow detector set either at the more sensitive
225 nm, if the solvent absorbance permits, or at the less sensitive 280 nm.
Effluent fractions should be collected and analyzed for the total protein, if a
flow absorbance detector is not available, and for the desired protein.
Finally, a gel-filtration standard should be injected into the column and
the effluent monitored again at the same wavelength. Comparison of the
elution profile for the gel-filtration standard with the profiles for the total
protein and the desired protein in the sample should facilitate selection of
a matrix that will optimize the purification (-fold) achievable by size-
exclusion chromatography.
If a high-performance analytical column is not available, then the
screening must be done with a conventional matrix having a broad
fractionation range. It is likely that the matrix selected will have to be
poured into a column. Instructions for pouring a column using a conven-
tional matrix are detailed below. Again, a conventional matrix that can
optimize the purification (-fold) obtained by size-exclusion chromatography
can be selected from the screening based on the elution profiles obtained
for the gel-filtration standard and for the total protein and the desired
protein in the sample.
382 Earle Stellwagen

2.5. Chromatography using conventional matrix

The volume of a conventional matrix used for protein purification should
be 30–100 times the volume of the sample to be fractionated. The amount
of matrix required to form the column is suspended in the chromatographic
solvent and brought to the temperature at which chromatography will be
performed. The volume of the suspension should be no more than twice the
volume of the column to be made. Fine particles should be removed by
gently swirling the suspension and the supernatant removed by suction
after about 90% of the beads have settled. Finally, the suspension should
be placed under negative pressure to reduce the volume of dissolved air.
A filter flask and a laboratory aspirator are useful for this purpose.
If the matrix is supplied as a dry powder, the matrix should be swollen
in the chromatographic solvent prior to removal of the fine particles. The
matrix may be swollen at either ambient temperature or at 100  C,
depending upon the time available to the investigator. As shown in
Table 23.2, swelling of a matrix proceeds much faster at 100  C without
damage to the matrix.
The chromatographic column should be made in a glass or transparent
plastic cylinder of either commercial design or laboratory improvisation. The
ratio of the length of the cylinder to its diameter may vary from 20 to 100.
When improvising, elements of the following procedure can be used. The
bottom of the column can be formed from a rubber stopper containing a short
length of a thick-walled capillary tube positioned flush with the narrow end
of the stopper. The cylinder is oriented vertically and clamped securely in the
location in which the chromatography will be performed. The stopper is
inserted into the bottom end of the cylinder. A short length of flexible tubing
is attached to the protruding glass tube and a clamping device attached to the
tubing to control the liquid flow through the cylinder. A nylon or Teflon
mesh is placed inside the cylinder and pushed to the bottom to fit snugly
against the stopper. The clamp is closed and the cylinder filled with the matrix
suspension. The excess suspension is placed in a vessel with a bottom exit and
stopcock, such as a separatory funnel, and the exit attached to the top of the
cylinder with a length of flexible tubing and a one-hole stopper containing a
short length of glass tubing. This assembled apparatus should be airtight
between the surface of the excess suspension in the separatory funnel and the
flexible tubing extending from the bottom of the cylinder. The flow rate is
controlled by the height of the separatory funnel relative to the column. The
column can be packed using a flow rate about five times greater than that listed
in Table 23.2. Once the desired column height is packed, the clamp and
stopcock are closed, the excess matrix suspension removed, and some chro-
matographic solvent passed through the column using the separatory funnel as
the reservoir. A pool of solvent several centimeters in height should be
continuously maintained at the top of the column to buffer the impact of the
Gel Filtration 383

chromatographic solvent as it enters the cylinder so as not to disturb the packing

at the top of the column. The packed column should never be allowed to run
dry, as it will produce channeling within the column which will severely
perturb protein resolution.
To apply a sample to the column, the stopcock should be closed, the
stopper at the top of the cylinder removed, and the solvent pooled above
the column drained through the column until the solvent just dips below
the top of the packed column. The clamp is then closed and the sample or
standard solution added carefully to minimally disturb the packing at the top
of the column. The clamp is then opened and the sample solution allowed
to enter the column until it just dips below the top of the column. The
clamp is then closed and a small amount of chromatographic solvent added
with minimal disturbance to the packing at the top of the column. This
solvent is then admitted to the column, the clamp again closed, and more
chromatographic solvent added to the column to form a pool of desired
height. A supply of chromatographic solvent is placed in the separatory
funnel and connected by an air-tight seal to the top of the cylinder with the
flexible tubing. The height of the separatory funnel is then adjusted to
achieve and maintain the desired flow rate.
The absorbance of the column effluent can be continuously monitored
at a desired wavelength using a flow monitor. It is important that the tubing
at the bottom of the column and the flow optical cell in the monitor have a
small diameter to prevent convective mixing of the liquid emerging from
the column. It is also important that a flexible tubing be used which does not
contribute ultraviolet-absorbing material to the chromatography solvent.
Alternatively, the column effluent can be directed to a fraction collector and
the fractions assayed for both total protein and desired protein. A drop
counter is ideal for this purpose.

2.6. Scaling upward

Size-exclusion chromatography using conventional matrices can be easily
scaled upward by increasing the volume of the column appropriate to the
volume of the sample to be fractionated. Very large sample volumes may be
best handled with repetitive chromatography as opposed to construction of
columns of monumental dimensions. Semipreparative and preparative scale
high-performance columns are available as indicated in Table 23.4 and some
suppliers will provide bulk material for packing by the investigator.
Although these larger high-performance columns can be quite expensive
it should be remembered that they represent a considerable saving in
investigator time and that the investment can be amortized over many
different uses.
384 Earle Stellwagen

2.7. Trouble shooting

2.7.1. Poor resolution
This is a common lament because size-exclusion chromatography has an
inherent low resolution. Nonetheless, changes in some operational
parameters may improve resolution. First, since flow rate and resolution
are inversely related, decreasing the flow rate may improve the resolution.
Second, use of a bead size having a smaller diameter should improve
resolution. Third, use of a matrix having a narrower fractionation range
may be helpful.

2.7.2. Low flow rate

This usually results from plugging of the filters or the matrix with particulate
material in the samples. The column should be first washed by reverse flow
with a solubilization agent such as a nonionic or ionic detergent, a protein
denaturant such as urea or guanidinium hydrochloride, an organic solvent
such as methanol or, within the stability of the matrix, brief exposure to a
strong acid or base. If this does not succeed for a conventional matrix, then
the column should be disassembled, the individual components cleaned,
and the column repacked. If this does not succeed for a high-performance
matrix, either the column may be sent to Phenomenex or another supplier
for cleaning and repacking at a fee or the column may be simply replaced.
Laboratories which have facilities for repacking columns at pressure can
clean and repack high-performance columns themselves.

2.7.3. Skewed peaks

A primary cause is poor sample application. For a conventional column, the
quality of sample application can be observed by placing an inert colored
component in the protein sample such as blue dextran or potassium dichro-
mate. If the sample has an irregular appearance in the column it will likely
generate an asymmetrical peak in the elution profile. For a high-performance
column, the injector can be disassembled and cleaned. Tailing of peaks
generally results from adsorption of proteins to the matrix. This situation
can be improved by using a more potent lyotropic salt, such as sodium
perchlorate instead of sodium chloride, as the principal ionic component in
the chromatographic solvent. In the case of a high-performance column,
tailing may indicate the loss of the coating on the silica beads, a situation
requiring replacement of the column. Skewed peaks may also result from a
reversible equilibrium between different states of polymerization of the
protein. For example, hemoglobin can exhibit a dynamic equilibrium
between the di- and tetrameric forms of the protein. Since polymerization
involves a change in molecular weight, the matrix will favor dissociation
while chemical equilibrium will favor association. These opposing forces can
result in the appearances of a skewed peak characteristic for a dynamic
Gel Filtration 385

exchange. Changes in the pH, temperature, or chemical composition of the

chromatographic solvent may shift the chemical equilibrium such that only
one polymeric form is significantly populated.

2.7.4. Disappearance of desired protein

This may occur for at least two reasons. The desired protein may be
moderately adsorbed to the column so that its elution occurs after Vt in a
very broad peak that is difficult to distinguish from noise in the baseline.
If this is the case, a protein solubilization agent such as a nonionic detergent
or a modest concentration of a protein denaturant should be added to the
chromatographic solvent. A second possibility involves the dissociation of a
functional protein complex into discrete proteins of different molecular
weight in which none of the dissociated proteins retains the function.
Mixing aliquots from different fractions should facilitate complexation of
the component proteins and restoration of the function.

2.8. Further information

Virtually all the suppliers of size-exclusion matrices and customized chro-
matographic columns have prepared detailed instructions regarding the use
of their products. These instructions are quite helpful and generally free of
charge.
 C H A P T E R T W E N T Y- F O U R

Protein Chromatography
on Hydroxyapatite Columns
Larry J. Cummings, Mark A. Snyder, and Kimberly Brisack

Contents
1. Introduction 388
2. Mechanisms 389
3. Chemical Characteristics 392
3.1. Cationic and anionic modification of HA surfaces 392
3.2. Metal adsorption 393
3.3. HA solubility 394
4. Purification Protocol Development 396
5. Packing Laboratory-Scale Columns 397
6. Process-Scale Column Packing 399
7. Applications 401
References 402

Abstract
The introduction of spherical forms of hydroxyapatite has enabled protein
scientists to separate and purify proteins multiple times with the same packed
column. Biopharmaceutical companies have driven single column applications
of complex samples to simpler samples obtained from upstream column purifi-
cation steps on affinity, ion exchange or hydrophobic interaction columns.
Multiple column purification permits higher protein loads to spherical forms
of hydroxyapatite and improved reduction in host cell protein, aggregates,
endotoxin, and DNA from recombinant proteins. Adsorption and desorption
mechanisms covering the multimodal properties of hydroxyapatite are dis-
cussed. The chemical interactions of hydroxyapatite surface with common
ions, metals, and phosphate species affect column lifetimes. Adsorbed hydrox-
onium ions from low ionic strength buffers are noted by a shift in effluent pH
during column equilibration. Hydroxonium ion desorption is observed by acidic
shifts in the column effluent with the magnitude and duration surprisingly
extreme. Buffering reagents with high buffering capacity reduce both the
magnitude and duration of the acidic shift. Column packing methods for the

Bio-Rad Laboratories, Inc., Hercules, California, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63024-X All rights reserved.

387
388 Larry J. Cummings et al.

robust spherical particles as well as the microcrystalline hydroxyapatite

particles are reviewed. Applications covering extracted proteins and recombi-
nant protein purification, especially monoclonal antibodies, with multiple
chromatography media in concert with hydroxyapatite are reviewed.

1. Introduction
The use of hydroxyapatite, a hydroxylated calcium phosphate mate-
rial, in columns for protein chromatography has increased substantially
between 1991 and 2009 primarily due its use for recombinant protein
purification. The methods for using hydroxyapatite (HA) follow those
discussed by Tiselius et al. (1956) and reviewed by Gorbunoff (1985).
Although microcrystalline HA is still manufactured in production-scale
quantities, it is generally used in batch mode operations rather than in
packed columns because the crystals are mechanically unstable in applica-
tions requiring multiple column cycles. Suppliers of HA developed
manufacturing processes to make spherical particles that allow multiple
column cycles. One process forms microcrystalline HA in agarose gel
particles, HA UltrogelÒ sorbent as described in French patent 2231422,
British patent 1468592, Swiss patent 597893, and German patent 2426501.
Another described by Thompson and Miles (1973) forms large spheroids,
approximately 50–600 mm in diameter utilizing microcrystalline HA and
a ceramic hardening process. The spheroidal ceramic particles were com-
mercialized in the early 1970s by BDH Ltd. The adsorption capacity for
protein on agarose–HA is similar if not lower than the microcrystalline HA.
The elution profile of bovine serum albumin is similar between the sphe-
roidal ceramic particles and microcrystalline HA. A granular, irregularly
shaped form of porous ceramic HA is produced by Clarkson Chromatogra-
phy. The most commonly used porous spherical ceramic HA for down-
stream processing of recombinant proteins (CHTTM Ceramic
Hydroxyapatite, Bio-Rad Laboratories). The spherical particles are manu-
factured by spray drying nanoparticles to obtain porous particles in 20, 40,
or 80 mm sizes. The particles are mechanically sized to obtain a narrow size
distribution. Mechanical stability and porosity is imparted to the sized
particles by sintering them at high temperatures. Packed beds of these
particles have 12–20 times the surface area of microcrystalline HA particles
as discussed by Kadoya et al. (1986) and Ichitsuka et al. (1992). The increase
in surface area results in increased protein adsorption to the commercialized
porous CHTTM Types I and II. Type I adsorbs greater than 25 mg of
lysozyme per dry gram of resin from a 10-mM phosphate buffer, pH 6.8
and Type II between 12 and 19 mg/g due to its lower surface area. This
relationship was confirmed by Zhu et al. (2009) studying protein adsorption
to porous and nonporous bioceramics.
Protein Chromatography on Hydroxyapatite Columns 389

The newer, robust forms of hydroxyapatite such as agarose–HA and

ceramic HA are used to manufacture production-scale quantities of recombi-
nant proteins. Large columns ranging from 20 to 600 l are used for many
column cycles and thus exposed to very large quantities of buffers prepared
from USP grade reagents and water for injection. These larger scale processes
revealed some new challenges concerning HA such as adsorption of metals
and excursions in effluent pH. The remainder of this chapter covers a review
of the mechanism of protein adsorption and desorption, a discussion of
hydroxyapatite chemical characteristics, a description of purification protocol
development, lab-scale and process-scale column preparation, and a brief
review of recent protein chromatography applications utilizing HA.

2. Mechanisms
The mechanism of protein adsorption and desorption to and from HA
surfaces has been reviewed periodically since 1971 (Bernardi, 1971;
Gorbunoff, 1990). A recent contribution cites (Kandori et al., 2004) earlier
mechanisms in explaining that acidic proteins bind through C (calcium)-
sites while basic proteins bind through P (phosphate)-sites. C- and P-sites
were previously described by Kawasaki et al. (1986) and later further
discussed by Gagnon (1996) with an emphasis on the importance of
C-sites in the purification of monoclonal antibodies; monoclonal antibody
binding occurs in part through calcium coordination complexes with
carboxyl clusters in the antibody. The mechanisms of HA–protein interac-
tions were simplified to the following description. Amino groups are
adsorbed to P-sites but repelled by C-sites. The situation is opposite and
more complex for carboxyl groups. Although amine binding to P-sites and
the initial attraction of carboxyl groups to C-sites are electrostatic, the
binding of carboxyl groups to C-sites involves formation of much stronger
chelation, coordination bonds than anion exchange. Phosphoryl groups on
proteins and other solutes interact more strongly with C-sites than do
carboxyl groups as discussed by Kawasaki (1991).
Further work has involved bone–protein interactions by Dong et al.
(2007) to study the mechanistic role of zero net charge –OH and –NH2
groups in protein for the surface of HA. Steered molecular dynamic simu-
lation demonstrated the strong interaction between the HA surface and
protein –OH and –NH2 groups of bone morphogenetic proteins (growth
factors for bone formation) through a water-bridged H-bond. Their obser-
vations were confirmed by Shen et al. (2008) using a fibronectin fragment
(FN-III10). These and similar investigations support the C- and P-site
model. Hydroxyapatite surfaces have been proposed to interact differentially
with hexahistidine tagged proteins and histidine-rich regions of
390 Larry J. Cummings et al.

immunoglobulin G (IgG) (Ng et al., 2007). In this case, calcium–histidine

interaction appears to be insignificant based on the coelution of hexahisti-
dine tagged and untagged fusion protein. The adsorption mechanism for
proteins remain consistent with those described previously by Gorbunoff
(1990) and has been advanced by the addition of the water-bridged H-bond
mechanism.
Some proteins do not adsorb to HA if the loading buffer contains
phosphate. Good’s buffers developed by Ferguson et al. (1980) are
frequently used under these circumstances. For example, serine hydroxy-
methyl transferase from Escherichia coli has been purified by Schirch et al.
(1985) to homogeneity by applying the anion-exchange fraction in 20 mM
N,N-bis(2-hydroxyethyl)-2-aminoethanesulfonic acid (BES), pH 7.0 to a
column of HA equilibrated with the loading buffer followed by elution
with a shallow linear phosphate gradient (Schirch et al., 1985).
Protein desorption is conducted by reversing P-site interactions, the two
types of C-site interactions and/or the H-bond. Linear and step gradients of
phosphate buffer are the most often used desorption agent. In addition,
Freitag and Breier (1995) noted that proteins retained by electrostatic
interaction of their net positive charge with negative net charge on the
HA surface are desorbed by anions added to the phosphate buffer or by
cations to desorb those retained through their net negative charge with the
positive net charge on the HA surface. Anions such as Cl , F , ClO4 ,
SCN , and phosphate have been discussed by Gorbunoff (1984a,b) and
Gorbunoff and Timasheff (1984) in series of publications. Fluoride ion was
used by Schlatterer et al. (2006) to desorb prostaglandin D synthase from a
column of ceramic hydroxyapatite. The enzyme was adsorbed from bovine
cerebrospinal fluid followed by sequential desorption with 150 and 375 mM
NaF using a linear gradient to 750 mM NaF in 10 mM sodium phosphate
buffer, pH 6.25.
Desorption is also highly influenced by buffer pH. According to the
study by Ogawa and Hiraide (1996), proteins desorbed more quickly at
lower phosphate concentration above pH 7 than below pH 7 and the
elution sequence was similar for Types I and II porous ceramic HA but
desorption of proteins from Type I required stronger solutions of sodium
phosphate buffer. Similarly, buffer pH also affects protein desorption from
porous ceramic fluoroapatite for the pH range 5.5–6.8. Table 24.1
compares the elution time in minutes of several proteins in this range
obtained on ceramic HA Type I and ceramic fluoroapatite using
1.1 cm  15 cm packed columns. The columns were packed using 0.2 M
phosphate buffer, pH 10 and equilibrated with 5 mM sodium phosphate at
pH 5.5, 6.0, or 6.8. The elution order of the proteins is not dependant solely
on the pI of the proteins. With some proteins desorption appears to depend
on their structure and carboxyl group content as explained by Gorbunoff
(1984a,b).
Table 24.1 Elution time in minutes of proteins on ceramic hydroxyapatite Type I (CHT I) and ceramic fluoroapatite (CFT)

CHT I CFT CHT I CFT CHT I CFT

Protein pI pH 6.8 pH 6.8 pH 6.0 pH 6.0 pH 5.5 pH 5.5
a-Lactalbumin 4.5 24.3 22.3 27.4 23.7 33.2 29.2
Transferrin 4.7 24.2 23.7 30.8 31.8 46.6 50.0
Bovine serum albumin 4.8 24.3 22.3 29.7 24.2 43.0 39.3
Ovalbumin 5.0 22.0 22.0 24.3 23.8 30.7 27.3
Carbonic anhydrase 5.3 24.3 22.6 29.1 27.6 43.0 39.8
Catalase 5.4 23.5 22.0 29.7 24.1 35.1 35.7
Conalbumin 6.8 32.6 33.7 38.7 37.9 46.4 46.1
Myoglobin 7.0 30.3 29.4 34.8 33.1 46.5 44.7
Ribonuclease a 9.7 30.8 31.4 34.1 35.9 40.5 42.5
a-Chymotrypsinogen 9.8 34.9 31.1 37.9 37.2 51.3 50.0
Lysozyme 10.5 31.6 30.7 34.9 34.9 41.8 42.1
Cytochrome c 10.6 45.0 43.4 49.9 49.2 63.1 60.6
Proteins were applied in 5 mM phosphate application buffer. The columns were rinsed at 1 ml/min for 10-min with application buffer then eluted with 50-min linear
gradient to 0.4 M sodium phosphate at pH 5.5, 6.0, or 6.8, respectively. The retention time of each protein excludes the 10-min rinse time with 5 mM phosphate
application buffer.
392 Larry J. Cummings et al.

3. Chemical Characteristics
3.1. Cationic and anionic modification of HA surfaces
Calcium and magnesium cations change the surface of HA through the
formation of a phosphate–Ca or Mg bridge. Although initially commented
upon over 20 years ago, Gorbunoff (1984a,b) and Gagnon et al. (2009) have
recently used Ca-modified ceramic HA surfaces to improve the purification
of Fab from Mab and Fc antibody components. These workers showed that
Ca-modified ceramic HA is restored to its native form after treatment with
20 column volumes of dilute phosphate buffer.
HA also has a high affinity for pyrophosphate (PPi) and polyphosphates
as explained by Krane and Glimcher (1962). PPi is a contaminant in
anhydrous salts of mono- and disodium phosphate and may be a contami-
nant in its potassium salt counter parts. There are advantages and disadvan-
tages to PPi in phosphate buffers relative to the adsorption and desorption of
proteins to the HA. Figure 24.1 shows the separation of bovine serum
albumin, ovalbumin, a-chymotrypsinogen A, and cytochrome c with

1.0 250

0.5 2 3 100
ODA280

mS/cm

0.0 0

0 min 30 min 60 min

Figure 24.1 Separation of protein mixture using phosphate buffers prepared from
anhydrous phosphate salts: (1) bovine serum albumin, (2) ovalbumin, (3) a-chymo-
trypsinogen, and (4) cytochrome c. Column: 10 mm  40 mm. Flow rate: 2.5 ml/min.
Temperature: 20  C. Program: sample inject, 5-min isocratic A; 20-min 0–80% linear
gradient to B; 10-min isocratic 80% B; 10-min isocratic 100% B; 15-min isocratic
A. Buffer A: 10 mM sodium phosphate, pH 6.8. Buffer B: 400 mM sodium phosphate,
pH 6.8. Reagents: sodium phosphate monobasic anhydrous and sodium phosphate
dibasic anhydrous. Dashed line indicates the relative conductivity of the column effluent.
Protein Chromatography on Hydroxyapatite Columns 393

phosphate buffers prepared from anhydrous sodium phosphate. Figure 24.2

shows the separation of these proteins when applied and desorbed with
sodium phosphate buffer prepared from dibasic sodium phosphate hepta
hydrate and monobasic sodium phosphate monohydrate. The column used
with sodium phosphate buffer prepared from anhydrous sodium phosphate
required four program cycles in order to diminish the effects of the PPi
(see Program: in Fig. 24.1). As mentioned previously, PPi may be used
advantageously in the chromatography of proteins on HA particularly for
the exclusion of otherwise weakly adsorbed proteins. Care must be taken,
however, to ensure that the levels of higher-order phosphates are kept the
same between all experiments to avoid artifacts.

3.2. Metal adsorption

Columns used multiple times can discolor due to the accumulation of
metals. The causes of discoloration are found in studies conducted by dental
scientists relative to caries in teeth, medical scientists relative to bone

1.0 250

1
0.5 2 100
ODA280

mS/cm
4

0.0 0

0 min 30 min 60min

Figure 24.2 Separation of protein mixture using phosphate buffers prepared from
hydrated phosphate salts: (1) bovine serum albumin, (2) ovalbumin, (3) a-chymotryp-
sinogen, and (4) cytochrome c. Column: 10 mm  40 mm. Flow rate: 2.5 ml/min.
Temperature: 20  C. Program: sample inject, 5-min isocratic A; 20-min 0–80% linear
gradient to B; 10-min isocratic 80% B; 10-min isocratic 100% B; 15-min isocratic
A. Buffer A: 10 mM sodium phosphate, pH 6.8. Buffer B: 400 mM sodium phosphate,
pH 6.8. Reagents: sodium phosphate monobasic monohydrate and sodium phosphate
dibasic, heptahydrate. Dashed line indicates the relative conductivity of the column
effluent.
394 Larry J. Cummings et al.

integrity and tissue fluids for artificial bone or cadaver implants, and material
scientists examining soil remediation and metals absorption.
According to Shepard et al. (2000), a column of ceramic HA discolored
after several recombinant protein purification cycles. The column of
ceramic HA was the third column in the process following cation and
anion exchange. They concluded that the metals that discolored the ceramic
HA came from various sources: process equipment, process reagents, pro-
cess water, and fermentation nutrients. However, the discoloration did not
impair the purification of their recombinant protein. Agronomists explain
the preferential absorption of multivalent metals to HA (Ma et al., 1994) and
show that HA has a high capacity for divalent heavy metals. Many of the
described metals were observed in the discolored ceramic HA column,
especially Fe, Al, Zn, and Mn.

3.3. HA solubility
Voids in the column headspace or the formation of channels can occur with
HA resulting from the massive amounts of buffer applied to the column.
The solubility constant for HA is 2.4  10 59 which is equivalent to about
4 ppm. Phosphate ion and basic pH suppress the solubility. However,
adsorbing protein to HA from very dilute phosphate buffer, 1 mM
(pH 6.8), in unbuffered alkali metal solutions, 1 mM NaCl or KCl or in
zwitterionic buffers is common. Some acidic proteins absorb only in water
as determined by Schirch et al. (1985). The low to unbuffered loading
conditions reduce the control over the surface pH of the HA as explained
by Scopes (1993). Some researchers (Schroder et al., 2003) determined that
cobuffers such as 4-morpholinepropanesulfonic acid (MES) could be used
to permit phosphate elution while maintaining the pH of the loading and
elution buffers. However, HA solubility is a risk even with phosphate
suppression. For example, one column was equilibrated with 20 mM
sodium phosphate, pH 6.5, loaded with protein applied in the equilibration
buffer then sequentially eluted with increased sodium chloride buffered
with 20 mM sodium phosphate. In this case, voids formed and the column
displayed increasing pressure from cycle to cycle. The voids and increased
pressure were caused by chemical damage to the HA largely due to the
liberation of hydroxonium ion that accumulated on the surface of the HA.
Figure 24.3 shows the pH profile of the column effluent and the zones
where hydroxonium ion adsorbed (A and B) and desorbed (C and D).
Harding et al. (2005) discuss the adsorption and desorption of hydro-
xonium ion as a function of the zero point charge of HA. The results of
these research (Skartsila and Spanos, 2007) show the relationship of zero
point charge, pH, hydroxonium adsorption, desorption, and calcium
content. The zero point charge determined by Skartsila and Spanos
differs, 7.3  0.1 and 6.5  0.2. These research contributions explain the
Protein Chromatography on Hydroxyapatite Columns 395

8.50
400
8.00

7.50 A B C D 300
7.00

mS/cm
200
pH

6.50

6.00
100
5.50

0
1 2 3 4 5 6 7 8 9 10

0 60 120 180
Min

Fig. 24.3 Effluent pH changes relative to buffer sequence. Column 11 mm  230 mm.
Flow rate: 7.0 ml/min. Temperature: 20  C. Program: (1) 3-min 20 mM sodium
phosphate, pH 6.5 to rinse 0.1 M NaOH from the column; (2) 17-min pH adjustment
with 500 mM sodium phosphate/0.1 M NaCl, pH 6.5; (3) 29-min 20 mM sodium
phosphate/low concentration NaCl, pH 6.5; (4) 29-min 20 mM sodium phosphate, pH
6.5; (5) 34-min 20 mM sodium phosphate/low concentration NaCl, pH 6.5; (6) 18-min
20 mM sodium phosphate/low concentration NaCl, pH 6.5; (7) 3-min 20 mM sodium
phosphate to rinse calcium ion from the column; (8) 29-min regeneration with 500 mM
sodium phosphate/0.1 M NaCl, pH 6.5; (9) 3-min 20 mM sodium phosphate to rinse
concentrated phosphate from the column; (10) 29-min sanitization with 0.5 N NaOH.
Points A and B indicate the change to a more basic effluent than the pH 6.5 input buffer.
Points C and D indicate the change to more acidic effluent than the pH 6.5 input buffer.
Dashed line indicates the relative conductivity of the column effluent. The solid line
indicates the pH of the column effluent.

differences between input buffer pH and deviations in the pH of column

effluent. It is reasonable to assume that changing the pH of the input buffers
or increasing their buffer capacity could reduce the intensity of the pH shift
and perhaps the duration of the shift. However, as noted previously, some
proteins do not adsorb to HA when the phosphate concentration is too
high. In some cases, even 5 mM phosphate prevents protein adsorption
as noted by these process development engineers (McCue et al., 2007).
Alternative buffering agents were used to preserve the integrity of the target
protein while allowing adsorption to HA. This was best accomplished
without any phosphate in the equilibration and load buffers; however, the
packed column failed after a few cycles. Studies showed that as little as 2 mM
phosphate increased the column life without compromising the purity of the
target protein. Phosphate in combination with MES or 3-(N-morpholino)
propanesulfonic acid (MOPS) and minor amounts of calcium ion was
396 Larry J. Cummings et al.

500

8.50
400
8.00

7.50 A B C D 300
7.00

mS/cm
pH

200
6.50

6.00
100
5.50

0
1 2 3 4 5 6 7 8 9 10

0 60 120 180
Min

Figure 24.4 Effluent pH changes relative to buffer sequence. The buffer sequence is
the same as in Fig. 24.3 except that buffers 1, 3, 4, and 5 contain 75 mM MES. Dashed
line indicates the relative conductivity of the column effluent. The solid line indicates
the pH of the column effluent.

sufficient to suppress the solubility of HA. pH shifts were not explicitly

mentioned by these development engineers (McCue et al., 2007).
Figure 24.4 shows the benefits of MES as a cobuffer in terms of minimizing
pH shifts and durations of the shift observed in Fig. 24.3. Calcium ion was
measured in the high sodium chloride fraction for the two elution buffers.
It was 36 ppm for the 0.02 M sodium phosphate buffered solution and 5 ppm
for the 75 mM MES/0.02 M phosphate buffered solution.

4. Purification Protocol Development

Most proteins are adsorbed to HA low ionic strength buffers with
phosphate at concentrations as low as 1 mM. Adsorbed proteins are often
eluted with higher concentrations of phosphate but also with NaCl or KCl
(salts). It is uncommon to see other types of desorbing agents for used for
these purposes.
If the protein sample contains a high amount of salt, dilute it sufficiently
so that it adsorbs to the HA but not below 1 mM in phosphate. A sufficient
concentration of an alternative buffer should be incorporated for pH con-
trol. Elute the column with a linear salt gradient to determine if the protein
desorbs. If retained then repeat the experiment using a linear phosphate
Protein Chromatography on Hydroxyapatite Columns 397

gradient as eluant. Test the protein for purity. Determine the concentration
of the salt or phosphate necessary to elute the protein and apply it as a
desorption step in the purification protocol. Compare the purity of the step
desorbed protein to the gradient desorbed protein. Make adjustment to the
desorption protocol if necessary. Desorb other bound proteins and
biological components from the column by rinsing with 3 column volumes
of phosphate-containing, salt-free, low ionic strength (LIS) buffer followed
by 3–5 column volumes of 0.5 M phosphate buffer, pH 7, 3 column
volumes LIS and 3–5 column volumes of 0.5–1 M NaOH. Prepare the
column for subsequent column cycles by rinsing with 3 column volumes
LIS, 3 column volumes of 0.5 M phosphate buffer at the pH of the
equilibration buffer then with 3 column volumes load buffer.

5. Packing Laboratory-Scale Columns

Packing protocols for laboratory-scale HA columns differ depending
on the source of HA. To pack HA Ultrogel sorbent, use a column with a
height/diameter ratio between 1 and 6 with diameters ranging from 1.6 to
5.0 cm. Prepare the sorbent for packing by mildly agitating the product
container by inverting it or by stirring the contents with a plastic paddle.
Transfer a suitable volume of slurry to a vacuum flask with a volume
capacity about twice the planned packed volume. Dilute the slurry with
about 40% additional water. Mix gently, then connect a vacuum pump to
remove dissolved gas. Mildly stir the degassed sorbent to obtain a homoge-
neous suspension. Pour it into the column in one continuous motion
minimizing entrapping air into the sorbent stream. Settle the suspension
for 5–10 min until a 1-cm clear supernatant layer is visible at the top of the
column. Insert the inlet adapter into the column, connect it to a liquid
chromatography then flow pack with water or equilibration buffer at
between 55 and 250 cm/h depending on the desired packed bed height;
55 cm/h for a 20-cm height bed, 250 cm/h for a 5-cm height packed bed.
Adjust the inlet adapter so that it contacts the surface of the packed bed
excluding any trapped air while lowering the adapter. If air accidently
intrudes the packed bed, the column requires repacking.
CHT is a spherical porous ceramic media. Suitable columns include
Millipore VantageTM L and Waters AP with variable height adapters with
pressure ratings of at least 3 bar. Calculate the weight of CHT required for
the packed column from the packed bed dimensions and the density of
CHT, 0.63 g/ml. For example, a 1.1-cm ID  20.0 cm height packed bed
has a volume of 19 ml or 12 g of CHT. Suspend the CHT in 3.5 volumes of
packing buffer (0.2 M sodium phosphate dibasic heptahydrate, pH 9–10) in
a suitably size beaker using a plastic spatula. Alternative packing solutions
398 Larry J. Cummings et al.

should be at least 0.1 M in ionic strength and contain at least 20 mM buffer

component that buffer in the range of pH 6.8–10. Allow the suspension to
settle to allow occluded air to separate from the particles. Attach a packing
extension tube to the column so that all of the suspended ceramic HA can
be dispensed in a single step. Pour packing buffer into the column/exten-
sion assembly to a height of 1–2 cm. Suspend the particles in the beaker with
the plastic spatula and pour it into the column/extension assembly. Rinse
any remaining ceramic HA into the column/extension assembly with
packing buffer. Open the column outlet and allow the column to pack by
gravity flow to constant height. Close the column outlet, remove the
extension tube then insert the inlet adapter and lower to just touch the
surface of the gravity packed ceramic HA. Connect the column to a liquid
chromatography, open the column outlet, and equilibrate with 3 column
volumes of packing buffer at 250 cm/h to flow pack the ceramic HA. Adjust
the adapter so that it just touches the surface of the flow packed ceramic HA.
If air accidently intrudes the bed, it is usually eliminated during column
conditioning steps. Condition the column by equilibrating with 3 column
volumes of equilibration buffer, 3 column volumes of 0.5–1.0 M NaOH, 3
column volumes of 0.4–0.5 M sodium phosphate, pH 6.5–7.2 then 3
column volumes equilibration buffer. Equilibration buffers are generally
low ionic strength buffers such as 2–20 mM phosphate with Good’s
cobuffers added when the phosphate concentration is 2–10 mM.
Microcrystalline hydroxyapatite is available from various companies.
Bio-GelÒ HTP and HT require preparation before packing them into a
column. Decant the supernatant from the HT then add an equal volume of
packing buffer. Suspend the HT by rotating the container then pour
approximately twice the volume of the suspension for the desired packed
volume it into a beaker. Settle the suspension for 30 min then decant the
supernatant which may contain fines. For HTP, dispense 1 g for each 2.5 ml
of desired packed volume. Add the powder to a beaker containing two to
six times the desired packed volume while mixing slowly with a plastic
stirring rod. Settle the suspension for 30 min then decant the supernatant
which may contain fines. Add packing buffer equal to the desired packed
volume to make a 50% (v/v) suspension. Gently suspend the HT or HTP
with a plastic stirring rod. Pour the suspension into the column using a
funnel or column extension to contain the total volume of the suspension.
Wait for 10 min then open the column outlet and pack the column by
gravity flow. Do not allow air to intrude into the packed bed. Close the
column outlet when approximately 2 cm of buffer remains above the
packed bed, remove the extension tube or funnel then insert the inlet
adapter and lower to eject air from the adapter inlet line. Continue lowering
until the adapter just touches the surface of the gravity packed bed. If using
DNA Grade Bio-Gel HTP, prepare it and pack it similarly to HTP. The
maximum flow rate for HT and HTP is 100 and 40 cm/h for the DNA
Protein Chromatography on Hydroxyapatite Columns 399

Table 24.2 Hydroxyapatite sources

Source Product name Particle size(s)

Bio-Rad Laboratories Bio-Gel HT Microcrystalline
Bio-Gel HTP Microcrystalline
DNA-Grade Microcrystalline
Bio-Gel HTP
CHT ceramic Type I 20, 40, and 80 mm
CHT ceramic Type II 20, 40, and 80 mm
Clarkson Ceramic 53–124 mm
Chromatography hydroxyapatite
Hypatite C Microcrystalline
Pall Corporation HA Ultrogel 60–80 mm,
microcrystalline in
agarose gel particles
Sigma-Aldrich Hydroxyapatite, Type I Microcrystalline
GFS Chemicals Calcium Microcrystalline
Hydroxyapatite
Spectrum Chemicals Calcium Hydroxyapatite Microcrystalline
Microcrystalline sizes of hydroxyapatite are not described. The product specification describes the
maximum flow rate under gravity flow where the maximum column height is 100 mm and the
hydrostatic head is 200 mm. All of the microcrystalline sizes are produced by the method of Tiselius
(Tiselius et al., 1956).

Grade. The preparation and packing instructions for calbiochem hydroxyl-

apatite fast flow and high resolution are similar to HTP and DNA Grade
HTP, respectively. Clarkson’s ceramic hydroxyapatite and Hypatite C
should be prepared and packed similarly to HTP and HT.
Suppliers of HA do not often list their entire HA product line.
Table 24.2 lists current suppliers of HA and ceramic HA.

6. Process-Scale Column Packing

Well-packed large-scale columns, in which the beds are homogeneous
and continuous from top to bottom, exhibit the best chromatographic
performance. Several methods exist for packing columns with CHT that
depend on the type of column and equipment used. Review the relevant
instruction manuals for columns, media transfer devices, and media packing
devices prior to packing the columns.
The maximum packed bed height of open columns such as EasyPackTM
(Bio-Rad Laboratories), BPGTM (GE Healthcare), and Moduline 2TM
(Millipore) cannot be greater than 50% of the height between the surfaces
400 Larry J. Cummings et al.

of the media retention plates (frits or nets). For example, if the distance is
54 cm, the maximum packed height is 27 cm. Calculate the volume of the
packed column. For each liter of packed bed volume, use 630 g of dry
powder and 1.79 l of packing buffer to prepare a 50% (v/v) slurry. With the
column outlet closed, dispense the packing buffer to the column followed
by the dry powder. Agitate the CHT-buffer mixture with a plastic paddle to
hydrate the powder and blend the components to a homogenous slurry.
Reverse the direction of agitation to minimize the motion of the slurry.
Assemble the top adapter following the manufacturer’s instruction and
insert into the column tube. Wait 5 min to allow for a resin free zone
then lower the adapter allowing air to vent through the top process inlet and
to purge the adapter flow distributor and inlet line with packing buffer.
Flow pack at 200–300 cm/h with 2 column volumes of packing buffer.
Once the bed is consolidated, lower the adapter, leaving a headspace of
1–5 mm between the media retention plate and the top of the packed bed.
Do not lower the adapter into the packed bed to avoid irreversible damage
to the CHT particles.
Closed columns such as the InPlaceTM (Bio-Rad Laboratories) and Bio-
ProcessTM LPLC (GE Healthcare) columns are packed with externally
prepared slurries. As with open columns, the maximum packed bed height
cannot be greater than 50% of the height between the surfaces of the media
retention plates (frits). Calculate the volume of the packed column. For each
liter of packed bed volume, use 630 g of dry powder and 1.79 l of packing
buffer to prepare a 50% (v/v) slurry. Dispense the packing buffer to the media
slurry tank followed by the dry powder. Agitate the CHT-buffer mixture to
with a low-shear hydrofoil impeller (Type A3) to hydrate the powder and
blend the components to a homogenous slurry. Transfer all of the slurry to the
column using a media transfer device. Apply up flow at 50 cm/h to expel air
bubbles from the transferred slurry. Stop the flow and wait 5 min to allow a
resin free zone to form. Lower the adapter, allowing air to vent through the
top process inlet and to purge the adapter flow distributor and inlet line with
packing buffer. Axially, flow pack at 200–300 cm/h to consolidate the packed
bed. Continue lowering until the adapter media retention plate is 1–5 mm
above the surface of the packed bed. When packing stainless steel InPlace or
LPLC columns axial flow pack until the adapter reaches 2 cm above the target
height of the column then lower at 10 cm/h until the adapter sensor signals
contact with the top of the bed.
Process-scale columns utilizing microcrystalline HA is not achievable.
HA Ultrogel is utilized in shallow bed columns with large diameters.
Packing this type of column and media requires assistance from the column
and media suppliers.
The evaluation of packed process columns is often required by end users
per the recommendation of regulatory agencies. The United States Food
and Drug Administration and its counterparts in Europe, Canada, Japan,
Protein Chromatography on Hydroxyapatite Columns 401

and Asia issued guidelines for evaluating packed columns. Chromatography

media suppliers have issued simple methods for lab-scale and process-scale
columns but do not necessarily correlate the results between the two.
However, innovators (Teeters and Quiñones-Garcı́a, 2005) within the
biopharmaceutical companies have been resourceful and proposed resi-
dence time distribution (RTD) to track the condition of a packed column
during the column’s lifetime.

7. Applications
Phosphate gradient or step elution remains a dominant desorbing
strategy in recombinant purification protocols. Recombinant human cata-
lase expressed in Pichia pastoris was purified to 95% in a three-step process—
ammonium sulfate precipitation, anion-exchange chromatography, and
hydroxyapatite chromatography. The recombinant catalase was desorbed
from the HA column with a linear gradient of 0.05–0.3 M phosphate,
pH 6.8. Shi et al. (2007) obtained a 60% yield of the catalase that was
secreted into the culture medium. Numerous proteins purified by chroma-
tography on HA columns utilize phosphate linear or step gradients as
described by Hsieh et al. (2003), Stránská et al. (2007), Luellau et al.
(1998), and Nuss et al. (2008).
Clones of cDNA for human and two E. coli pyridoxal kinase genes were
used by di Salvo et al. (2004) to express pyridoxal kinases in E. coli. The
ammonium sulfate precipitated cell lysate pellet containing the enzyme was
dissolved with phosphate buffer then purified using three types of chroma-
tography media—hydrophobic interaction gel, anion-exchange resin, and
ceramic HA. The HA purification step for the human pyridoxal kinase
utilized a 20-mM sodium N,N-bis(2-hydroxyethyl)-2-aminoethanesulfonic
acid (BES), pH 7.3 adsorption buffer and a linear gradient to 100 mM
potassium phosphate, pH 7.3 to desorb the enzyme. Methyl jasmonate
hydrolyzing esterase was purified from cell cultures of Lycopersicon esculentum
by Stuhlfelder et al. (2002) using anion-exchange chromatography,
gel-filtration, and chromatography on ceramic HA. The HA step utilized
a 20-mM potassium phosphate, 20-mM b-mercaptoethanol, 0.3-mM
calcium chloride, pH 6.8 adsorption buffer and a linear gradient to
500 mM potassium phosphate, 20 mM b-mercaptoethanol, pH 6.8 to
desorb the esterase. Recombinant erythropoietin (rEPO) is manufactured
with a multistep chromatography process using affinity, hydrophobic inter-
action, HA, and anion-exchange chromatography as explained by these
inventors in a recent application for a patent (Schumann et al., 2007). The
ceramic HA step utilized a 20-mM Tris, 5-mM calcium chloride, 250-mM
sodium chloride, 9% isopropanol, pH 6.9 adsorption buffer. The rEPO is
402 Larry J. Cummings et al.

desorbed with 10 mM Tris, 0.5 mM calcium chloride, 10 mM potassium

phosphate, pH 6.8. In another purification process for rEPO, the inventors
(Kawasaki et al., 2008) use HA Ultrogel sorbent equilibrated with 0.05 M
Tris buffer (pH 7.5) containing 1 M sodium chloride and 2 mM calcium
chloride to adsorb the protein. The rEPO desorbed with 20 mM phosphate
buffer solution (pH 7.5) containing 0.005% (w/v) of polysorbate 80.
Immunoglobulins have been isolated and purified utilizing multiple
chromatography steps. For example, human immunoglobulin A (IgA)
from Cohn fraction III of human serum was further purified by Leibl
et al. (1996) used heparin-affinity chromatography followed by fraction-
ation with ceramic HA. The ceramic HA was equilibrated with 10 mM
phosphate, 137 mM NaCl, pH 7.4, and the IgA heparin fraction applied.
The adsorbed IgA monomer was desorbed with 15.3 mM phosphate,
287 mM NaCl, pH 6.8 then further purified by anion exchange, size
exclusion, and affinity chromatography to remove IgG from the IgA com-
ponent. In another example, monoclonal IgG (Mab) was obtained from
mouse ascites fluid and isolated by hydrophobic charge induction chroma-
tography. Guerrier et al. (2001) applied the Mab fraction to HA gel in
10 mM sodium phosphate, pH 8 then desorbed it with 0.5 M potassium
chloride, 10 mM sodium phosphate, pH 6.8. HA was used as a polishing
step by Sinacola and Robinson (2002) to remove aggregates of single-chain
antibodies (scFv) and inactive monomeric scFv from active scFv. The active
scFv did not adsorb to HA equilibrated with 200 mM NaCl, 1 mM ethy-
lenediaminetetraacetic acid buffered with 100 mM Tris–HCl, pH 8.3.
Mab expressed in high titers often contain clinically significant amounts of
immunoglobulin aggregates. The selectivity of the monomer and polymer
aggregates is often the same for HA surfaces in phosphate desorption systems.
Sodium chloride was used to purify monomer from aggregates on HA
(Sun, 2003). The details are given in US patent application 20050107594.
Aggregates, endotoxin and DNA were desorbed with 0.5 M sodium phos-
phate buffer, pH 6.8. A significant improvement in aggregate removal with
phosphate desorption was demonstrated by Gagnon (2008) with buffers
supplemented with polyethylene glycol (PEG). With higher concentrations
of PEG in the adsorption buffer, monomer was not adsorbed but aggregates,
endotoxin and DNA were.

REFERENCES
Bernardi, G. (1971). Chromatography of proteins on hydroxyapatite. Methods Enzymol. 22,
325–339.
di Salvo, M. L., Hunt, S., and Schirch, V. (2004). Expression, purification, and kinetic
constants for human and Escherichia coli pyridoxal kinases. Protein Expr. Purif. 36, 300–306.
Dong, X., Pan, H.-H., Wang, Q., and Wu, T. (2007). Understanding adsorption–
desorption dynamics of BMP-2 on hydroxyapatite (001) surface. Biophys. J. 93, 750–759.
Protein Chromatography on Hydroxyapatite Columns 403

Ferguson, W. J., Bell, D. H., Braunschweiger, K. I., Braunschweiger, W. R., Good, N. E.,
Jarvis, N. P., McCormick, J. J., Smith, J. R., and Wasmann, C. C. (1980). Hydrogen ion
buffers for biological research. Anal. Biochem. 104, 300–310.
Freitag, R., and Breier, J. (1995). Displacement chromatography in biotechnological down-
stream process. J. Chromatogr. A. 691, 101–112.
French patent 2231422, British patent 1468592, Swiss patent 597893, and German patent
2426501.
Gagnon, P. (1996). Purification tools for monoclonal antibodies. p. 87. Validated Biosystems
Inc., Tucson, AZ.
Gagnon, P. (2008). Improved antibody aggregate removal by hydroxyapatite chromatogra-
phy in the presence of polyethylene glycol. J. Immunol. Methods 336, 222–228.
Gagnon, P., Cheung, C.-W., and Yazaki, P. (2009). Reverse calcium affinity purification of
Fab with calcium derivatized hydroxyapatite. J. Immunol. Methods 342, 115–118.
Gorbunoff, M. J. (1984a). The interaction of proteins with hydroxyapatite: I. Role of
protein charge and structure. Anal. Biochem. 136, 425–432.
Gorbunoff, M. J. (1984b). The interaction of proteins with hydroxyapatite: II. Role of acidic
and basic groups. Anal. Biochem. 136, 433–439.
Gorbunoff, M. J. (1985). Protein chromatography on hydroxyapatite columns. Methods
Enzymol. 117, 370–380.
Gorbunoff, M. J. (1990). Protein chromatography on hydroxyapatite columns. Methods
Enzymol. 182, 329–339.
Gorbunoff, M. J., and Timasheff, S. (1984). The interaction of proteins with hydroxyapatite:
III. Mechanism. Anal. Biochem. 136, 440–445.
Guerrier, L., Boschetti, E., and Flayeux, I. (2001). A dual-mode approach to the selective
separation of antibodies and their fragments. J. Chromatogr. B 755, 37–46.
Harding, I. S., Hing, K. A., and Rashid, N. (2005). Surface charge and the effect of excess
calcium ions on the hydroxyapatite surface. Biomaterials 26, 6818–6826.
Hsieh, H.-Y., Calcutt, M. J., Chapman, L. F., Mitra, M., and Smith, D. S. (2003).
Purification and characterization of a recombinant alpha-N-acetylgalactosaminidase
from Clostridium perfringens. Protein Expr. Purif. 32, 309–316.
Ichitsuka, T., Kawamura, K., Ogawa, T., Sumita, M., and Yokoo, A. (1992). Packing
material for liquid chromatography US Patent 5039408.
Kadoya, T., Ebihara, M., Ishikawa, T., Isobe, T., Kobayashi, A., Kuwahara, H., Ogawa, T.,
Okuyama, T., and Sumita, M. (1986). A new spherical hydroxyapatite for high perfor-
mance liquid chromatography of proteins. J. Liq. Chromatogr. 9, 3543–3557.
Kandori, K., Ishikawa, T., and Miyagawa, K. (2004). Adsorption of immunogamma globulin
onto various synthetic calcium hydroxyapatite particles. J. Colloid Interface Sci. 273, 406–413.
Kawasaki, T. (1991). Hydroxyapatite as a liquid chromatographic packing. J. Chromatogr.
544, 147–184.
Kawasaki, T., Ikeda, K., Kuboki, Y., and Takahashi, S. (1986). Further study of hydroxyap-
atite high-performance liquid chromatography using both proteins and nucleic acids, and
a new technique to increase chromatographic efficiency. Eur. J. Biochem. 155, 249–257.
Kawasaki, A., Kirihara, S., and Mukai, K. (2008). Method for production of erythropoietin
WO/2008/068879.
Krane, S. M., and Glimcher, M. J. (1962). Transphosphorylation from nucleoside di and
triphosphates by apatite crystals. J. Biol. Chem. 237, 2991–2998.
Leibl, H., Eibl, M. M., Mannhalter, J. W., Tomasits, R., and Wolf, H. (1996). Method for
the isolation of biologically active monomeric immunoglobulin A from plasma fraction.
J. Chromatogr. B 678, 173–180.
Luellau, E., Freitag, R., Vogt, S., and von Stockar, U. (1998). Development of a down-
stream process for isolation and separation of immunoglobulin monomers, dimers, and
polymers from cell culture supernatant. J. Chromatogr. A 796, 165–175.
404 Larry J. Cummings et al.

Ma, Q. Y., Logan, T. J., and Traina, S. J. (1994). Effects of aqueous Al, Cd, Cu, Fe(II), Ni,
and Zn on Pb immobilization by hydroxyapatite. Environ. Sci. Tech. 28, 1219–1228.
McCue, J. T., Cecchini, D., Dolinski, E., and Hawkins, K. (2007). Use of an alternative scale-
down approach to predict and extend hydroxyapatite column lifetimes. J. Chromatogr. A
1165, 78–85.
Ng, P. K., Gagnon, P., and He, J. (2007). Mechanistic model for adsorption of immuno-
globulin on hydroxyapatite. J. Chromatogr. A 1142, 13–18.
Nuss, J. E., Choksi, K. B., DeFord, J. H., and Papaconstantinou, J. (2008). Decreased
enzyme activities of chaperones PDI and BiP in aged mouse livers. Biochem. Biophys.
Res. Commun. 365, 355–361.
Ogawa, T., and Hiraide, T. (1996). Effect of pH on gradient elution of different proteins on
two types of Macro-Prep ceramic hydroxyapatite. Am. Lab. 28, 31–34.
Schirch, V., Angelaccio, S., Hopkins, S., and Villar, E. (1985). Serine hydroxymethyltrans-
ferase from Escherichia coli: Purification and properties. J. Bacteriol 163, 1–7.
Schlatterer, J. C., Baeker, R., Kehler, W., Klose, J., Schlatterer, B., and Schlatterer, K.
(2006). Purification of prostaglandin D synthase by ceramic and size exclusion chroma-
tography. Prostaglandins Other Lipid Mediat. 81, 80–89.
Schroder, E., Jonsson, T., and Poole, L. (2003). Hydroxyapatite chromatography: Altering
the phosphate-dependent elution profile of protein as a function of pH. Anal. Biochem.
313, 176–178.
Schumann, C., Hesse, J.-O., and Mack, M. (2007). Method for purifying erythropoietin US
Patent Application Publication, US2007/0293420 A1.
Scopes, R. K. (1993). Protein Purification Principles and Practice. 3 edn. pp. 172–175.
Springer, New York.
Shen, J. W., Pan, H.-H., Wang, Q., and Wu, T. (2008). Molecular simulation of protein
adsorption and desorption on hydroxyapatite surfaces. Biomaterials 29, 513–532.
Shepard, S. R., Brickman-Stone, C., Koch, G., and Schrimsher, J. L. (2000). Discoloration of
ceramic hydroxyapatite used for protein chromatography. J. Chromatogr. A 891, 93–98.
Shi, X.-L., Feng, M.-Q., Shi, J.-H., Shi, J., Zhong, J., and Zhou, P. (2007). High-level
expression and purification of recombinant human catalase in Pichia pastoris. Protein Expr.
Purif. 54, 24–29.
Sinacola, J. R., and Robinson, A. S. (2002). Rapid refolding and polishing of single-chain
antibodies from Escherichia coli inclusion bodies. Protein Expr. Purif. 26, 301–308.
Skartsila, K., and Spanos, N. (2007). Surface characterization of hydroxyapatite: Potentiometric
titrations coupled with solubility measurements. J. Colloid Interface Sci. 308, 405–412.
Stránská, J., Chmelı́k, J., Peč, P., Popa, I., Řehulka, P., Šebela, M., and Tarkowski, P.
(2007). Inhibition of plant amine oxidases by a novel series of diamine derivatives.
Biochimie 89, 135–144.
Stuhlfelder, C., Lottspeich, F., and Mueller, M. J. (2002). Purification and partial amino acid
sequences of an esterase from tomato. Photochemistry 60, 233–240.
Sun, S. (2003). Removal of high molecular weight aggregates from an antibody preparation
using ceramic hydroxyapatite chromatography. In ‘‘Third International Hydroxyapatite
Conference’’, Lisbon, Portugal.
Teeters, M. A., and Quiñones-Garcı́a, I. (2005). Evaluating and monitoring the packing
behavior of process-scale chromatogragphy columns. J. Chromatogr. A 1069, 53–64.
Thompson, A. R., and Miles, B. J. (1973). New materials, especially for chromatography.
Methodol. Dev. Biochem. Prep. Tech. 2, 95–101.
Tiselius, A., Hjertén, S., and Levin, O. (1956). Protein chromatography on calcium phos-
phate columns. Arch. Biochem. Biophys. 65, 132–155.
Zhu, X. D., Fan, H. S., Li, D. X., Luxbacher, T., Xiao, Y. M., Zhang, H. J., and
Zhang, X. D. (2009). Effect of surface structure on protein adsorption to biphasic
calcium-phosphate ceramics in vitro and in vivo. Acta Biomater. 5, 1311–1318.
 C H A P T E R T W E N T Y- F I V E

Theory and Use of Hydrophobic

Interaction Chromatography in
Protein Purification Applications
Justin T. McCue

Contents
1. Theory 406
2. Latest Technology in HIC Adsorbents 408
3. Procedures for Use of HIC Adsorbents 409
3.1. Introduction 409
3.2. Choice of adsorbent 409
3.3. Feed/load preparation 410
3.4. Adsorbent preparation 410
3.5. Product elution 410
3.6. Gradient elution 412
3.7. Stepwise (isocratic) elution 412
3.8. Adsorbent regeneration and sanitization 413
References 413

Abstract
Hydrophobic interaction chromatography (HIC) is a valuable tool used in protein
purification applications. HIC is used in the purification of proteins over a broad
range of scales—in both analytical and preparatory scale applications. HIC is
used to remove various impurities that may be present in the solution, including
undesirable product-related impurities. In particular, HIC is often employed to
remove product aggregate species, which possess different hydrophobic proper-
ties than the target monomer species and can often be effectively removed using
HIC. In this chapter, we provide a description of the basic theory of HIC and how it
is used to purify proteins in aqueous-based solutions. Following the theoretical
background, the latest in HIC adsorbent technology is described, including a list
of commonly used and commercially available adsorbents. The basic procedures
for using HIC adsorbents are described next, in order to provide the reader with
useful starting points to apply HIC in protein purification applications.

Biogen Idec Corporation, Cambridge, Massachusetts, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63025-1 All rights reserved.

405
406 Justin T. McCue

1. Theory
Hydrophobic proteins will self-associate, or interact, when dissolved
in an aqueous solution. This self-association forms the basis for a variety of
biological interactions, such as protein folding, protein–substrate interac-
tions, and transport of proteins across cellular membranes ( Janson and
Rydén, 1997). Hydrophobic interaction chromatography (HIC) is used
in both analytical and preparatory scale protein purification applications.
HIC exploits hydrophobic regions present in macromolecules that bind
to hydrophobic ligands on chromatography adsorbents. The interaction
occurs in an environment which favors hydrophobic interactions, such as
an aqueous solution with a high salt concentration.
By itself, water (a polar solvent) is a poor solvent for nonpolar molecules.
Under such an environment, proteins will self-associate, or aggregate, in
order to achieve a state of lowest thermodynamic energy. Prior to self-
association, water molecules form highly ordered structures around each
individual macromolecule (Fig. 25.1A). The self-association of nonpolar
molecules (such as proteins) in the polar solvent is driven by a net increase in
entropy of the environment. During the aggregation process, the overall
surface area of hydrophobic sites of the protein exposed to the polar solvent

Legend

Protein self- Protein

association
Water molecule
Entropy
increase Hydrophobic
ligand

Protein-ligand
binding
Entropy
increase
Base matrix Base matrix

Figure 25.1 Schematic diagram showing hydrophobic interaction between proteins in

an aqueous solution (A) and between proteins and a hydrophobic ligand on an HIC
adsorbent (B).
Hydrophobic Interaction Chromatography 407

is decreased, which results in a less structured (higher entropy) condition,

which is the favored thermodynamic state.
This same concept is responsible for the interaction (association)
between hydrophobic ligands attached to an adsorbent and the proteins of
interest (Fig. 25.1B). Association, or hydrophobic interaction, between the
protein and the hydrophobic ligand is driven primarily by an increase in the
overall entropy (compared with the condition when no interaction is
occurring between the protein and the adsorbent).
The polarity of the solvent can be controlled through the addition of
salts or organic solvents, which can strengthen or weaken hydrophobic
interactions between the HIC adsorbent and the protein. The influence of
ions on hydrophobic interaction follows the well-known Hofmeister series
(Hofmeister, 1988). Anions which promote hydrophobic interaction the
greatest are listed (in decreasing strength of interaction) from left to right
(Påhlman et al., 1977):








4 > SO4 > CH3 COO > Cl > Br > NO3 > CLO4 > I > SCN
PO3
2

Ions which promote hydrophobic interactions are called lyotropes,

while those which disrupt (weaken) hydrophobic interactions are called
chaotropes. In the above series, phosphate ions promote the strongest
hydrophobic interaction, while thiocyanate ions disrupt hydrophobic
interactions.
For cations, the Hofmeister series consists of the following (listed in
order of decreasing lyotropic strength):
NHþ þ þ þ þ þ
4 > Rb > K > Na > Cs > Li > Mg
2þ
> Ca2þ > Ba2þ
Two of the most common Lyotropic salts used to promote hydrophobic
interaction in aqueous solution are ammonium sulfate and sodium chloride.
These salts are commonly employed when using HIC for protein
purification.
In addition to salts, organic solvents can also be used to alter the strength
of hydrophobic interactions (Fausnaugh and Regnier, 1986; Melander and
Horvath, 1977). Organic solvents commonly used to weaken, or disrupt
hydrophobic interactions include glycols, acetonitrile and alcohols. The
organic solvents alter the polarity of the mobile phase, thereby weakening
potential interactions that may occur. They may be added to the solution
during the elution process, in order to disrupt hydrophobic interactions and
elute the strongly bound protein of interest.
Protein hydrophobicity is a complex function of several properties,
which include the amino acid sequence, as well as protein tertiary and
quaternary structure in a given solution (Ben-Naim, 1980; Tanford,
1980). Hydrophobicity scales have been created for particular amino
acids, which are based upon the solubility in water and organic solvents
408 Justin T. McCue

( Jones, 1975; Nozaki and Tanford, 1971; Tanford, 1962; Zimmerman et al.,
1968). Empirical hydrophobic scales for proteins have also been created (Chotia,
1976; Krigbaum and Komoriya, 1979; Manavalan and Ponnuswamy,
1978; Rose et al., 1985; Wertz and Scheraga, 1978) which are based upon the
fraction of amino acids exposed on the protein surface, as well as the degree of
amino acid hydrophobicity. The ability to predict the hydrophobicity of
complex proteins has been only semiquantitative to date, and experiments are
usually required to accurately understand protein hydrophobicity in a given
aqueous solution.

2. Latest Technology in HIC Adsorbents

HIC adsorbents consist of a base matrix which is coupled to a hydro-
phobic ligand. The base matrix, which typically consists of porous beads
with diameters ranging from 5 to 200 mm, provides high surface area for
ligand attachment and protein binding. Common base matrices include
agarose, methacrylate, polystyrene–divinylbenzene and silica (Table 25.1).
For analytical applications, the bead size of the adsorbent is in the lower
range (5–20 mm). Small beads are used in order to maximize resolution
when performing analytical separations. For preparatory scale applications,
larger bead sizes are usually required (20 mm). Larger bead sizes are
required for preparatory scale columns due to pressure drop limitations
associated with the column hardware.
HIC adsorbents containing hydrophobic ligands with various degrees of
hydrophobicity are available. The ligands consist of alkyl or aryl chains. As a

Table 25.1 Properties of commercially available HIC adsorbents

Base matrix Available ligand types Adsorbent manufacturers

Cross-linked agarose Butyl GE Healthcare
Octyl
Phenyl
Polystyrene– Phenyl GE Healthcare
divinylbenzene Applied Biosystems
Methacrylate Butyl
Ether TosoHaas
Phenyl EM Industries
Hexyl
Silica Propyl J. T. Baker
Diol Synchrom
Pentyl Supelco
YMC
Hydrophobic Interaction Chromatography 409

general rule, the strength of hydrophobic binding of the ligand will increase
with the length of the organic chain. Several of the most common ligands
include butyl, octyl, and phenyl, which are linked to the base bead support
through several different coupling approaches (Hjertén et al., 1974; Ulbrich
et al., 1964). Aromatic ligands, such as phenyl, can also interact with the
adsorbed compounds through so-called ‘‘p–p interactions,’’ which can
further strengthen the hydrophobic interaction (Porath and Larsson, 1978).
The hydrophobic interaction strength of the ligand can also be influ-
enced by the ligand loading (ligand density) on the base matrix. The
strength of interaction can increase with higher ligand densities. In order
to have reproducible performance, manufacturers of HIC adsorbents must
often produce adsorbents with narrow ranges of ligand density to ensure
consistent performance from lot to lot.

3. Procedures for Use of HIC Adsorbents

3.1. Introduction
The application of HIC adsorbents for use in the purification of protein
compounds is fairly straightforward. The purification process is composed
of a series of subsequent steps, which are described in this section.

3.2. Choice of adsorbent

Since the hydrophobic properties of a given protein are often unknown,
several HIC adsorbents may need to be screened prior to selection of a final
adsorbent. The screening process is used to determine how strongly com-
pounds bind to the adsorbent, and determine which adsorbents are viable
candidates to purify the protein. Adsorbents with a broad range of hydro-
phobic binding properties should be included in the initial screening studies
to determine the hydrophobicity of the compound, as well as which
adsorbents effectively purify the protein.
Adsorbents may be available in the form of prepacked columns from the
adsorbent manufacturer or may have to be packed by the user. In the case
when columns need to be packed, the packing instructions from the
manufacturer should be followed. Height equivalent to the theoretical
plate (HETP) measurements can be used to verify that the column is packed
correctly. Once packed, the adsorbent manufacturer often provides recom-
mended operating ranges for use, which include allowable flow rate
ranges, column conditioning, and column cleaning procedures. Acceptable
operating conditions will vary for different adsorbents, and the vendor
documentation should be consulted prior to use.
410 Justin T. McCue

3.3. Feed/load preparation

Prior to column loading, the salt concentration of the protein mixture
(which will be purified using the HIC adsorbent) must be increased to a
level in which the target protein binds to the adsorbent using a high salt
buffer. Proteins may precipitate in high salt solutions, so the compound
solubility in a given salt solution should be evaluated prior to its selection.
The influence of buffer pH on compound binding to HIC adsorbents has
no general trend, but can influence the strength of interaction (Hjertén
et al., 1974). A buffer pH should be chosen in which the protein and the
adsorbent are stable (e.g., avoid the use of extreme pH solutions). During
the protein load adjustment step, the salt concentration may range from
approximately 0.5 to 2.0 M, and will be increased high enough to ensure the
protein binds effectively to the adsorbent. The salt concentration required to
bind the protein to the adsorbent will depend greatly on the choice of salt, as
described in Section 3.2. Selection of the appropriate salt concentration in
which the protein binds to the adsorbent will require experimental work
in most cases.

3.4. Adsorbent preparation

Prior to loading the protein feed, the column should first be equilibrated in a
high salt buffer solution which possesses a similar composition (salt concen-
tration) and pH as the feed solution to ensure the protein will bind tightly to
the adsorbent. This step is referred to as the equilibration step.
Following the equilibration step, the adjusted feed (which contains the
protein of interest) is loaded onto the HIC column at an appropriate
velocity. After the protein-containing solution is loaded onto the column,
the column is often washed with the equilibration buffer prior to product
elution. Additional wash steps may be implemented prior to the elution step
to remove undesirable impurity species. The wash steps may contain a salt
concentration at an intermediate salt concentration—less than the load step,
but greater than the elution step.

3.5. Product elution

After being bound to the adsorbent, the desired protein must be eluted and
then collected in the column effluent. In many cases, the elution process is
used to separate, or resolve, unwanted species from the desired protein. The
unwanted species may bind less tightly to the adsorbent, and will be eluted
prior to the product. In other cases, undesirable species bind more tightly to
the adsorbent and will remain bound after the product is eluted. This is
usually the case when HIC is used to separate protein aggregate species,
which bind more tightly to the adsorbent than the desired protein monomer
Hydrophobic Interaction Chromatography 411

species. During the elution step, a portion (or fraction) of the eluate may
contain highly purified product, while fractions before and after contain
higher levels of undesirable impurities. A schematic of an elution process
(during a gradient elution) is shown in Fig. 25.2. Figure 25.2 illustrates that
the column effluent collected during the elution step may need to be
fractionated in order to achieve acceptable product purity when using
HIC. The gradient elution process is described in more detail in Section 3.6.
The elution process can be done using either a stepwise (isocratic) or a
gradient approach. The four most common methods (listed from most
common to least common) used to elute the bound protein include the
following:
1. Decrease in the salt concentration (relative to the binding conditions). A decrease
in the salt concentration will decrease the strength of hydrophobic
interaction between the protein and the ligand, and the protein will be
desorbed and eluted from the column.
2. Addition of organic solvents. Addition of an organic solvent (such as
ethylene or propylene glycol) changes the solvent polarity, which dis-
rupts the hydrophobic interaction.
3. Increase in the salt concentration (using a chaotropic salt). Addition of a
chaotropic salt will disrupt the hydrophobic interaction.
4. Detergent addition. Detergents are used as protein displacers, and have
been used mainly for the purification of membrane proteins when using
HIC ( Janson and Rydén, 1997).

Target
product Strongly bound
Weakly impurities
bound (e.g. aggregates)
impurities

High salt
concentration
Protein concentration

Salt concentration

Salt
gradient Low salt
elution concentration

Elution volume

Figure 25.2 Schematic chromatogram showing a gradient elution of a protein mixture

using hydrophobic interaction chromatography. In the diagram, the salt concentration
is linearly decreased (from high salt to low salt), which results in elution of both
impurities and the target protein.
412 Justin T. McCue

This most common approach used to elute proteins from HIC adsor-
bents is by lowering the salt concentration during the elution step. This
should be the first method that is attempted when using HIC for purifica-
tion of a new protein compound. The other approaches described above
have the disadvantage that an additional component (such as a chaotropic
salt or an organic solvent) needs to be added, which may impact protein
stability. However, such agents may be required in order to effectively elute
a strongly bound protein species from the adsorbent. Each protein must be
evaluated case by case to determine which elution method is appropriate.
The HIC adsorbent used in the purification may also influence which
elution method is effective.

3.6. Gradient elution

Gradient elutions are an extremely effective method useful for screening
different HIC adsorbents in protein purification. During the gradient elution
process, the salt concentration is decreased gradually (in a linear fashion) from
a high salt concentration to a low salt concentration over a defined volume.
During the initial screening of a bound compound on an adsorbent, the salt
concentration may be decreased to as low as 0 mM to determine the salt
concentration when the product elutes. As a starting point, a typical gradient
elution process is performed over 10 column volumes, during which frac-
tions are collected and evaluated for product purity. The gradient in salt
concentration may be decreased (performed over a larger volume) in order
to improve protein resolution (Yamamoto et al., 1988).
In the event that the protein remains bound to the adsorbent following
the gradient elution process, this may indicate that either a weaker lyotropic
salt should be selected to bind the protein to the adsorbent or that a stronger
elution condition is required to elute the protein. Stronger elution solutions
may include the use of an organic solvent. For preparatory scale applica-
tions, nonflammable organic solvents (such as propylene or ethylene glycol)
are often selected. Organic solvents in analytical scale applications may
include such solvents as acetonitrile and alcohols. Alternatively, an adsor-
bent with weaker hydrophobic binding strength may need to be selected to
decrease the strength of hydrophobic interaction.

3.7. Stepwise (isocratic) elution

After identifying the appropriate adsorbent and salt concentration to effec-
tively elute the protein of interest, an isocratic elution can be used if desired.
An advantage of using isocratic elution is its simplicity—it requires a simple
switch in the inlet buffer (from a high to a low salt concentration). Use of an
isocratic elution is a preferable approach to simplify the equipment
Hydrophobic Interaction Chromatography 413

requirements, as gradient elution requires multiple pumps and additional

process control to generate a linear change in the buffer salt concentration.

3.8. Adsorbent regeneration and sanitization

HIC adsorbents are reusable for multiple cycles and have a relatively long
lifetime before having to be replaced. However, adsorbents must be cleaned
and regenerated between uses in order to ensure reproducible performance
over many cycles. The adsorbent manufacturers’ provide regeneration
procedures for the adsorbents, which should be consulted prior to use. In
general, the cleaning procedures depend upon the stability of the base
matrix and the hydrophobic ligand. For strongly bound proteins, 6 M
guanidine hydrochloride is often recommended. If detergents have been
used during the process, ethanol or methanol can be used to as part of the
regeneration procedure (GE Healthcare, 2006). For sanitization, a caustic
solution (1.0 M NaOH) can be used for most of the adsorbents (with the
exception of silica). The manufacturer should also provide information on
the appropriate storage conditions. Storage solution should be selected that
prevents microbial growth, but does not impact ligand or base matrix
stability.

REFERENCES
Ben-Naim, A. (1980). Hydrophobic Interactions. Plenum Press, New York.
Chotia, C. (1976). Surface of monomelic proteins. J. Mol. Biol. 105, 112.
Fausnaugh, J. L., and Regnier, F. E. (1986). Solute and mobile phase contributions to
retention in hydrophobic interaction chromatography of proteins. J. Chromatogr. 359,
131–146.
GE Healthcare (2006). Data File No. 18-1127-63 AC.
Hjertén, S., Rosengren, J., and Påhlman, S. (1974). Hydrophobic interaction chromato-
graphy. J. Chromatogr. 101, 281–288.
Hofmeister, F. (1988). On regularities in the albumin precipitation reactions with salts and
their relationship to physiological behavior. Arch. Exp. Pathol. Pharmakol. 24, 247–260.
Janson, J.-C., and Rydén, L. (eds.) (1997). Protein Purification: Principles, High-Resolution
Methods, and Applications, 2nd edn., p. 284. Wiley-VCH, New York.
Jones, D. D. (1975). Amino acid properties and side-chain orientation in proteins. J. Theor.
Biol. 50, 167–183.
Krigbaum, W. R., and Komoriya, A. (1979). Local interactions as a structure determinant for
protein molecules. Biochim. Biophys. Acta 576, 204–248.
Manavalan, P., and Ponnuswamy, P. K. (1978). Hydrophobic character of amino acid
residues in globular proteins. Nature 275, 673–674.
Melander, W., and Horvath, C. (1977). Salt effects on hydrophobic interactions in precipi-
tation and chromatography of proteins: An interpretation of the lyotropic series. Arch.
Biochem. Biophys. 183, 200–215.
Nozaki, Y., and Tanford, C. (1971). The solubility of amino acids and two glycine peptides
in aqueous ethanol and dioxane solutions. Establishment of a hydrophobicity scale.
J. Biol. Chem. 246, 2211–2217.
414 Justin T. McCue

Påhlman, S., Rosengren, J., and Hjerten, S. (1977). Hydrophobic interaction chromatography
on uncharged Sepharose derivatives. J. Chromatogr. 131, 99–108.
Porath, J., and Larsson, B. (1978). Charge-transfer and water-mediated chromatography.
I. Electron-acceptor ligands on cross-linked dextran. J. Chromatogr. 155, 47–68.
Rose, G. D., Geselowitz, A. R., Lesser, G. J., Lee, R. H., and Zehfus, M. H. (1985).
Hydrophobicity of amino acid residues in globular proteins. Science 229, 834–838.
Tanford, C. (1962). Contribution of hydrophobic interactions to the stability of the globular
conformation of proteins. J. Am. Chem. Soc. 84, 4240–4247.
Tanford, C. (1980). In The Hydrophobic Effect 2nd edn. Wiley, New York.
Ulbrich, V., Makes, J., and Jurecek, M. (1964). Identification of giycidyl ethers. Bis(phenyl-)
and bis(a-naphthylurethans) of glycerol a-alkyl (aryl)ethers. Collect. Czech. Chem.
Commun. 29, 1466–1475.
Wertz, D. H., and Scheraga, H. A. (1978). Influence of water on protein structure.
Macromolecules 11, 9–15.
Yamamoto, S., Nakanishi, K., and Matsuno, R. (1988). Ion-Exchange Chromatography of
Proteins Mercel Dekkar, New York.
Zimmerman, J.M, Eliezer, N., and Simha, R. (1968). The characterization of amino acid
sequences in proteins by statistical methods. J. Theor. Biol. 21(2), 170–201.
 C H A P T E R T W E N T Y- S I X

Affinity Chromatography:
General Methods
Marjeta Urh, Dan Simpson, and Kate Zhao

Contents
1. Introduction 418
2. Selection of Affinity Matrix 419
2.1. General features of the support material 419
2.2. Selectivity 420
2.3. Stability 422
2.4. Magnetic affinity beads 422
3. Selection of Ligands 423
3.1. General considerations for ligands design and selection 423
3.2. Characterization of immobilized ligand 424
3.3. Affinity matrices carrying specific ligands 425
3.4. Immunoglobulin binding proteins 425
3.5. Lectins 426
3.6. Biomimetic ligands 427
3.7. Covalent affinity chromatography 428
4. Attachment Chemistry 429
4.1. Activation of surface 430
4.2. Ligand attachment 430
5. Purification Method 433
5.1. Sample preparation 433
5.2. Binding and wash 433
5.3. Elution 434
References 435

Abstract
Affinity chromatography is one of the most diverse and powerful chro-
matographic methods for purification of a specific molecule or a group of
molecules from complex mixtures. It is based on highly specific biological
interactions between two molecules, such as interactions between enzyme
and substrate, receptor and ligand, or antibody and antigen. These interactions,
which are typically reversible, are used for purification by placing one of the

Promega Corporation, Madison, Wisconsin, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63026-3 All rights reserved.

417
418 Marjeta Urh et al.

interacting molecules, referred to as affinity ligand, onto a solid matrix to create

a stationary phase while the target molecule is in the mobile phase. Successful
affinity purification requires a certain degree of knowledge and understanding
of the nature of interactions between the target molecule and the ligand to
help determine the selection of an appropriate affinity ligand and purification
procedure. With the growing popularity of affinity purification, many of the
commonly used ligands coupled to affinity matrices are now commercially
available and are ready to use. However, in some cases new affinity chro-
matographic material may need to be developed by coupling the ligand onto
the matrix such that the ligand retains specific binding affinity for the molecule
of interest. In this chapter, we discuss factors which are important to consider
when selecting the ligand, proper attachment chemistry, and the matrix. In
recent years, matrices with unique features which overcome some of the
limitations of more traditional materials have been developed and these are
also described. Affinity purification can provide significant time savings and
several hundred-fold or higher purification, but the success depends on the
method used. Thus, it is important to optimize the purification protocol to
achieve efficient capture and maximum recovery of the target.

1. Introduction
Affinity chromatography is a method for selective purification of a
molecule or group of molecules from complex mixtures based on highly
specific biological interaction between the two molecules. The interaction
is typically reversible and purification is achieved through a biphasic inter-
action with one of the molecules (the ligand) immobilized to a surface while
its partner (the target) is in a mobile phase as part of a complex mixture. The
capture step is generally followed by washing and elution, resulting in
recovery of highly purified protein. Highly selective interactions allow for
a fast, often single step, process, with potential for purification in the order
of several hundred to thousand-fold. Additional uses of affinity chromatog-
raphy include the ability to concentrate substances present at low concen-
tration and the ability to separate proteins based on their biological function
where an active form can be separated from the inactive form or a form with
different biological function.
Recent decades have seen tremendous advancements in the utility of
affinity chromatography, with developments in support materials such as
flow-through beads, magnetic beads and monolithic materials as well as new
ligands with a variety of interesting biological properties. In addition, this
approach is no longer used only for purification of specific biomolecules.
It is also quickly becoming a method of choice to study biological interac-
tions and can be used for preparation of samples for mass spectrometry or for
specific removal of contaminants.
Overview of Affinity Chromatography Methods 419

While the first examples of affinity chromatography exploited binding of

enzyme amylase to insoluble starch (Starkenstein, 1910), it was the develop-
ment of solid support materials and the chemistry used to attach ligands to
these solid supports that truly enabled affinity chromatography (Campbell
et al., 1951; Cuatrecasas et al., 1968; Lerman, 1953a,b). Since these seminal
advances, there has been an explosion in the development of more support
materials, affinity ligands, and improvements in the methods that make
affinity purification an attractive approach to purify biomolecules from com-
plex mixtures. While affinity chromatography can be used to purify any
biological molecule with a specific interacting ligand, this chapter will focus
on purification of proteins. Several excellent reviews and books have been
written on this subject (Hage et al., 2006; Hirabayashi, 2008; Labrou, 2003;
Ostrove, 1990; Zachariou, 2007); thus, we will not review all the proteins
and methods that have been developed up to date, but will rather give general
guidelines and considerations important for the selection of different support
materials, ligands, attachment chemistry, and optimization of the purification
and discuss some of the latest developments in affinity chromatography.

2. Selection of Affinity Matrix

Successful affinity purification depends on the selection of a suitable
solid support and a suitable immobilized ligand. The affinity matrix (i.e.,
a solid support onto which ligand is immobilized) should ideally be macro-
porous with high chemical and physical stability, and should selectively
capture target of interest while at the same time exhibit low nonspecific
adsorption and maintain good flow properties throughout processing. They
are preferably inexpensive, readily available, and simple to use. The affinity
matrix may be commercially available or can be made by attaching a suitable
ligand to a solid support via the appropriate chemistry (see Section 4). There
are many detailed reviews on how to select a suitable affinity matrix
(GE Healthcare, 2007; Gustavsson and Larsson, 2006; Zachariou, 2007), a
summary of some of the key consideration points are given in the following
sections.

2.1. General features of the support material

The matrix should be macroporous with uniform particle and pore size and
with good flow properties. Pore size of a matrix is inversely correlated to its
surface area, which in turn, directly affects the amount of immobilized
ligand and thus the capacity. The pore size correlates to exclusion limit,
which is the size (molecular weight) or the size range of proteins that cannot
enter the pore. Large pores do not suffer from size exclusion effect and allow
420 Marjeta Urh et al.

unhindered access of large molecules to the immobilized ligands, but have

reduced surface area and lower ligand density that can result in lower
capacity. According to the Renkin equation (Renkin, 1954), the size of
the pores should be at least five times larger than the average size of a
biomolecule for its easy access to immobilized ligands. The size of the pores
should be at least 300 Å or greater if assuming an average protein size of

60 Å. Most commonly available supports satisfy this requirement
(Table 26.1). Some soft gel-based matrices are cross-linked to increase
their mechanical stability, but this process can reduce porosity (pore vol-
ume), thus reducing the amount of attached ligand and binding capacity.
In most cases, a compromise is made between pore size and surface area to
fit a particular application.
Other important considerations are diameter of particles and particle size
distribution. In theory, smaller particles are better because they allow for
faster mass transfer between outer flow and interior of the particle, making
higher flow rates possible while maintaining efficient affinity capture.
However, they lead to higher flow resistance, greater potential for particle
collapse, and increased sensitivity to contaminants such as particles and
denatured proteins in the sample which can lead to high back pressure.
The selection of the particle size will largely depend on the purpose and
method of the separation where particles of 10 mm are best suited for HPLC
separation and 400 mm for preparative purposes. The particle size distribu-
tion should be as uniform as possible so that smaller particles do not fill the
void volume and restrict the flow.

2.2. Selectivity
One key feature of an affinity matrix is its selectivity. It should be specific for
a protein of interest as determined by the specific ligand coupled to the
matrix and inert to all other compounds present in the complex sample.
Since most applications are performed in aqueous solutions, often with low
ionic strength, the support should be hydrophilic and contain limited charge
that may lead to undesirable ionic interaction. Many commercially available
supports fulfill these requirements, either with their native structure or by
coating with suitable materials. Common supports include beaded agarose
and cellulose, available commercially from a number of vendors including
Sepharose from GE Healthcare and Affigel from Bio-Rad (Table 26.1).
Nonspecificity can come from the support itself, such as hydrophobicity
associated with polystyrene beads and negative charge on the surface of
silica. It can also be introduced when modifying a matrix to accept a
particular ligand. In these cases, the attachment chemistry, the ligand, and
the spacer between the ligand and the matrix should be carefully designed,
screened and optimized for selectivity, capacity for target of interest, and
low nonspecific binding.
Table 26.1 Examples of commercially available matrices

Name Vendor Matrix material Particle size (mm) Exclusion limit (Da)
SepharoseTM GE Healthcare Agarose 40–165 10,000–1,000,000
CL6B
SepharoseTM GE Healthcare Agarose 40–165 30,000–5,000,000
CL4B
Bio-gel A-5m Bio-Rad Agarose 75–300 (50–200 mesh) 10,000–5,000,000
medium
Perloza MT100 Iontosorb Cellulose 100–250 2,000,000
medium
Perloza MT50 Iontosorb Cellulose 100–250 100,000
medium
AllTech Grace Silica 7 Pore size
Macrosphere 60–300 Å
Bio-gel P-100 Bio-Rad Polyacrylamide 90–180 5000–100,000
medium
SephacrylTM GE Healthcare Cross-linked ally dextrose 50 1000–100,000
Poros 50 Applied Cross-linked poly(styrene- 50 Pore size 50–100 nm
Biosystems divinylbenzene)
422 Marjeta Urh et al.

2.3. Stability
The affinity matrix must also be chemically and physically stable during the
process, such that the support material itself as well as the attached ligand
should not react to the solvents used in the process, nor should they be
degraded or damaged by enzymes and microbes that might be present in the
sample. The chemical compatibilities of commercially available affinity
matrices are usually supplied by the manufacturer, which should be used
as guidance for developing a successful purification protocol. Cross-linked
agarose can usually withstand a wide pH range (e.g., pH 3–12), most
aqueous solvents (including denaturants), many organic solvents or modi-
fiers, and enzymatic treatments. Materials such as glass and silica are not
stable at alkaline conditions due to hydrolysis; therefore, coating of these
surfaces is often needed before attaching ligands. The matrix should also
withstand physical stress, such as pressure, especially when packed into a
column, and remain intact during the purification process. High pressure
can compress the matrix, causing it to collapse. Agarose beads and other soft
gel matrices are more susceptible to pressure, relative to stronger supports,
such as silica, polystyrene and other highly cross-linked materials.
Monoliths are macroporous, nonbeaded single matrix that can be made
from different materials such as agarose, silica, GMA/EDMA, and cryogel
(Mallik and Hage, 2006; Plievaa et al., 2009). It possesses many of the
desired properties of an affinity support and has become popular due to
the presence of large flow-through pores with no void volume permitting
convective flow instead of diffusion and high flow rate for shortened run
time. In addition, the matrix will not compress and has less pressure drop
during column chromatography and can be made to withstand larger pH
ranges and harsh chemicals. There are several limitations of monoliths as
compared to traditional matrix, such as lower capacity, special processes
required for making each type of monolith affinity supports and the current
limited range of available affinity types (Mallik and Hage, 2006).

2.4. Magnetic affinity beads

Magnetic separation can significantly shorten the purification process by
quick retrieval of affinity beads at each step (e.g., binding, wash, and
elution), and reduce sample dilution usually associated with traditional
column-based elution. The method can be used on viscous materials that
will otherwise clog traditional columns and can therefore simplify the
purification process by eliminating sample pretreatment, such as centri-
fugation or filtration to remove insoluble materials and particulates. The
capability of miniaturization and parallel screening of multiple condi-
tions, such as growth conditions for optimal protein expression and
buffer conditions for purification, makes magnetic separation amenable
Overview of Affinity Chromatography Methods 423

to high-throughput analysis which can significantly shorten the purification

process (Saiyed et al., 2003).
Paramagnetic particles are available as unmodified, modified with com-
mon affinity ligands (e.g., streptavidin, GSH, Protein A, etc.), and conju-
gated particles with specific recognition groups such as monoclonal and
polyclonal antibodies (Koneracka et al., 2006). In addition to target protein
purification, they can also be used to immobilize a target protein which then
acts as a bait to pull down its interaction partner(s) from a complex
biological mixture. See Chapter 16.

3. Selection of Ligands
Selection of the appropriate ligand requires a certain degree of knowl-
edge and understanding of the nature of interactions between the ligand and
the target molecule; where the ligand must specifically bind the target
molecule and should be stable in different binding and elution conditions.
Additionally, when developing affinity purification scheme it is important
to consider whether the ligand is commercially available or de novo devel-
opment of the ligand and affinity matrix will be required. The success of the
second scenario will largely depend on existing knowledge of protein
structure and the nature of interaction, requiring the use of molecular
modeling and combinatorial organic synthesis coupled with immobilization
chemistry and selective binding analysis. The time and effort to design a
novel ligand and to develop appropriate coupling chemistry and matrix may
prove to be too lengthy and costly, whereas the use of nonaffinity-based
purification technique such as ion exchange and hydrophobic interaction
schemes may be a better choice.

3.1. General considerations for ligands design and selection

Affinity between ligand and target molecule is one of the most important
considerations when developing a new affinity purification material; low
affinity can reduce the binding efficiency resulting in poor yield while high
affinity can lead to inefficient elution or inactivation of the target protein by
harsh elution conditions also giving low yield. Generally, an affinity con-
stant in the range of 106–108 M 1 can be used for affinity-based purifica-
tion. It is preferred that binding affinity of ligand and target protein be
evaluated prior to the creation of an affinity matrix, but one has to be aware
that affinity in solution may differ from affinity after immobilization and in
some cases immobilization may even result in a change in specificity.
When attaching ligand to a matrix, covalent coupling is preferred due to
the reduced risk of ligand leaching during purification. However,
424 Marjeta Urh et al.

noncovalent attachment strategies such as nonspecific adsorption, biospeci-

fic interaction (biotin–streptavidin), entrapment, and most recently devel-
oped, molecular imprinting (Alexander et al., 2006) may also be utilized in
the absence of reactive groups or to reduce the risk of ligand denaturation.
See Section 4 for more details.
Besides weak affinity, reduced binding between ligand and target can
also be a result of steric hindrance caused either by the matrix or by other
ligands. Introduction of a spacer between the matrix and the ligand can
reduce the steric hindrance caused by the matrix (Hage et al., 2006). When
designing a spacer arm, consideration should be given to both its length and
the hydrophobic character, as a more hydrophilic nature is desirable to
reduce the nonspecific interactions with molecules other than the target.
Optimization of spacer length is also important, since shorter spacers may
not relieve the steric hindrance effect of the matrix and longer ones may
promote nonspecific interaction or may fold back onto themselves thereby
limiting specific interactions. The density of ligand on the surface should
also be optimized. Very high density of a ligand may have an adverse effect
and lead to loss of binding capacity either because of close proximity of
binding sites causing steric hindrance or strong binding that may prevent
effective elution (Hage et al., 2006).
Other factors that may influence selection of the best ligand are ability to
be sterilized, stability of the ligand, proper storage conditions and cost.

3.2. Characterization of immobilized ligand

An important factor determining the utility of the affinity resin is the ability
to reproducibly make the resin with similar performance characteristics.
Thus, it is important to determine optimal reaction conditions including the
amount of the ligand needed for the synthesis to assure consistency in
performance. Determining the optimal amount of ligand in the synthesis
may also reduce the cost by preventing wasteful use of excess ligand.
For covalent attachment it is often desirable to determine how much
ligand is attached on the matrix and to determine how coupling conditions
affect the amount of immobilized ligand. A simple approach to determine
ligand density is to analyze the amount of the ligand left in the reaction after
coupling is complete and subtract it from the total starting amount.
Depending on the nature of the ligand, different detection methods can
be applied, such as spectrophotometric detection, BCA assay for proteins,
Ellman’s reagent can be used to detect sulfhydryl groups, fluorescent detec-
tion and others (Guilbault, 1988; Langone, 1982).
In some instances, such as amino acid or elemental analysis, the ligand
can be measured directly on the support, but this leads to destruction of the
material and should therefore be done on a small fraction. An indirect
estimation of immobilized ligand can be achieved by estimating the binding
Overview of Affinity Chromatography Methods 425

capacity by determining the amount of target molecule retained by the

affinity matrix under desired conditions. This may be the most relevant
measurement of the ligand immobilization since it will directly correlate to
the performance of the affinity matrix. Maximum binding capacity should
be performed at equilibrium, usually at a slow flow rate or in batch binding
mode. Note that dynamic binding capacity may differ from optimal binding
capacity as it will be affected by the flow rate, mass transfer through matrix
pores and affinity constant.

3.3. Affinity matrices carrying specific ligands

In recent years, we have witnessed an explosion in the development of
affinity ligands and ready-to-use affinity matrices specific for different target
molecules. An overview of commonly used, commercially available ligands
is given in Table 26.2, which is by no means a comprehensive list of all the
available ligands.
Most available affinity matrices carry what are often called ‘‘group
specific ligands’’ that exhibit binding affinity for a group of structurally or
functionally related proteins and thus can be used for purification of differ-
ent target proteins with similar functionality. In the following sections, we
describe some commonly used ligands for affinity matrices.

3.4. Immunoglobulin binding proteins

Purification of antibodies, one of the most effective and frequently used
application of affinity purification methods is based on binding between the
constant region (Fc) of many types of immunoglobulins and protein A or
protein G. Protein A from Staphylococcus aureus and protein G made by
Streptococcus are related bacterial proteins that bind the IgG class of anti-
bodies, with differences in subclass specificity and source organism. Detailed
lists of different binding affinities of protein A and G can be found in many
sources including several commercial suppliers (Guss et al., 1986; Hage et al.,
2006). Other immunoglobulin binding proteins that have lately become
popular are protein B, a surface protein of group A Streptococci bacteria,
which binds several subclasses of human IgA antibodies, and protein L
(Peptostreptococcus magnus), which has ability to interact with kappa light
chains without affecting the antigen-binding site of antibodies (Faulmann
et al., 1991; Hermanson, 1992). This makes protein L uniquely suited for
purification of antibodies lacking Fc regions.
In most cases, good binding to antibodies can be achieved at or near
neutral pH, but the optimal pH for sample application can differ depending
on the protein used. Protein A binds antibodies strongest at pH 8.2,
protein L at pH 7.5 and protein G at pH 5 but protein G can also be used
at pH 7–7.5. For elution, solutions at acidic pH (2.5–3.0) are often applied
426 Marjeta Urh et al.

Table 26.2 Commonly used ligands and specificity

Ligand Specificity
Cibacron Blue Albumin, kinases, dehydrogenases, enzymes requiring
F3G-A adenylyl-containing cofactors, NADþ
Blue B Kinases, dehydrogenases, nucleic acid binding proteins
Orange A Lactate dehydrogenase
Green A HAS, dehydrogenases
Polymixin Endoproteins
Benzamidine Serine proteases (thrombin, trypsin, kallikrein)
Biotin Streptavidin, avidin
Gelatin Fibronectin
Heparin DNA binding proteins, serine protease inhibitors
(antithrombin III), growth factors, lipoproteins,
hormone receptors, coagulation factors, DNA, RNA
Lysine Plasminogen, rRNA, dsDNA
Arginine Serine proteases with affinity for arg, fibronectin,
prothrombin
ADT Enzymes with affinity for NADPþ
AMP NAD-dependent dehydrogenases and ATP-dependent
kinases
NAD, NADP Dehydrogenases
Lectins Glycoproteins, polysaccharides, glycolipids
Calmodulin Calmodulin binding proteins, ATPase, adenylate cyclase,
kinases, phosphodiesterase,
Protein A Fc regions of many IgG subtypes, species dependent, weak
interactions with IgA, IgM, IgD
Protein G Fc region of many IgG subtypes, species dependent
Protein L Kappa light chains of antibodies (Fab, single chain variable
region scFv)

and samples are collected into buffers with neutral or slightly basic pH to
avoid denaturation and loss of activity. In cases where biological stability is
lost during elution at low pH, elution with a pH gradient in combination
with salt can be explored.

3.5. Lectins
Lectins are a diverse group of proteins which bind carbohydrates with
high degree of specificity where each lectin has its own specificity profile.
They are often used in affinity purification or enrichment of carbohydrate
moieties of complex glycoconjugates, such as polysaccharides, glycolipids,
Overview of Affinity Chromatography Methods 427

and glycoproteins. They also allow for a specific isolation of different glycol-
forms of a specific protein depending on the nature of glycosylation. Recently,
lectins have been used not only for the purification of sugar-containing
molecules but also for the enrichment of subgroups of glycoproteins for
analysis in mass spectrometry (Hirabayashi, 2008). See Chapter 34.
Most commercially available lectins are of plant origin, and there are
over 100 different lectins available either in conjugated or free form. Lectin
from Canavalia ensiformis, known as Conacanavalin A (ConA) has the
affinity for a-D-mannose, a-D-glucose, and N-acetylglucosamine and is
probably the most frequently used lectin (Hermanson, 1992). Two other
popular lectins are wheat germ agglutinin (WGA) and Jacalin; WGA binds
to sialic acid and molecules containing N-acetyl-D-glucosamine residue and
Jacalin binds to a-D-galactosyl groups.
Coupling of lectin onto resin is often performed at neutral pH and in the
presence of sugar to preserve sugar binding site. Binding to the target
molecule is also usually carried out in neutral pH; note that some lectins
require the presence of divalent metal ions, Ca2þ and Mn2þ in the case of
ConA, for optimal binding. Elution is accomplished by adding an access of
the specific sugar molecule to the elution buffer and this can be done as
stepwise or gradient elution. After elution, the free sugar should be removed
by dialysis or size exclusion chromatography.

3.6. Biomimetic ligands

Most of affinity chromatography ligands are naturally occurring and the
appeal of these ligands is high selectivity and capacity of binding, but they
often suffer from several drawbacks such as low stability, need for purifica-
tion in their own right, variability in performance from lot to lot, contami-
nation with other biomolecules, sensitivity to sterilization methods and high
cost. To overcome some of these limitations and accelerate the use of
affinity chromatography in purification of therapeutic proteins, there is an
increased focus on the development of synthetic or altered affinity ligands,
biomimetics, which mimic the structure of and binding of natural biological
ligands.
One of the most popular group of biomimetic ligands are reactive textile
dyes which gained popularity because of their flexibility and ability to
assume polarity and geometry of the surface of a variety of competitive
biomolecules and can function in solution as a competitive inhibitor,
coenzyme or effector of many proteins (Madoery and Minchiotti, 2006;
Stellwagen, 1990). The majority of these ligands, including the most popu-
lar dye, Cibacron blue F3G-A, contain a triazine scaffold which can be
modified for improved specificity, forming the basis for the biomimetic
dye–ligands concept. Over the past decades, a plethora of dyes were devel-
oped and used in the purification of a broad spectrum of proteins including
428 Marjeta Urh et al.

blood proteins such as albumin, oxidoreductase, decarboxylases, glycolitic

enzymes, nucleases, hydrolases, lyases, synthetases, and transferases, and
many dye–ligand-based affinity chromatographic materials are now com-
mercially available materials (see Table 26.2) (Labrou and Clonis, 1994,
2002). In addition to the expansion of commercially available materials,
there is a growing popularity in creating tailor-made biomimetics for better
performance. These ligands are designed to target specific proteins by
mimicking peptide templates, natural biological recognition motifs, or
complementary surface-exposed residues and have been generated by a
combination of rational design, combinatorial library synthesis, and
subsequent screening (Cecı́lia et al., 2005; Labrou, 2003; Lowe et al.,
2001). The immobilization of triazine containing dye onto affinity matrices
such as agarose, dextran, and cellulose can be achieved under alkaline
conditions by using nucleophilic displacement of the dye’s chlorine by
hydroxyl groups on the support surface (Labrou, 2000; Labrou and
Clonis, 1994; Labrou et al., 1995).

3.7. Covalent affinity chromatography

Affinity chromatography typically relies on reversible interaction between
the target molecule and the ligand; however, covalent interactions have also
been utilized for isolation of specific target molecules (Blumberg and
Strominger, 1972; GE Healthcare, 2007; Hage, 2006).
One such approach is based on the covalent binding between a thiol-
containing target molecule and activated thiol immobilized onto purifica-
tion matrix. Bound protein can be eluted by reducing the cysteine disulfide
with 2-mercaptoethanol, TCEP, or dithiothreitol.
Recently, another covalent binding-based purification has been devel-
oped on the basis of specific and covalent interaction between a chloroalk-
ane ligand and a protein called HaloTag, Figure 26.1 (Los et al., 2008;
Ohana et al., 2009). HaloTag is a 34-kDa monomeric protein fusion tag
which can be genetically fused to any protein of interest either at its C- or
N-terminus. HaloTag protein was created by first modifying the active site
of bacterial haloalkane dehydrolase so that a permanent covalent bond can
be formed with the specific chloroalkane ligand. This modification was
followed by additional mutagenesis to increase stability of the protein and
the binding rate, which is similar to that of the biotin–streptavidin. To make
use of HaloTag for purification, a chromatographic matrix, the HaloLink
resin carrying the chloroalkane ligand, was created. The covalent nature of
the HaloTag-fusion protein capture, combined with rapid binding kinetics,
overcome the equilibrium-based limitations associated with traditional
affinity purification and enables efficient capture of target proteins even at
very low abundance. Furthermore, it allows for extensive and stringent
washing without losing the bound protein. While the covalent association
Overview of Affinity Chromatography Methods 429

A B
N

Solid support
Cl O O N
C H Fluorophore
Chloroalkane
ligand

HaloTag protein

Figure 26.1 HaloTag technology comprises two components: (A) The HaloTag
protein shown on the left with covalently bound HaloTag-TMR ligand. N- and
C-termini are indicated. (B) The chloroalkane ligand. Different functional groups
including, but not limited to, fluorescent dyes or solid support surfaces can be attached
to the chloroalkane ligand, which covalently binds to HaloTag protein and imparts
different functionalities including protein immobilization and fluorescent labeling.

clearly has its advantages, it also creates a challenge in eluting the protein of
interest. Because the covalent bond between HaloTag and chloroalkane
cannot be reversed, traditional approaches to elute proteins from the resin
cannot be utilized. Instead the protein of interest can be released from
HaloTag by specific protease (TEV) as the TEV recognition site is present
between the two moieties (the HaloTag and fusion target protein). Upon
cleavage, HaloTag stays bound to the resin while the fusion partner is
released yielding highly pure protein free of tag (Urh et al., 2008).
Besides using HaloTag for purification of fusion proteins described
earlier, the immobilized HaloTag fusions can also be considered as affinity
ligands in their own right, and can, similarly to protein G, be used to capture
antibodies or other proteins which specifically bind to the proteins fused to
HaloTag. The advantage of this system is that unlike other covalent immo-
bilization techniques, where binding of protein is random and may lead to
multiple attachment sites and improper orientation, immobilization using
HaloTag is a single point attachment through active site and therefore
oriented. Single point, oriented attachment increases capacity, effectiveness
and reproducibility of the system, and HaloTag fusions covalently bound to
the matrix can therefore improve purification of specific antibodies or
isolation of binding partners.

4. Attachment Chemistry
This section will briefly review some of the more common chemistry
for the covalent attachment of affinity ligands to conventional surfaces such
as agarose, cellulose, silica, glass, and synthetic polymeric supports. Strate-
gically, the process can be divided into three components: (1) the surface or
430 Marjeta Urh et al.

resin (matrix); (2) the linkage or resin-activation component (spacer); and

(3) the affinity ligand. The constraint that most often dictates how the ligand
is attached to a surface is the type and availability of reactive moieties on the
ligand. Common moieties used are primary amine, thiol, and carboxyl
groups. Alternatively, modification of a ligand to create a reactive group,
such as the oxidation of a carbohydrate or the covalent attachment of an
orthogonal reactive group, provides additional options for ligand immobi-
lization. Attachment strategies for small synthetic molecules tend to follow
the same chemical pathway as biomolecules but often with somewhat
different design considerations.

4.1. Activation of surface

The process of preparing a covalent affinity-based surface generally begins
with activation of the surface to either directly accept the ligand or to accept
an activated linker to physically ‘‘space’’ the ligand away from the surface.
Representative attachment chemistries are summarized in Table 26.3, listed
with established surfaces and shown in contrast with the ligand entity used
to attach to the activated surface. A variety of preactivated surfaces are
commercially available for most of the chemistries listed in Table 26.3. If
the appropriate chemical entity is readily displayed on the ligand, the
remaining chemistry is reduced to defining the optimal conditions for the
coupling reaction to occur weighed against the stability constraints of the
ligand. Most ligand coupling reactions are done in aqueous pH-controlled
buffers for periods between 1 and 20 h at temperatures between 4  C and
room temperature. Since the reaction is heterogenous, both concentration
and mixing are important to ensure an even distribution of reactants.
Whenever possible keep the solution phase of the reaction to no more
than twice the resin volume. Ligand concentration will vary somewhat by
objective but for proteins a 2.5–10 mg/ml is a reasonable starting point.
Depending on the activation chemistry and ligand reactivity, protocols exist
for a fairly broad range of pH and buffer conditions (Hermanson, 1992).
Efficiencies will vary from system to system and as a consequence it is
essential to cap left over surface reactive groups to neutralize or reduce
nonspecific interactions.

4.2. Ligand attachment

For a peptide, protein, or nucleic acid, the covalent attachment is most
commonly done through an amine group on the ligand to an amine reactive
functionality on the resin surface. The most common amine-reactive
attachment is an imidocarbonate that results from the reaction of cyanogen
bromide activated surface with a primary amine, most typically a lysine side
chain (Hermanson, 1992; Porath et al., 1967) (see Chapter 28). Advantages
Overview of Affinity Chromatography Methods 431

Table 26.3 Commonly used attachment chemistries

Affinity ligand
Bead or surface Activation and/or linkage reactive group
Soft gels: agarose, cellulose Cyanogen bromide Amine
Synthetic supports: Aldehyde (reductive Amine
Polyacrylamide beads, amination)
Trisacryl, Sephacryl, Activated carboxyl ester Amine
Ultragel, Azlactone beads, (succinimidyl ester)
Methacrylate (TSK gel), Carbonyldiimidazole Amine
Eupergit, Polystyrene Carboxyl (activation Amine
(Poros supports) concurrent with
coupling)
FMP activation Amine, thiol
Divinyl sulfone Amine, thiol
Azlactone Amine, thiol
Epoxy (bisoxirane, Amine, thiol
epichlorohydrin)
Tresyl chloride Amine, thiol
Haloacetyl (iodo or Thiol
bromo)
Maleimide Thiol
Pyridyl disulfide Thiol
Amine Carboxyl
(activation
concurrent
with
coupling)
Hydrazide Carbohydrate
(periodate
reduced)
Inorganics: controlled pore 3-(Glycidyloxypropyl) Amine
glass, silica, alumina, trimethoxy-silane
zeolites, etc. 3-(Aminopropyl) Carboxyl (after
trimethoxysilane activation),
aldehyde

of this activation include commercially available preactivated resin and very

efficient protein coupling, near 100%. While this is historically a popular
method, it suffers a number of drawbacks including: (1) linkage leading to
leaching of the ligand off of the resin, (2) attachment directly to the surface
with no spacer, and (3) required additional safety precautions due to the
toxicity of cyanogen bromide. In addition, cross-linking may be needed to
help limit the leakiness of this immobilization (Korpela and Hinkkanen,
1976; Kowal and Parsons, 1980).
432 Marjeta Urh et al.

Covalent attachment through the formation of an amide bond is an

alternative to the cyanogen bromide chemistry but requires either an
activated carboxyl surface (such as a N-hydroxysuccinimidyl ester) or
in situ activation of the carboxyl group with a coupling agent such as
N-ethyl-N0 -(3-dimethylaminopropyl)carbodiimide (EDC) (Wilchek et al.,
1984, 1994). Another stable alternative is the formation of a secondary
amine linkage resulting from a reductive alkylation of a Schiff-base inter-
mediate formed between a primary amine (lysine or N-terminus) to an
aldehyde activated surface. This reductive alkylation attachment mechanism
is a popular method to immobilize enzymes to carbohydrate surfaces
(agarose or cellulose) as the conditions for coupling are mild and the
immobilized enzyme has been reported to retain more activity than
observed with other methods of attachment (Hermanson, 1992).
Amine reactive linkages are also a common approach to attachment of
small molecule ligands to surfaces. The primary requirement being that the
ligand has the predesigned amine group for attachment. A number of the
chemistries that are applied to amine immobilization also apply to sulfhydryl
(or thiol)-reactive immobilization which is in concept the same as amine-
reactive immobilization but relies on the reaction of a cysteine with the
reactive surface. For thiol specific immobilization, the two most prominent
strategies are through haloacetyl (iodo or bromo) and maleimide activated
surfaces (Mallik et al., 2007). Sulfhydryl linkage may be advantageous in that
it can also be made reversible through the use of a disulfide linkage that can
be removed from the surface by treatment with a reducing agent such as
DTT or TCEP (Brena et al., 1993).
Immobilization of antibodies or glycosylated protein to a surface can be
done by the methods described earlier but an additional option that offers
a potentially more oriented attachment is also available through modifica-
tion of the carbohydrate (Oates et al., 1998; Vijayendran and Leckband,
2001). Immobilization through carbohydrate requires a mild oxidation
(i.e., periodate) of the carbohydrate to form reactive aldehydes with the
sugar residues. The aldehydes produced by oxidation are then used to
immobilize the protein or antibody to a hydrazide reactive surface. This
approach is commonly used for antibodies, glycoproteins, glycopolymers,
and ribonucleic acids (O’Shannessy and Wilchek, 1990).
In most cases, it is sufficient to rely on the attachment of a ligand through
one of the reactive groups discussed earlier, but in some cases, it may be
desirable to space the linker further from the surface. One option for doing
so is to orthogonally modify the ligand of interest with a spacer and reactive
group selective for a second group that will preferentially recognize the
orthogonal label attached to a surface. This requires modification of both
the surface and the ligand with the appropriate reactive groups. A current
popular embodiment of this approach is the copper(I)-catalyzed 1,3-dipolar
cycloaddition of azides to terminal alkynes to form 1,2,3-triazoles known
Overview of Affinity Chromatography Methods 433

commonly now as ‘‘click chemistry’’ (Chandran et al., 2009; Gauchet et al.,

2006). The advantage of this approach is that the chemistry is mild, the
selectivity is unique and reactive groups can be easily switched between
the ligand and the surface. Because the reaction itself requires catalysis, the
reactive groups on their own are much more stable than reactive groups
commonly used with thiol and amine labeling.

5. Purification Method
Purification by affinity starts with proper handling of the sample and
the matrix, followed by selective binding (capture) of the target, washing to
remove nonspecific background, and, finally, elution of the bound target.
Successful affinity purification depends on a number of notable factors
including, the amount and accessibility of the ligand on the resin, the
strength of the interaction, and the integrity of protein to be immobilized.
Usually, conditions are optimized to maximize the interaction between a
target and immobilized ligand during the binding and wash process, then,
switched to substantially weaken the interaction thus allowing for release of
the target. It is recommended to perform small scale trials to select for the
best purification conditions. A short summary of some common practical
issues and considerations affecting affinity purification is given in the
following sections.

5.1. Sample preparation

When preparing sample for purification, conditions should be selected to
retain the proper fold and functionality of the target of interest. It is also
highly recommended to remove insoluble materials and to reduce viscosity
because both of these factors could clog the column, reduce the flow rate,
and increase back pressure.
Some proteins tend to aggregate at high concentrations, which results in
increased apparent molecular weight, decreased diffusion rate, and reduced
capture by the affinity matrix. Dilution of the sample or cell lysate in a larger
volume may be needed to reduce aggregation and increase protein capture
and recovery under this circumstance. Conversely, if the sample is too
dilute, binding rate and capture efficiency can be reduced especially for
low-affinity binders.

5.2. Binding and wash

Efficiency of binding is related to the strength and the kinetics of protein–
ligand interaction which can be affected by the nature of the interaction, the
concentration of applied target, the amount of immobilized ligand, and
434 Marjeta Urh et al.

the flow rate used for binding. The binding process can be simplified
as Eq. (26.1), assuming a 1:1 molar ratio where Ka is the association
equilibrium constant, [L] is ligand concentration, [T ] is the concentration
of target protein, and [LT ] is the concentration of the complex. Ka equals
[LT ]/([L]  [T ]), which can also be expressed as ka/kd, where ka is second
order association rate constant that depends on the concentrations of both
L and T, and kd is the first order dissociation constant that does not depend
on ligand concentration. Higher Ka usually leads to a higher adsorption
ratio, defined as the ratio of bound to total applied target, thus, better
binding. Normally, the ligand concentration on the matrix is around
10 2–10 4 M and to achieve efficient binding, the Ka value should be in
the range of 104–106 M 1.
ka
L þ T ! LT ð26:1Þ
kd

Affinity immobilization onto solid support can be achieved by passing

the sample through a column packed with the affinity matrix, usually under
ambient pressure and a slow flow rate. Generally, a higher flow rate will
reduce the binding efficiency, especially, when the interaction between the
ligand and protein is weak or the mass-transfer rate in the column is slow.
The binding process can also be performed in batch, where the resin and
sample are constantly mixed. Batch binding promotes effective contact
between target and immobilized ligand and often saves time, especially
when dealing with large sample volumes; however, nonspecific binding
can also increase. It is a good practice to optimize the amount of resin used
during purification, where saturation of the resin with target during binding
is preferred since excess resin can result in an increase in nonspecific binding
as well as reduced target recovery due to readsorption, unless the latter is
required under special circumstances.
Following binding, protein bound by nonspecific interactions can be
removed by washing. For example, ionic interactions can be reduced by
increasing salt (0.1–0.5 M) or changing pH values, and hydrophobic inter-
actions can be removed by decreasing salt, altering pH, or adding surfactants
(such as Triton X-100). Low amounts of competitive reagents can also be
used to remove contaminants with weak affinity to the ligands. The flow
rate and the volume (e.g., 5–10 column volumes) of the wash buffer should
also be carefully determined for maximum removal of contaminants with
minimum loss of target.

5.3. Elution
Elution of bound target from the resin is essentially the reverse process of
binding, where conditions are optimized to reduce the Ka, that is, weaken-
ing the interaction between target and ligand. The elution condition should
Overview of Affinity Chromatography Methods 435

not denature the target protein, unless such conditions are compatible with
downstream applications.
There are two different types of elution methods, namely, specific and
nonspecific elution. In specific elution, the target protein–ligand complex is
challenged by agents that will compete for either the ligand or the target
thereby releasing the target protein into solution. The concentration and
amount (volume) of competitive reagent used for elution will depend
on their affinity relative to that of the immobilized complex, where weaker
competitive reagents require higher concentration and more volume
as compared to higher affinity additives. A good starting point for weak
competitors is to use concentration 10-fold higher than that of the ligand.
The specific elution is usually milder and proteins are more likely to retain
their activity, but the slow elution, broad elution peaks, and the need to
remove competing agent from the recovered protein are some of the
drawbacks of this approach. For nonspecific elution, solvent conditions
are manipulated to reduce the association rate constant (Eq. (26.1)), which
ideally should approach zero, and to increase the dissociation rate constant,
thus, weakening the overall affinity (Ka) resulting in dissociation of the
complex. Elution conditions can be optimized according to the mechanism
of interaction between the ligand and protein, such as increasing salt
concentration to reduce ionic interactions or by altering pH to change the
protonation/ionization state, thus modulating the strength of hydrogen
bonds, hydrophobic interactions as well electrostatic interactions. An exam-
ple of elution by changing pH is the elution of antibodies from immobilized
protein A or protein G, yet because the affinity of proteinA/G to antibodies
is very strong, with Ka in the 108 M 1 range, a combination of different
elution conditions may be required for maximum antibody release. In many
cases, affinity requires proper three-dimensional folding of a protein so that
chaotropic reagents or reagents that will affect protein folding can be used to
elute target of interest; however, care must be taken to maintain proper
folding of the target after elution by quickly returning to native conditions.
When the affinity is weak, binding is achieved at high concentration of the
target molecule which is then eluted by dissociation of the complex through
dilution. This approach is known as isocratic elution.

REFERENCES
Alexander, C., Andersson, H. S., Andersson, L. I., Ansell, R. J., Kirsch, N., Nicholls, I. A.,
O’Mahony, J., and Whitcombe, M. J. (2006). Molecular imprinting science and tech-
nology: A survey of the literature for the years up to and including 2003. J. Mol. Recognit.
19, 106–180.
Blumberg, P. M., and Strominger, J. L. (1972). Isolation by covalent affinity chromatogra-
phy of the penicillin-binding components from membranes of Bacillus subtilis. Proc. Natl.
Acad. Sci. USA 69, 3751–3755.
436 Marjeta Urh et al.

Brena, B., Ovsejevi, K., Luna, B., and Batista-Viera, F. (1993). Thiolation and reversible
immobilization of sweet potato beta-amylase on thiolsulfonate-agarose. J. Mol. Catal. 84,
381–390.
Campbell, D. H., Luescher, E., and Lerman, L. S. (1951). Immunologic adsorbents: I.
Isolation of antibody by means of a cellulose-protein antigen. Proc. Natl. Acad. Sci.
USA 37, 575–578.
Cecı́lia, A., Roque, A., and Lowe, C. R. (2005). Advances and applications of de novo
designed affinity ligands in proteomics. Biotechnol. Adv. 24, 17–26.
Chandran, S. P., Hotha, S., and Prasad, B. L. V. (2009). Tunable surface modification of
silica nanoparticles through ‘‘click’’ chemistry. Current Science 95, 1327–1333.
Cuatrecasas, P., Wilchek, M., and Anfinsen, C. B. (1968). Selective enzyme purification by
affinity chromatography. Proc. Natl. Acad. Sci. USA 61, 636–643.
Faulmann, E. L., Duvall, J. L., and Boyle, M. D. (1991). Protein B: A versatile bacterial
Fc-binding protein selective for human IgA. Biotechniques 10, 748–755.
Gauchet, C., Labadie, G. R., and Poulter, C. D. (2006). Regio- and chemoselective
covalent immobilization of proteins through unnatural amino acids. J. Am. Chem. Soc.
128, 9274–9275.
GE Healthcare (ed.) (2007). Affinity chromotography, principle and methods. GE
Healthcare.
Guilbault, G. G. (ed.) (1988). In ‘‘Analytical uses of immobilized biological compounds for
detection, medical and industrial uses,’’ Dordrecht, Boston, MA.
Guss, B., Eliasson, M., Olsson, A., Uhlen, M., Frej, A. K., Jornvall, H., Flock, J. I., and
Lindberg, M. (1986). Structure of the IgG-binding regions of streptococcal protein G.
EMBO J. 5, 1567–1575.
Gustavsson, P.-E., and Larsson, P.-O. (2006). Support materials for affinity chromatography.
In ‘‘Handbook of Affinity Immobilization,’’ (D. S. Hage, ed.), pp. 15–34. CRC press/
Taylor and Francis Group, Boca Raton, FL.
Hage, D. S. (ed.) (2006). In ‘‘Handbook of affinity chromatography,’’ CRC press/Taylor
and Francis Group, Boca Raton, FL.
Hage, D. S., Bian, M., Burks, R., Karle, E., Ohnmachi, C., and Wa, C. (2006). Bioaffinity
Chromatography. In ‘‘Handbook of Affinity Immobilization,’’ (D. S. Hage, ed.),
pp. 101–126. CRC press/Taylor and Francis Group, Boca Raton, FL.
Hermanson, G. T. (ed.) (1992). In ‘‘Immobilized affinity ligand techniques,’’ Academic
Press, San Diego, CA.
Hirabayashi, J. (2008). Concept, strategy and realization of lectin-based glycan profiling.
J. Biochem. 144, 139–147.
Koneracka, M., Kopcansky, P., Timbo, M., Ramchand, C. N., Saiyed, Z. M., and Trevan, M.
(2006). Immobilization of enzymes on magnetic particles. In ‘‘Immobilization of Enzymes
and Cells,’’ ( J. M. Guisan, ed.), pp. 217–228. Humana Press.
Korpela, T., and Hinkkanen, A. (1976). A simple method to introduce aldehydic function to
agarose. Anal. Biochem. 71, 322–323.
Kowal, R., and Parsons, R. G. (1980). Stabilization of proteins immobilized on Sepharose
from leakage by glutaraldehyde crosslinking. Anal. Biochem. 102, 72–76.
Labrou, N., and Clonis, Y. D. (1994). The affinity technology in downstream processing.
J. Biotechnol. 36, 95–119.
Labrou, N. E. (2000). Dye-ligand affinity chromatography for protein separation and
purification. Methods Mol. Biol. 147, 129–139.
Labrou, N. E. (2003). Design and selection of ligands for affinity chromatography.
J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 790, 67–78.
Labrou, N. E., and Clonis, Y. D. (2002). Immobilized synthetic dyes in affinity chromatog-
raphy. In ‘‘Theory and Practice of Biochromatograph,’’ (M. A. Vijayalakshimi, ed.),
pp. 235–251. Taylor and Francis Group, London.
Overview of Affinity Chromatography Methods 437

Labrou, N. E., Karagouni, A., and Clonis, Y. D. (1995). Biomimetic-dye affinity adsorbents
for enzyme purification: Application to the one-step purification of Candida boidinii
formate dehydrogenase. Biotechnol. Bioeng. 48, 278–288.
Langone, J. J. (1982). Applications of immobilized protein A in immunochemical techni-
ques. J. Immunol. Methods 55, 277–296.
Lerman, L. S. (1953a). A biochemically specific method for enzyme isolation. Proc. Natl.
Acad. Sci. USA 39, 232–236.
Lerman, L. S. (1953b). Antibody chromatography on an immunologically specific adsorbent.
Nature 172, 635–636.
Los, G. V., Encell, L. P., McDougall, M. G., Hartzell, D. D., Karassina, N., Zimprich, C.,
Wood, M. G., Learish, R., Ohana, R. F., Urh, M., Simpson, D., Mendez, J., et al.
(2008). HaloTag: A novel protein labeling technology for cell imaging and protein
analysis. ACS Chem. Biol. 3, 373–382.
Lowe, C. R., Lowe, A. R., and Gupta, G. (2001). New developments in affinity chroma-
tography with potential application in the production of biopharmaceuticals. J. Biochem.
Biophys. Methods 49, 561–574.
Madoery, R., and Minchiotti, M. (2006). Cibacron Blue-Eupergit, an affinity matrix for
soybean (Glycine max) phospholipase A2 purification. Enzyme and Microb. Technol. 38,
869–872.
Mallik, R., and Hage, D. S. (2006). Affinity monolith chromatography. J. Sep. Sci. 29,
1686–1704.
Mallik, R., Wa, C., and Hage, D. S. (2007). Development of sulfhydryl-reactive silica for
protein immobilization in high-performance affinity chromatography. Anal. Chem. 79,
1411–1424.
O’Shannessy, D. J., and Wilchek, M. (1990). Immobilization of glycoconjugates by their
oligosaccharides: Use of hydrazido-derivatized matrices. Anal. Biochem. 191, 1–8.
Oates, M. R., Clarke, W., Marsh, E. M., and Hage, D. S. (1998). Kinetic studies on the
immobilization of antibodies to high-performance liquid chromatographic supports.
Bioconjug. Chem. 9, 459–465.
Ohana, R. F., Encell, L. P., Zhao, K., Simpson, D., Slater, M. R., Urh, M., and Wood, K.V
(2009). Halotag7: A genetically engineered tag that enhances bacterial expression of
soluble protein and improves protein purification. Protein Expr. Purif. 68, 110–120.
Ostrove, S. (1990). Affinity chromatography: General methods. Methods Enzymol. 182,
357–371.
Plievaa, F. M., De Seta, E., Galaevb, I. Y., and Mattiasson, B. (2009). Macroporous elastic
polyacrylamide monolith columns: Processing under compression and scale-up. Sep.
Purif. Technol. 65, 110–116.
Porath, J., Axen, R., and Ernback, S. (1967). Chemical coupling of proteins to agarose.
Nature 215, 1491–1492.
Renkin, E. M. (1954). Filtration, diffusion, and molecular sieving through porous cellulose
membranes. J. Gen. Physiol. 38, 225–243.
Saiyed, Z., Telang, S., and Ramchand, C. (2003). Application of magnetic techniques in the
field of drug discovery and biomedicine. Biomagn. Res. Technol. 1, 2.
Starkenstein, E. (1910). Ferment action and the influence upon it of neutral salts. Biochem. Z.
24, 210–218.
Stellwagen, E. (1990). Chromatography on immobilized reactive dyes. Methods Enzymol.
182, 343–357.
Urh, M., Hartzell, D., Mendez, J., Klaubert, D. H., and Wood, K. (2008). Methods for
detection of protein-protein and protein-DNA interactions using HaloTag. Methods Mol.
Biol. 421, 191–209.
Vijayendran, R. A., and Leckband, D. E. (2001). A quantitative assessment of heterogeneity
for surface-immobilized proteins. Anal. Chem. 73, 471–480.
438 Marjeta Urh et al.

Wilchek, M., Knudsen, K. L., and Miron, T. (1994). Improved method for preparing
N-hydroxysuccinimide ester-containing polymers for affinity chromatography.
Bioconjug. Chem. 5, 491–492.
Wilchek, M., Miron, T., and Kohn, J. (1984). Affinity chromatography. Methods Enzymol.
104, 3–55.
Zachariou, M. (ed.) (2007). In ‘‘Affinity Chromatography, Methods and Protocols,’’
Humana Press, Totowa, NJ.
 C H A P T E R T W E N T Y- S E V E N

Immobilized-Metal Affinity
Chromatography (IMAC): A Review
Helena Block,* Barbara Maertens,* Anne Spriestersbach,*
Nicole Brinker,* Jan Kubicek,* Roland Fabis,* Jörg Labahn,†
and Frank Schäfer*

Contents
1. Overview on IMAC Ligands and Immobilized Ions 440
2. IMAC Applications 444
2.1. Detection and immobilization 445
2.2. Purification of protein fractions 446
2.3. Purification of His-tagged proteins 448
2.4. General considerations of protein purification by IMAC 451
2.5. Copurifying proteins on IMAC and what to do about it 454
2.6. IMAC for industrial-scale protein production 458
2.7. High-throughput automation of IMAC 460
2.8. Special applications: Purification of membrane proteins 461
2.9. Special applications: Purification of zinc-finger proteins 463
2.10. Protein purification protocols 464
2.11. Cleaning and sanitization 466
2.12. Simplified metal-ion stripping and recharging protocol 467
3. Conclusions 467
Acknowledgments 468
References 468

Abstract
This article reviews the development of immobilized-metal affinity chromatog-
raphy (IMAC) and describes its most important applications. We provide an
overview on the use of IMAC in protein fractionation and proteomics, in protein
immobilization and detection, and on some special applications such as purifi-
cation of immunoglobulins and the Chelex method. The most relevant applica-
tion—purification of histidine-tagged recombinant proteins—will be reviewed

* QIAGEN GmbH, Qiagen Strasse 1, Hilden, Germany

{
Institute of Structural Biology (IBI-2), Research Center Jülich, Jülich, Germany

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63027-5 All rights reserved.

439
440 Helena Block et al.

in greater detail with focus of state-of-the-art materials, methods, and proto-

cols, and the limitations of IMAC and recent advances to improve the technology
and the methods will be described.

1. Overview on IMAC Ligands

and Immobilized Ions
The concept of immobilized-metal affinity chromatography (IMAC)
has first been formulated and its feasibility shown by Porath et al. (1975).
It was based on the known affinity of transition metal ions such as Zn2þ,
Cu2þ, Ni2þ, and Co2þ to histidine and cysteine in aqueous solutions
(Hearon, 1948) and extended to the idea to use metal ions ‘‘strongly fixed’’
to a support to fractionate protein solutions. As the chelating ligand used to fix
the metal to agarose Porath used iminodiacetic acid (IDA) which is still in use
today in many commercial IMAC resins. Already in 1975, Porath speculated
that the affinity of immobilized metals to histidine-containing proteins might
not be the only application for IMAC—and he was right after all.
In the years following Porath’s publications, the new IMAC technology
was successfully evaluated by purification of a variety of different proteins
and peptides summarized in a first IMAC review by Sulkowski (1985).
What started as a crude fractionation of serum soon became what today is
the most widely used affinity chromatography technique (Biocompare,
2006; Derewenda, 2004), if not chromatography technique in general.
This development was accelerated by the fast maturation of recombinant
techniques and modern molecular biology in the late 1970s and by the
invention of an improved chelating ligand, nitrilotriacetic acid (NTA) in the
1980s (Hochuli et al., 1987). In the meantime, the purification of recombi-
nant proteins genetically modified on the DNA template level in order to
generate oligohistidine extended (His-tagged) polypeptides using NTA-
based supports (Hochuli et al., 1988) represents the most important applica-
tion of IMAC. The principal mechanism of the interaction of a His-tagged
protein to an immobilized metal ion is presented in Fig. 27.1.
A possible model is that of an interaction of a metal ion with histidine
residues n and nþ2 of a His tag. This is confirmed by the fact that IMAC
ligands can bind to His tags consisting of consecutive histidine residues as
well as to alternating tags. As at least consecutive His tags are usually
unstructured and thus flexible, other interaction patterns such as n:nþ 1,
n:nþ3, n:nþ4, and so on could be imagined as well, even within a single
molecule ( Jacob Piehler, personal communication).
The NTA ligand coordinates the Ni2þ with four valencies (tetradentate,
coordination number 4) highlighted spherically in Fig. 27.1B, and two
valencies are available for interaction with imidazole rings of histidine residues.
IMAC 441

A
Protein

N N N
N N N
N N N
N N N

Ni-IDA
Ni2+

O
CH2
O
H2O CH2
N

B
Protein

N N N
N N N
N N N
N N N

Ni-NTA
Ni2+
O

CH2
O CH2
N

CH
O

C
Protein

N N N
N N N
N N N
N N N
O

Ni-TED
CH2 Ni2+
CH2
N CH2 N
CH2
CH2

Figure 27.1 Model of the interaction between residues in the His tag and the metal ion
in tri- (IDA), tetra- (NTA), and pentadentate IMAC ligands (TED).
442 Helena Block et al.

This ratio has turned out to be most effective for purification of His-tagged
proteins. Another tetradentate ligand is carboxymethyl aspartate (CM-Asp;
Chaga et al., 1999), commercially available as cobalt-charged Talon resin.
In contrast to tetradentate ligands, IDA coordinates a divalent ion with three
valencies (tridentate, coordination number 3, Fig. 27.1A) leaving three
valencies free for imidazole ring interaction while it is unclear whether
the third is sterically able to participate in the interaction. The coordination
number seems to play an important role regarding the quality of the purified
protein fraction. While protein recovery is usually similar between IDA-
and NTA-based chromatography (Fig. 27.2D), a higher leaching of metal
ions from IDA ligands compared to NTA is observed in general (Hochuli,
1989) and even increased under reducing conditions (Fig. 27.2C). Although
the metal content in the elution fractions (E in Fig. 27.2C) is higher but still
within the same order of magnitude, significantly more Ni2þ ions leach
from the IDA resin in the equilibration and wash steps (W in Fig. 27.2C).
Besides considerable metal leaching, purification of His-tagged proteins
using an IDA matrix frequently results in lower purity compared to
NTA-based purification (Fig. 27.2A and D).

A Ni-IDA Ni-NTA Ni-TED B

M E1 E2 E1 E2 E1 E2
His6-HIV-RT

150
[FU]

Ni-IDA
Ni-NTA
100 Ni-TED

50

C 0
10,000
24 26 28 30
1300
1000 [s]
470
log [Ni] (ppb)

100
100
54

10 4
2
1
W E W E W E
Ni-IDA Ni-NTA Ni-TED

Figure 27.2 (Continued)

IMAC 443

D
EMG1 FYN1 JNK1 MAPKAP5 p38a PIM1

NTA IDA NTA IDA NTA IDA NTA IDA NTA IDA NTA IDA
L F WE FWE LFWEFWE LFWEFWE LFWEFWE LFWEFWE LFWEFWE

E a b g d
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 M
200
97
66
55

36
31

21

14
6

Figure 27.2 Purification of His-tagged proteins with NTA, IDA, and TED.
(A) H6-HIV-RT was expressed in E. coli BL21(DE3) and purified via Ni-IMAC in
the presence of 1 mM DTT under standard conditions (IDA, NTA; see Section 2.10 for
standard conditions) or according to the manufacturer’s recommendations (TED).
Corresponding aliquots of the IMAC elution fractions (E1, E2) were analyzed by
SDS–PAGE and Coomassie staining. (B) Bioanalyzer 2100 lab-on-a-chip analysis of
pooled elution fractions. The peaks from the electropherograms corresponding to H6-
HIV-RT were overlayed. Peak areas directly correlate to the protein amount in the
respective pool fraction. (C) Determination of the nickel content in wash (W) and
pooled elution fractions (E) of the H6-HIV-RT purifications described in (A) and (B).
Nickel was measured by ICP-MS (intercoupled plasma mass spectrometry) at Wessling
Laboratories, Bochum, Germany, and values are provided in mg/l (parts per billion,
ppb). (D) QIAgene constructs carrying optimized human genes were used for expres-
sion of the indicated proteins in E. coli BL21(DE3) LB cultures. Cleared lysates were
divided for purification of His-tagged proteins via Ni-NTA (NTA) and Ni-IDA (IDA),
respectively. Fractions were analyzed by SDS–PAGE and Coomassie staining as fol-
lows: L, cleared lysate; F, IMAC flow-through fraction; W, wash fraction; E, peak
elution fraction. (E) Fragments of H6-tagged proteins named a, b, g, and d expressed in
E. coli were purified under standard conditions using NTA and Cm-Asp tetradentate
ligands loaded with Ni2þ or Co2þ as follows: 1, Ni-NTA; 2, Co-NTA; 3, Co-CmAsp;
4, Ni-CmAsp. Aliquots from the peak elution fractions (2 ml each) were analyzed by
SDS–PAGE and Coomassie staining.
444 Helena Block et al.

The reason for the lower purity may be that leaching of metal from the
tridentate ligand generates charged groups which could act as a cation
exchanger and bind positively charged groups on the surface of proteins.
The lowest metal leaching is obtained if a pentadentate ligand is used
(Fig. 27.2C) which coordinates the ions extremely tightly, and such resins
may represent a valid alternative if low metal ion leaching into the protein
preparation is very important. However, in this case only one coordination
site remains for His tag binding and recovery of His-tagged protein is usually
considerably lower than with IDA or NTA (Fig. 27.2B).
The choice of the metal ion immobilized on the IMAC ligand depends
on the application. While trivalent cations such as Al3þ, Ga3þ, and Fe3þ
(Andersson and Porath, 1986; Muszynska et al., 1986; Posewitz and Tempst,
1999) or tetravalent Zr4þ (Zhou et al., 2006) are preferred for capture of
phosphoproteins and phosphopeptides, divalent Cu2þ, Ni2þ, Zn2þ, and
Co2þ ions are used for purification of His-tagged proteins. Combinations of
a tetradentate ligand that ensures strong immobilization, and a metal ion that
leaves two coordination sites free for interaction with biopolymers (Ni2þ,
Co2þ) has gained most acceptance and leads to similar recovery and purity
of eluted protein. As a typical result using such combinations, Fig. 27.2E
shows the purification of several protein fragments derived from different
genes that have been expressed and purified as His-tagged proteins by Ni2þ
and Co2þ immobilized on NTA and Cm-Asp tetradentate ligands.

2. IMAC Applications
Initially developed for purification of native proteins with an intrinsic
affinity to metal ions (Porath et al., 1975), IMAC has turned out to be a
technology with a very broad portfolio of applications. On the chro-
matographic purification side, the range of proteins was expanded from the
primary metalloproteins to antibodies, phosphorylated proteins, and recom-
binant His-tagged proteins. IMAC is being used in proteomics approaches
where fractions of the cellular protein pool are enriched and analyzed
differentially (phosphoproteome, metalloproteome) by mass spectrometrical
techniques; here, IMAC formats can be traditionally bead based or the ligand
can be used on functionalized surfaces such as SELDI (surface-enhanced laser
desorption/ionization) chips. Other chip-based applications include surface
plasmon resonance (SPR) and allow the immobilization of His-tagged
proteins for quantitative functional and kinetic investigations. In addition,
the IMAC principle has been used as an inhibitor depletion step prior to
PCR amplification of nucleic acids from complex samples such as blood in a
technology called Chelex (Walsh et al., 1991). The most relevant of the
IMAC 445

numerous IMAC applications will be discussed briefly in the following

sections. The main application of IMAC—the purification of His-tagged
proteins—will then be discussed in detail. This section will include problems
and limitations of IMAC, solutions and recent advances in this field.

2.1. Detection and immobilization

In efforts to make use of the specificity and high affinity of His-tagged
proteins to immobilized metal ions, IMAC ligands have been employed for
applications such as protein:protein interactions where proteins need to be
stably immobilized on surfaces. Two applications—ELISA as a diagnostic
tool and chip-based technologies for functional investigations—shall be
briefly described.
Ni-NTA ligands attached to surfaces of microtiter plates are used to
immobilize His-tagged antigens in its soluble and structurally intact form for
serological studies. The directed immobilization via the His tag can be an
advantage to standard ELISA where proteins are randomly adsorbed to
plastic surfaces which destroys the protein structure and hides part of the
protein surface and possible antibody-binding sites. In contrast, IMAC-
based ELISA allows the screening of conformation-dependent monoclonal
antibodies (Padan et al., 1998) and immunosorbent assays with increased
sensitivity ( Jin et al., 2004).
Immobilization of His-tagged proteins on chip surfaces for interaction
studies with other molecules, for example by SPR, is a widely used protein
characterization method. The factor ‘‘stability’’ is important to reduce
‘‘bleeding’’ of the immobilized molecule, and therefore, due to its favorable
binding features described above the NTA ligand is frequently used for
immobilization applications (Knecht et al., 2009; Nieba et al., 1997). How-
ever, interaction of biotinylated proteins to supports coated with streptavi-
din is still considerably stronger. A significant improvement in the stability
of functional immobilization of His-tagged proteins on glass-type surfaces
even at low concentration was achieved by the concept of multivalent
chelator heads where a single ligand molecule carries three NTA moieties
(tris-NTA; Lata and Piehler, 2005; Zhaohua et al., 2006). This development
represents a valid alternative to streptavidin/biotin-based protein immobili-
zation and allows the use of His-tagged proteins without the need for
biotin-labeling of proteins following purification. We have synthesized
the tris-NTA ligand and coupled it to magnetic agarose beads in order to
test whether this approach may be transferred to purification of His-tagged
proteins and allow even more specific single-step recovery from complex
samples. Initial data (Fig. 27.3) indicate that this may indeed be the case;
AKT1 kinase separated from Spodoptera frugiperda-derived cell-free lysate
reactions could be purified with both bead types and purity was slightly
446 Helena Block et al.

tris-Ni-NTA Ni-NTA
F W1 W2 W3 40 250 F W1 W2 W3 40 250 mM imidazole

97
66 His6-AKT1

36
21
14

Figure 27.3 Purification of His6-tagged AKT1 kinase using tris-Ni-NTA and Ni-NTA
magnetic beads. Human C-terminally His6-tagged AKT1 was expressed cell-free in an
insect cell-based lysate (EasyXpress Insect II system) in a 100 ml reaction volume using
a pIX4.0 vector construct; purification was performed using magnetic agarose beads
functionalized with tris-Ni-NTA (left panel) or Ni-NTA (right panel) under standard
conditions (see Section 2.10) in the presence of 0.05% (v/v) Tween-20 using a
magnet-equipped tube holder. Aliquots of purification fractions as follows were
analyzed by SDS–PAGE and silver staining: F, unbound protein; W1, 2, wash fractions
1 and 2 (10 mM imidazole); W3, wash fraction 3 (20 mM imidazole); 40 ! 250, elution-
fractions (40, 60, 100, and 250 mM imidazole, respectively). Identity of the purified His6-
tagged AKT1 protein was verified by anti-His Western blot analysis (data not shown).

higher using tris-Ni-NTA beads. Whether these findings are of general

relevance and whether tris-NTA-based chromatography can be economic
in larger scales remains to be evaluated.
IMAC ligands have also been used successfully as reporter in immuno-
blot type of applications replacing an antibody. His-tagged proteins trans-
ferred to nitrocellulose membrane (Western or dot blot) can be detected by
probing with Ni-NTA conjugated to alkaline phosphatase or horseraddish
peroxidase reporter enzymes (Lv et al., 2003) in a chromogenic or chemi-
luminescence reaction or to quantum dots for fluorescent detection (Kim
et al., 2008). This represents an attractive fast and economic alternative to
antibody-based detection reactions in cases where the high specificity of an
antibody is not required. The specificity of NTA conjugate-based detection
has been increased by generation of tris-NTA conjugates (Lata et al., 2006;
Reichel et al., 2007).

2.2. Purification of protein fractions

IMAC had originally been developed as a group separation method for
metallo- and histidine-containing proteins (Porath et al., 1975). Today,
these features are made use of in proteome-wide studies where the reduc-
tion of the complexity of the system (the proteome) is indispensable for
sensitive analyses of low-abundance proteins. Consequently, preseparation
methods such as liquid, reverse-phase, ion-exchange, and affinity
IMAC 447

chromatography—such as IMAC—have gradually been used in proteomics

to enrich proteins that may otherwise be lost in detection (Loo, 2003; Stasyk
and Huber, 2004). The application of IMAC in proteomics has recently
been reviewed (Sun et al., 2005) and is focused on the enrichment of
phosphoproteins and phosphopeptides and on metal-binding proteins.
In the enrichment step, a complex sample such as a cell lysate or blood is
passed over the IMAC matrix, washed and the fraction of interest eluted by
variation of pH or with high concentrations of imidazole. This fraction is
then analyzed by mass spectrometry (MS) or fractionated further by two-
dimensional gel electrophoresis followed by MS or by additional liquid
chromatography coupled to MS (LC–MS).
Whereas Fe3þ, Al3þ, and Ga3þ are the preferred ions for phospho-
protein research and are usually immobilized to IDA, the ions useful for
IMAC-based analysis of the metalloproteome are the elements copper,
nickel, zinc, and iron which are essential for life. The metalloproteome is
defined as a set of proteins that have metal-binding ability and several aspects
of this proteomics discipline have been reviewed recently (Shi and Chance,
2008; Sun et al., 2005). Proteins with metal-binding affinity can be enriched
by either making use of their ability to bind to certain immobilized Me2þ
ions (e.g., to Me2þ-NTA) or by making use of their bound Me2þ ion by
catching as a Me2þ-protein on an uncharged IMAC ligand (e.g., NTA).
Also, chip-based proteome profiling IMAC methods have been reported
(Slentz et al., 2003) and are in use as a tool in clinical screening applications
for phospho group- and histidine-containing proteins and peptides
(SELDI–IMAC).
IMAC can also be used to bind and separate at least mono- and dinu-
cleotides based on a complex phenomenon accounted for by differential
interplay of affinities of the potential binding sites (oxygen in the phosphate
group, nitrogen and oxygen on the bases, hydroxyl groups on the ribose) to
the immobilized metal (Hubert and Porath, 1980, 1981).
A quite different group-specific separation application of IMAC is
represented by the affinity of antibodies to immobilized metal ions. As the
molecular basis for this interaction an endogenous metal-binding site on the
heavy chain (Hale and Beidler, 1994) and an arrangement of histidine
residues on the antibody (Porath and Olin, 1983) have been discussed.
Adsorption of immunoglobulins from different sources on IMAC matrices
has been reported by many authors (human IgG, Porath and Olin, 1983;
humanized murine IgG, Hale and Beidler, 1994; goat IgG, Boden et al.,
1995). Antibody purification has been successfully performed using various
IMAC formats including gels (Hale and Beidler, 1994; Vancan et al., 2002),
methacrylate polymer (Mészárosová et al., 2003), and membraneous hollow
fibers (Serpa et al., 2005). The mild elution of the protein with salts, costs,
and the robustness of IMAC matrices have been identified as advantageous
over traditional protein A or G chromatography (Serpa et al., 2005).
448 Helena Block et al.

The Chelex method shall be mentioned here as well in order to com-

plete the list of applications which make use of IMAC ligands. Unlike
classical IMAC, however, where an immobilized metal ion is used to purify
a (poly)peptide by its affinity for this transition metal, Chelex represents a
nucleic acid sample preparation method that depletes metal ion inhibitors of
PCR as the downstream application (Walsh et al., 1991). As such, Chelex
resins such as Bio-Rad’s Chelex 100 are uncharged ligands like IDA
coupled to a usually agarose-based matrix. In brief, the procedure works
as follows: a blood or tissue sample is incubated with Chelex resin in the
presence or absence of proteinase K followed by separation of beads from
supernatant containing the nucleic acids. The resulting nucleic acid fraction
is not pure but suitable for amplification by PCR because metal ions have
been removed from the sample which may otherwise catalyze rupture of the
DNA at high temperature during PCR and thus cause PCR inhibition.
Chelex is mainly applied as a fast and inexpensive method to prepare small
samples obtained from biopsies and puncture aspirations for amplification of
small DNA fragments by PCR (Garcı́a González et al., 2004; Gill et al.,
1992).

2.3. Purification of His-tagged proteins

2.3.1. His tags and their effects on protein expression
The most important application of IMAC is purification of recombinant
proteins expressed in fusion with an epitope containing six or more histi-
dine residues, the His tag (Fig. 27.1). Due to the relatively high affinity and
specificity of the His tag a single IMAC purification step in most cases leads
to a degree of purity of the target protein preparation that is sufficient for
many applications. The structure of the tag, that is its position, sequence,
and length, can influence production of a protein on several levels: expres-
sion rate, accessibility for binding to the IMAC ligand, protein three-
dimensional structure, protein crystal formation, and—although to a
minor extent—solubility and activity. The most common form of a His
tag consists of six consecutive histidine residues (H6) which provides a
number of six metal-binding sites high enough to shift the association/
dissociation equilibrium more to the association side leading to stable
binding in most cases (Table 27.1; Knecht et al., 2009). In Biacore experi-
ments, the dissociation rate of a hexahistidine-tagged protein to Ni-NTA
has been determined to approximately 1  10 6 to 1.4  10 8 M at pH 7.0
to 7.4 (Knecht et al., 2009; Nieba et al., 1997). However, the situation on a
planar chip surface is significantly different from a porous agarose bead with
respect to flow characteristics, ligand density, and protein concentration.
Also, the stability of the interaction of the His-tagged protein to the IMAC
ligand is influenced by the grade of accessibility of the tag and by the overall
number of chelating residues (histidine, cysteine, aspartate, and glutamate)
IMAC 449

Table 27.1 Reported His tag sequences (single letter amino acid sequence code)

His tag Reference/vector

H HH HH H H6 Hochuli et al. (1988)
H HH HH HH H H8 pQE-TriSystem,
pTriEx
H HH HH HH HH H H10 pQE-TriSystem-5, -6
H HH HH HH HH HH HH H H14
H QH QH QH QH QH Q (HQ)6 Pedersen et al. (1999)
H NH NH NH NH NH N (HN)6 US patent 7176298
H GH GH GH GH GH Q (H G/Q)6 Pedersen et al. (1999)
H HQ HH AH HG (HHX)3 Pedersen et al. (1999)
K DH LI HN VH KE H HAT Imai et al., 2001 andUS
A HA HN K patent 7176298
(H X*)3–6 (HXa)n US patent 5594115
H X1 H R H X2 Hb (HXH)2R US patent application
2004/0029781
a
X can be D, E, P, A, G, V, S, L, I, and T.
b
X1 can be A, R, N, D, Q, E, I, L, F, P, S, T, W, V; X2 can be A, R, N, D, C, Q, E, G, I, L, K, M, P, S,
T, Y, V.

on the surface of a protein (Bolanos-Garcia and Davies, 2006; Jensen et al.,

2004) and is thus individual to a significant extent. In most cases, that is, if
the His tag is accessible, its affinity—or better termed avidity here—to Ni-
NTA is high enough for column chromatography even under stringent
conditions. Table 27.1 lists the tag sequences reported in the literature and
some unpublished ones tested recently in our laboratories.
A different situation compared to a ‘‘standard’’ purification of a soluble
protein (see Section 2.10) can be encountered when a membrane protein is
to be recovered in the presence of detergents because the deterent micelle
may cover part of or the complete His tag. In such cases, the use of longer
tag sequences or the use of a linker can be helpful to allow binding of the
protein to the IMAC resin (Mohanty and Wiener, 2004). A range of
variations of the His tag including alternating sequences has been proposed
(Table 27.1) and improved binding to IMAC resins postulated but in our
and in the hands of others (Knecht et al., 2009) they have no practical
advantage over the classical Hn tags. What is found to be of greater impor-
tance than the sequence of the His tag itself is its position (N- or C-terminal)
and the amino acids at the N-terminus. The nature of the amino acid
following the N-terminal methionine has been reported to prevent N-
terminal methionine processing and to have a positive effect on the general
protein expression rate in Escherichia coli (Dalb
ge et al., 1990; Hirel et al.,
1989). We and others have evaluated these reports and confirmed that
450 Helena Block et al.

namely, lysine and arginine in position 2 of a protein N-terminus show this

effect (Pedersen et al., 1999; Schäfer et al., 2002a; Svensson et al., 2006). The
high success rate of using N-terminal His (and Strep II) tags is, however, not
only based on the stimulatory effect of the second amino acid on expression
but also seems to have stabilizing impact on the mRNA structure in the
translation initiation region. We compared bacterial and eukaryotic expres-
sion of N- versus C-terminal positioning of the His (and Strep II) tag on
several proteins and analyzed expression level and solubility (Fig. 27.4). N-
terminal tags improved protein expression in most cases.
Systematic investigation of the 50 region of mRNAs showed that hairpin
loops forming in the translation initiation area frequently are the reason for
low expression as they prevent the ribosome:mRNA binding (Cèbe and
Geiser, 2006). Sequence optimization can be performed to destabilize
hairpin formation and improve expression, and a similar result was obtained
when the proteins were expressed with an N-terminal H6 tag. The effects
described in Fig. 27.4 can be explained with this initiator effect of prevent-
ing secondary structures on the mRNA in the translation initiation region.
Similar observations were also made by others (Busso et al., 2003; Svensson
et al., 2006) and this may be the reason for the attractiveness of N-terminal

A E.coli cell-free B Insect cell-free

-S ll /C-H 6 -S ll -H 6
/C /C /C
H6 H6 S ll S ll H6 S 11 H6 H6 S ll S ll H6 S II
N- C- N- C- N - N- N- C- N- C- N- N-
M T S T S T S T S T S T S T S T S T S T S T S T S
30 TNFa
50
MKK3
30
50 TFIIAab

15 TFIIAg
50
TBP
30
75
IRAK4
50

Figure 27.4 Effect of His tag position on protein expression. Proteins were expressed
from PCR product templates generated by two-step PCR using the EasyXpress Linear
Template Kit in E. coli- (A) and insect cell-derived (B) lysates (EasyXpress Protein
Synthesis and EasyXpress Insect II Kits, respectively). Initiator (adapter) primers in the
Linear Template Kit were designed in order to prevent formation of secondary struc-
ture in the translation initiation region on the mRNA and introduce the following tag
sequence(s): N-H6, N-terminal His6 tag; C-H6, C-terminal His6 tag; N-SII, N-terminal
Strep II tag; C-SII, C-terminal Strep II tag; N-H6/C-SII, N-terminal His6 and
C-terminal Strep II tags; N-SII/C-H6, N-terminal Strep II and C-terminal His6 tags.
Corresponding aliquots of total (T) and soluble protein (S, supernatant after centrifu-
gation at 15,000  g for 10 min) were loaded on a SDS gel. Protein bands were
visualized by Western blot analysis using a mixture of Penta anti-His and anti-Strep
tag antibodies. Protein sizes are (kDa) TNFa, 21; TBP, 38; TFIIAab, 55; TFIIAg, 12.5;
MKK3, 39; IRAK4, 55. M, His-tagged protein size markers (kDa).
IMAC 451

His tags. However, in some cases such as the one of IRAK4 the C-terminal
His tag has a more pronounced effect on both expression rate and solubility
(Fig. 27.4). Recently, expression of an insect toxin in E. coli was reported
where the similar observation of higher solubility and thermostability of the
C-terminally His-tagged form was made (Xu et al., 2008). The authors
discussed that the C-terminal tag stabilized the overall protein structure.
Other groups found the His tag to contribute slightly negative to solubility
when compared to the untagged protein but to improve yield when fused to
the C-terminus (Woestenenk et al., 2004). All in all, these data suggest that an
evaluation of at least N- and C-terminally tagged variants of a protein will
increase the chance to obtain reasonable expression and quality of a recom-
binant protein. For expression of proteins to be secreted, tags should be
placed at the C-terminus to prevent interference with membrane trafficking.

2.4. General considerations of protein purification by IMAC

There are several advantages of IMAC for purification of His-tagged
proteins compared to other affinity chromatography principles as the reason
for being the most widely used chromatographic technique (Biocompare,
2006; Derewenda, 2004). Besides low costs and the simplicity of use the
robustness of IMAC is certainly its most striking feature: (i) the His-tag:
ligand interaction works under both native and denaturing conditions such
as 8 M urea or 6 M guanidinium hydrochloride (Hochuli et al., 1988)
enabling subsequent on-column refolding ( Jungbauer et al., 2004) as well
as (ii) both oxidizing and reducing conditions, (iii) protein binding with-
stands a broad spectrum of various chemicals of different types (Table 27.1
summarizes chemical compatibilities for Ni-NTA IMAC and some limita-
tions), (iv) its relatively high affinity and specificity allows high capturing
efficiency even in the presence of high protein titers, and (v) the scalability
of the purification procedure.
Despite the broad compatibility, IMAC has its limitations. Obviously,
the use of chelating agents has to be avoided which can be a disadvantage as
EDTA, a potent inhibitor of metalloproteases, can only be applied in low
concentrations. Care should also be taken with the use of other potentially
chelating groups such as Tris, ammonium salts, and certain amino acids
(Table 27.2).
Until recently, the use of strong reducing agents such as DTT in IMAC
processing has been regarded as problematic because of reduction of nickel
and, as a consequence, suspected increase of nickel concentrations in
protein preparations. However, we found that moderate concentrations of
DTT are fully compatible with NTA-based purification. This is shown, for
example, in Fig. 27.5A for HIV-1 reverse transcriptase (RT) purified with
unaffected efficiency in the presence of up to 10 mM DTT. Also, RT
activity is not influenced by these conditions and both end-point
 Table 27.2 Chemical compatibility of purification of His-tagged protein using agarose-based IMAC (Ni-NTA) resins and its limitations
452

IMAC chemical compatibility

Component Limitation (up to) Component Limitation (up to)
Buffers Salts
Na-phosphate Recommended, limit NaCl 4M
not known
Phosphate citrate Limit not known MgCl2 4M
Tris–HCl, HEPES, MOPS 100 mM CaCl2 5 mMc
Citrate 60 mM NaHCO3 Not recommended
Detergents (in 300 mM NaCl) Ammonium salts Not recommended
n-Hexadecyl-b-D-maltoside 0.0003% (w/v) Protease inhibitors
n-Tetradecyl-b-D-maltopyranoside 0.005% (w/v) EDTA 1 mMa
n-Tridecyl-b-D-maltopyranoside 0.016% (w/v) Commonly used protease Compatible in effective
inhibitorsd concentrations
Brij 35 0.1% (v/v) Complete cocktail 1 concentrated
(EDTA-free)
Digitonin 0.6% (w/v) Denaturants
Cymal 6 1% (w/v) Urea 8M
n-Nonyl-b-D-glucopyranoside (NG) 1% (w/v) Gu–HCl 6M
n-Decyl-b-D-maltopyranoside (DM) 2% (w/v) Amino acids
n-Dodecyl-b-D-maltoside (DDM) 2% (w/v) Histidine 1–2 mMb
C12–E9 1% (w/v) Glycine Not recommended
n-Octyl-b-D-glucopyranoside (OG) 1.5% (w/v) Cysteine Not recommended
Triton X-100, Tween, NP-40 2% (v/v) Glutamate Not recommended
Triton X-114 2% (v/v) Aspartate Not recommended
Fos-Cholines 0.05% (w/v) Arginine 500 mM
 Dodecyldimethyl-phosphineoxide 0.15% (w/v) Organic solvents
N,N-Dimethyldodecylamine-N-oxide 0.7% (w/v) Isopropanol 60% (v/v)e
(LDAO)
CHAPS 1% (w/v) Ethanol 20% (v/v)
Laurosyl-sarcosine 1% (w/v) Reducing reagents
SDS 0.3% (w/v)a b-ME 20 mM
Other TCEP 20 mM
EGTA 1 mMa DTT 10 mM
Imidazole 10–20 mMb DTE 10 mM
Hemoglobin Not recommended
Glycerol 50% (v/v)
a
Has been used successfully in the indicated concentration but should be avoided whenever possible.
b
Dissociates His-tagged proteins at high concentrations.
c
Should be avoided in combination with Na-phosphate.
d
Include, for example, Aprotinin, Leupeptin, PMSF, and related serine protease inhibitors: Pepstatin, Antipain, Bestatin, E64, Benzamidine.
e
Compatible with Ni-NTA purification and endotoxin removal according to Franken et al. (2000), but in this concentration is incompatible with reuse of the
chromatographic media (data not shown).
This table provides some of the most relevant tested substances and concentrations and may not be complete or represent the maximal concentrations compatible with
purification of His-tagged proteins.Abbreviations: b-ME, b-mercaptoethanol; TCEP, tris(2-carboxyethyl)phosphine hydrochloride; DTT/DTE, dithio-threitol/-
erythritol; Gu–HCl, guanidinium hydrochloride.
453
454 Helena Block et al.

A B
0 1 5 10 mM DTT
M L F W E1E2 E3 E4 E2 E3 E4 E2 E3 E4 E2 E3 E4
200
116/97
66 His6- M C NTC
55 HIV-RT 0 1 5 10

36
31
b -actin
21
14
6

[Ni] (ppb) 150 100 250 150

Figure 27.5 IMAC under reducing conditions. (A) His6-tagged HIV-1 reverse tran-
scriptase (RT) was purified under standard conditions in the presence of the indicated
DTT concentrations. Aliquots of each chromatographic fraction of the purification
without DTT (M, markers; L, lysate; F, flow-through; W, wash; E, elution fraction)
and of corresponding volumes of elution fractions 2–4 containing DTT were analyzed
by SDS–PAGE and Coomassie staining. Elution fractions were pooled and analyzed for
nickel content by ICP-MS at Wessling Laboratories, Bochum, Germany; nickel
concentrations are given in mg/l (parts per billion, ppb). (B) Equal amounts of HIV-1
RT purified in the presence of the indicated concentrations of DTT were used in
duplicates to reverse transcribe a 1.5 kb b-actin cDNA which was subsequently ampli-
fied by PCR and the PCR products were analyzed by agarose gel electrophoresis and
ethidium bromide staining. Reducing conditions of at least 1 mM DTT were found to
be required for full RT activity (compare weaker amplification with RT purified in the
absence of DTT). C, Omniscript positive control; an equal amount of Omniscript RT
protein was used; NTC, no template negative control.

(Fig. 27.5B) and quantitative real-time RT-PCR (data not shown) show no
inhibitory effect which could origin from high heavy-metal-ion concentra-
tions. Even though nickel ions may become reduced by DTT leading to a
color change of the resin bed they do not increasingly leach from the ligand
(Fig. 27.5A), and resins processed under reducing conditions can be repeat-
edly reused and regenerated (data not shown). These findings suggest that
despite a color change as a consequence of nickel reduction by DTT the
resin is still functional. TCEP, a different, nonthio-based reducing reagent is
more and more replacing DTT and b-ME in protein purification by IMAC
as it seems to be more selective to reduction of disulfide bonds, is odorless
and more stable in aqueous solution. We recommend the use of TCEP with
Ni-NTA chromatography in a concentration of 1–5 mM.

2.5. Copurifying proteins on IMAC and what to do about it

IMAC leads to high protein purity upon single-step chromatographic
purification in many cases (Figs. 27.2A, D, E and 27.5A; Bornhorst and
Falke, 2000; Schmitt et al., 1993). The higher the purity is the closer the
amount of IMAC resin applied in the chromatographic step has been
IMAC 455

correlated to the amount of recombinant His-tagged protein present in the

sample to be processed. The reason is that proteins naturally displaying
surface motifs suitable to interact with an immobilized metal ion may bind
to the resin although usually with slightly lower affinity than a flexible His
tag. Most His-tagged proteins will therefore displace proteins with natural or
accidental surface motifs. However, there are some proteins where the local
density of chelating amino acids such as histidine is so high that they will bind
to immobilized metal ions almost inevitably. In general, mammalian systems
have a higher natural abundance than bacterial systems of proteins containing
consecutive histidines (Crowe et al., 1994), and a prominent example in
human cells is the alpha subunit of transcription factor TFIIA which has
seven consecutive and surface-exposed histidine residues and can be purified
via IMAC from natural sources under native conditions (DeJong and
Roeder, 1993; Ma et al., 1993) and which is observed frequently as a signal
band of 55 (the ab precursor) or 35 kDa (the a subunit) of TFIIA in Western
blots using an anti-His tag antibody. Another example is the human tran-
scription factor YY1 with 11 consecutive histidines (Shi et al., 1991). In E.
coli, the proteins observed to copurify with His-tagged target proteins can be
divided into four groups: (i) proteins with natural metal-binding motifs, (ii)
proteins with histidine clusters on their surfaces, (iii) proteins that bind to
heterologously expressed His-tagged proteins, for example by a chaperone
mechanism, and (iv) proteins with affinity to agarose-based supports
(Bolanos-Garcia and Davies, 2006). Whether or not one of the E. coli
proteins is copurified is not easily predictable. For example, a protein from
group (ii) sometimes reported to copurify with Ni-NTA is the 21 kDa SlyD,
however, we have never observed this one in our lab upon purification from
E. coli BL21(DE3), DH5a, M15(pREP4), and other strains. This may be
explained by the fact that many of these impurities are stress-responsive
proteins, suggesting that the cultivation conditions and the bacterial strain
have an influence on their abundance and, a consequence, their appearance
as a contaminating species in the target protein preparation; it is therefore
recommended to induce as little stress as possible during cultivation of E. coli
cells (e.g., by using shake flasks without baffles). Furthermore, some copuri-
fying proteins seem to have a binding preference for Co over Ni (or other
ions) and others vice versa.
Several options to get rid of these copurified proteins or prevent their
adsorption early on have been evaluated and some of them shall be discussed in
the following section. These options include (i) performing additional purifi-
cation steps, (ii) adjusting the His-tagged protein to resin ratio, (iii) to use an
engineered host strain that does not express certain proteins, (iv) using an
alternative support, and (v) tag cleavage followed by reverse chromatography.
Suitable additional purification steps include classical chromatographic
techniques such as ion-exchange (IEX) and size exclusion chromatography
(SEC) whereas IEX has the higher separation power. However, as SEC not
456 Helena Block et al.

only serves to separate molecules by size and helps to remove aggregates of

high molecular weight but also can be used to desalt the preparation and
provide conditions suitable for certain downstream applications, it is
frequently used as a standardized procedure by, for example, high-throughput
labs such as structural biology consortia (Acton et al., 2005; Gräslund et al.,
2008) who perform protein crystallization or NMR spectroscopy. Although
IMAC–SEC (as opposed to IMAC–IEX) can be performed as a standar-
dized procedure without having to take into account the protein biochem-
istry such as pI the separation range of a given SEC column will not be
suitable for all separation tasks and a range of SEC columns may have to be
held in place. Another issue with the application of IEX and SEC is that in
order to make use of the full power provided by these technologies, costly
equipment such as an automated chromatography system is required which
frequently precludes multiparallel processing leading to low throughput.
Affinity purification such as IMAC can usually be done as a bind–wash–
elute procedure in a bench top/gravity flow mode. By introduction of a
second affinity tag (e.g., StrepII, GST, Flag) into the expression construct a
two-step purification leading to high-purity protein preparations is enabled
as a bench-top two-step affinity chromatography procedure (Cass et al.,
2005; Prinz et al., 2004).
As mentioned earlier, adjusting the amount of His-tagged protein to be
recovered to the binding capacity of the used IMAC resin can help to
improve protein purity by preventing copurification of proteins with certain
affinity to IMAC resins. However, the amount of His-tagged protein is
usually unknown unless a pilot experiment is performed to estimate the
content of target protein. While this is an option, the better way would be
to exclude the presence of such copurifying proteins by expressing the target
protein in an engineered strain where the respective genes have been
knocked out. However, to our knowledge results from work with such
strains have not yet been reported and the experience with knockout strains
in protein production is still low. Also, it does not seem realistic that an E. coli
strain can be generated which lacks 17 proteins reported to bind to IMAC
resins including as important functions as superoxide dismutase and iron-
uptake regulation (Bolanos-Garcia and Davies, 2006) and which is still well
viable under stress situations such as protein overproduction.
A different approach to improve the purity of proteins recovered from
IMAC has been reported that made use of dextran-coating of an agarose
matrix, the constituting material of the most widely used chromatographic
supports (Sepharoses, Superflow, Agaroses), and which prevented copur-
ification of proteins with affinity to these matrices (Mateo et al., 2001).
Dextran-coated beads, however, are not readily commercially available as
IMAC resins and this measure only helps to preclude proteins with affinity
to agarose and not to the immobilized metal or to the target protein. Silica-
based IMAC supports also prevent adsorption of proteins with affinity for
IMAC 457

agarose and, in addition, have good pressure stability which makes them
suitable for high-resolution HPLC applications but the silica resins
frequently suffer from low binding capacity and limited resistance to high
pH sanitization procedures. A very recent method that avoids the need to
use solid chromatographic supports completely is called affinity precipita-
tion (Hilbrig and Freitag, 2003) and has been applied to IMAC (Matiasson
et al., 2007). Here, the IMAC ligand is chemically coupled to a responsive
polymer which, after binding to the His-tagged protein, can be aggregated
upon change of environmental conditions such as pH or temperature and
can thus be precipitated by centrifugation. Protocols for its use are still
relatively complicated but as soon as robust and easy to use commercial
materials are available this method may have the potential to play an
important role in IMAC applications. Using a ligand in solution could
overcome steric hindrance of the binding of some His-tagged proteins to
an immobilized ligand as well as mass transport limitations of porous
chromatographic media. Moreover, it is in line with a trend in industrial-
scale chromatography toward single-use disposable materials.
Yet another approach has recently been reported that can be applied for
protein separation from lysates after cell-free expression (Kim et al., 2006).
An E. coli-derived lysate was preincubated with Ni-NTA magnetic agarose
beads to remove proteins with affinity to Ni-NTA prior to template
addition and protein expression; the expression capacity of the S30 extract
was found to remain unaltered and the Ni-NTA purified His-tagged
protein fractions to be of higher purity than without pretreatment.
While the aforementioned strategies to improve the purity of MAC
protein preps have proven useful in many cases they are not generally
applicable and successful. There is a method, however, that almost meets
the criterium of universal applicability regarding improvement of purity,
and it has the additional benefit of resulting in a protein native or near-
native structure: proteolytic His tag cleavage using a His-tagged protease
followed by reverse IMAC (Block et al., 2008). This strategy overcomes the
copurification issue by passing the proteolytically processed protein under
similar or identical conditions over the same column, and the proteins that
bound to the IMAC resin as impurities in the initial purification step will
bind to the same resin again while the cleaved, that is untagged, target
protein is collected in the flow-through fraction (reverse or subtractive
IMAC mode). It can be performed with both exo- and endoproteases that
themselves carry an (uncleavable) His-tag (Nilsson et al., 1997; Polayes et al.,
2008) but the exoproteolytical approach has the advantage that it is faster
and results in a protein with native structure with no vector-derived amino
acids (Arnau et al., 2006; Block et al., 2008; Pedersen et al., 1999). This
approach is especially suitable for demanding downstream applications such
as protein crystallization or biopharmaceutical production. Notably, the
method requires only a single chromatography column to achieve an
458 Helena Block et al.

extremely high degree of purity. An application is shown in Fig. 27.6 where

TNFa was crystallized in its IMAC one-step purified form (A, B, C, D) or
in its subtractive-IMAC processed form as described above (A, E, F, G).
Although already His-tagged TNFa purified on Ni-NTA appeared highly
pure upon SDS–PAGE and Coomassie staining (Fig. 27.6A, lane IMAC)
analysis on a silver-stained 2D gel showed impurities (Fig. 27.6B). These
impurities are removed by reverse mode IMAC resulting in a protein
preparation of extremely high purity (Fig. 27.6E). Figure 27.6 also demon-
strates the influence the His tag can have on protein crystallization. While
both the tagged and the mature native form of TNFa eluted from a SEC
column as a trimer (Fig. 27.6C and F), they crystallized under significantly
different conditions and resulted in different crystal forms (His6-TNFa:
tetragonal, Fig. 27.6D; TNFa: rhombohedral, Fig. 27.6G). Nevertheless,
calculated structures both were in accordance with the one deposited in the
pdb (data not shown). However, when we attempted to perform the same
experimental workflow with His-tagged and native mature IL-1b, we were
unable to crystallize the His-tagged cytokine (Block et al., 2008; and data
not shown). Our data confirm the observation also made by many others
that it is frequently possible to crystallize tagged proteins but suggest that it
makes sense to provide a tag cleavage option when designing an expression
construct.

2.6. IMAC for industrial-scale protein production

IMAC for production of proteins in industrial scale, for example for use as
biopharmaceutical, has not been used until quite recently due to worries
regarding the potential immunogenicity of a His tag sequence and because
of allergenic effects of nickel leaching from an IMAC matrix. However,
nickel concentrations typically observed in protein preparations obtained
from tetradentate IMAC resins are low and content in expected daily doses
of a biopharmaceutical will be far below the typical daily intake of nickel
and the permanent nickel body burden (Block et al., 2008). Removal of
artificial sequences from recombinant proteins by the use of protease has
been discussed above as the way to go for use in humans but proteins
carrying a His tag have been successfully used for vaccination (Kaslow and
Shiloach, 1994; Stowers et al., 2001) or are presently commercialized as
drugs (unpublished). IMAC is a chromatography method that can simply be
scaled linearly from milliliter to liter bed volumes (Block et al., 2008;
Hochuli et al., 1988; Kaslow and Shiloach, 1994; Schäfer et al., 2000) and
Ni-NTA Superflow columns in dimensions up to 50 l are in use for
biopharmaceutical production processes (F. Schäfer, personal communica-
tion). Compatibility of IMAC matrices with a wide range of chemicals such
as chaotropics, salts, organic solvents, and detergents (Table 27.2) facilitates
adaption to the specific needs of the production of the individual protein.
 e
m
Zy
C G C
A A TA M
A
I M sI
M C F W 10′ 30′
170
130
100
72
55
43
D G
34
26
0.4 mm 0.3 mm
His6-TNFa
17 TNFa

C His6-TNFa TNFα
B E F
mAU
pH 3 pH 10 pH 3 pH 10 mAU
800 82.36
800 85.12
600
600
400
400
His6-
200 TNFa TNFa 200
0
0
0 20 40 60 80 100 ml
0 20 40 60 80 100 ml

Figure 27.6 Removal of copurifying proteins by His tag cleavage and reverse IMAC. (A) His6-TNFa expressed in E. coli was purified via Ni-NTA
Superflow and processed using the TAGZyme exoproteolytic system as described (Sch€afer et al., 2002a). (B, E) 2D gel electrophoresis and silver
staining of His6-TNFa and TNFa, respectively, was performed as described (Block et al., 2008). The subband pattern in the first dimension between
pI 6.7 and 5.8 for both His6-TNFa and TNFa is in accordance with the report for TNFa produced in yeast (Eck et al., 1988). (C, F) Analytical size
exclusion chromatography (SEC) on HR 10/30 Superdex 200 was run with 1 TAGZyme buffer (Sch€afer et al., 2002a). (D, G) His6-TNFa
tetragonal crystal (D) formed in 2.7 M MgSO4, MES, pH 5.5 and diffracted to 2.5 Å (homelab X-ray source, FR591 Nonius Bruker);
TNFa rhombohedral crystal (G) formed in 1.8 M NH4SO3, 200 mM Tris–HCl, pH 7.8 and diffracted to 2.0 Å (ESRF synchrotron). The size of
typical crystals (mm) is indicated by the bar in (D) and (G).
460 Helena Block et al.

For example, bacterial endotoxins (lipopolysaccharides) can be eliminated

from the protein product during chromatography by including a wash step
using a detergent (Triton X-114, Block et al., 2008) or an organic solvent
(60% isopropanol; Franken et al., 2000). The suitability of IMAC for
industrial production purposes has been demonstrated and it can be
expected that IMAC-based processes become increasingly used in the future
due to its robustness and relatively low requirements for individual
optimization.

2.7. High-throughput automation of IMAC

Due to its robustness, universal applicability, and widespread use, IMAC is
also an ideal tool for multiparallel screening for protein expression and
solubility. This is mostly done in the convenient 96-well format using
agarose- or magnetic bead-based IMAC resins in SBS footprint filter or
microplates on 96-well magnets or plate centrifuges (Braun et al., 2002;
Büssow et al., 2000). As the complete expression and the simple bind–wash–
elute IMAC purification workflow is easily miniaturizable in microplate
formats it was also suitable for hands-on free automation using liquid
handling laboratory robots (see Lesley, 2001, for an overview). The auto-
mated steps covered the workflow to various extents, ranging from protein
purification from manually generated E. coli lysates (Lanio et al., 2000),
E. coli or insect cell lysis, lysate clarification, and protein purification (Garzia
et al., 2003; Schäfer et al., 2002b; Scheich et al., 2003), to automation of
complete workflows from construct cloning to protein analysis (Acton et al.,
2005; Hunt, 2005; Koehn and Hunt, 2009). Recently, we added another
series of protocols and consumables for purification of His-tagged proteins
from E. coli or eukaryotic cells or cell-free lysates to the list of options: a
conceptually new lab automation instrument allows isolation of microgram
to milligram amounts of proteins from a variable number of samples (with a
random-access sample feeding option) using ready-to-go prefilled cartridges
that provide enzymes, buffers, and Ni-NTA magnetic beads for lysis and
purification. Figure 27.7 describes an expression and purification screening
using a set of 24 optimized-gene constructs for production of human
proteins. Between 1.4 and 35 mg of highly pure protein was obtained
under native conditions. Proteins that could not be purified under native
conditions were obtained upon purification under denaturing conditions,
and Western blot analyses using an anti-His antibody showed the absence of
cross-contaminations between wells (data not shown). Protein resulting
from such high-throughput purification experiments can be used for
functional assays (e.g., interaction studies), characterization of protein
random mutagenesis, solubility analyses, and clone screening.
The next step following an expression screening is frequently a scale-up with
a limited number of proteins or clones for production of milligram amounts
IMAC 461

5
SMS1 PK

1 1
p3 EB D

TN g F
IF -6 1A
A

R C

2
N S

D
IL G1

C R
ET PK

Ju C1

Y C2
IF -C

FY Fa
IL B

BI 2

C K1
PI a
EM 1

A
A

C N
N 8a

SM1
JN 1
Fκ
M

IL -4

N
M -7

M
Len
A

D
Y
f
M

PI
200
116/97
66
55

36
31

21

14

Figure 27.7 High-throughput protein purification screening on a new automation

platform. E. coli BL21 (DE3) cells were transformed with QIAgene constructs carrying
optimized human genes for expression of the indicated proteins in 1 ml LB cultures in a
96-deep well block. Cells were harvested by centrifugation and the pellet block placed
onto the sample input drawer of the QIAsymphony SP instrument. Cells were
resuspended and lysed and His6-tagged proteins purified from the crude lysates using
Ni-NTA magnetic agarose beads and buffer solutions provided in the QIAsymphony
cartridge setup. A 5 ml of each elution fraction was analyzed by SDS–PAGE and
Coomassie staining. Expected protein sizes (kDa) for the individual proteins were
as follows: PIM1, 40; EMG1,30; IL-4, 15.5; IL-7,18; MAPKAPK5,55; ETS1,55;
SMARCD1,60; CREB1,50; p38a, 40; NFkB1A, 40; IL-6,21; IFNa, 20; PIM2, 40;
BIRC5, 30; Jun, 50; Lef1, 55; JNK1, 45; CM-CSF, 15; IFNg, 17; TNFa, 17; FYN, 40;
CDC2, 35; YY1, 66; SMAD2, 55. M, markers (kDa).

of protein for animal immunization, structural, or pharmacokinetic studies.

Single proteins can be purified using standard ÄkTA or FPLC systems,
and an ÄkTA system for slightly increased throughput has been developed
(ÄkTAxpress). However, systems with a significantly higher throughput,
lower complexity, and more dedicated to one-step (mainly IMAC) affinity
purification have been reported (Steen et al., 2006; Strömberg et al., 2005)
and are in use in high-throughput projects such as the human protein atlas
project (Hober and Uhlén, 2008).

2.8. Special applications: Purification of membrane proteins

Membrane proteins have received the highest attention of all protein classes
in the past few years due to their enormous importance as drug targets.
In fact, more than 50% of all currently commercialized drugs as well as those
under development target membrane proteins (Drews, 2000). Furthermore,
membrane proteins account for approximately 30% of the human
462 Helena Block et al.

proteome. However, in contrast to soluble proteins, very little is known

about the biology and structure of membrane proteins which is reflected by
the underrepresentation of membrane protein structures in public databases
such as the pdb (<1%). While structure determination remains a challenge,
purification of membrane proteins by a more or less standardized IMAC
procedure has recently become significantly simpler by the development of
new detergents and consequent screening for the most suitable detergent for
both resolubilization and purification (Eshaghi et al., 2005; Klammt et al.,
2005; Lewinson et al., 2008). We have screened more than 50 of the most
frequently used detergents for their IMAC compatibility and their resolubili-
zation and Ni-NTA purification efficiency testing more than 10 bacterial and
human membrane proteins and have reduced the number to seven powerful
detergents as follows: Cymal 6 (Cy6), n-nonyl-b-D-glucopyranoside (NG),
n-octyl-b-D-glucopyranoside (OG), n-decyl-b-D-maltopyranoside (DM),
n-dodecyl-b-D-maltoside (DDM), FOS-choline-16 (FC16), and N,
N-dimethyldodecylamine-N-oxide (LDAO). In screenings, we achieved a
positive result for efficient resolubilization from E. coli or insect cell
membrane fractions and IMAC purification for at least one detergent in
each case. In Fig. 27.8, the detergent screening and Ni-NTA purification of
His-tagged human Caveolin 1 is shown as an example.
IMAC compatible detergents are listed in Table 27.2 whereas this list
may not be complete. Nevertheless, binding of a His-tagged membrane
protein to IMAC resins seems not to work in combination with certain
O

A B
M

16
A
M

M L R F W E1 E2
G
D

y6
LD
G

FC

M
N
D

D
O

80 72
70
60 55
50 36
40 28 His6-
30
CAV1
His6-
20 CAV1 17

15 11

Figure 27.8 Detergent screening and IMAC purification of His6-tagged human

Caveolin 1 expressed in E. coli C41(DE3) in the presence of LDAO. The gene optimized
for expression in E. coli was synthesized and cloned into pQE-T7 with an N-terminal His6
tag. (A) Detergent screening; the membrane fraction from 200 ml E. coli culture divided
into seven portions was mixed with detergent as indicated, centrifuged at 20,000  g and
aliquots of detergent-soluble protein analyzed by Western blot (Penta anti-His antibody).
Detergents are given in the text. (B) Caveolin 1 purification upscale; His6-tagged
Caveolin 1 (CAV1) was purified from 200 ml E. coli culture volume via Ni-NTA
Superflow in the presence of 30 mM LDAO. Fractions were analyzed by SDS–PAGE
and Coomassie staining: M, markers (kDa); L, E. coli total lysate; R, supernatant from
resolubilization (loaded onto IMAC column); F, IMAC flow-through fraction; W, wash
fraction; E, elution fractions.
IMAC 463

detergents, including some of the listed ones even though resolubilization is

fine. These phenomena depend on the protein–detergent combination and
seem to arise from the detergent micelle around the protein partially or
completely hiding the His tag. The use of longer tag sequences such as a H10
tag has become popular and seems to overcome such limitations of mem-
brane protein recovery and improve affinity to IMAC resins (Byrne and
Jormakka, 2006; Grisshammer and Tucker, 1997; Mohanty and Wiener,
2004; Rumbley et al., 1997).

2.9. Special applications: Purification of zinc-finger proteins

Another big protein superfamily which—due to technical issues—shall be
mentioned in the context of IMAC is the group of proteins containing zinc-
finger motifs. The C2H2 zinc-finger transcription factors alone with over 600
members represent more than 2% of the human proteome (Knight and
Shimeld, 2001). In these proteins, zinc ions are coordinated by a defined spatial
organization of two cysteine and two histidine residues each per finger, and a
single polypeptide usually contains four or five of these motifs. Purification of
such a metalloprotein via a metal ion chelated in a similar, that is tetradentate,
manner deserves some reflection regarding the best way of IMAC purification.
Despite the strong interaction of metal ions with tetradentate IMAC
ligands, it may not be excluded that a metal from the resin could exchange
with the zinc in the zinc finger. Nickel, the most frequently used metal in
IMAC protein purification has quite similar physicochemical properties to
zinc and might therefore well replace it in such a metal-binding motif, and
the nickel concentration in an IMAC resin (approximately 15 mM ) is
usually considerably higher than the concentration of a His-tagged target
protein (mM range). In order to analyze a potential metal ion exchange
between matrix and metalloprotein, we expressed the His6-tagged C2H2
zinc-finger containing transcription factor YY1 in E. coli and purified the
protein in parallel using Ni-NTA and Zn-NTA. YY1 could be purified to
high purity by both nickel- and zinc-based IMAC (Fig. 27.9A, lane E).
Both YY1 preparations were subjected to determination of nickel and
zinc content by ICP-MS (intercoupled plasma mass spectrometry), a quan-
titative method frequently applied to detect trace metals in biological
systems (Shi and Chance, 2008). The zinc-IMAC purified protein prepara-
tion contained 35.6 mM zinc corresponding to approximately six Zn2þ ions
per YY1 polypeptide and some Ni2þ which can be attributed to residual
traces from the buffers (Fig. 27.9B, right). The protein recovered by nickel-
IMAC, however, contained more than 25 mM Ni2þ and 14.2 mM Zn2þ
(Fig. 27.9B, left), corresponding to again approximately six Me2þ ions per
YY1 polypeptide. The Me2þ:polypeptide molar ratio of 6 is higher than the
four zinc-finger motifs in YY1 reported in databases (http://www.uniprot.
org/uniprot/P25490) which may in part be explained with additional metal
464 Helena Block et al.

A B

40
66 35.6
His6-YY1
55
30
36 25.6

[Me2+] (mM)
31
20
14.2
21
10
14
0.9
0
M L FWE FWE Analyzed Me2+ ion: Ni Zn Ni Zn
Ni-NTA Zn-NTA IMAC resin : Ni-NTA Zn-NTA

Figure 27.9 Purification of His6-tagged zinc-finger protein YY1 by Ni- and

Zn-IMAC. (A) IMAC purification. His6-YY1 was expressed in E. coli Bl21 (DE3) and
a cleared lysate was generated as described below. NTA resin was treated with ZnCl2 as
described below to generate Zn-charged NTA (Zn-NTA). Protein purifications were
performed using the same cleared lysate preparation under standard conditions
(see Section 2.10) and aliquots of the chromatographic fractions (L, cleared lysate; F,
flow-through fraction; W, wash fraction; E, elution fraction) analyzed by SDS–PAGE
and Coomassie staining. Protein concentration of elution fractions were determined
by the Bradford method to 0.3 mg/ml (6.7 mM, Ni-NTA) and 0.25 mg/ml (5.6 mM,
Zn-NTA). (B) Quantitative detection of metals. Both elution fractions were then
dialyzed for 16 h at 4  C versus three changes of buffer NPI-10 to remove metal ions
from the solution. Dialyzed proteins and fresh dialysis buffer were analyzed for
nickel and zinc content by ICPMS at Wessling Laboratories (Bochum, Germany) and
the values determined for the buffer subtracted from the protein analyses. Values
(ppm, mg/l) were converted to molar units using the mean relative masses for Ni
(58.7) and Zn (65.4).

ions binding elsewhere on the protein. Nevertheless, the data from

Fig. 27.9B suggest that in the case of YY1 there is a significant exchange
of metal ions between the charged IMAC ligand and the zinc fingers.
Metal-affinity purification of proteins with C2H2 zinc-finger motifs can
be effectively performed but an IMAC ligand charged with zinc, if applica-
ble, should be chosen in order to obtain a preparation with an intact and
homogenous zinc-finger composition.

2.10. Protein purification protocols

In this section, the standard IMAC protocols will be described briefly. The
procedures have been optimized for the use of tetradentate (i.e., Ni-NTA)
resins but should be transferable to tridentate (IDA-based) resins as well. TED
resins behave different and the manufacturers’ recommendations should be
followed (e.g., proteins elute at significantly lower imidazole concentrations
and consequently imidazole level in wash buffers should be kept low).
IMAC 465

Here, we will provide recommendations for agarose-based Ni-NTA

resins (Superflow, Agarose) regarding purification, cleaning, and rechar-
ging. These can be regarded as an update of the more detailed description
provided in the Ni-NTA handbook (QIAGEN, 2003). Buffers provided in
the following may be supplemented according to the needs of the individual
protein (e.g., in order to generate reducing conditions, to stabilize the
protein with glycerol, or to provide the presence of cofactors).

2.10.1. Purification of His-tagged proteins under native conditions

1. Lyse cells using the basis buffer NPI-10 (50 mM NaH2PO4, 300 mM NaCl,
10 mM imidazole, pH 8.0) supplemented with a suitable lysis reagent.
For E. coli lysis, we recommend to use standard hen egg white lysozyme
at 1 mg/ml final concentration because lysis is extremely efficient (if cells
had been frozen) and lysozyme is inexpensive. Lysozyme is reliably washed
from IMAC resins and does not occur in elution fractions (Figs. 27.5A,
27.6A, and 27.8; Block et al., 2008). Other suitable methods are based on
detergents (e.g., 1 % (v/v) CHAPS or proprietary solutions) or physical
treatment (sonication, high pressure/depressurization homogenization).
For cultures derived from insect or mammalian cells, 1% Igepal CA-630
(former name NP-40) is recommended. For reduction of lysate viscosity,
the addition of a nuclease is helpful, and Benzonase (3 units/ml bacterial
culture) has been shown to be robust and its removal in wash steps from
IMAC resins works reliably and can be checked by a commercial ELISA
(Block et al., 2008). Incubate the lysate on ice for 30 min.
2. Generate a cleared lysate by centrifugation for 30 min at 10,000  g
and 2–8  C. Collect the supernatant.
3. Load the cleared lysate onto the resin equilibrated with 5 bed volumes
(bv) of NPI-10 and allow to flow through at approximately 1 ml/min
(column of 1 ml bv) or let flow through (gravity flow application).
A suitable linear flow rate during binding is 155 cm/h corresponding
to 1 ml/min of a column with a bed diameter of ~ 7 mm (1 ml HisTrap
and Ni-NTA Superflow Cartridges). If applicable to the workflow, we
recommend to perform binding in batch mode as this is most efficient
with agarose-based resins in general; for this, add the required volume of
lysate to the equilibrated resin and incubate for 1 h rotating end-over-
end at 2–8  C.
4. Wash the Ni-NTA column with 10 bv wash buffer NPI-20 (50 mM
NaH2PO4, 300 mM NaCl, 20 mM imidazole, pH 8.0).
If binding had been performed in batch, pour the binding suspension
into a suitable flow-through column.
5. Elute His-tagged protein with 5 bv of elution buffer NPI-500 (50 mM
NaH2PO4, 300 mM NaCl, 500 mM imidazole, pH 8.0).
466 Helena Block et al.

2.10.2. Purification of His-tagged proteins under

denaturing conditions
1. Lyse cells using the basis buffer B (100 mM NaH2PO4, 10 mM
Tris–HCl, 8 M urea, pH 8.0).
E. coli as well as most eukaryotic cells are efficiently lysed by 7–9 M
urea but occasionally, His-tagged proteins forming inclusion bodies do
not completely dissolve; in such cases, we recommend to replace the
chaotrope urea by guanidinium hydrochloride (buffer A: 100 mM
NaH2PO4, 10 mM Tris–HCl, 6 M Gu–HCl, pH 8.0). Reduction of
lysate viscosity by the addition of Benzonase (3 units/ml final concen-
tration) is also possible and effective under denaturing conditions but
here the maximum urea concentration is 7 M (use of Gu–HCl is not
possible in combination with Benzonase). Incubate the lysate for 30 min
at ambient temperature.
2. Generate a cleared lysate by centrifugation for 30 min at 10,000g and
room temperature. Collect the supernatant.
3. Load the cleared lysate onto the resin equilibrated with 5 bv of buffer B
(or buffer A where applicable) and allow to flow through at approxi-
mately 1 ml/min (column of 1 ml bv) or by gravity flow.
A suitable linear flow rate during binding is 155 cm/h corresponding
to 1 ml/min of a column with a bed diameter of ~ 7 mm (1 ml HisTrap
or Ni-NTA Superflow Cartridges). If applicable to the workflow and
the resin format used, we recommend to perform binding in batch mode
as this is most efficient with agarose-based resins in general; for this, add
the required volume of resin slurry to the lysate and incubate for 1 h
rotating end-over-end.
4. Wash the Ni-NTA column with 10 bv wash buffer C (100 mM
NaH2PO4, 10 mM Tris–HCl, 8 M urea, pH 6.3).
If binding had been performed in batch, pour the binding suspension
into a suitable flow-through column prior to the wash step. If lysate was
generated using buffer A, wash and elution steps may be performed by
switching to urea-based buffers or by continuing with Gu–HCl-based
buffers with pH values adjusted accordingly.
5. Elute His-tagged protein with 5 bv of elution buffer E (100 mM
NaH2PO4, 10 mM Tris–HCl, 8 M urea, pH 4.5).

2.11. Cleaning and sanitization

A simple and effective cleaning-in-place (CIP) method for Ni-NTA IMAC
resins used to purify proteins from ‘‘standard’’ samples such as E. coli or human
cell lysates or supernatants is contacting the resin with 0.5 M NaOH for 30 min
(Schäfer et al., 2000). The resins have been stored in up to 1 M NaOH
for several months and shown to withstand these conditions without
IMAC 467

compromising its performance even upon >100 cycles of purification/CIP

cycles (data not shown). This CIP procedure reliably denatures and desorbs
proteins originating from the loaded sample that might have bound unspeci-
fically to agarose during purification and is generally suitable for sanitization
(depyrogenation, viral, and microbial clearance; Levison et al., 1995). Cleaning
protocols may have to be adapted if more ‘‘unusual’’ samples such as lysates rich
in lipids are loaded onto a column. Bases, acids, and other reagents that may be
used for cleaning include ethanol (100%), isopropanol (30%, v/v), SDS
(2%, w/v), acetic acid (0.2 M ), NaOH (1 M ), or detergents (see Table 27.2).
For repeated reuse of a Ni-NTA column, we recommend to perform the
CIP description provided above followed by reequilibration. For long-term
storage (several years), resin may be kept in either 30% (v/v) ethanol or, if an
inflammable reagent is preferred, in 0.01–0.1 M NaOH. Storage in 10 mM
NaN3 is possible as well. It is usually not required to strip off the metal ions
and recharge Ni-NTA, even after repeated reuse or long-term storage.

2.12. Simplified metal-ion stripping and recharging protocol

However, in cases where the resin has been seriously damaged or if binding
capacity decreased over time, for example, by repeated loading of lipid-rich
or samples containing chelating components, Ni-NTA may be easily
stripped and recharged with nickel or a different metal ion. Starting with
step 3, this simplified protocol is also suitable to initially charge NTA resin
purchased uncharged.
1. Wash cleaned (see above) resin with 10 bv of deionized H2O (dH2O).
2. Strip off metal ions by passing 5 bv of 100 mM EDTA, pH 8.0 over the
resin bed.
3. Wash resin with 10 bv of dH2O.
4. Pass 2 bv of a 100 mM metal ion aqueous solution (e.g., NiSO4 or
NiCl2) over the resin bed.
Other metal ions that have been successfully and stably immobilized
to NTA include copper (CuCl2, CuSO4), zinc (ZnCl2, ZnSO4), cobalt
(CoCl2, CoSO4), and iron (FeCl3, Fe2(SO4)3).
5. Wash resin with 10 bv of dH2O to remove any unbound metal ions.
6. Add storage buffer or equilibrate the column with at least 5 bv of starting
buffer for immediate use.

3. Conclusions
We have presented the wide variety of applications the IMAC principle
offers for research in general and for production of His-tagged proteins
in particular. Its robustness and versatility are the reasons why IMAC has
468 Helena Block et al.

become one of the most broadly used chromatographic methods. Modifica-

tions in production procedures for both resin and ligand materials as well as
optimized application protocols led to a significant improvement of IMAC
performance in the recent past, well reflected by, for example, the increase
of binding capacity of both NTA- and IDA-based matrices for His-tagged
proteins from 5–10 to 50 mg per ml resin bv. We anticipate a continued
methodological improvement and dissemination of the use of IMAC in the
field of purification of recombinant proteins, for example, regarding indus-
trial-scale production of biopharmaceuticals.

ACKNOWLEDGMENTS
The authors thank Jacob Piehler for valuable discussions regarding the manuscript and his
support during the transfer of the tris-NTA ligand synthesis procedures. Furthermore, we
thank Annette Zacharias-Koch for excellent technical assistance and her support in prepara-
tion of the manuscript. Part of the work presented here was performed under a grant of the
German Ministry of Education and Research (BMBF, grant no. 0313965B).

REFERENCES
Acton, T. B., Gunsalus, K. C., Xiao, R., Ma, L. C., Aramini, J., Baran, M. C., Chiang, Y.-W.,
Climent, T., Cooper, B., Denissova, N. G., Douglas, S. M., Everett, J. K., et al. (2005).
Robotic cloning and protein production platform of the northeast structural genomics
consortium. Methods Enzymol. 394, 210–243.
Andersson, L., and Porath, J. (1986). Isolation of phosphoproteins by immobilized metal
(Fe3þ)affinity-chromatography. Anal. Biochem. 154, 250–254.
Arnau, J., Lauritzen, C., Petersen, G. E., and Pedersen, J. (2006). Current strategies for the
use of affinity tags and tag removal for the purification of recombinant proteins. Protein
Expr. Purif. 48, 1–13.
Biocompare (2006). Protein chromatography: Tools for protein expression and purification.
Biocompare Surveys and Reports, Biocompare Inc., July 10.
Block, H., Kubicek, J., Labahn, J., Roth, U., and Schäfer, F. (2008). Production and
comprehensive quality control of recombinant human Interleukin-1b: A case study for
a process development strategy. Protein Expr. Purif. 27, 244–254.
Boden, V., Winzerling, J. J., Vijayalakshmi, M., and Porath, J. (1995). Rapid one-step
purification of goat immunoglobulins by immobilized metal ion affinity chromatogra-
phy. J. Immunol. Methods 181, 225–232.
Bolanos-Garcia, V. M., and Davies, O. R. (2006). Structural analysis and classification of
native proteins from E. coli commonly co-purified by immobilized metal affinity
chromatography. Biochim. Biophys. Acta 1760, 1304–1313.
Bornhorst, J. A., and Falke, J. J. (2000). Purification of protein using polyhistidine affinity
tags. Methods Enzymol. 326, 245–254.
Braun, P., Hu, Y., Shen, B., Halleck, A., Koundinya, M., Harlow, E., and LaBaer, J. (2002).
Proteome-scale purification of human proteins from bacteria. Proc. Natl. Acad. Sci. USA
99, 2654–2659.
Busso, D., Kim, R., and Kim, S.-H. (2003). J. Biochem. Biophys. Methods 55, 233–240.
IMAC 469

Büssow, K., Nordhoff, E., Lübbert, C., Lehrach, H., and Walter, G. (2000). A human
cDNA library for high-throughput protein expression screening. Genomics 65, 1–8.
Byrne, B., and Jormakka, M. (2006). Solubilization and purification of membrane proteins.
In ‘‘Structural Genomics on membrane proteins’’ (K. Lundstrom, ed.), pp. 179–198.
CRC Press, Taylor and Francis, Boca Raton.
Cass, B., Pham, O. L., Kamen, A., and Durocher, Y. (2005). Purification of recombinant
proteins from mammalian cell culture using a generic double-affinity chromatography
scheme. Protein Expr. Purif. 40, 77–85.
Cèbe, R., and Geiser, M. (2006). Rapid and easy thermodynamic optimization of the 50 -end
of mRNA dramatically increases the level of wild type protein expression in Escherichia
coli. Protein Expr. Purif. 45, 374–380.
Chaga, G., Hopp, J., and Nelson, P. (1999). Immobilized metal ion affinity chromatography
on Co2þ-carboxymethylaspartate–agarose Superflow, as demonstrated by one-step
purification of lactate dehydrogenase from chicken breast muscle. Biotechnol. Appl.
Biochem. 29, 19–24.
Crowe, J., Döbeli, H., Gentz, R., Hochuli, E., Stüber, D., and Henco, K. (1994). 6xHis-
Ni-NTA chromatography as a superior technique in recombinant protein expression/
purification. Methods Mol. Biol. 31, 371–381.
Dalb
ge, H., Bayne, S., and Pedersen, J. (1990). In vivo processing of N-terminal methionine
in E. coli. FEBS Lett 266, 1–3.
DeJong, J., and Roeder, R. G. (1993). A single cDNA, hTFIIA/a, encodes both the p35 and
p19 subunits of human TFIIA. Genes Dev. 7, 2220–2234.
Derewenda, Z. S. (2004). The use of recombinant methods and molecular engineering in
protein crystallization. Methods 34, 354–363.
Drews, J. (2000). Drug discovery: A historical perspective. Science 287, 1960–1964.
Eck, M. J., Beutler, B., Kuo, G., Merryweather, J. P., and Sprang, S. R. (1988). Crystalliza-
tion of trimeric recombinant human tumor necrosis factor (catechin). J. Biol. Chem. 263,
12816–12819.
Eshaghi, S., Hedrén, M., Ignatushchenko Abdel Nasser, M., Hammarberg, T., Thomell, A.,
and Nordlund, P. (2005). An efficient strategy for high-throughput expression of recom-
binant integral membrane proteins. Protein Sci. 14, 676–683.
Franken, K. L. M. C., Hiemstra, H. S., van Meijgaarden, K. E., Subronto, Y., den
Hartigh, J., Ottenhoff, T. H. M., and Drijfthout, J. W. (2000). Purification of His-tagged
proteins by immobilized chelate affinity chromatography: The benefits from the use of
organic solvents. Protein Expr. Purif. 18, 95–99.
Garcı́a González, L. A., Rodrigo Tapia, J. P., Sánchez Lazo, P., Ramos, S., and Suárez
Nieto, C. (2004). DNA extraction Using Chelex resin for the oncogenic amplification
analysis in head and neck tumors. Acta Otorrinolaringol. Esp. 55, 139–144.
Garzia, L., André, A., Amoresano, A., D’Angelo, A., Martusciello, R., Cirulli, C.,
Tsurumi, T., Marino, G., and Zollo, M. (2003). Method to express and purify nm23–
H2 protein from baculovirus-infected cells. BioTechniques 35, 384–391.
Gill, P., Kimpton, C. P., and Sullivan, K. (1992). A rapid polymerase chain reaction method
for identifying fixed specimens. Electrophoresis 13, 173–175.
Gräslund, S., Nordlund, P., Weigelt, J., Bray, J., Gileadi, O., Knapp, S., Oppermann, U.,
Arrowsmith, C., Hui, R., Ming, J., dhe-Paganon, S., Park, H.-W., et al. (2008). Protein
production and purification. Nat. Methods 5, 135–146.
Grisshammer, R., and Tucker, J. (1997). Quantitative evaluation of neurotensin receptor
purification by immobilized metal affinity chromatography. Protein Expr. Purif. 11,
53–60.
Hale, J. E., and Beidler, D. E. (1994). Purification of humanized murine and murine
monoclonal antibodies using immobilized metal-affinity chromatography. Anal. Biochem.
222, 29–33.
470 Helena Block et al.

Hearon, J. (1948). The configuration of cobaltihistidine and oxy-bis (cobalthistidine).

J. Natl. Cancer Inst. 9, 1–11.
Hilbrig, F., and Freitag, R. (2003). Protein purification by affinity precipitation. J. Chroma-
togr. B 790, 79–90.
Hirel, P.-H., Schmitter, J.-M., Dessen, P., Fayat, G., and Blanquet, S. (1989). Extent of
N-terminal methionine excision from Escherichia coli proteins is governed by side-chain
length of the penultimate amino acid. Proc. Natl. Acad. Sci. USA 86, 8247–8251.
Hober, S., and Uhlén, M. (2008). Human protein atlas and the use of microarray
technologies. Curr. Opin. Biotechnol. 19, 30–35.
Hochuli, E. (1989). Genetically designed affinity chromatography using a novel metal
chelate adsorbent. Biologically Active Mol. 411, 217–239.
Hochuli, E., Döbeli, H., and Schacher, A. (1987). New metal chelate adsorbent for proteins
and peptides containing neighbouring histidine residues. J. Chromatogr. 411, 177–184.
Hochuli, E., Bannwarth, W., Döbeli, H., Gentz, R., and Stüber, D. (1988). Genetic
approach to facilitate purification of recombinant proteins with a novel metal chelate
adsorbent. Biotechnology 6, 1321–1325.
Hubert, P., and Porath, J. (1980). Metal chelate affinity chromatography. I. Influence of
various parameters on the retention of nucleotides and related compounds. J. Chromatogr.
198, 247–255.
Hubert, P., and Porath, J. (1981). Metal chelate affinity chromatography. II. Group
separation of mono- and dinucleotides. J. Chromatogr. 206, 164–168.
Hunt, I. (2005). From gene to protein: A review of new and enabling technologies for
multi-parallel protein production. Protein Expr. Purif. 40, 1–22.
Imai, T., Tokunaga, A., Yoshida, T., Hashimoto, M., Mikoshiba, K., Weinmaster, G.,
Nakafuku, M., and Okano, H. (2001). The neural RNA-binding protein Musashi1
translationally regulates mammalian numb gene expression by interacting with its
mRNA. Mol. Cell. Biol. 21, 3888–3900.
Jensen, M. R., Lauritzen, C., Dahl, S. W., Pedersen, J., and Led, J. J. (2004). Binding ability
of a HHP-tagged protein towards Ni2þ studied by paramagnetic NMR relaxation: The
possibility of obtaining long-range structure information. J. Biomol. NMR 29, 175–185.
Jin, S., Issel, C. J., and Montelaro, R. C. (2004). Serological method using recombinant S2
protein to differentiate equine infectious anemia virus (EIAV)-infected and
EIAV-vaccinated horses. Clin. Diagn. Lab. Immunol. 11, 1120–1129.
Jungbauer, A., Kaar, W., and Schlegl, R. (2004). Folding and refolding of proteins in
chromatographic beds. Curr. Opin. Biotechnol. 15, 487–494.
Kaslow, D. C., and Shiloach, J. (1994). Production, purification and immunogenicity of a
Malaria transmission-blocking vaccine candidate: TBV25H expressed in yeast and
purified using Ni-NTA Agarose. Biotechnology 12, 494–499.
Kim, T.-W., Oh, I.-S., Ahn, J.-H., Choi, C.-Y., and Kim, D.-M. (2006). Cell-free synthesis
and in situ isolation of recombinant proteins. Protein Expr. Purif. 45, 249–254.
Kim, M. J., Park, H.-Y., Kim, J., Ryu, J., Hong, S., Han, S.-J., and Song, R. (2008).
Western blot analysis using metal–nitrilotriacetate conjugated CdSe/ZnS quantum dots.
Anal. Biochem. 379, 124–126.
Klammt, C., Schwarz, D., Fendler, K., Haase, W., Dötsch, V., and Bernhard, F. (2005).
Evaluation of detergents for the soluble expression of a-helical and b-barrel-type integral
membrane proteins by a preparative scale individual cell-free expression system. FEBS J.
272, 6024–6038.
Knecht, S., Ricklin, D., Eberle, A. N., and Ernst, B. (2009). Oligohis-tags: Mechanisms of
binding to Ni2+-NTA surfaces. J. Mol. Recognit. 22, 270–279.
Knight, R. D., and Shimeld, S. M. (2001). Identification of conserved C2H2 zinc-finger
gene families in the bilateria. Genome Biol. 2, 1–8.
Koehn, J., and Hunt, I. (2009). High-throughput protein production (HTPP): A review of
enabling technologies to expedite protein production. Methods Mol. Biol. 498, 1–18.
IMAC 471

In ‘‘High Throughput Protein Expression and Purification’’ (S. A. Doyle, ed.). Humana
Press, Springer.
Lanio, T., Jeltsch, A., and Pingoud, A. (2000). Automated purification of His6-tagged
proteins allows exhaustive screening of libraries generated by random mutagenesis.
BioTechniques 29, 338–342.
Lata, S., and Piehler, J. (2005). Stable and functional immobilization of histidine-tagged
proteins via multivalent chelator headgroups on a molecular poly(ethylene glycol) brush.
Anal. Chem. 77, 1096–1105.
Lata, S., Gavutis, M., Tampé, R., and Piehler, J. (2006). Specific and stable fluorescence
labeling of histidine-tagged proteins for dissecting multi-protein complex formation.
J. Am. Chem. Soc. 128, 2365–2372.
Lesley, S. A. (2001). High-throughput proteomics: Protein expression and purification in
the postgenomic world. Protein Expr. Purif. 22, 159–164.
Levison, P. R., Badger, S. E., Jones, R. M. H., Toome, D. W., Streater, M.,
Pathirana, N. D., and Wheeler, S. (1995). Validation studies in the regeneration of
ion-exchange celluloses. J. Chromatogr. A 702, 59–68.
Lewinson, O., Lee, A. T., and Douglas, C. R. (2008). The funnel approach to the
precrystallization production of membrane proteins. J. Mol. Biol. 377, 62–73.
Loo, J. A. (2003). The tools of proteomics. Adv. Protein Chem. 65, 353–369.
Lv, G. S., Hua, G.C, and Fu, X. Y. (2003). Expression of milk-derived antihypertensive
peptide in Escherichia coli. J. Dairy Sci. 86, 1927–1931.
Ma, D., Watanabe, H., Mermelstein, F., Admon, A., Oguri, K., Sun, X., Wada, T.,
Imai, T., Shiroya, T., Reinberg, D., and Handa, H. (1993). Isolation of a cDNA
encoding the largest subunit of TFIIA reveals functions important for activated transcrip-
tion. Genes Dev. 7, 2246–2257.
Mateo, C., Fernandez-Lorente, G., Pessela, B. C. C., Vian, A., Carrascosa, A. V.,
Garcia, J. L., Fernandez-Lafuente, R., and Guisan, J. M. (2001). Affinity chromatography
of polyhistidine tagged enzymes: New dextran-coated immobilized metal ion affinity
chromatography matrices for prevention of undesired multipoint adsorptions. J. Chro-
matogr. A 915, 97–106.
Matiasson, B., Kumar, A., Ivanov, A. E., and Galaev, I. Y. (2007). Metal-chelate affinity
precipitation of proteins using responsive polymers. Nat. Protocols 2, 213–220.
Mészárosová, K., Tishchenko, G., Bouchal, K., and Bleha, M. (2003). Immobilized-metal
affinity sorbents based on hydrophilic methacrylate polymers and their interaction with
immunoglobulins. React. Funct. Polym. 56, 27–35.
Mohanty, A. K., and Wiener, M. C. (2004). Membrane protein expression and production:
Effects of polyhistidine tag length and position. Protein Expr. Purif. 33, 311–325.
Muszynska, G., Andersson, L., and Porath, J. (1986). Selective adsorption of phosphopro-
teins on gel-immobilized ferric chelate. Biochemistry 25, 6850–6853.
Nieba, L., Nieba-Axamann, S. E., Persson, A., Hämäläinen, M., Edebratt, F., Hansson, A.,
Lidholm, J., Magnusson, K., Karlsson, A. F., and Plückthun, A. (1997). Biacore analysis
of histidine-tagged proteins using a chelating NTA sensor chip. Anal. Biochem. 252,
217–228.
Nilsson, J., Ståhl, S., Lundeberg, J., Uhlén, M., and Nygren, P.-A. (1997). Affinity fusion
strategies for detection, purification, and immobilization of recombinant proteins. Protein
Expr. Purif. 11, 1–16.
Padan, E., Venturi, M., Michel, H., and Hunte, C. (1998). Production and characterization
of monoclonal antibodies directed against native epitopes of NhaH, the Naþ/Hþ anti-
porter of E. coli. FEBS Lett. 441, 53–58.
Pedersen, J., Lauritzen, C., Madsen, M. T., and Dahl, S. W. (1999). Removal of N-terminal
polyhistidine tags from recombinant proteins using engineered aminopeptidases. Protein
Expr. Purif. 15, 389–400.
472 Helena Block et al.

Polayes, D.A., Parks, T.D., Johnston, S.A., and Dougherty, W.G. (2008). Application of
TEV protease in protein production. Methods Mol. Med. 13, 169–183. In ‘‘Molecular
Diagnosis of Infectious Diseases’’ (U. Reischl. ed.). Humana Press, Springer.
Porath, J., and Olin, B. (1983). Immobilized metal ion affinity adsorption and immobilized
metal ion affinity chromatography of biomaterials. Serum protein affinities for
gel-immobilized iron and nickel ions. Biochemistry 29, 1621–1630.
Porath, J., Carlsson, J., Olsson, I., and Belfrage, G. (1975). Metal chelate affinity chroma-
tography, a new approach to protein fractionation. Nature 258, 598–599.
Posewitz, M. C., and Tempst, P. (1999). Immobilized gallium (III) affinity chromatography
of phosphopeptides. Anal. Chem. 71, 2883–2892.
Prinz, B., Schultchen, J., Rydzewski, R., Holz, C., Boettner, M., Stahl, U., and Lang, C.
(2004). Establishing a versatile fermentation procedure for human proteins expressed in
the yeasts Saccharomyces cerevisiae and Pichia pastoris for structural genomics. J. Struct. Funct.
Genomics 5, 29–44.
QIAGEN (2003). The QIAexpressionist. Handbook for High-Level Expression and Purifi-
cation of 6xHis-Tagged Proteins. 5th edn. Hilden, Germany: QIAGEN, Hilden,
Germany.
Reichel, A., Schaible, D., Al Furoukh, N., Cohen, M., Schreiber, G., and Piehler, J. (2007).
Noncovalent, site-specific biotinylation of histidine-tagged proteins. Anal. Chem. 79,
8590–8600.
Rumbley, J. N., Furlong Nickels, E., and Gennis, R. B. (1997). One-step purification of
cytochrome bo3 from Escherichia coli and demonstration that associated quinone is not
required for the structural integrity of the oxidase. Biochim. Biophys. Acta 1340, 131–142.
Schäfer, F., Blümer, J., Römer, U., and Steinert, K. (2000). Ni-NTA for large-scale
processes—systematic investigation of separation characteristics, storage and CIP condi-
tions, and leaching. QIAGEN. QIAGEN News 4, 11–15. Available from http://www1.
qiagen.com/literature/qiagennews/0400/Ni-NTA%20for%20large-scale.pdf
Schäfer, F., Schäfer, A., and Steinert, K. (2002a). A highly specific system for efficient
enzymatic removal of tags from recombinant proteins. J. Biomol. Tech. 13, 158–171.
Schäfer, F., Römer, U., Emmerlich, M., Blümer, J., Lubenow, H., and Steinert, K. (2002b).
Automated high-throughput purification of 6xHis-tagged proteins. J. Biomol. Tech. 13,
131–142.
Scheich, C., Sievert, V., and Büssow, K. (2003). An automated method for high-throughput
protein purification applied to a comparison of His-tag and GST-tag affinity chromatog-
raphy. BMC Biotechnol. 3, 12. Available from www.biomedcentral.com/1472-6750/3/12
Schmitt, J., Hess, H., and Stunnenberg, H. G. (1993). Affinity purification of histidine-
tagged proteins. Mol. Biol. Rep. 18, 223–230.
Serpa, G., Augusto, E. F. P., Tamashiro, W. M. S. C., Ribeiro, M. B., Miranda, E. A., and
Bueno, S. M. A. (2005). Evaluation of immobilized metal membrane affinity chroma-
tography for purification of an immunoglobulin G1 monoclonal antibody. J. Chromatogr. B
816, 259–268.
Shi, W., and Chance, M. R. (2008). Metallomics and metalloproteomics. Cell. Mol. Life Sci.
65, 3040–3048.
Shi, Y., Seto, E., Chang, L.-S., and Shenk, T. (1991). Transcriptional repression by YY1,
a human GLI-Krüppel-related protein, and relief of repression by adenovirus E1A
protein. Cell 67, 377–388.
Slentz, B. E., Penner, N. A., and Regnier, F. E. (2003). Protein proteolysis and the
multi-dimensional electrochromatographic separation of histidine-containing peptide
fragments. J. Chromatogr. A 984, 97–103.
Stasyk, T., and Huber, L. A. (2004). Zooming in: Fractionation strategies in proteomics.
Proteomics 4, 3704–3716.
IMAC 473

Steen, J., Uhlén, M., Hober, S., and Ottosson, J. (2006). High-throughput protein purifica-
tion using an automated set-up for high-yield affinity chromatography. Protein Expr.
Purif. 46, 173–178.
Stowers, A. W., Zhang, Y., Shimp, R. L., and Kaslow, D. C. (2001). Structural conformers
produced during malaria vaccine production in yeast. Yeast 18, 137–150.
Strömberg, P., Rotticci-Mulder, J., Björnestedt, R., and Schmidt, S. R. (2005). Preparative
parallel protein purification (P4). J. Chromatogr. B 818, 11–18.
Sulkowski, E. (1985). Purification of proteins by IMAC. Trends Biotechnol. 3, 1–7.
Sun, X., Chiu, J.-F., and He, Q.-Y. (2005). Application of immobilized metal affinity
chromatography in proteomics. Expert Rev. Proteomics 2, 649–657.
Svensson, J., Andersson, C., Reseland, J. E., Lyngstadaas, P., and Bülow, L. (2006). Histidine
tag fusions increases expression levels of active recombinant amelogenin in Escherichia coli.
Protein Expr. Purif. 48, 134–141.
Vancan, S., Miranda, E. A., and Bueno, S. M. A. (2002). IMAC of human IgG: Studies with
IDA-immobilized copper, nickel, zinc, and cobalt ions and different buffer systems.
Process Biochem. 37, 573–579.
Walsh, P. S., Metzger, D. A., and Higuchi, R. (1991). Chelex 100 as a medium for simple
extraction of DNA for PCR-based typing from forensic material. BioTechniques 10,
506–513.
Woestenenk, E. A., Hammarström, M., van den Berg, S., Härd, T., and Berglund, H.
(2004). His tag effect on solubility of human proteins produced in Escherichia coli:
Comparison between four expression vectors. J. Struct. Funct. Genomics 5, 217–229.
Xu, C.-G., Fan, X.-J., Fu, Y.-J., and Liang, A.-H. (2008). Effect of the His-tag on the
production of soluble and functional Buthus martensii Karsch insect toxin. Protein Expr.
Purif. 59, 103–109.
Zhaohua, H., Park, J. I., Watson, D. S., Hwang, P., and Szoka, F. C. (2006). Facile synthesis
of multivalent nitrilotriacetic acid (NTA) and NTA conjugates for analytical and drug
delivery applications. Bioconjug. Chem. 17, 1592–1600.
Zhou, H., Xu, S., Ye, M., Feng, S., Pan, C., Jiang, X., Li, X., Han, G., Fu, Y., and Zou, H.
(2006). Zirconium phosphonate-modified porous silicon for highly specific capture of
phosphopeptide and MALDI-TOF MS analysis. J. Proteome Res. 5, 2431–2437.
 C H A P T E R T W E N T Y- E I G H T

Identification, Production, and Use

of Polyol-Responsive Monoclonal
Antibodies for Immunoaffinity
Chromatography
Nancy E. Thompson, Katherine M. Foley, Elizabeth S. Stalder,
and Richard R. Burgess

Contents
1. Introduction 476
2. Polyol-Responsive Monoclonal Antibodies 477
2.1. Properties of a PR-mAb 478
2.2. Source of mAbs 479
2.3. Identification of PR-mAbs by ELISA-elution assay 480
2.4. Producing mAbs in continuous culture 482
2.5. Purification of the antibody 485
2.6. Immobilization of PR-mAbs on a chromatography support 487
2.7. Purification of proteins with PR-mAbs 488
2.8. Purification of proteins using cross-reacting PR-mAbs 490
2.9. Use of epitopes of PR-mAbs as purification tags 491
3. Conclusions 492
Disclosure 493
References 493

Abstract
Immunoaffinity chromatography is a powerful tool for purification of proteins
and protein complexes. The availability of monoclonal antibodies (mAbs) has
revolutionized the field of immunoaffinity chromatography by providing a con-
tinuous supply of highly uniform antibody. Before the availability of mAbs, the
recovery of the target protein from immobilized polyclonal antibodies usually
required very harsh, often denaturing conditions. Although harsh conditions are
often still used to disrupt the antigen–antibody interaction when using a mAb,
various methods have been developed to exploit the uniformity of the antigen–
antibody reaction in order to identify agents or conditions that gently disrupt

McArdle Laboratory for Cancer Research, University of Wisconsin–Madison, Madison, Wisconsin, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63028-7 All rights reserved.

475
476 Nancy E. Thompson et al.

this interaction and thus result in higher recovery of active protein from immu-
noaffinity chromatography. We discuss here the use of a specific type of
monoclonal antibody that we have designated ‘‘polyol-responsive monoclonal
antibodies’’ (PR-mAbs). These are naturally occurring mAbs that have high
affinity for the antigen under binding conditions, but have low affinity in the
presence of a combination of low molecular weight hydroxylated compounds
(polyols) and nonchaotropic salts. Therefore, these PR-mAbs can be used for
gentle immunoaffinity chromatography. PR-mAbs can be easily identified and
adapted to a powerful protein purification method for a target protein.

1. Introduction
All forms of affinity chromatography are defined by a specific interac-
tion between two components that allows the purification of one of the
components. Immunoaffinity chromatography is a subset of the affinity
chromatography principle where the specific interaction of an antigen
with an antibody is employed (for review, see Subramanian, 2002). Immu-
noaffinity chromatography is really a scaled up extension of an immunopre-
cipitation procedure, except that one of the components (generally the
antigen) is recovered after the chromatography as an active protein.
The ability to produce monoclonal antibodies (mAbs) has revolutio-
nized the field of immunochemistry (for review, see Nelson et al., 2000).
When considering immunoaffinity chromatography, two features give
mAbs an advantage over polyclonal antibodies derived from immune
serum. First, the mAb is a reproducible reagent that can be prepared in
large quantities. Second, a mAb is a homogeneous population that responds
uniformly to an eluting reagent. Generally, the purified antibody is conju-
gated to some type of bead and the antigen-containing solution is applied to
the bead (in batch or in a column). After washing away unbound or loosely
bound material, the antigen is eluted from the bead.
The elution step is usually the most difficult obstacle to overcome in
developing an immunoaffinity chromatography procedure. Antigen–anti-
body interactions are generally a result of a combination of ionic, hydro-
phobic, and hydrogen bonds formed between amino acids in the specific
antigenic determinant of the antigen (epitope) and the protein loops of
the complementarity determining regions (CDRs), which are located in the
variable regions of the heavy and light chain of the antibody molecule.
The ideal way to gently elute the antigen is by competition with a peptide
containing the epitope for the antibody. However, the epitope is not always
known, and a peptide is not always available or is too expensive to synthe-
size in the quantities needed. In addition, sometimes the antibody reacts
with a discontinuous epitope, and the epitope cannot be mimicked by a
Immunoaffinity Chromatography 477

synthetic peptide. In these cases, the antigen is usually eluted with very
harsh conditions (high or low pH values, denaturants such as urea, or an
ionic detergent.) that can inactivate the protein.

2. Polyol-Responsive Monoclonal Antibodies

We have pioneered the use of a specific type of monoclonal antibody
for use in immunoaffinity chromatography. These mAbs can be used for
‘‘gentle’’ immunoaffinity chromatography because the elution conditions
require only a combination of a nonchaotropic salt and a low molecular
weight polyhydroxylated compound (polyol), conditions that are regarded
as nondenaturing to proteins. We have referred to this type of antibody as a
‘‘polyol-responsive’’ mAb (PR-mAb).
Our laboratory studies proteins involved in transcription. Therefore,
most of the PR-mAbs that we have isolated have been mAbs that react
with proteins involved in transcription in either prokaryotic or eukaryotic
systems. Eukaryotic transcription systems pose a significant challenge for the
separation scientist because many of the factors are actually multisubunit
proteins. For example, eukaryotic RNA polymerase II (RNAP II) contains
12 subunits (12 different gene products); however, for initiation from a
promoter, RNAP II also requires transcription factors TFIIA, TFIIB,
TFIID, TFIIE, TFIIF, and TFIIH (for review, see Woychik and
Hampsey, 2002). With the exception of TFIIB, all of these transcription
factors are comprised of two or more subunits. Large protein complexes are
ideal subjects for the use of PR-mAb immunoaffinity chromatography. We
have developed PR-mAbs for E. coli RNAP (Thompson et al., 1992) and
eukaryotic RNAP II (Thompson et al., 1990), as well several eukaryotic
transcription factors (Table 28.1).
Perhaps the most notable use of PR-mAb technology was the use of our
PR-mAb 8WG16 for use to purify large amounts of yeast RNAP II
(Edwards et al., 1990) for use in crystallization of this protein complex
(Cramer et al., 2000). Other laboratories have successfully isolated PR-
mAbs for purifying their protein of interest ( Jiang et al., 1995; Lynch
et al., 1996; Nagy et al., 2002). It is also possible to screen already established
collections of mAbs for the polyol-responsive property.
Several reviews have been written concerning PR-mAbs (Burgess and
Thompson, 2002; Thompson and Burgess, 1996, 2001; Thompson et al.,
2006). In this report, we will update some of these methods. We will also
describe our method for producing large quantities of PR-mAbs by a relatively
inexpensive cell culture system, thus minimizing the need to prepare the mAbs
in mouse ascites fluid. Finally, because polyol-responsiveness is a property of
the antibody, an unrelated protein of interest can be tagged with the epitope for
478 Nancy E. Thompson et al.

Table 28.1 Polyol-responsive monoclonal antibodies that have been used to purify
proteins involved with transcription

PR-mAb Antigen Epitope Reference

8WG16 Heptapeptide YSPTSPSYSPTSPS Edwards et al.
repeat on the (1990) and
C-terminus Thompson
of the largest et al. (1990)
subunit of
RNAP II
NT73 Far C-terminus SLAELLNAGLGGS Thompson et al.
of the b0 (1992, 2003)
subunit of
E. coli RNAP
(1392–1404)
8RB13 b-flap region of PEEKLLRAIFGEKAS Bergendahl et al.
the b subunit of (2003),
E. coli RNAP Probasco et al.
(2007), and
E. S. Stalder
et al.
(unpublished)
4RA2 C-terminal half of Unknown Anthony et al.
the a subunit of (2003)
E. coli RNAP
1TBP22 N-terminus of Unknown Thompson et al.
human TBP (2004)
(1–199)
IIB8 N-terminus of TKDPSRVG Duellman et al.
human TFIIB (2004) and
(61–68) Thompson and
Burgess (1994)
1RAP1 N-terminus of Unknown Thompson and
RAP30 Burgess (1999)
(1–118)

a PR-mAb by recombinant DNA methodologies, and the protein purified by

this gentle polyol-elution method.

2.1. Properties of a PR-mAb

1. By screening a large number of mAbs (218 antigen-specific wells), we
have estimated that PR-mAbs represent 5–10% of mAbs (Thompson
et al., 1992).
Immunoaffinity Chromatography 479

2. Screening can be performed at the master-well stage in order to imme-

diately identify PR-mAbs.
3. PR-mAbs are not limited to any subisotype of mouse IgG. PR-mAbs
have also been identified in a collection of rat mAbs (R. R. Burgess,
unpublished data).
4. Most PR-mAbs respond to a variety of different combinations of salt and
polyol. Most commonly we use 0.75 M ammonium sulfate or 0.75 M
NaCl combined with 30–40% propylene glycol.
5. PR-mAbs can be high-affinity antibodies. In fact, high affinity is pre-
ferred in order for the mAb to be effective in binding the antigen in a
dilute solution. These mAbs become low-affinity antibodies under the
elution conditions.
6. Most (but not all) PR-mAbs respond to a variety of combinations of salt
and polyol. Salts that we have tested are ammonium sulfate, sodium
chloride, sodium acetate, and potassium glutamate. Polyol that we have
tested are propylene glycol, ethylene glycol, 2,3-butanediol, and in some
cases, glycerol.
7. Because these salts and polyols are often used as protein stabilizers, the
gentle elution results in antigens that retain biological properties and
structural integrity, even for multisubunit complexes (Cramer et al.,
2000; Thompson et al., 1990).

2.2. Source of mAbs

We produce our mAbs by a typical hybridoma procedure (Harlow and
Lane, 1988); that is by the fusion of antigen-stimulated plasma cells from a
mouse spleen and a myeloma cell line. We find that a hyperimmunized
mouse is preferred; only about 5–10% of mAbs are PR-mAbs, a large
number of original hybridomas will make it more likely that a PR-mAb
will be isolated. However, it is possible to use mAbs that are produced
by other methods, such as by retroviral infection of plasma cells
(Largaespada et al., 1996) or construction of antibody libraries by recombi-
nant methods.
Regardless of the source of antibodies, antigen-specific antibodies must
be identified. We find that a standard enzyme-linked immunosorbent assay
(ELISA) works well for this. We then screen the antibodies by a modified
ELISA, which we have termed ‘‘ELISA-elution assay.’’ This procedure is
based on a standard ELISA except that the specific antigen–antibody com-
plex is treated with a combination of salt and polyol before the enzyme-
conjugated secondary antibody is added. The subsequent reaction with the
substrate gives an estimation of the amount of antigen–antibody complex
that is dissociated by the salt/polyol treatment. A general procedure for the
ELISA-elution assay is as follows.
480 Nancy E. Thompson et al.

2.3. Identification of PR-mAbs by ELISA-elution assay

1. Coat a polystyrene microtiter plate with antigen. We typically use
30–100 ng of antigen contained in 50 ml of phosphate buffered saline
(PBS, pH 7.4) per well. The antigen is allowed to incubate at room
temperature for 1 h to allow sufficient time for the antigen to bind to
the polystyrene.
2. Each well is then blocked with 200 ml of 1% nonfat dry milk in PBS
(1% BLOTTO). We typically block overnight at 4  C, but 2 h at room
temperature is sufficient.
3. The test antibody (50 ml) is then added to two adjacent wells, and
incubated at room temperature for 1 h. If the antibody is contained in
cell culture fluid, it usually can be used directly. However, very high
affinity antibodies, or very high titer antibody preparations should be
diluted so that nonsaturating levels are used.
4. The wells are washed five times with PBS containing 0.1% Tween 20
(PBST) to remove the unbound antibody.
5. To the control well, 100 ml of TE buffer (50 mM Tris–HCl, pH 7.9 and
0.1 mM EDTA) is added, and to the test well, 100 ml of TE buffer
containing 0.75 M ammonium sulfate and 40% propylene glycol is
added. The plate is incubated at room temperature for 20 min, with
occasional (about every 5 min) tapping of the side of the plate to mix
the contents.
6. The wells are washed five times with PBST.
7. A commercially available horseradish peroxidase-conjugated secondary
antibody is diluted into 1% BLOTTO (generally 1:2000) and 50 ml is
added to each well. The plate is incubated at room temperature for 1 h.
8. The wells are washed 10 times with PBST.
9. The appropriate substrate is added to the wells. We use 100 ml of 0.03%
H2O2 and 0.4 mg/ml o-phenylenediamine (OPD) contained in 0.05 M
citrate buffer (pH 5.0).
10. The plate is incubated 5–15 min at room temperature. The reactions
are stopped in pairs (TE and TE þ salt/polyol for each mAb) with 50 ml
of 1 M H2SO4 per well.
11. The absorbance is read on a microtiter plate reader; for OPD we use
490 nm. Treatment with the polyol and salt will reduce the absorbance
of a PR-mAb in the test well by about 50% compared to the control well
that received just TE buffer (Fig. 28.1A) If a plate reader is not available,
the approximately 50% reduction is usually evident by visual inspection.

2.3.1. Comments
1. Screening for PR-mAbs can be performed at the master-well stage,
immediately after the hybridomas are screened for specific antibody
Immunoaffinity Chromatography 481

A
1
mAb NT63 mAb 8RB13 mAb NT73
0.9
0.8
Absorbance value

0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
PG
NaCl

NaCl/PG

TE buffer
NaCl
PG
NaCl/PG

TE buffer
NaCl
PG
NaCl/PG
TE buffer

B
1.4
0% PG
1.2
Absorbance value

1
20% PG
0.8
0.6
0.4
0.2 30% PG
40% PG
0
0 0.2 0.4 0.6 0.8 1
NaCl (M )

Figure 28.1 ELISA-elution assay to identify and characterize PR-mAbs. Each well of
the microtiter plate was coated with 100 ng core RNAP. (A) ELISA-elution assay of
previously identified PR-mAb. mAb NT63 is a control mAb that reacts with E. coli
RNAP (Thompson et al., 1992). mAbs 8RB13 and NT73 are PR-mAbs (Bergendahl
et al., 2003; Thompson et al., 1992). mAb NT63 did not elute from the antigen with
0.75 M NaCl and 40% propylene glycol (NaCl/PG), but mAbs 8RB13 and NT73 did
elute with the salt/polyol combination. None of the mAbs eluted well with 0.75 M
NaCl or 40% propylene glycol (PG) alone. (B) ELISA-elution assay using mAb 8RB13
and varying concentrations of both NaCl (0–1.0 M) and propylene glycol (0–40%).

production. The preliminary screening can be performed with 100 ml of

cell culture fluid (50 ml for the buffer control and 50 ml for the buffer
containing polyol and salt). In one study (Thompson et al., 1992), we
screened over 200 hybridomas for PR-mAb from a single fusion at the
master-well stage.
2. Binding of the antigen to the microtiter plate can result in distortion of
the antigen structure. This can expose buried epitopes that are not
482 Nancy E. Thompson et al.

accessible in the protein when it is in solution. This can result in a ‘‘false-

positive’’ because the mAb does not react with the target when it is in
solution and is not a useful mAb for immunoaffinity chromatography.
3. When a few milliliters of cell culture fluid are available, it is possible to
examine the response of a mAb to different concentrations of polyol and
salt on a single plate (Fig. 28.1B).

2.4. Producing mAbs in continuous culture

The practice of producing mouse mAbs in ascites fluid has been discouraged
in many settings. Therefore, we have explored alternatives methods to
produce the large amounts of mAbs that are needed for immunoaffinity
chromatography. We describe here one method that we have found to be
particularly useful for this scale of antibody production. This protocol uses a
commercially available cell culture chamber called a CELLine Flask 350
(CL 350) manufactured by Integra Biosciences AG (Switzerland). In the
United States, this product is distributed by Argos Technologies, Inc (Elgin,
IL) and Bioraco International (Framingham, MA). We found this product
to be easy to use, capable of producing 10–50 mg of antibody, and reusable
several times for the same hybridoma. Standard cell culture aseptic techni-
ques are required, along with a cell culture hood and a 37  C humidified
CO2 incubator (5%). A schematic of this culture flask is shown in
Fig. 28.2A. The general protocol is as follows:
1. Preparing the cell culture media: Two slightly different media are prepared
for the two chambers of the CL 350 flask. Dulbecco’s Modified Eagle
Medium (DMEM) containing glutamine and high glucose (Gibco/Invi-
trogen #11965), supplemented with 1 mM sodium pyruvate (Sigma),
100 units of penicillin/ml, and 100 mg/ml of streptomycin (Gibco/
Invitrogen) is used as the base for the two media. For the cell growth
chamber, DMEM with the supplements indicated above is also supple-
mented with 15% heat-inactivated fetal bovine serum (Hyclone). The
nutrient chamber uses DMEM with the supplements indicated above
and 5% fetal bovine serum. We refer to the medium for the growth
chamber as ‘‘complete medium’’ and the medium for the nutrient
chamber as ‘‘nutrient medium.’’
2. Preparing the inoculum for seeding the flask: Remove a vial containing the
hybridoma from liquid N2 and thaw rapidly at 37  C. Plate the hybrid-
oma in complete medium at about 2  104 cells/ml. We use 10-cm cell
culture plates containing 20 ml of medium.
3. Seeding the flask: Prepare 8  106 to 20  106 viable cells in log growth
phase in 5 ml of fresh complete medium. Place 25 ml of nutrient
medium in the nutrient compartment (green cap) to wet the membrane
between the nutrient chamber and the cell chamber before the cells are
Immunoaffinity Chromatography 483

A
Port to nutrient chamber
Green cap
White cap

Port to growth chamber

Nutrient medium Semi-permeable

compartment membrane
Silicone Growth chamber
membrane
Support platform
with air ports
AS ppt (solubilized)

B
Cell supernatant
AS supernatant
Markers

Flow-through fractions from DE52

F10
F11
F2
F3
F4
F5
F6
F7
F8
F9
KDa

160
110
80
60
HC
50
40

30
LC
20
15

10

3.5
1 2 3 4 5 6 7 8 9 10 11 12 13 14

Figure 28.2 mAb production in CELLine flasks. (A) Diagrammatic representation of

the CL 350 flask (adapted from the manufacturer’s literature). (B) The hybridoma that
produces the 8RB13 mAb was cultured in the flask for 3 weeks. The cell supernatants
that were collected every 3 days were pooled (Lane 2) and purified using ammonium
sulfate precipitation (Lanes 3 and 4) and chromatography on DE52 (Lanes 5–14) as
described in the text. The fractions were run on a 4–12% NuPAGE gel (Invitrogen),
using MES buffer; the gel was stained with GelCode (Pierce). The immunoglobulin
heavy chain (HC) and light chain (LC) are indicated.

placed in the cell chamber. Suspend the cell preparation and aspirate into
a 10-ml serological pipette. With the green cap loose, inoculate 5 ml
suspension into cell compartment (white cap) by inserting the pipette
484 Nancy E. Thompson et al.

firmly into the cell compartment port. Remove trapped air bubbles by
pipetting the fluid up and down slowly to allow the bubbles to rise prior
to returning only the fluid to the cell chamber. Replace and tighten the
white cap. Add 350 ml of nutrient medium to the nutrient compartment
and completely tighten the green cap.
4. Cell compartment harvest and culture maintenance: Every 3–7 days remove
nutrient medium from nutrient medium compartment. With green cap
loose, insert 10 ml pipette into the cell compartment and pipette fluid up
and down to thoroughly mix the cells. Then remove entire cell com-
partment liquid volume to a centrifuge tube. Volume may be greater
than 5 ml due to osmotic flux. Take a sample for a cell count and cell
viability using a hemocytometer and trypan blue. Centrifuge the mate-
rial removed from the cell compartment to pellet the cells. Remove and
save the medium, which contains the mAbs; this can be frozen for
purification at a later date. Suspend the cells with fresh complete
medium. Depending upon the initial inoculation density, growth rate,
and maintenance frequency the cells may need to be split back (usually
1:2–1:4) at this point. With the green cap loosened, return 5 ml of cells
back to the cell compartment (white cap). Remove air bubbles and
completely tighten white cap. Add 350 ml of nutrient medium to
nutrient compartment and completely tighten green cap.
5. The Integra CL 350 flask is harvested every 3–7 days after the culture has
been established. Harvest intervals depend on the growth rate of the
hybridoma and the ability of the hybridoma to adapt to the flask
environment. This trait seems to be cell line dependent.

2.4.1. Comments
1. For hybridomas that grow well in the CL 350 flask, approximately 0.5–
1 mg of mAb is present in each ml of harvested cell culture supernatant.
The continuous culture can usually be maintained for about 1 month.
2. Liquid handling: Warm medium to 37  C in a water bath. This helps to
prevent condensation on the flask and temperature shocking the cells.
When adding or removing liquid from the cell compartment (white
cap), always loosen the green cap of the nutrient medium compartment
first to prevent air lock. Always tighten both white and green caps before
placing the flask into the incubator. The use of 10 ml serological pipettes
is recommended for all cell compartment manipulations. The medium
in the nutrient medium chamber can be exchanged by aspirating the
medium out and pouring fresh medium in.
3. The minimum cell concentration is 1.5  106 cells/ml for the inoculum
(Step 2). We have not been successful in reducing the need for fetal
bovine serum in the nutrient medium. Some of the new serum-free
media on the market may be used as the nutrient medium.
Immunoaffinity Chromatography 485

4. It is important to track cell numbers and health (Step 3). This helps to
determine whether to split cells to reduce numbers once the total viable
cell count is greater than 100  106 cells. If the cell viability is greatly
reduced, the maintenance frequency needs to be increased.
5. Because the cells are split back every time the culture is harvested, the
percentage of viable cells (Step 4) will decrease during the continuous
culture due to death of the cells. It is not unusual to end up with only
30–40% viability at the end of the culture period.
6. Some hybridomas are unstable in this long-term, continuous culture and
lose the ability to produce the mAb. Therefore, the antibody production
should be monitored during the culturing. We do this by a standard
ELISA.
7. Information on CELLine flasks can be obtained at: www.integra-
biosciences.com/celline.

2.5. Purification of the antibody

It is usually advantageous to at least partially purify the antibody in order to
maximize the binding of the antibody to the available reactive sites on the
support during immobilization. Antibodies can be purified by many differ-
ent methods. Some of the most common, low-cost methods are by size-
exclusion or ion-exchange chromatography. A common, but more costly
method of purification is by affinity chromatography on a column contain-
ing either protein A or protein G, or a mixture of the two. Mouse mAbs
that belong to the IgG class of immunoglobulins belong to one of four
subclasses: IgG1, IgG2a, IgG2b, and IgG3. Mouse IgG2a, IgG2b, and IgG3 can
be purified on a column containing protein A. Mouse mAbs of the IgG1
subclass do not bind to protein A very well. These mAbs bind better to
protein G, but not as well as mAbs of the mouse IgG2a and IgG2b bind to
either protein A or protein G.
In our experience, the majority of the mAbs identified in a typical fusion
are of the IgG1 subclass. We describe here a simple method for purifying
mouse IgG mAbs on an inexpensive DEAE column (Whatman DE52).
This method is particularly useful for IgG1 mAbs. In fact, every IgG1 mAb
that we have tested can be purified to about 90% purity. In addition, many
mouse IgG2a and IgG2b antibodies can be purified by this method.
1. Antibodies contained in either ascites fluid or concentrated cell superna-
tant from the CELLine Flask are precipitated by ammonium sulfate.
A saturated solution of ammonium sulfate is added to the mAb until 45%
saturation is reached. The slurry is allowed to stir for 20 min on ice. This
precipitates the mAb but leaves most of the serum albumin in solution.
2. The precipitate is removed by centrifugation (about 6000g, 10 min).
Antibody buffer (50 mM Tris–HCl, pH 6.9, 25 mM NaCl) is added to
486 Nancy E. Thompson et al.

the pellet to obtain one-fourth (for the CELLine flask-prepared anti-

body) to one-half (for ascites fluid) of the volume of the original
antibody preparation.
3. The precipitate is allowed to solubilize for about 15 min, and clarified by
centrifugation (6000g, 10 min).
4. The supernatant from Step 3 is then dialyzed against 1 l of antibody
buffer overnight at 4  C.
5. The clarified supernatant is applied to a column of DE52 that has been
prepared as discussed in ‘‘comments’’ below. The column is equilibrated
in antibody buffer (pH 6.9), and at this pH most of the mAb flows
through the column while most impurities bind. For 10 ml of the
starting material, we use a 5-ml column of DE52.
6. Fractions (about 1 ml) of material that does not bind to the column are
collected, and samples of each fraction are subjected to SDS–PAGE. The
SDS–PAGE in Fig. 28.2B shows the fractions from the purification of
mAb 8RB13 (an IgG1 mAb) that was produced in a CL 350 flask.
Fractions are pooled according to purity of the antibody product.
7. The column is then eluted with 5 ml of antibody buffer containing
0.5 M NaCl. Save this eluate until the fractions are analyzed by PAGE.

2.5.1. Comments
1. The preparation of the DE52 is particularly important. Although the
manufacturer states that precycling is not necessary, we have found that
precycling greatly improves the performance of the chromatography.
Ten grams of resin should be resuspended in 100 ml of water and
washed several times with 100 ml of water, removing the fines (the
small particles that do not settle easily) with each washing. The resin is
then treated with 100 ml of 0.1 M HCl for 30 min. The acid is decanted
and the resin washed at least three times with 100 ml of water for each
wash. After the last wash is decanted, the resin is treated with 0.1 M
NaOH for 30 min. The base is decanted, and the resin is washed at least
three times with 100 ml of water for each wash. The resin is then
washed three times with 100 ml of the antibody buffer and resuspended
in the same buffer. The pH is checked with pH paper, and NaN3 is
added to 0.02%. Resin is distributed to disposable tubes and stored in
the refrigerator.
2. If the mAb is made in ascites fluid, the product is not as clean (approxi-
mately 80–90% pure), but the impurities do not seem to interfere with
the performance of the immunoaffinity resin.
3. Under the conditions described above, most mouse mAbs will flow
through the DE52. However, a few mAbs will bind under these condi-
tions. Therefore, the high salt elution (Step 7) will elute the mAb. It will
Immunoaffinity Chromatography 487

be necessary to use a salt gradient to purify a mAb that binds to the DE52
column.
4. If the mAb is made in the CELLine flask, the DE52 step might not be
necessary.

2.6. Immobilization of PR-mAbs on a chromatography

support
Numerous resins and coupling chemistries are available through commer-
cial suppliers. We have tested a number of these resins, but have not found
that any of them work better than cross-linked agarose that has been
derivatized with cyanogen bromide (CNBr).
1. The purified mAb is dialyzed against coupling buffer (100 mM sodium
bicarbonate, 500 mM NaCl, pH 8.3). The antibody solution is removed
from the dialysis tube and the volume adjusted to about 10 ml with
coupling buffer for each gram of dried CNBr-activated Sepharose used.
A sample (about 100 ml) is reserved for protein concentration analysis.
2. Dried CNBr-activated agarose is swelled in 0.1 mM HCl for about
20 min. Each gram of dried CNBr-activated resin makes 3.5 ml of
swollen gel. The resin is then washed on a fritted glass filter, using
about 100 ml of 0.1 mM HCl for a gram of resin.
3. The resin is quickly washed with about 20 ml of coupling buffer, and the
resin is transferred to the antibody solution. This slurry is mixed end-
over-end on a laboratory rotator for 2 h at 23  C.
4. The resin is collected on the fritted glass filter, and the filtrate is saved for
determining the amount of protein that did not couple.
5. The resin is transferred to about 10 ml of 1 M ethanolamine, pH 8.3 and
mixed end-over-end to 2 h at 23  C, to allow the ethanolamine to react
with any remaining reactive CNBr groups.
6. The resin is again collected on the fritted glass filter, and washed with the
coupling buffer (about 50 ml) and then washed with 100 mM sodium
acetate buffer (pH 4.0).
7. Repeat the two washings in Step 6 at least two to three times.
8. Store the conjugated Sepharose in 10 ml of coupling buffer containing
0.02% NaN3 at 4  C.
9. Perform a protein concentration analysis on the antibody solutions saved
before and after conjugation (samples from steps 1 and 4). A coupling
efficiency can be determined from these samples.

2.6.1. Comments
1. One gram of this resin yields 3.5 ml of swollen resin. For most purposes,
we find making 0.5–2.0 g at a time to be convenient. We have found
488 Nancy E. Thompson et al.

that coupling 2.5 mg of mAb to 1 ml of swelled resin works well.

Therefore, 8.75 mg of mAb is needed to conjugate 3.5 ml (1 g) of resin.
2. The CNBr is activated by pH values over 8. Therefore, the transfer of
the resin to the antibody solution described in Step 3 above should be
performed as quickly as possible.
3. The blocking agent (Step 5) can vary with the use. For example, yeast
produces an ethanolamine-binding protein that will be copurified with
the target protein if ethanolamine is used for blocking. In this case 0.1 M
glycine is a more appropriate blocking agent.
4. The resin is stored at 4  C in 0.02% NaN3. The resin is stable for about
6 months under these conditions. We have noticed that some antibody
leaches off after about 6 months of storage. Storage of the resin in the
above buffer but at 50 % glycerol allows storage at  20  C (where it
does not freeze) and seems to allow for a longer half life.

2.7. Purification of proteins with PR-mAbs

As an example, we describe here the method of purifying RNA polymerase
from E. coli, using either mAb NT73 or 8RB13. We have presented a one-step
immunoaffinity chromatography procedure that will yield RNAP that is
about 90% pure. Some of the extra proteins that coelute are RNAP-binding
proteins. This protocol assumes that the starting material is a bacterial pellet
containing E. coli cells (2–3 g wet weight) from 1 l of late-log phase culture.
The SDS–PAGE shown in Fig. 28.3 is a composite gel of this purification.
1. The pellet is allowed to partially thaw on ice and resuspended in 20 ml of
TEN buffer (50 mM Tris–HCl, pH 7.9, 0.1 mM EDTA, and 100 mM
NaCl).
2. Lysozyme is added to a concentration of 250 mg/ml and the cells are
incubated on ice for 20 min. Alternatively, 1500 kU of rLysozyme
(EMD/Novagen #71110) can be used.
3. The cells are sonicated on ice four times for 15 s each, with 15 s rests
between each sonication burst.
4. The lysate is centrifuged at 15,000 rpm (27,000g) for 15 min.
5. The supernatant is applied to the immunoaffinity column (1–2 ml) at
23  C. A flow-through fraction is saved.
6. The column is washed with TE containing 100 mM NaCl (about 20 ml)
and then with TE containing 500 ml NaCl (about 5 ml). The column is
the reequilibrated in TE containing 100 mM NaCl (about 10 ml).
7. The column is eluted with TE containing 0.75 M NaCl and 40%
propylene glycol (at room temperature). The fractions are placed on
ice as they are collected.
8. The peak fractions are analyzed on a SDS–PAGE gel. The SDS–PAGE
in Fig. 28.3 shows the fractions obtained from the one-step
Immunoaffinity Chromatography 489

0.5 M NaCl wash

Flow-through

Elution 3

Elution 4
Markers

Lysate
kDa

260
160 b ′/b
110
80 s 70
60

50
40 a

30

20

15

10

3.5
1 2 3 4 5 6

Figure 28.3 SDS–PAGE (4–12%) of RNAP purified from E. coli by a one-step immu-
noaffinity chromatography procedure, using mAbNT73-Sepharose. The cell lysate was
prepared as described in the text, and the soluble material (Lane 2) was applied to a NT73-
Sepharose column (2 ml). The flow-through material (Lane 3) was collected. After
washing with TE buffer containing 100 mM NaCl, the column was washed briefly with
TE containing 0.5 M NaCl (Lane 4). The column was re-equilibrated in TE containing
100 mM NaCl, and then the RNAP was eluted from the column with TE buffer contain-
ing 0.75 M NaCl and 40% propylene glycol, and 1-ml fractions collected (Lanes 5 and 6).
RNAP subunits are indicated on the right side. Most of the extra bands are RNAP-
binding proteins.

chromatography. The appropriate fractions (Lanes 5–6 in Fig. 28.3) are

pooled and dialyzed against a suitable storage buffer. (For transcription
proteins, we use 50 mM Tris–HCl, pH 7.9; 0.1 mM EDTA, 0.1 mM
DTT, 50 mM NaCl, and 20–50% glycerol).

2.7.1. Comments
1. The lysate (Step 3) can be treated with Benzonase (EMD/Novagen
#70746, Madison, WI) to help digest nucleic acids and decrease viscosity
(see Chapter 18 in this volume).
490 Nancy E. Thompson et al.

2. The 500 mM NaCl wash (Step 6) helps to remove nucleic acids and
NusA, a RNA polymerase-binding protein.
3. For reasons that are unknown to us, the elution of the target protein
(Step 8) is significantly more effective if performed at room temperature
than at 4  C.

2.8. Purification of proteins using cross-reacting PR-mAbs

It is often desirable to purify proteins or protein complexes from biological
materials that have not been genetically manipulated. In this case, we have
found that the use of PR-mAbs that react with an epitope that is highly
conserved makes the immunoadsorbent more versatile. Two of our most
successful PR-mAbs react with the same enzyme across many species.
mAb 8WG16 reacts with the heptapeptide repeat on the C-terminus of
the largest subunit of eukaryotic RNAP II (Table 28.1). This sequence,
commonly referred to as the C-terminal domain (CTD) of RNAP II, is
accessible on the surface of this large protein complex. The CTD is highly
conserved on RNAP II from almost all species. mAb 8WG16 has been used
to purify RNAP II from calf thymus (Thompson et al., 1990), yeast
(Edwards et al., 1990), and human cells (Maldonado et al., 1996). In fact,
this PR-mAb was used to purify the yeast RNAP II that was used in the first
crystallization studies of RNAP II (Cramer et al., 2000). We have described
a detailed procedure for using this PR-mAb to purify RNAP II (Thompson
and Burgess, 1996). In most cases, the purification of RNAP II from crude
material requires some bulk purification steps before the material is applied
to the immunoaffinity resin.
mAb 8RB13 is a highly cross-reactive PR-mAb that has proved to be
very useful for purification of bacterial core RNAP (Bergendahl et al., 2003;
Probasco et al., 2007). This PR-mAb reacts with the highly conserved
‘‘b-flap’’ region of the bacterial RNAP from most bacteria (E. S. Stalder,
unpublished data). Because the b-flap is one of the major binding sites for
the sigma subunit (Kuznedelov et al., 2002), core RNA polymerase (that
lacks a sigma subunit) is purified on the 8RB13-agarose column. Lane 3 in
Fig. 28.4 shows core RNAP purified from E. coli. Because of the cross-
reactivity of this mAb, core RNAP can also be purified from Bacillus subtilis
(Lane 4). Also shown in Fig. 28.4 is RNAP purified from E. coli using mAb
NT73 (Lane 2); this preparation is a mixture of the holoenzyme and the
core enzyme, and these two forms of the enzyme can be separated from this
fraction by ion-exchange chromatography (Thompson et al., 1992). The
peptide band corresponding to the major sigma factor is missing from core
RNAP purified from E. coli and B. subtilis.
Immunoaffinity Chromatography 491

B. subtilis core RNAP (8RB13)

Epitope-tagged GFP (NT73)

E. coli core RNAP (8RB13)
E. coli RNAP (NT73)
Markers
kDa

160 b /b ′

s 70
60
50
40 a

30 GFPep
20

15
10

3.5
1 2 3 4 5

Figure 28.4 SDS–PAGE (4–12%) of RNAP and epitope-tagged GFP purified on PR-
mAbs. Lane 1 contains Novex Sharp Standard markers; Lane 2 contains E. coli RNAP
purified on mAb NT73; Lane 3 contains E. coli core RNAP purified on mAb 8RB13;
Lane 4 contains B. subtilis core RNAP purified on mAb 8RB13; Lane 5 contains
epitope-tagged GFP purified on mAb NT73.

2.9. Use of epitopes of PR-mAbs as purification tags

Although we believe that about 5–10% of the mAbs that are isolated from
standard hybridoma techniques are potentially useful PR-mAbs, another
approach is to epitope-tag a protein with the epitope for an established PR-
mAb. The concept of purification tags is described in Chapter 16 in this
volume. Therefore, we will only briefly describe this topic, and only as it
relates to PR-mAbs and their epitopes.
The ability to tag a protein with an epitope and use the PR-mAb to
purify the protein established that polyol-responsiveness is indeed a prop-
erty of the antibody, and not a property of the environment in which the
epitope is presented. The three PR-mAbs for which we have developed
epitope-tagged systems are NT73, IIB8, and 8RB13 (Duellman et al., 2004;
492 Nancy E. Thompson et al.

Thompson et al., 2003; E. S. Stalder, unpublished data). We have designated

these epitope tags as ‘‘Softags’’ (Burgess and Thompson, 2002). All of these
epitopes are listed in Table 28.1, and a patent has been issued on the use of
these tags (Burgess et al., 2007).
All of our PR-mAbs have been isolated using the full-length protein as
the immunogen; therefore, for some of these mAbs, the epitope mapping has
been a laborious process. We have not attempted to isolate a PR-mAb using
a synthetic peptide as an immunogen, although this should be possible. IIB8,
the PR-mAb that reacts with human TFIIB, was mapped using phage
display followed by site-directed mutagenesis (Duellman et al., 2004).
The two PR-mAbs that react with the largest subunits of E. coli RNA
polymerase (mAbs NT73 and 8RB13) were roughly mapped using the
ordered fragment ladder method (Burgess et al., 2000; Rao et al., 1996),
followed by fine deletion analysis and oligonucleotide tagging of an unre-
lated protein for which we have an antibody (Thompson et al., 2003;
E. S. Stalder, unpublished data). The tags are constructed by fusing an
oligonucleotide containing the coding sequence for the epitope in-frame
with the target protein.
We have used the green fluorescent protein (GFP) as a target protein as
proof-of-principle for the technique (Duellman et al., 2004; Thompson
et al., 2003). The mAb NT73 epitope tag can be fused to either the N- or
C-terminus of GFP, but for other target proteins, this might be dependent
upon the accessibility of the termini. In the case of GFP, both of the termini
are accessible in the crystal structure (Tsien, 1998). The epitope for mAb
8RB13 has been used as a tag in both the E. coli and mammalian cell culture
systems (E. S. Stalder, unpublished).
While we have used the epitope tags derived from E. coli RNAP for
purification of epitope-tagged GFP produced in the E. coli expression
system, the endogenous RNAP is still in the lysate and some RNAP is
also purified. Lane 5 in Fig. 28.4 shows GFP tagged with the epitope for
NT73 expressed in E. coli and purified with NT73-Sepharose. The RNAP
can be removed from this material by bringing the lysate to 300 mM NaCl
and adding 0.3% polyethyleneimine (PEI) (see Chapter 20 in this volume).
The resulting heavy precipitate is removed by centrifugation, and the
epitope-tagged GFP remains in solution. The supernatant is then applied
to the immunoaffinity column.

3. Conclusions
We have described the isolation, identification, and use of PR-mAbs
for use in gentle immunoaffinity chromatography. Although the examples
that we have presented here apply to our PR-mAbs and their use to purify
Immunoaffinity Chromatography 493

transcription factors, the methods outlined in this chapter can be applied to

any PR-mAb or to a protein that has been genetically tagged with an
epitope for a PR-mAb. The most expensive component of this procedure
is the isolation of the PR-mAb. Because there are now many panels of
mAbs for a given protein, and these existing mAbs can be screened for
polyol-responsiveness by the ELISA-elution assay, it is possible that a PR-
mAb that reacts with a protein that you are interested in already exists.

DISCLOSURE
N. Thompson and R. Burgess are required by the University of Wisconsin–Madison
Conflict of Interest Committee to disclose that they have financial interests in the company
NeoClone which markets many of the mAbs mentioned in this chapter.

REFERENCES
Anthony, J. R., Green, H. A., and Donohue, T. J. (2003). Purification of Rhodobacter
sphaeroides RNA polymerase and its sigma factors. Meth. Enzymol. 370, 54–65.
Bergendahl, V., Thompson, N. E., Foley, K. M., Olson, B. M., and Burgess, R. R. (2003).
A cross-reactive polyol-responsive monoclonal antibody useful for isolation of core
RNA polymerase from many bacterial species. Protein Expr. Purif. 31, 155–160.
Burgess, R. R., and Thompson, N. E. (2002). Advances in gentle immunoaffinity chroma-
tography. Curr. Opin. Biotechnol. 13, 304–308.
Burgess, R. R., Arthur, T. A., and Pietz, B. C. (2000). Mapping protein-protein interaction
domains using ordered fragment ladder far-Western analysis of hexahistidine-tagged
fusion proteins. Meth. Enzymol. 328, 141–157.
Burgess, R. R., Thompson, N. E., and Duellman, S. J. (2007). Immunoaffinity chromatog-
raphy using epitope tags to polyol-responsive monoclonal antibodies. US Patent No.
7,241,580.
Cramer, P., Bushnell, D. A., Fu, J., Gnatt, A. L., Maier-Davis, B., Thompson, N. E.,
Burgess, R. R., Edwards, A. M., David, P. R., and Kornberg, R. D. (2000). Architecture
of RNA polymerase II and implications for the transcription mechanism. Science 288,
640–649.
Duellman, S. J., Thompson, N. E., and Burgess, R. R. (2004). An epitope tag derived from
human transcription factor IIB that reacts with a polyol-responsive monoclonal antibody.
Protein Expr. Purif. 35, 147–155.
Edwards, A. M., Darst, S. A., Feaver, W. J., Thompson, N. E., Burgess, R. R., and
Kornberg, R. D. (1990). Purification and lipid layer crystallization of yeast RNA
polymerase II active in transcription initiation. Proc. Natl. Acad. Sci. USA 87, 2122–2126.
Harlow, E., and Lane, D. (1988). Antibodies: A Laboratory Manual. Cold Spring Harbor
Press, Cold Spring Harbor, NY11724.
Jiang, Y., Zhang, S. J., Wu, S. M., and Lee, M. Y. (1995). Immunoaffinity purification of
DNA polymerase delta. Arch. Biochem. Biophys. 230, 297–304.
Kuznedelov, K., Minakhin, L., Niedziela-Majka, A., Dove, S. L., Rogulja, D.,
Nickels, B. E., Hochschild, A., Heyduk, T., and Severinov, K. (2002). A role for the
interaction of the RNA polymerase flap domain with the sigma subunit in promoter
recognition. Science 295, 855–857.
494 Nancy E. Thompson et al.

Largaespada, D. A., Jackson, M. W., Thompson, N. E., Kaehler, D. A., Byrd, L. G., and
Mushinski, J. F. (1996). The ABL-MYC retrovirus generates antigen-specific plasmacy-
tomas by in vitro infection of activated B lymphocytes from spleen and other murine
lymphoid organs. J. Immunol. Methods 197, 85–95.
Lynch, N. A., Jiang, H., and Gibson, D. T. (1996). Rapid purification of the oxygenase
component of toluene dioxygenase from a polyol-responsive monoclonal antibody.
Appl. Environ. Microbiol. 62, 2133–2137.
Maldonado, E., Drapkin, R., and Reinberg, D. (1996). Purification of human RNA
polymerase II and general transcription factors. Meth. Enzymol. 274B, 72–100.
Nagy, P. L., Griesenbeck, J., Kornberg, R. D., and Cleary, A. (2002). A trithorax-group
complex purified from Saccharomyces cerevisiae is required for methylation of histone H3.
Proc. Natl. Acad. Sci. USA 99, 90–94.
Nelson, P. N., Reynolds, G. M., Waldron, E. E., Ward, E., Giannopoulos, K., and
Murray, P. G. (2000). Demystified: Monoclonal antibodies. J. Clin. Pathol. Mol. Pathol.
53, 111–117.
Probasco, M. D., Thompson, N. E., and Burgess, R. R. (2007). Immunoaffinity purification
and characterization of RNA polymerase from Shewanella oneidensis. Protein Expr. Purif.
55, 23–30.
Rao, L., Jones, D. P., Nguyen, L. H., McMahan, S. A., and Burgess, R. R. (1996). Epitope
mapping using histidine-tagged protein fragments: Application to Escherichia coli RNA
polymerase sigma70. Anal. Biochem. 241, 141–157.
Subramanian, A. (2002). Immunoaffinity chromatography. Mol. Biotechnol. 20, 41–47.
Thompson, N. E., and Burgess, R. R. (1994). Purification of recombinant human transcription
factor IIB (TFIIB) by immunoaffinity chromatography. Protein Expr. Purif. 5, 469–475.
Thompson, N. E., and Burgess, R. R. (1996). Immunoaffinity of RNA polymerase and
transcription factors using polyol-responsive monoclonal antibodies. Meth. Enzymol.
274B, 513–526.
Thompson, N. E., and Burgess, R. R. (1999). Immunoaffinity purification of the RAP30
subunit of the human transcription factor IIF (TFIIF). Protein Expr. Purif. 17, 260–266.
Thompson, N. E., and Burgess, R. R. (2001). Preparation and use of specialized antibodies.
Identification of polyol-responsive monoclonal antibodies for use in immunoaffinity
chromatography. Curr. Protoc. Mol. Biol. Sect. VI (Suppl. 54), 11.18.1–11.18.9.
Thompson, N. E., Aronson, D. B., and Burgess, R. R. (1990). Purification of eukaryotic
RNA polymerase II by immunoaffinity chromatography: Elution of active enzyme with
protein stabilizing agents from a polyol-responsive monoclonal antibody. J. Biol. Chem.
265, 7069–7077.
Thompson, N. E., Hager, D. A., and Burgess, R. R. (1992). Isolation and characterization of
a polyol-responsive monoclonal antibody useful for gentle purification of E. coli RNA
polymerase. Biochemistry 31, 7003–7008.
Thompson, N. E., Arthur, T. M., and Burgess, R. R. (2003). Development of an epitope tag
for the gentle purification of proteins by immunoaffinity chromatography: Application to
epitope-tagged green fluorescent protein. Anal. Biochem. 323, 171–179.
Thompson, N. E., Foley, K. M., and Burgess, R. R. (2004). Antigen-binding properties of
monoclonal antibodies reactive with human TATA-binding protein (TBP) and use in
immunoaffinity chromatography. Protein Expr. Purif. 36, 186–197.
Thompson, N. E., Jensen, D. B., Lamberski, J. A., and Burgess, R. R. (2006). Purification of
protein complexes by immunoaffinity chromatography: Application to transcription
machinery. Genet. Eng. 27, 81–100.
Tsien, R. Y. (1998). The green fluorescent protein. Annu. Rev. Biochem. 67, 509–544.
Woychik, N. A., and Hampsey, M. (2002). The RNA polymerase II machinery: Structure
illuminates function. Cell 108, 453–463.
 C H A P T E R T W E N T Y- N I N E

One-Dimensional Gel Electrophoresis1

David E. Garfin

Contents
1. Background 498
2. Polyacrylamide Gels 500
3. Principle of Method 501
4. Procedure 502
4.1. Stock solutions 502
4.2. Catalyst 503
4.3. Electrode buffer 503
4.4. Casting gels 503
4.5. Sample preparation 505
4.6. Electrophoresis 505
4.7. Comments on method 506
4.8. Variations of method 507
5. Detection of Proteins in Gels 508
5.1. Dye staining with Coomassie Brilliant Blue R-250 509
5.2. Silver staining 509
5.3. Copper staining 510
6. Marker Proteins 510
7. Molecular Weight Determination 511
8. Preparative Electrophoresis 511
References 513

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS–

PAGE) is an excellent method with which to identify and monitor
proteins during purification and to assess the homogeneity of purified
fractions. SDS–PAGE is routinely used for the estimation of protein
subunit molecular weights and for determining the subunit compositions
of purified proteins. SDS–PAGE can also be scaled up, for use in a
preparative mode, to yield sufficient protein for further studies. In addi-
tion, two-dimensional analysis, combining isoelectric focusing with
SDS–PAGE (Dunbar, 1987, this volume), is a very high resolution
method for protein fractionation, enabling thousands of polypeptides to

Chemical Division, Research Products Group, Bio-Rad Laboratories, Incorporated, Richmond, California
1
Reprinted from Methods in Enzymology, Volume 182 (Academic Press, 1990)

Methods in Enzymology, Volume 463 Copyright # 1990 by Academic Press

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63029-9 All rights reserved.

497
498 David E. Garfin

be resolved in a single gel. When used in conjunction with blotting

methods (Timmons and Dunbar, this volume), SDS–PAGE provides
one of the most powerful means available for protein analysis.
A great many electrophoretic systems have been developed and no
attempt is made to summarize them here. In particular, the distinctions
between the various ‘‘continuous’’ and ‘‘discontinuous’’ buffer systems are
not discussed, nor are alternative support matrices considered. Gradient gels
(gels whose pore sizes vary) are also omitted from discussion, since these can
be prepared by relatively straightforward adaptation of any of a number of
well-known methods for forming gradients. Rather, only the most com-
mon (and most reliable) analytical SDS–PAGE procedure (Laemmli, 1970)
is described. Those wishing further information on the practical or theoret-
ical aspects of electrophoretic processes can use (Allen et al., 1984; Andrews,
1986; Chrambach, 1985; Hames, 1981) to gain access to the large volume of
literature in the field. Some problems may require adoption of alternative
procedures (Allen et al., 1984; Andrews, 1986; Blackshear, this series; Bury,
1981; Chrambach, 1985; Hames, 1981; Neville, 1971; Neville and
Glossmann, this series), but for most applications the SDS–PAGE method
presented here will perform satisfactorily.

1. Background
Although the detailed theory of gel electrophoresis is complicated and
at present incomplete (Bier et al., 1983; Chrambach and Jovin, 1983; Jovin,
1973), the fundamental concepts are easily understood. Briefly, in an
electrophoretic separation, charged particles are caused to migrate toward
the electrode of opposite sign under the influence of an externally applied
electric field. The movements of the particles are retarded by interactions
with the surrounding gel matrix, which acts as a molecular sieve. The
opposing interactions of the electrical force and molecular sieving result in
differential migration rates for the constituent proteins of a sample.
In general, fractionation by gel electrophoresis is based on the sizes,
shapes, and net charges of the macromolecules. Systems designed to frac-
tionate proteins in their native configurations cannot distinguish between
the effects of size, shape, and charge on electrophoretic mobility. As a
consequence, proteins with differing molecular weights can have the same
mobility in these systems. Thus, while PAGE methods for native proteins
are valuable for separating and categorizing protein mixtures, they should
not be used to assess the purity of a preparation or the molecular weight of
an unknown.
SDS–PAGE overcomes the limitations of native PAGE by imposing
uniform hydrodynamic and charge characteristics on all the proteins in a
One-Dimensional Gel Electrophoresis 499

sample mixture. During sample preparation, proteins are treated with

hot SDS. The anionic detergent binds tightly to most proteins at about
1.4 mg of SDS/mg of protein, imparting a negative charge to the resultant
complexes (Nielsen and Reynolds, this series). Interaction with SDS
disrupts all noncovalent protein bonds, causing the macromolecules to
unfold. Concomitant treatment with a disulfide-reducing agent, such as
2-mercaptoethanol or dithiothreitol, further denatures proteins, breaking
them down to their constituent subunits. The electrophoretic mobilities
of the resultant detergent–polypeptide complexes all assume the same
functional relationship to their molecular weights. Migration of SDS deri-
vatives is toward the anode at rates inversely proportional to the logarithms
of their molecular weights (Neville, 1971; Neville and Glossmann, this
series; Shapiro et al., 1967; Weber and Osborn, 1969). SDS polypeptides,
thus, move through gels in a predictable manner, with low-molecular-
weight complexes migrating faster than larger ones. This means that the
molecular weight of a protein can be estimated from its relative mobility in
a calibrated SDS–PAGE gel and that a single band in such a gel is a criterion
of purity.
Most electrophoresis is done in vertical chambers in gel slabs formed
between two glass plates (Andrews, 1986; Hames, 1981). The slab format
provides uniformity, so that different samples can be directly compared in
the same gel. Gel thicknesses are established by spacers placed between the
glass plates and sample wells are formed in the gels during polymerization
with plastic, comb-shaped inserts. Electrophoresis cells provide means for
sealing the assemblies during gel formation and for maintaining contact with
the electrode buffers during runs. The better cells provide means for heat
dissipation, because uneven heat distribution in the gel slab can cause band
distortions.
Conventional gels are of the order of 16–20 cm long, 16 cm wide, and
0.5–3.0 mm in thickness and can accommodate about 25 samples. Thick
gels have greater total protein capacity than thin ones, but are correspond-
ingly less efficient at dissipating electrically generated heat and more difficult
to stain and destain. Gel thicknesses of 0.75 or 1 mm are good compromise
sizes, combining adequate protein loads and good staining speeds with
minimal heat-related distortions. Typical runs take 4–5 h.
Small-format cells (minicells) allow rapid analyses and are adequate for
relatively uncomplicated samples. The design of these cells allows analyses
to be completed two to three times faster than is possible with conventional
cells. The gels are about 7 cm long
8 cm wide and are very easily
manipulated. Each gel can hold up to about 15 samples and a typical run
can be completed in less than 1 h (not counting setup and polymerization
time). The resolution of complex samples may be better in conventional
gels than with minicells, since the separation of protein bands is improved
by increasing the lengths of SDS–PAGE gels.
500 David E. Garfin

2. Polyacrylamide Gels
Polyacrylamide gels are formed by copolymerization of acrylamide
monomer, CH2¼CH–CO–NH2, and a cross-linking comonomer, N,N 0 -
methylenebisacrylamide, CH2¼CH–CO–NH–CH2–NH–CO–CH¼CH2,
(bisacrylamide) (Allen et al., 1984; Andrews, 1986; Chrambach, 1985;
Chrambach and Rodbard, 1971; Hames, 1981). The mechanism of gel
formation is vinyl addition polymerization and is catalyzed by a free radical-
generating system composed of ammonium persulfate (the initiator) and an
accelerator, tetramethylethylenediamine (TEMED). TEMED causes the
formation of free radicals from persulfate and these in turn catalyze polymeri-
zation. Oxygen, a radical scavenger, interferes with polymerization, so that
proper degassing to remove dissolved oxygen from acrylamide solutions is
crucial for reproducible gel formation.
The sieving properties of a gel are established by the three-dimensional
network of fibers and pores which is formed as the bifunctional bisacryla-
mide cross-links adjacent polyacrylamide chains (Rodbard and Chrambach,
1970). Within limits, as the acrylamide concentration of a gel increases, its
effective pore size decreases. The effective pore size of a gel is operationally
defined by its sieving properties; that is, by the resistance it imparts to the
migration of protein molecules. By convention, a given gel is physically
characterized by the pair of figures (%T, %C), where %T is the weight
percentage of total monomer (acrylamide þ cross-linker, in g/100 ml), and
%C is the proportion of cross-linker (as a percentage of total monomer) in
the gel. The practical limits for %T lie between 3% and 30%. The factors
governing pore size are complicated, but, in general, the pore size of a gel
decreases as %T increases. For any given fixed %T, pore size is at a minimum
at about 5% C, increasing at both higher and lower cross-linker concentra-
tions (Allen et al., 1984; Andrews, 1986; Chrambach, 1985; Chrambach and
Rodbard, 1971; Hames, 1981).
The use of high-quality reagents is a prerequisite for reproducible, high-
resolution gels. This is particularly true of acrylamide, which constitutes the
most abundant component in the gel–monomer mixture. Residual acrylic
acid, linear polyacrylamide, and ionic impurities are the major contaminants
of acrylamide preparations. Moreover, buffer components should be of
reagent grade and only distilled or deionized water should be used for all
phases of gel electrophoresis.
In SDS–PAGE, the quality of the SDS is of prime importance.
Differential protein-binding properties of impurities such as C10, C14, and
C16 alkyl sulfates can cause single proteins to form multiple bands in gels
(Margulies and Tiffany, 1984). Even with pure SDS, very basic proteins,
very acidic proteins, various glycoproteins, and lipoproteins, because of
their unusual compositions, migrate ‘‘anomalously’’ during electrophoresis
(Allen et al., 1984; Andrews, 1986; Hames, 1981).
One-Dimensional Gel Electrophoresis 501

3. Principle of Method
The most popular electrophoretic method is the SDS–PAGE system
developed by Laemmli (Allen et al., 1984; Andrews, 1986; Blackshear, this
series; Hames, 1981; Laemmli, 1970). This is a discontinuous system con-
sisting of two contiguous, but distinct gels: a resolving or separating (lower)
gel and a stacking (upper) gel. The two gels are cast with different porosities,
pH, and ionic strength. In addition, different mobile ions are used in the gel
and electrode buffers. The buffer discontinuity acts to concentrate large
volume samples in the stacking gel, resulting in better resolution than is
possible using the same sample volumes in gels without stackers. Proteins,
once concentrated in the stacking gel, are separated in the resolving gel.
The Laemmli SDS–PAGE system is made up of four components. From
the top of the cell downward, these are the electrode buffer, the sample, the
stacking gel, and the resolving gel. Samples prepared in low-conductivity
buffer (0.06 M Tris–Cl, pH 6.8) are loaded between the higher conductivity
electrode (0.025 M Tris, 0.192 M glycine, pH 8.3) and stacking gel
(0.125 M Tris–Cl, pH 6.8) buffers. When power is applied, a voltage
drop develops across the sample solution which drives the proteins into
the stacking gel. Glycinate ions from the electrode buffer follow the proteins
into the stacking gel. A moving boundary region is rapidly formed with the
highly mobile chloride ions in the front and the relatively slow glycinate
ions in the rear (Allen et al., 1984; Andrews, 1986; Blackshear, this series;
Bury, 1981; Hames, 1981; Wyckoff et al., 1977). A localized high-voltage
gradient forms between the leading and trailing ion fronts, causing the SDS–
protein complexes to form into a thin zone (stack) and migrate between the
chloride and glycinate phases. Within broad limits, regardless of the height
of the applied sample, all SDS–proteins condense into a very narrow region
and enter the resolving gel as a well-defined, thin zone of high protein
density. (The stacking phenomenon is strikingly demonstrated with
prestained protein standards, which are mixtures of proteins derivatized
with reactive dyes.) The large-pore stacking gel (4% T ) does not retard
the migration of most proteins and serves mainly as an anticonvective
medium. At the interface of the stacking and resolving gels, the proteins
experience a sharp increase in retardation due to the restrictive pore size of
the resolving gel. (Proteins too large to enter the resolving gel will stop
at the interface.) Once in the resolving gel, proteins continue to be slowed
by the sieving of the matrix. The glycinate ions overtake the proteins,
which then move in a space of uniform pH (pH 9.5) formed by the Tris
and glycine. Molecular sieving causes the SDS–polypeptide complexes to
separate on the basis of their molecular weights.
502 David E. Garfin

4. Procedure
Equipment and reagents for SDS–PAGE can be obtained from a
variety of suppliers. Electrophoresis cells vary in design, but their operation
generally follows the steps outlined below. Since the many available cells
differ in size, formulations are presented in conveniently sized units for
simplicity. Required volumes can be prepared using multiples of these unit
sizes. Except where noted, reagents for SDS–PAGE can be prepared as
concentrated stock solutions.

4.1. Stock solutions

1. Acrylamide concentrate (30% T, 2.7% C): Dissolve 29.2 g of acrylamide
and 0.8 g of bisacrylamide in 70 ml of deionized water. When the
acrylamide is completely dissolved, add water to a final volume of
100 ml. Filter the solution under vacuum through a 0.45-mm mem-
brane. Store stock acrylamide at 4  C in a dark bottle for no more than
1 month. Caution: Acrylamide monomer is a neurotoxin. Avoid breath-
ing acrylamide dust, do not pipette acrylamide solutions by mouth, and
wear gloves when handling acrylamide powder or solutions containing
it. For disposal of unused acrylamide, add bisacrylamide (if none is
present), induce polymerization, and discard the solidified gel.
2. 1.5 M Tris–Cl, pH 8.8, concentrated resolving gel buffer: Dissolve 18.2 g Tris
base in 80 ml of water, adjust to pH 8.8 with HCl, and add water to a
final volume of 100 ml. Store at 4  C.
3. 0.5 M Tris–Cl, pH 6.8, concentrated stacking gel buffer: Dissolve 6.1 g Tris
base in 80 ml of water, adjust to pH 6.8 with HCl, and add water to a
final volume of 100 ml. Store at 4  C.
4. 10% (w/v) sodium dodecyl sulfate (SDS): Dissolve 10 g SDS in 60 ml of
water and add water to a final volume of 100 ml.
5. Stock sample buffer (0.06 M Tris–Cl, pH 6.8, 2% SDS, 10% glycerol,
0.025% Bromphenol Blue):
Water 4.8 ml
0.5 M Tris–Cl, pH 6.8 1.2 ml
10% SDS 2.0 ml
Glycerol 1.0 ml
0.5% Bromphenol Blue (w/v water) 0.5 ml

Store at room temperature. SDS-reducing buffer is prepared by

adding 50 ml of 2-mercaptoethanol to each 0.95 ml of stock sample
buffer before use.
One-Dimensional Gel Electrophoresis 503

4.2. Catalyst
1. 10% ammonium persulfate (APS): Dissolve 100 mg APS in 1 ml of water.
Make the APS solution fresh daily.
2. TEMED (N,N,N0 ,N0 -tetramethylethylenediamine): Use TEMED undi-
luted from the bottle. Store cool, dry, and protected from light.

4.3. Electrode buffer

Electrode buffer: 0.025 M Tris, 0.192 M glycine, 0.1% (w/v) SDS, pH 8.3
(0.3 g Tris base, 1.4 g glycine, 1 ml 10% SDS/100 ml electrode buffer).
Do not adjust the pH of the electrode buffer; just mix the reagents together
and confirm that the pH is near 8.3 ( 0.2). Electrode buffer can be made as
a 5
concentrate consisting of 15 g Tris base, 72 g glycine, and 5 g SDS/l.
5
electrode buffer concentrate must be stored in glass containers. To use
5
concentrate, dilute it with four parts water.

4.4. Casting gels

Thoroughly clean the glass plates, spacers, combs, and upper buffer reservoir
of the gel apparatus with detergent and rinse them well. Wear gloves while
assembling the equipment. The resolving gel is cast first, then overlaid with
the stacking gel.
1. Assemble the casting apparatus and determine the gel volume from the
manufacturer’s instructions or by calculation. A 1- to 2-cm stacking gel
is used above the resolving gel. Determine the height to which the
resolving gel is to be poured by inserting a well-forming comb between
the glass plates and marking the outer plate 1–2 cm below the teeth of
the comb.
2. Prepare the monomer solution for the appropriate resolving gel by
combining all of the reagents in Table 29.1 except the ammonium
persulfate (APS) and TEMED; a disposable, plastic beaker is a conve-
nient mixing vessel. The two gel recipes given in Table 29.1 cover the
molecular weight ranges usually encountered. Gels of any other acryl-
amide concentration desired (Andrews, 1986; Blackshear, this series;
Hames, 1981), can be prepared by adjusting (only) the amounts of
30% monomer stock and water used in the recipes. Deaerate the solution
under vacuum (e.g., in a bell jar or desiccator) for at least 15 min.
3. Gently mix the APS and TEMED (Table 29.1) into the deaerated
monomer solution. Using a pipet and bulb, add the monomer solution
between the gel plates up to the mark delimiting the resolving gel.
Immediately overlay the monomer solution with water-saturated
2-butanol or tert-amyl alcohol to exclude air, which might inhibit
504 David E. Garfin

Table 29.1 Formulations of SDS–PAGE resolving gelsa

Component 7.5% T b 12% T c

Water 4.85 ml 3.35 ml
1.5 M Tris–Cl, pH 8.8 2.5 ml 2.5 ml
10% SDS 0.1 ml 0.1 ml
Acrylamide/bis (30% T, 2.7% C) 2.5 ml 4.0 ml
10% ammonium persulfated 50 ml 50 ml (0.05%)
TEMED 5 ml 5 ml (0.05%)
a
Any desired volume of monomer solution can be prepared by using multiples of the 10-ml recipes.
Combine the first four items and deaerate the solution under vacuum for 15 min. Start polymerization
by adding ammonium persulfate and TEMED.
b
For SDS-treated proteins in the approximate molecular weight range between 40 and 250 K.
c
For SDS-treated proteins in the approximate molecular weight range between 10 and 100 K.
d
To make 10% ammonium persulfate (APS), dissolve 100 mg APS in 1 ml of water. Make the APS
solution fresh daily.

polymerization, from the surface of the monomer mixture. Allow the

gel to polymerize for 45 min to 1 h. Polymerization is evidenced by the
appearance of a sharp interface beneath the overlay, which will start to
become visible in about 15 min. Polymerization is essentially complete
in about 90 min, but the stacking gel can be poured after about an hour
(Bio-Rad Lab., Bull. No. 1156). Allow unused monomer to polymerize
in the beaker and discard the gel.
4. Prepare 10 ml of stacking gel monomer solution (4% T, 2.7% C ), by
combining
Water 6.1 ml
0.5 M Tris–Cl, pH 6.8 2.5 ml
Acrylamide stock solution (30% T) 1.3 ml
10% SDS 0.1 ml
Deaerate the monomer solution under vacuum for at least 15 min.
5. Thoroughly rinse the top of the resolving gel with water and dry the area
above it with filter paper. Place a well-forming comb between the gel
plates and tilt it at a slight angle to provide a way for bubbles to escape.
6. Add 50 ml of 10% APS and 10 ml of TEMED to each 10 ml of degassed
monomer solution and pour the stacking gel solution on top of the
resolving gel. Align the comb in its proper position, being careful not to
trap bubbles under the teeth. Visible polymerization of the stacking gel
should occur in about 10 min. No overlay is required, because the
comb excludes oxygen from the surfaces of the wells. Allow the gel to
polymerize for 30–45 min. Allow unused monomer to polymerize in the
beaker before disposing of it.
One-Dimensional Gel Electrophoresis 505

In some situations, it may be necessary or convenient to let the gel stand

overnight before it is used. When this is the case, it is best to pour the stacking
gel on the day of the run to maintain the ion discontinuities at the interface
between the two gels. For storage, the top of the resolving gel should be
rinsed thoroughly and covered with resolving gel buffer (0.375 M Tris–Cl,
0.1% SDS, pH 8.8) to avoid dehydration and ion depletion. Also, the tops of
the gel sandwiches should be covered with plastic wrap during storage.

4.5. Sample preparation

The common biochemical buffers are usually tolerated in SDS–PAGE, so
that pretreatment of samples is not generally required. Distorted band
patterns, such as pinching or flaring of lanes, can be caused by excessive
amounts of salt in the samples. These distortions can often be remedied by
desalting the samples.
1. Prepare the volume of SDS-reducing buffer required for the number of
samples to be run by adding 50 ml of 2-mercaptoethanol to each 0.95 ml of
stock sample buffer (to a final concentration of 5% 2-mercaptoethanol).
This step may be omitted, if reduction of disulfide bonds is not desired.
2. Dilute samples with at least 4 vol of complete SDS-reducing buffer
(although as little as twofold dilution may be adequate for some samples).
Sample volumes are of the order of 20–50 ml for conventional gels and
5–30 ml for minicells, depending on the widths of the wells and the
thicknesses of the gels. Detection in gels requires on the order of 1 mg of
protein per band for easy visibility when staining with Coomassie Blue
R-250 or 0.1 mg of protein per band with silver staining (see below).
3. Heat the diluted samples at 95  C for 4 min by suspending the sample
tubes in hot water. Do not store prepared samples.

4.6. Electrophoresis
Assemble the electrophoresis cell, fill the upper and lower reservoirs with
electrode buffer, and remove the comb from the stacking gel. Load the
prepared samples into the wells in the stacking gel by layering them under
electrode buffer using a microliter syringe or micropipet. The glycerol in
the samples provides the necessary density for them to sink to the bottoms of
the wells and the Bromphenol Blue tracking dye enables the samples to be
seen during loading. Finally, attach leads to the unit and connect them to a
power supply. The lower electrode is the anode and the upper one is the
cathode, in SDS–PAGE.
During an electrophoresis run, electrical energy is converted to heat
which can cause band distortion and diffusion. In general, electrophoresis
should be carried out at power settings at which the run proceeds as rapidly
506 David E. Garfin

as the chamber’s ability to draw off heat will allow. In other words, the run
should be as fast as possible without exceeding desired resolution and
distortion limits.
Many of the power supplies which are available allow control of any
electrical quantity and the choice is almost a matter of preference. Constant
current conditions, as a rule, result in shorter but hotter runs than does
constant voltage (Allen et al., 1984). In the early stages of a run, the resistance
of the gel increases as the chloride ions migrate out of it. Accordingly,
voltage will rise or current will fall, depending on whether constant current
or constant voltage operation is in use.
Small-format minicells, with their thin glass plates, are better able to
efficiently dissipate the heat generated by the initially high currents at the
beginnings of runs than are standard-sized cells. Thus, the recommendation
is that gels should be run under constant current conditions (16–24 mA/mm
of gel thickness) in conventional apparatus and at constant voltage
(20–30 V/cm of gel length) in minicells. The use of recirculated coolant,
where possible, allows higher voltages and currents to be used for shortened
run times. Electrophoresis should be started immediately after the samples
are loaded and is generally continued until the Bromphenol Blue tracking
dye has reached the bottom of the gel.

4.7. Comments on method

The Laemmli SDS–PAGE system (Allen et al., 1984; Andrews, 1986;
Blackshear, this series; Hames, 1981; Laemmli, 1970) is an adaptation of
an earlier method devised by Ornstein (1964) and Davis (1964) for frac-
tionation of native serum proteins. The different (discontinuous) buffers
used in the stacking and resolving gels are required for the proper function-
ing of the Ornstein–Davis system (Ornstein, 1964; Jovin, 1973). However,
inclusion of SDS modifies the rationale of the Ornstein–Davis technique in
important ways, since the properties of the detergent dominate the system
(Allen et al., 1984; Chrambach, 1985; Wyckoff et al., 1977).
The necessary components of the Laemmli SDS–PAGE system are a
Tris–Cl gel buffer, the Tris–glycine–SDS electrode buffer, and the SDS-
reducing sample buffer. As a consequence of SDS in the system, it is actually
not necessary to cast the stacking gels at different pH or ionic strength than
the resolving gels. Similar resolution is obtained whether the stacking gel is
cast as above or in resolving gel buffer (0.375 M Tris–Cl, pH 8.8). This is
because the mobilities of SDS–polypeptide complexes are insensitive to pH
in this range (Allen et al., 1984). When many gels are being cast at one time
for storage and later use, it is convenient to cast the stacking and resolving
gels in the same buffer.
Total SDS load, on the other hand, has considerable influence on
resolution (Wyckoff et al., 1977). Inclusion of more than 200 mg of SDS
One-Dimensional Gel Electrophoresis 507

in 30–50 ml samples in the minigel configuration can lead to broadening and

spreading of protein bands. With dilute, large volume samples, it may prove
advantageous to limit the total SDS in the system by dropping the final SDS
concentration of the treated sample to about 0.5% and casting the gels
without SDS. Because the mobility of SDS is greater than those of proteins,
SDS from the electrode buffer quickly overtakes the proteins during
electrophoresis. The gel is thus supplied and continuously replenished
with SDS from the electrode buffer at a level sufficient to maintain the
saturation of the proteins (Chrambach, 1985).

4.8. Variations of method

The complete denaturation and dissociation of proteins with the Laemmli
SDS–PAGE system (Allen et al., 1984; Andrews, 1986; Blackshear, this series;
Hames, 1981; Laemmli, 1970) are not always desirable. For some analyses, it
might be of interest to estimate the molecular weights of particular proteins in
their intact, oligomeric forms. In other experiments, interest might center on
the biological activities of proteins in their native, nondenatured states.
Through selective use of the two denaturants, 2-mercaptoethanol and SDS,
conditions can be adjusted as needed to separate proteins in the completely
denatured, partially denatured, or native states.
Covalent associations between protein units can be maintained by
omitting 2-mercaptoethanol from the sample buffer. In the absence of the
reducing agent, the intra- and interchain disulfide bonds of sample proteins
remain intact. The electrophoretic mobilities of the resultant SDS–protein
complexes are correspondingly altered relative to those obtained under
dissociating conditions. During electrophoresis, the mobilities of oligomeric
SDS–proteins are lower than those of their fully denatured SSDS-polypeptide
components. Further, the electrophoretic behaviors of single-chain poly-
peptides can also be affected by reduction. The intrachain disulfide bridges
of single-chain proteins can hold them in compact configurations that are
more or less retained in the presence of SDS. Thus, some SDS-proteins
migrate faster electrophoretically in the absence of 2-mercaptoethanol than
when in the extended structures brought on by reduction. Proteins often
show characteristic, individual responses to reduction, so that comparisons
of SDS–PAGE gels run with and without 2-mercaptoethanol can be very
informative (Marshall, 1984).
To separate proteins without reduction, carry out the SDS–PAGE
procedure described above, omitting 2-mercaptoethanol from the sample
buffer. Note that oligomeric SDS–protein complexes migrate more slowly
than their SDS-polypeptide subunits. It may, therefore, be necessary to use
lower concentration (%T ) gels than with the fully denaturing method to get
oligomers to move adequate distances into the matrices. In addition, non-
reduced proteins may not be completed saturated with SDS and, hence,
508 David E. Garfin

may not bind the detergent in a constant weight ratio. This makes molecular
weight determinations of these molecules by SDS–PAGE less straightfor-
ward than analyses of fully denatured polypeptides, since it is necessary that
both standards and unknown proteins be in similar configurations for valid
comparisons.
When both SDS and 2-mercaptoethanol are left out of the Laemmli
procedure, what remains is the classical Ornstein–Davis PAGE system
(Davis, 1964; Ornstein, 1964) for native proteins. This is a high-resolution
native PAGE method designed for separation of the full spectrum of serum
proteins. Because the system was meant to separate a wide variety of
proteins, resolution may not be optimal for some restricted ranges of protein
mobilities. Although there are a number of high-resolution native PAGE
systems available to meet differing requirements (Allen et al., 1984;
Andrews, 1986; Blackshear, this series; Chrambach, 1985; Hames, 1981),
the Ornstein–Davis method should perform adequately for the fractionation
of the majority of commonly encountered protein mixtures. Molecular
weights are more difficult to determine by native PAGE than by SDS–
PAGE, since a single native system cannot distinguish the effects of charge
and conformation on protein electrophoretic mobilities (Allen et al., 1984;
Andrews, 1986; Chrambach, 1985; Hames, 1981).
The procedure described here is readily modified for native PAGE.
Merely omit 2-mercaptoethanol from the sample buffer and replace the
10% SDS in the recipes for the gel, sample, and electrode buffers with
equivalent volumes of water. Follow the procedure as otherwise presented,
except for sample treatment. Samples should be diluted in nondenaturing
buffer (0.06 M Tris–Cl, pH 6.8, 10% glycerol, 0.025% Bromphenol Blue)
following the same guidelines as for denaturing gels, but they should not be
heated.

5. Detection of Proteins in Gels

Three of the simplest and most reliable methods for the detection of
proteins in SDS–PAGE gels are presented. They should be adequate to
cover the requirements of most situations. Coomassie Brilliant Blue R-250
is the most common protein stain and is recommended for routine work.
Silver staining is the most sensitive method for staining proteins in gels and
should be employed when electrophoresis is used to assess the purity of a
preparation; for example, an antigen preparation. Copper staining is a recent
development allowing rapid and sensitive staining. Discussions of other
detection methods, including radiolabeling and means for quantitating
proteins in gels, can be found by Dunbar (1987), Andrews (1986), Allen
et al. (1984), Hames (1981), and Merril (this volume).
One-Dimensional Gel Electrophoresis 509

After electrophoresis, remove the gel assembly and separate the glass
plates. The gel will probably stick to one of the two plates. Remove the
spacers and cut off and discard the stacking gel. Place the glass plate holding
the gel into fixative or staining solution and float the gel off of the plate.
All of the steps in gel staining are done at room temperature with gentle
agitation (e.g., on an orbital shaker platform) in any convenient container,
such as a glass casserole or a photography tray. Always wear gloves when
staining gels, since fingerprints will stain. Permanent records of stained gels
can be obtained by photographing them or by drying them on filter paper
using commercially available drying apparatus.

5.1. Dye staining with Coomassie Brilliant Blue R-250

This is the standard method of protein detection (Allen et al., 1984;
Andrews, 1986; Hames, 1981; Wilson, this series). Easy visibility requires
on the order of 0.1–1 mg of protein per band.
1. Prepare the staining solution: 0.1% Coomassie Brilliant Blue R-250 (w/v)
in 40% methanol (v/v), 10% acetic acid (v/v). Filter the staining solution
after the dye has dissolved. The staining solution is reusable. Store it at
room temperature.
2. Soak the gel in an excess of staining solution for 30 min.
3. Destain with a large excess of 40% methanol, 10% acetic acid. Change
the destaining solution several times, until the background has been
satisfactorily removed.
The acid–alcohol solutions used in this procedure do not completely
fix proteins in the gel. This can lead to losses of some low-molecular-
weight proteins during the staining and destaining of thin gels. Permanent
fixation is obtainable by incubating the gel in 40% methanol (v/v), 10%
trichloroacetic acid (w/v) for 1 h before it is immersed in the staining
solution.

5.2. Silver staining

This method, developed by Merril and coworkers, can be as much as
100 times more sensitive than dye staining (Allen et al., 1984; Merril et al.,
1981, this series). Bands containing 10–100 ng of protein can be easily seen.
The reagents are available in kit form from Bio-Rad Laboratories.
Reaction times vary with the thicknesses of the gels.
1. Fix the proteins in the gel in about 400 ml of 40% methanol, 10% acetic
acid (v/v) (or 40% methanol, 10% trichloroacetic acid) for 30 min to
overnight.
2. Fix twice in 400 ml 10% ethanol, 5% acetic acid (v/v) for 15–30 min.
510 David E. Garfin

3. Soak the gel for 3–10 min in 200 ml of fresh oxidizer solution
(0.0034 M potassium dichromate, 0.0032 N nitric acid).
4. Wash the gel three or four times for 5–10 min in 400 ml water, until the
yellow color has been washed out.
5. Soak the gel in 200 ml fresh silver reagent (0.012 M silver nitrate) for
15–30 min.
6. Wash the gel with 400 ml water for 1–2 min.
7. Wash the gel for about 1 min in developer (0.28 M sodium carbonate,
1.85% paraformaldehyde).
8. Replace the developer with fresh solution and incubate for 5 min.
9. Replace the developer a second time and allow development to
continue until satisfactory staining has been obtained.
10. Stop development with 5% acetic acid (v/v).
Vertical streaks and sample-independent bands in the 50–70-kDa region
are sometimes seen in silver-stained gels. These artifacts have been attributed
to reduction of contaminants inadvertently introduced into the samples
(Ochs, 1983). They can be eliminated by adding excess iodoacetamide to
sample solutions after treatment with SDS-reducing buffer (Görg et al., 1987).

5.3. Copper staining

Rapid, single-step staining of SDS–PAGE gels is achieved by incubating
gels in copper chloride (Lee et al., 1987). The resultant, negatively stained
image of the electrophoretogram is intermediate in sensitivity between
Coomassie blue and silver staining.
1. Wash the gel briefly in water.
2. Soak the gel in 0.3 M CuCl2 for 5 min.
3. Wash the gel for 2–3 min in water.
The method yields negatively stained gels showing clear protein bands
on an opaque, blue–green background. The protein bands can be easily seen
and photographed with the gel on a black surface. Proteins are not perma-
nently fixed by this method and can be quantitatively eluted after chelating
the copper (Lee et al., 1987). The electrophoretic pattern is lost when
copper-stained gels are dried so they must be photographed, restained
with Coomassie Blue, or stored in water.

6. Marker Proteins
Mixtures of marker proteins are available for calibrating gels. PAGE
standards are mixtures of proteins with precisely known molecular weights
blended for uniform staining. They are obtainable in various molecular
One-Dimensional Gel Electrophoresis 511

weight ranges. Concentrated stock solutions of the standards are diluted in

sample buffer just prior to electrophoresis and treated in the same manner as
the sample proteins. These proteins are suitable as reference markers for
molecular weight determinations.
Prestained SDS–PAGE standards have recently become available. The
coupling of dye molecules to the marker proteins changes their molecular
weights significantly and unpredictably and they should not be used for
molecular weight determinations. However, prestained standards are very
useful for following the course of an electrophoretic run and are valuable for
assessing the efficiencies of protein transfers when gels are blotted.

7. Molecular Weight Determination

Molecular weights of proteins are determined by comparison of their
mobilities with those of several marker proteins of known molecular weight
(Allen et al., 1984; Andrews, 1986; Blackshear, this series; Chrambach,
1985; Hames, 1981). After the gel has been run, but before it has been
stained, mark the position of the Bromphenol Blue tracking dye to identify
the leading edge of the electrophoretic ion front. This can be done by
cutting notches in the edges of the gel or by inserting a needle soaked in
India ink into the gel at the dye front. After staining, measure the migration
distances of each protein (markers and unknowns) from the top of the
resolving gel. Divide the migration distance of each protein by the distance
traveled by the tracking dye. The normalized migration distances so
obtained are called the relative mobilities of the proteins (relative to the
dye front) and conventionally denoted as Rf. Construct a (semilogarithmic)
plot of the logarithms of the molecular weights of the protein standards as
functions of the Rf values. Note that the graphs are slightly sigmoid. As long
as the extremities of a molecular weight range are avoided, unknown
molecular weights can be estimated by linear regression analysis or interpo-
lation from the curves of log Mr versus Rf. Keep in mind that the molecular
weights obtained using SDS–PAGE are those of the polypeptide subunits
and not those of native, oligomeric proteins.

8. Preparative Electrophoresis
The most satisfactory way to recover proteins separated by SDS–
PAGE for further study is to extract them from bands excised from the
gels. Many attempts have been made to design continuous elution devices
suitable for routine protein purification, in which bands emerging from the
bottoms of electrophoresis gels are swept away to fraction collectors
512 David E. Garfin

(Andrews, 1986; Chrambach, 1985; Chrambach and Nguyen, 1979). The

scarcity of preparative gel devices is evidence of the disappointing lack of
success in developing generally useful instruments. Preparative gel electro-
phoresis would ideally be capable of yielding high-milligram to gram
quantities of individual proteins recovered cleanly with the resolution
anticipated from the corresponding analytical gels. In general, though,
band distortion and poor elution have limited the resolution attainable
with most apparatus so that they have only worked well with relatively
simple protein mixtures. The difficulties in scaling gel electrophoresis up to
preparative levels has tended to result in devices which are rather cumber-
some and which require much technical skill for best results. As a conse-
quence, proteins are usually obtained by extraction from analytical type gels
(Harrington, this volume).
Gels to be run for the isolation of proteins (Andrews, 1986; Chrambach,
1985) can be cast using special preparative combs. These combs form wide
sample wells spanning the widths of the gels and usually provide a separate,
narrow reference well for marker proteins. The maximum amount of sample
which can be loaded on a gel ultimately depends on how well the proteins of
interest are separated from their neighbors in the sample mixture. Since bands
become wider as the amount of material increases, as sample load is raised, the
corresponding loss of resolution will eventually become unacceptable. Pro-
tein loads 10- to 50-fold greater per unit of cross-sectional area than are
usually run in analytical gels are easily tolerated. Thus, with some large slab
gels, proteins can be recovered in tens-of-milligram amounts.
Copper staining (Lee et al., 1987) is advisable for the visualization of the
bands in preparative SDS–PAGE, since this method does not employ
fixative solvents. Desired bands are cut from the gel and destained by
incubation in three changes (for 10 min each) of 0.25 M EDTA, 0.25 M
Tris–Cl, pH 9. After destaining, gel slices are incubated in the appropriate
elution buffer.
Proteins are often extracted from macerated gel slices by simple diffusion
into appropriate buffers or by solubilization of the gel (Andrews, 1986;
Harrington, this volume). In the latter method, cross-linkers other
than bisacrylamide are copolymerized into the gels (Allen et al., 1984;
Andrews, 1986). For example, gels cross-linked with N,N 0 -bisacrylylcysta-
mine (BAC) are dissolvable in 2-mercaptoethanol or dithiothreitol, while
both N,N 0 -dihydroxyethylenebisacrylamide (DHEBA) and N,N 0 -diallyltar-
tardiamide (DATD) result in gels which can be solubilized with periodic acid.
Once gels have been dissolved, proteins must be separated from the large
excess of gel material by gel filtration or ion-exchange chromatography.
Electrophoretic elution is an efficient method for recovering proteins
from gel slices (Andrews, 1986; Chrambach, 1985; Dunbar, 1987). In the
simplest versions of this method, proteins are electrophoresed out of gel
pieces into dialysis sacks in the types of apparatus used for running
One-Dimensional Gel Electrophoresis 513

cylindrical gel rods. Devices are available for the rapid recovery of proteins
in small volumes with yields of greater than 70% in most cases. Elution takes
about 3 h at 10 mA/tube in 0.025 M Tris, 0.192 M glycine, 0.1% SDS, pH 8.3
(standard SDS–PAGE electrode buffer). SDS can be removed from the
eluted samples by dialysis or ion-exchange chromatography (Furth, 1980).

REFERENCES
Allen, R. C., Saravis, C. A., and Maurer, H. R. (1984). Gel Electrophoresis and Isoelectric
Focusing of Proteins: Selected Techniques. de Gruyter, Berlin.
Andrews, A. T. (1986). Electrophoresis: Theory, Techniques, and Biochemical and Clinical Appli-
cations. 2nd edn. Oxford University Press, New York.
Bier, M., Palusinski, O. A., Mosher, R. A., and Saville, D. A. (1983). Science 219, 1281.
Blackshear, P. J. (this series). Vol. 104, p. 237.
Bury, A. F. (1981). J. Chromatogr. 213, 491.
Chrambach, A. (1985). The Practice of Quantitative Gel Electrophoresis. Weinheim, VCH.
Chrambach, A., and Jovin, T. M. (1983). Electrophoresis 4, 190.
Chrambach, A., and Nguyen, N. Y. (1979). In ‘‘Electrokinetic Separation Methods’’,
(P. G. Righetti, C. J. Van Oss, and J. W. Vanderhoff, eds.), p. 337. Elsevier, Amsterdam.
Chrambach, A., and Rodbard, D. (1971). Science 172, 440.
Davis, B. J. (1964). Ann. NY Acad. Sci. 121, 404.
Dunbar, B. S. (1987). Two-Dimensional Electrophoresis and Immunological Techniques. Plenum,
New York.
Dunbar, B. S., Kimura, H., and Timmons, T. M. (this volume). Chapter 34.
Furth, A. J. (1980). Anal. Biochem. 109, 207.
Görg, A., Postel, W., Weser, J., Günther, S., Strahler, J. R., Hanash, S. M., and Somerlot, L.
(1987). Electrophoresis 8, 122.
Hames, B. D. (1981). In ‘‘Gel Electrophoresis of Proteins: A Practical Approach’’,
(B. D. Hames and D. Rickwood, eds.), p. 1. IRL Press, Oxford.
Harrington, M. (this volume). Chapter 37.
Jovin, T. M. (1973). Biochemistry 12, 871, 879, 890.
Laemmli, U. K. (1970). Nature (London) 227, 680.
Lee, C., Levin, A., and Branton, D. (1987). Anal. Biochem. 166, 308.
Margulies, M. M., and Tiffany, H. L. (1984). Anal. Biochem. 136, 309.
Marshall, T. (1984). Clin. Chem. 30, 475.
Merril, C. R. (this volume). Chapter 36.
Merril, C. R., Goldman, D., Sedman, S. A., and Ebert, M. H. (1981). Science 211, 1437.
Merril, C. R., Goldman, D., and Van Keuren, M. L. (this series). Vol. 104, p. 441.
Neville, D. M. Jr. (1971). J. Biol. Chem. 246, 6328.
Neville, D. M., and Glossmann, H. (this series). Vol. 32, p. 92.
Nielsen, T. B., and Reynolds, J. A. (this series). Vol. 48, p. 3.
Ochs, D. (1983). Anal. Biochem. 135, 470.
Ornstein, L. (1964). Ann. NY Acad. Sci. 121, 321.
Rodbard, D., and Chrambach, A. (1970). Proc. Natl. Acad. Sci. USA 65, 970.
Shapiro, A. L., Vinuela, E., and Maizel, J. V. Jr. (1967). Biochem. Biophys. Res. Commun. 28,
815.
Timmons, T. M., and Dunbar, B. S. (this volume). Chapter 51.
Weber, K., and Osborn, M. (1969). J. Biol. Chem. 244, 4406.
Wilson, C. M. (this series). Vol. 91, p. 236.
Wyckoff, M., Rodbard, D., and Chrambach, A. (1977). Anal. Biochem. 78, 459.
 C H A P T E R T H I R T Y

Isoelectric Focusing and

Two-Dimensional Gel Electrophoresis
David B. Friedman,* Sjouke Hoving,† and Reiner Westermeier‡

Contents
1. Introduction 516
1.1. Isoelectric focusing—Basic principle 517
1.2. Types of pH gradients for isoelectric focusing 518
1.3. SDS polyacrylamide electrophoresis 521
1.4. Enhancement of resolution for alkaline pH gradients 523
1.5. Difference gel electrophoresis 524
2. Materials 527
2.1. Equipment 527
2.2. Solutions and reagents 527
3. Methods 528
3.1. Protein sample preparation 528
3.2. Sample cleanup and precipitation 530
3.3. Isoelectric focusing using acidic range IPG gels 531
3.4. Isoelectric focusing using alkaline range IPG gels 534
3.5. Equilibration of IPG gels 535
3.6. SDS–PAGE: The second dimension 536
References 538

Abstract
By far the highest resolution of all separation techniques for intact proteins in a
single analytical run continues to be by the combination of isoelectric focusing
(IEF) and SDS–PAGE, originally introduced by O’Farrell [(1975). J. Biol. Chem. 250,
4007–4021]. This analytical platform has seen a number of significant advances
and applications over the past 25 years, including reproducibility using immobi-
lized pH gradient (IPG) strips [Bjellqvist et al. (1982). J. Biochem. Biophys.
Methods 6, 317–339.], resolution in alkaline IEF using hydroxyethyldisulfide
(HED) [Olsson et al. (2002). Proteomics 2, 1630–1632], and quantification
for differential expression proteomics on intact proteins on a global scale

* Proteomics Laboratory, Mass Spectrometry Research Center, Vanderbilt University, Nashville,

Tennessee, USA
{
Novartis Institutes of BioMedical Research, Basel, Switzerland
{
Gelcompany GmbH, Paul-Ehrlich-Strasse, Tübingen, Germany

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63030-5 All rights reserved.

515
516 David B. Friedman et al.

[DIGE; Unlu et al. (1997). Electrophoresis 18, 2071–2077]. These major improve-
ments will be highlighted in this chapter alongside the principle and theory of 2D
gel electrophoresis, as well as detailed methods for general 2D gel electropho-
resis best use protocols.

1. Introduction
For complex mixtures such as whole cell lysates or enriched subcellu-
lar fractions, two-dimensional gel electrophoresis (2D gel) can typically
resolve hundreds-to-thousands of individual protein species using two
orthogonal separations1 (Fig. 30.1). The first separation is typically based
on charge using denaturing IEF, and the second separation by apparent
molecular mass using denaturing sodium dodecyl sulfate–polyacrylamide
gel electrophoresis (SDS–PAGE).2 2D gel experiments are particularly
powerful for visualizing protein isoforms that result from charged posttrans-
lational modification, such as phosphorylation and sulfation (which add
charge), or acetylation (which neutralize charge), among others. They are
also useful in detecting splice variants and proteolytic cleavages that result in
protein species with altered molecular weight (MW) and isoelectric point
(pI). More recently, 2D gels have been used extensively in differential
expression proteomics experiments, especially since the commercialization
of the difference gel electrophoresis (DIGE) technology in early 2000
(Friedman and Lilley, 2009; Lilley and Friedman, 2004; Unlu et al., 1997).
In modern 2D gel-based proteomics experiments, intact protein resolu-
tion by 2D gel electrophoresis is commonly coupled with protein (‘‘spot’’)
excision, in-gel digestion and mass spectrometry to provide for protein
identification using sophisticated database searching algorithms. As with
all high-end proteomics technologies, this method also requires many

1
Despite this resolution, proteins of extreme MW and pI (10 > kDa < 200; 3 > pI < 11), very hydropho-
bic proteins (e.g., integral membrane proteins with multiple trans-membrane domains), and low-abundance
proteins are typically difficult to resolve or detect on 2D gels. It is in these cases where complementary data
can be obtained from an orthogonal technology, such as liquid chromatography coupled with tandem mass
spectrometry (LC–MS/MS), that is capable of detecting many of these proteins. The added sensitivity of the
LC–MS/MS approach arises from the fact that proteins are detected based on surrogate peptides from an en
masse proteolytic digest that are resolved by liquid chromatography. But, for this very reason, the LC–MS/
MS technique is greatly challenged to identify the multiple protein isoforms that arise from charged or bulky
post-translational modifications that are readily detectable and quantifiable on a 2D gel (especially in a
nontargeted, discovery experiment). Furthermore, the quantification of intact protein expression using the
2D gel platform can now be done using the fluorescent multiplexing technology of DIGE, whereby the
requisite number of individual biological replicates from multiple experimental conditions can be easily
analyzed to provide for statistically powered discovery proteomics.
2
Although denaturing IEF is most commonly used for the highest resolution first-dimensional separations (first
described by O’Farrell, 1975), other techniques such as blue native PAGE (Schagger and von Jagow, 1991)
for analysis of protein complexes, and acidic PAGE in presence of a cationic detergent (Hartinger et al., 1996)
for separating hydrophobic membrane proteins, are often utilized for more specific investigations. However,
SDS–PAGE is almost always being used for the second-dimensional separation.
Isoelectric Focusing and Two-Dimensional Gel Electrophoresis 517

Figure 30.1 High resolution 2D gel (24 cm pH 4–7). 400 mg cellular proteins from a HeLa
cell extract were separated in two orthogonal dimensions, the first based on charge by
isoelectric focusing, and the second based on apparent molecular mass by sodium dodecyl
sulfate polyacrylamide gel electrophoresis. The gel was stained with the fluorescent dye
Sypro Ruby. Typically hundreds-to-thousands of individual protein species, including post-
translationally modified isoforms and proteolytic products, can be resolved and quantified.

technically challenging manual steps in addition to specialized instrumenta-

tion. Often the success of a 2D electrophoresis experiment is dependent on
the skill of the operator. By introducing stringent analysis conditions and
developing the first multiple gel systems, Anderson and Anderson (1978a,b)
made the technique markedly more reliable and reproducible.

1.1. Isoelectric focusing—Basic principle

Isoelectric focusing (IEF) is an electrophoretic separation method which
separates amphoteric molecules such as proteins and peptides according to
their charge as defined by the pKa values of proton-accepting sites within a
molecule. For proteins and peptides, these sites can be found in the free
amines and carboxylic acids located at the N- and C-termini as well as on
the side chains of arginine, lysine, histidine, aspartic acid, and glutamic acid
residues. The isoelectric point is a specific physicochemical parameter of an
amphoteric molecule defined as the pH that a molecule is in a net neutral
charge state. IEF can provide very high resolving power, including the
separation of protein posttranslational modifications that alter charge (e.g.,
phosphorylation, acetylation/deactylation). Although the method is applied
for several types of amphoteric compounds, this chapter will describe the
methodology for protein separations.
The sample has to be prepared and conditioned with a high chaotrope
concentration, a zwitterionic detergent, a reducing thiol, and carrier
518 David B. Friedman et al.

ampholytes to avoid the formation of aggregates and complexes between

proteins. Removal of nonprotein-based ions from the sample is also essential
for high-resolution IEF, since resistance must be kept high to enable a high
electric field (8000–10,000 V) while keeping the current at a minimum.
Protein precipitation methods are often used to reduce such contaminants
(see Section 3.2). Ideally, the only ions in a sample for IEF will be the
proteins themselves. The sample is then applied to the IEF gel, which
contains the same additives as the sample, and separated in the electric
field. After the proteins have reached their isoelectric points, the gels are
incubated in SDS buffer and applied on SDS–PAGE slab gels for the
separation according to apparent molecular mass (influenced primarily by
molecular weight but also by hydrophobicity and to a lesser extent by
protein shape).
Modern IEF methods for 2D gel electrophoresis use a thin polyacryl-
amide gel as a molecular sieve that contains an immobilized pH gradient
(IPG). The gels have a low concentration of acrylamide (usually 4–5% total
acrylamide), because the matrix should not be restrictive to high-molecular-
weight proteins. Proteins are introduced into this medium and then an
electric field is applied. Proteins that are located in a position in the pH
gradient that is lower (more acidic) than their pI will be positively charged
and migrate toward the cathode in an electric field. Conversely, proteins
located in a position more basic than their pI will migrate toward the anode.
Since the pI is defined as the pH at which a protein is net neutral, the electric
field has no effect on proteins when they are at the point in the pH gradient
which is equal to the pI. Thus, the proteins stop migrating and are
‘‘focused’’ at the pH equal to their pI. The method has an in-built focusing
effect: when a protein diffuses away from its pI, it will gain a charge and the
electric field, therefore, forces it to migrate back to the isoelectric point.
In the end of an IEF run, the proteins become highly concentrated at
their isoelectric points, resulting in a high sensitivity for detection. Even
small charge differences can be differentiated, and the resolution can be
increased by using longer separation distances and by employing narrow
gradient intervals (Hoving et al., 2000). If desired, the isoelectric points of
the proteins can be estimated with a calibration curve using marker proteins.

1.2. Types of pH gradients for isoelectric focusing

1.2.1. Carrier ampholytes
The concept of separating proteins in a pH gradient built by a mixture of
amphoteric buffers in free solution had been developed by Vesterberg and
Svensson (1966) before it had been converted into a real method by
Vesterberg a few years later (Vesterberg, 1972). Because naturally occurring
amino acids and peptides have very low buffering power at their own
isoelectric points, the buffers needed to be chemically synthesized.
Isoelectric Focusing and Two-Dimensional Gel Electrophoresis 519

The first reagents on the market for this purpose, AmpholinesÒ , were
mixtures of aliphatic oligo-amino/oligo-carbonic acids, of 600–700 differ-
ent homologues exhibiting a spectrum of isoelectric points between pH 3
and 10. These compounds have high buffering capacities at their isoelectric
points and form a pH gradient under the influence of the electric field.
Optimally, they have molecular weights below 1 kDa, and do not bind to
proteins, because they are highly hydrophilic. Mixtures with narrow inter-
vals are available for higher resolution using defined isoelectric point ranges.
More recently came the development of carrier ampholytes, which are
synthesized from different reagents than the AmpholinesÒ and are available
from a variety of commercial vendors. Although the function of forming a
pH gradient under the influence of the electric field is the same for all of
these products, the profiles of the pH gradients, the distribution of buffering
power, and the number of homologues are different (Righetti et al., 2007),
which leads to differences in the focusing pattern.
Carrier ampholytes-based IEF is mostly performed in polyacrylamide
gels, and the original 2D gel electrophoresis procedure by O’Farrell (1975)
accomplishes IEF in carrier ampholyte gradients in thin gel rods (‘‘tube gels’’
for the first dimension). However, these long and soft gels are not easy to
handle, and the pH gradients become unstable with increased running time,
resulting in cathodal drift. Although some laboratories still use the carrier
ampholyte technique, and this method has mostly been supplanted by the
introduction of IPG strips, which are described next.

1.2.2. Immobilized pH gradients

A major improvement in 2D gel electrophoresis methodology came from
the introduction of an IPG within a polyacrylamide matrix for IEF
(Bjellqvist et al., 1982). This major technological development overcame
several shortcomings of the carrier ampholytes system such as gradient drift
(especially in the cathodal region), mechanical instability, and technical
variation between runs and between laboratories. The IPG strip technology
has greatly facilitated the methodology and highly increased the reproduc-
ibility within a laboratory and across different working groups (Gorg et al.,
2000, 2004), and is now considered to be the method of choice for IEF and
is commercially available through several vendors.
For modern IPG strip technology, the pH gradient is formed by acidic
and alkaline buffering groups which are copolymerized with the polyacryl-
amide matrix during preparing the gel. Less than 10 different acrylamide
derivatives with acidic and basic pKa values are sufficient to create any
desired pH gradient. Two monomer solutions containing precalculated
mixtures of acrylamide derivatives are prepared: to the solution for the
acidic end of the gradient, a portion of glycerol is added to stabilize the
gradient during pouring. These pH gradient gels are cast on a covalently
bound film backing. After the gel has formed, polymerization catalysts and
520 David B. Friedman et al.

nonreacted compounds are washed out from the matrix with distilled
water to produce the very low electrical conductivity necessary for IEF
(Westermeier, 2005).
The IPG gels are then dried for long-term storage, and can be rehydrated as
necessary shortly before use. The major benefit of IPGs is the absence of the
cathodal drift: the gradient is fixed to the matrix. The major application of
IPGs is IEF under denaturing conditions, serving as the first dimension of
high-resolution two-dimensional electrophoresis. Carrier ampholytes are still
used, and for improved conductivity and protein solubility they are added to
the sample as well as to the rehydration solution of the gels with IPGs.
Precast IPG strips are now commercially available from a number of
vendors with a wide variety of strip lengths and pH gradients to enable the
optimal separation and display of selected protein groups within a protein
lysate. In particular, pH gradients are available from a very wide range (e.g., pH
3–11), to medium ranges (e.g., pH 4–7, pH 7–11), to narrow ranges (e.g., pH
4–5, pH 5.5–6.5). Strip lengths also vary from as short as 7 cm to as long as
24 cm. Strips typically are 3 mm in width, with an average thickness (after
reswelling) of approximately 0.5 mm (Fig. 30.2). Care should be taken in
selection of the optimal pH range and strip length for the desired experimental
goals (Hoving et al., 2000). For example, 7-cm pH 3–11 IPG strips may seem
best to provide the greatest range in a small gel format, but they provide overall
the lowest resolution and sensitivity for proteome-wide, discovery-based
experiments.

Figure 30.2 Immobilized pH gradient (IPG) strip, shown in the manual process of
preparing for rehydration loading. Sample is first dispensed into a reswelling try, and then
the IPG strip is carefully overlayed on the protein sample solution (face-down, with
plastic backing up) using a forceps to hold one end of the IPG (the plastic backing exceeds
the length of the IPG gel). The strips are then overlaid with 2–3 mL paraffin oil to
prevent urea crystallization during rehydration. IPG strips are provided from commercial
vendors with barcodes to aide sample-tracking.
Isoelectric Focusing and Two-Dimensional Gel Electrophoresis 521

A Dry IPG strip

Sample mixed with + Platinum electrode contacts −

rehydration solution

B
Filter paper pad
soaked with water + IPG strip pre-rehydrated −
with sample/rehydration solution

C
Filter paper pad Sample
soaked with water + −
Pre-rehydrated IPG strip

D
Paper bridge
soaked with sample
+ −
Pre-rehydrated IPG strip

Figure 30.3 Different configurations for sample loading into immobilized pH gradient
(IPG) strips for isoelectric focusing. (A) The IPG gel is placed face-down for sample loading
via active rehydration loading, after which isoelectric focusing is also possible. (B) The
sample is rehydrated into the IPG gel, which is then placed “sunny-side up” for isoelectric
focusing. (C) The sample is introduced into a previously rehydrated IPG gel via cup loading
at the anode. This configuration is especially beneficial for isoelectric focusing over alkaline
pH gradients. (D) Modification of cup loading, where the sample is introduced at the anode
via a paper-bridge. Reproduced with permission, Westermeier, Naven and Heopker
(2008), Proteomics in Practice, 2nd edition, Wiley-VCH Verlag GmbH & Co. KGaA.

The IPG strips offer several modes of sample application. Rehydration

loading and cup loading are the two major methods used, with modification
for active rehydration and paper-bridge loading (Fig. 30.3). All of these
methods are described in greater detail in Sections 3.3 and 3.4. The choice
of the sample application method depends on the type of pH gradient and
the sample. Because the IPG strips have a flat surface, they can also be
applied on horizontal flatbed SDS gels if desired.

1.3. SDS polyacrylamide electrophoresis

SDS–PAGE, described by Laemmli (1970), has long been the method of
choice for resolving intact proteins for a variety of biochemical analyses.
Mainly because SDS is the best solubilizing detergent also for very
522 David B. Friedman et al.

hydrophobic proteins, all proteins, also very basic ones, migrate into the
same direction, and a separation according to the apparent molecular mass
(commonly referred to as molecular weight) is obtained.

1.3.1. Separation principle

When SDS is added to a protein solution in excess, the proteins form
anionic micelles with a constant negative net charge per mass unit. Tertiary
and secondary structures are disrupted and the polypeptides become
unfolded. To achieve a complete unfolding, the disulfide bonds between
cysteines have to be cleaved with a reducing agent like 2-mercaptoethanol
or dithiothreitol. According to Ibel et al. (1990), these complexes look like
necklaces with partly covered and partly open polypeptide surfaces. During
electrophoresis all SDS–protein micelles migrate toward the anode, and
their electrophoretic mobilities are dependent on mostly molecular weight
but can also be influenced by protein hydrophobicity. When the migration
distances of proteins are plotted against the logarithm of the apparent
molecular mass, a sigmoidal curve results with a wide linear range. With
the help of comigrated molecular weight standard proteins, the molecular
weights of the polypeptides can be approximated.

1.3.2. Gel types

In principle, SDS–PAGE can be run in various basic buffer systems; how-
ever, the discontinuous buffer system according to Laemmli (1970) is the
most commonly used system, employing Tris–chloride pH 8.8 in the gel
and SDS–Tris–glycine in the running buffer. For one-dimensional separa-
tions, a stacking gel with pH 6.8 and lower gel concentration is applied to
provide a slow sample entry without protein aggregation, and to provide a
band sharpening effect at the start of the separation. However, a stacking gel
typically is not included for 2D gel electrophoresis, since the proteins are
preseparated and effectively stacked in the 3 mm width of the IPG strip (see
Section 3.6).
The standard matrix is a homogeneous gel layer with 12.5% T (total
acrylamide) and 2.6% C (bis-acrylamide for cross-linking) in the resolving
gel. But some applications require optimal resolution for a certain size
range, such that lower %T is required for high-molecular-weight proteins,
and a higher %T provides a better resolution for low-molecular-weight
proteins, so the concentration must be increased. Gradient SDS–PAGE gels
can provide a greater dynamic range of resolution, but are more challenging
to pour consistently and reproducibly, which is vital for quantitative pro-
teomic studies using 2D gels. Porosity gradient gels exhibit a wider linear
range of the molecular weights and improve the resolution for difficult
proteins, like glycoproteins.
SDS electrophoresis can be run in vertical setups in cassettes between
glass plates or on horizontal flatbed systems. The gel thicknesses vary
Isoelectric Focusing and Two-Dimensional Gel Electrophoresis 523

between 1.5 mm and 0.5 mm. Thicker gels have a better mechanical
stability, but are more difficult to stain and provide additional sample loss
and background noise for subsequent mass spectrometry. Thinner gels can
be cooled more efficiently, thus they can be run faster to obtain sharper
resolution.
Of the improvements made to the 2D gel process over the past few
decades, the second-dimension SDS–PAGE separations have remained
largely unchanged. Vertical electrophoresis systems are more commonly
used, and equipment to run individual or 12 or more gels simultaneously is
available. Horizontal flatbed configurations have also been available histori-
cally and are now beginning to see a resurgence due to the low consump-
tion of reagents (e.g., SDS running buffer), sample manipulation, as well as
the potential for increased resolution afforded by short run times. Some
improvements include using flexible plastic backing (including low-fluores-
cent media for DIGE and other fluorescence-based techniques), specialized
equipment for multicasting reproducible acrylamide gradients, and different
acrylamide formulations for increased shelf life and gel durability.

1.4. Enhancement of resolution for alkaline pH gradients

High-resolution IEF for alkaline pH ranges has always been more challeng-
ing than for acidic pH gradients, both due to issues of cathodic drift and also
due to the loss of reductant to keep some proteins soluble (DTT is a weak
acid that becomes charged at pH > 8 and, therefore, leaves the alkaline
region of the gradient). The first issue of cathodic drift was addressed by the
introduction of nonequilibrium pH gradient electrophoresis (NEpHGE;
O’Farrell et al., 1977). But the problem of reductant loss at pH > 8 still
remained, and leads to nonspecific reactions with urea, backfolding and
aggregation of polypeptides because the cysteines are no longer protected.
This results in protein spot artifacts and horizontal streaking of proteins on
the basic range of 2D gels.
The replacement of thiol reagents by phosphines like TBP or TCEP can
partly prevent these effects, but these reagents cause additional artifacts due
to their instability in the electric field. Prealkylation of proteins prior to IEF
with iodoacetamide, vinylpyridine, or monomeric acrylamide also leads to
additional artifacts due to incomplete alkylation of polypeptides and mod-
ifications of isoelectric points.
The introduction of the use of hydroxyethyldisulphide (HED) in excess,
commercialized under the trade name DeStreak (GE Healthcare), provided
a way to keep the sulfhydryls reduced (by mass action) at alkaline pH and
dramatically increased the resolution of alkaline IEF using IPG strips
(Olsson et al., 2002). This is espeically found in conjunction with using
cup loading or paper-bridge loading at the anode for sample entry, whereby
the alkaline proteins all enter the IPG strip with the same acidic charge and
524 David B. Friedman et al.

O CH2 S CH2 H
Protein-S− + H CH2 S CH2 O

S CH2 H O CH2
Protein-S CH2 O + H CH2 S−

Figure 30.4 Oxidation of protein thiol groups using hydroxyethyl disulfide (HED)
according to Olsson et al. (2002). The resulting mixed disulfides are stable and prevent
the reformation of disulfide bond during isoelectric focusing over alkaline pH ranges,
whereas the standard reductant DTT becomes charged and migrates away from this
region of the gradient, resulting in streaking and artifacts.

then have to migrate toward the cathode to reach their isoelectric points
(rather having them migrating in both directions).
When the proteins enter the IPG strip, the sulfhydryl groups on cysteine
side chains become immediately oxidized to highly stable mixed disulfides.
This is an equilibrium reaction with high specificity, preventing any
unwanted side reactions (Fig. 30.4). Small amounts of DTT can be tolerated
for the sample preparation when using DeStreak, so that proteins can stay
reduced prior to focusing. However, care must be taken because excessive
DTT will reduce HED to produce 2-mercaptoethanol, which results in
even more horizontal streaking.3

1.5. Difference gel electrophoresis

Even with the increased reproducibility of IPG strips and stringent sample
preparation and running conditions, gel–gel variations can occur. The
success of a differential expression proteomics experiment relies on repeti-
tive measurements across individually-procured (biological) replicates, and
technical replicates (repetitive measurements on the same biological sample)
are also necessary to control for analytical variation, such as sample procure-
ment, handling, and gel–gel variation. These challenges can be addressed by
applying the DIGE technique (Friedman and Lilley, 2009; Lilley and
Friedman, 2004; Unlu et al., 1997).

1.5.1. Staining and detection

A major advancement made recently for 2D gel electrophoresis has been in
the development and application of fluorescent dyes used for the detection
and quantification of the intact protein species that are resolved in the gels.

3
When using DeStreak for IEF in alkaline pH gradients (pH 7–11, 3–11), 100 mL sample may contain up to
20 mM reducing agent. DeStreak can also be used for other pH ranges, with different tolerances for
reductants (adapted from the DeStreak rehydration solution manual, GE Healthcare).
Isoelectric Focusing and Two-Dimensional Gel Electrophoresis 525

Fluorescent dyes typically provide detection sensitivity at least as sensitive as

silver stain (ca. 1 ng) if not greater, and significantly increase the linear
dynamic range of abundance changes to 3–4 orders of magnitude (silver and
coomassie-blue stains are typically provide less than 1 order of magnitude
dynamic range).

1.5.2. DIGE and quantitative analytical algorithms

The implementation of the DIGE method in 1997 (Unlu et al., 1997)
leverages the sensitivity and dynamic range of fluorescence labeling with
multiplexing samples into the same gel to remove analytical (gel–gel)
variation for the coresolved samples. It also utilizes a mixed-sample internal
standard present throughout a series of DIGE gels to enable registration of
migration patterns and, therefore, normalization of abundance ratios to
provide multivariable experiments with exceptional statistical power
(Karp and Lilley, 2005).
Briefly, the approach involves prelabeling several samples with spectrally
distinct fluorescent dyes (Cy2, Cy3, and Cy5), followed by sample multi-
plexing and coresolution on the same 2D gel. In this way gel–gel variations
are eliminated, and the mutually exclusive fluorescent images can be indi-
vidually recorded and used for direct quantification of protein abundance
changes for each resolved feature (Fig. 30.5). Two different labeling con-
cepts are used: ‘‘minimal’’ labeling the e-amino group of the lysine of about
5% of the total proteins and ‘‘saturation’’ labeling of all available cysteines of
a protein mixture.
The DIGE approach is most beneficial and statistically powered when a
Cy2-labeled internal standard is coresolved on a series of gels that each
contains individual samples labeled with Cy3 or Cy5. Importantly, this
Cy2-labeled internal standard is comprised of an equal mixture of all samples
in the experiment, and is present on every gel of a multigel experiment.
Because the individual samples (labeled with Cy3 or Cy5) are multiplexed
with an equal aliquot of the same Cy2-standard mixture, each resolved
feature can be directly related to the cognate feature in the Cy2-standard
mixture within that gel. Intragel Cy3:Cy2 and Cy5:Cy2 ratios are then
calculated without interference from gel–gel variation, and these ratios can
then be normalized to all other measurements for that feature from the other
samples across the experiment with extremely low technical (analytical)
noise and high statistical power (Karp and Lilley, 2005).
The DIGE approach is also directly amenable to multivariate statistical
analyses such as principle component analysis and hierarchical clustering.
These additional statistical tools can be extremely beneficial in visualizing
the variation within a set of experimental samples. Importantly, they can
help determine if the major source of variation is describing the biology or
indicating unanticipated variation between samples (or introduced during
sample preparation), as well as pinpoint subsets of proteins that respond
526 David B. Friedman et al.

Figure 30.5 Fluorescence image of a multiplexed DIGE gel, 24 cm pH 4–7 (12%

second dimension) in false-color representation. Samples had been labeled prior to
separation with Cy2 (blue; labeling 100 mg human AGS cells), Cy3 (green; labeling
100 mg human AGS cells infected with H. pylori strain 26695), and Cy5 (red; labeling
100 mg human AGS cells infected with H. pylori strain 26695 deleted for the cag
pathogenicity island). Individual Cydye signals were acquired using mutually exclusive
excitation and emission spectra at 100 mm resolution, and recorded at 16-bit depth using
a grayscale intensity from 1 to 100,000. Using Cy2 to label a mixed-sample internal
standard enables a statistically powered differential expression experiment by register-
ing independent-replicate samples (separately labeled with Cy3 and Cy5) from multiple
experimental conditions across several matched DIGE gels.

collectively to an experimental stimulus or classification (e.g., see Franco

et al., 2009; Friedman et al., 2006, 2007; Hatakeyama et al., 2006; Suehara
et al., 2006; and reviewed in Friedman and Lilley, 2009; Lilley and
Friedman, 2004).

1.5.3. Software analytical tools

To facilitate the quantitative analysis of 2D gel experiments (whether they
be DIGE or otherwise), several software programs are now available. These
programs differ mostly in the algorithms used to detect protein features
(boundaries) and gel–gel alignment/registration (e.g., vector-based image
warping). In general, they all provide powerful analytical tools coupled with
univariate (Student’s t-test, ANOVA) and multivariate statistical analyses
(e.g., principle components analysis and hierarchical clustering) that can be
extremely beneficial in evaluating abundance changes of individual protein
features as well as global expression patterns that can help discern changes
that describe the biological phenotype from those that arise from unantici-
pated variation in the experiment.
Isoelectric Focusing and Two-Dimensional Gel Electrophoresis 527

2. Materials
2.1. Equipment
1. Electrophoresis system for IEF. Several formats are available from a
number of commercial vendors. Most can focus up to 12 IPG strips
simultaneously.
2. Programmable Power Supply capable of producing at least 3500 V but
preferably 8,000–10,000 V.
3. Multigel system for second-dimension SDS–PAGE. Several formats are
available for running 2, 4, 6, 12 or more gels in a single run.
4. Thermostatic circulator (necessary for most second-dimension systems;
check specific vendors).
5. Immobiline DryStrip Reswelling Tray (GE Healthcare).
6. Sample-loading cups (required for IEF in gradients that include
pH > 8).
7. Rotary shaker.
8. Image capture devices (e.g. densitometry, flatbed scanners, fluorescent
imager (must be compatible with Cy dyes for best results with DIGE)).

2.2. Solutions and reagents

Most reagents and ready-made solutions are available from a variety of
suppliers. All solutions are prepared fresh before use, except where indi-
cated. All solutions should be prepared using water that has a resistivity of
18.2 mO-cm; this is referred to as ‘‘water’’ throughout the text.
1. IPG strips and accompanying ampholines and/or IPG buffers are
commercially available from a number of vendors in a large variety of
pH ranges. This includes wide range (pH 3–11, both as a linear and a
nonlinear gradient), medium range (e.g., pH 4–7, pH 7–11), and
narrow range (e.g., pH 5–6, pH 5.5–6.7). Strips are also available in a
variety of lengths, ranging from low-resolution 7 cm to high-resolution
24 cm.
2. Sample buffer according to Rabilloud (1998): 7 M urea, 2 M thiourea,
4% (w/v) CHAPS, 1% (w/v) DTT, 2% (v/v) Pharmalytes 3–10, and
cocktail of protease inhibitors (Complete, Roche, 1 tablet/50 mL
solution). For samples containing more hydrophobic proteins, the
following ASB14 lysis buffer may be beneficial: 7 M urea, 2 M thiourea,
2% amidosulfobetaine-14, and 50 mM Tris–HCl pH 8.0. Sample buffer
is stored at
80  C in small aliquots and should be thawed only once.
For DIGE experiments, use the Rabilloud buffer without DTT and
528 David B. Friedman et al.

without Pharmalytes or IPG buffers; other specifics to this technique

can be found in other detailed methods chapters (Friedman and Lilley,
2009).
3. Hydroxyethyldisulphide (DeStreak, GE Healthcare) is supplied either
as a liquid reagent or as a complete sample buffer formulation, for IEF
in alkaline pH ranges.
4. Cell lysis buffer. The above sample buffers with 7 M urea, 2 M thiourea
and 4% CHAPS may be used for protein solubilization. Alternatively,
TNE buffer (50 mM Tris–HCl pH 7.6, 150 mM NaCl, 2 mM EDTA
pH 8.0, 2 mM DTT, 1% (v/v) NP-40), and RIPA buffer (50 mM Tris–
HCl pH 8.0, 150 mM NaCl, 1% NP-40, 0.5% deoxycholic acid, 0.1%
SDS), can also be used if followed by a precipitation/cleanup step (see
Section 3.2).
5. Ready-made solutions for SDS–PAGE: 30% acrylamide:bis-acrylam-
ide (37.5:1), N,N,N,N0 -tetramethylethylenediamine (TEMED), and
ammonium persulfate, available from several suppliers (e.g., National
Diagnostics, Bio-Rad, GE Healthcare).
6. 4 separating gel buffer: 1.5 M Tris–HCl pH 8.8.
7. Water-saturated butanol.4
8. 10 SDS–PAGE running buffer (1 L): 30.25 g Tris–base, 144.13 g
glycine, and 10 g SDS (0.1%).
9. Equilibration solution: 50 mM Tris–HCl pH 8.8, 6 M urea, 30% (v/v)
glycerol, 2% (w/v) SDS, and 0.01% (w/v) bromophenol blue.
10. Dithiothreitol (store desiccated).
11. Iodoacetamide (store desiccated, keep in the dark).

3. Methods
3.1. Protein sample preparation
Robust sample preparation is vital for any successful analytical measure-
ment. Buffers and materials used should be of the highest quality, and care
should be taken during sample procurement to increase reproducibility and
to minimize unanticipated variation that might be measured in the analysis.
Small molecule protease and phosphatase inhibitors such as aprotinin,
leupeptin, pepstatinA, antipain, AEBSF, sodium orthovanadate, okadaic
acid, and microcystin, among others (or commercial kits that utilize these

4
Mix equal parts butanol and water and shake vigorously. Let the two phases separate overnight, and use the
butanol phase for overlay. Butanol that is not completely water saturated can extract water from the top of the
gel. A more recent modification is to use a 0.1% SDS solution in a conventional spray bottle, used to carefully
spray a fine mist over the top of the gels to thoroughly cover the top of the gel (the gel/overlay interface will
not be as obvious).
Isoelectric Focusing and Two-Dimensional Gel Electrophoresis 529

small molecule inhibitors), should be used. In particular, do not use inhibi-

tors (e.g., soybean trypsin inhibitor) that will resolve on the 2D gel.5
1. Essentially any protein extraction buffer can be used, provided that samples
are subsequently precipitated to remove nonprotein-based ionic compo-
nents (that can severely interfere with IEF, see Section 3.2). Samples are
then resuspended with a 2D gel-compatible buffer, and in many cases
protein extracts can be made directly in this buffer and analyzed without
precipitation. The following buffers contain sufficient chaotropic activity
without providing additional ionic composition to the sample:
A. Standard 2D gel electrophoresis lysis buffer (Rabilloud, 1998): 7 M
urea, 2 M thiourea, 4% CHAPS, 2 mg/mL DTT, and 50 mM
Tris–HCl pH 8.0.
B. Modification for membrane-associated proteins: 7 M urea, 2 M
thiourea, 2% amidosulfobetaine-14, and 50 mM Tris–HCl pH 8.0.
Buffer systems containing other salts and detergents, especially sodium
dodecylsulfate, may be more efficient at protein extraction, but must be
precipitated prior to IEF. For example:
C. TNE: 50 mM Tris–HCl pH 7.6, 150 mM NaCl, 2 mM EDTA pH
8.0, 2 mM DTT, and 1% (v/v) NP-40.
D. RIPA buffer: 50 mM Tris–HCl pH 8.0, 150 mM NaCl, 1% NP-40,
0.5% deoxycholic acid, and 0.1% SDS.
2. It is critical to resuspend the cells rapidly to prevent proteolytic activity.
Extracts should be centrifuged for 10 min at 15,000 or 100,000g to
remove insoluble proteins and phospholipids that can interfere with IEF.
3. Sonication can, in some cases, improve sample quality by disrupting
nucleic acids which are subsequently removed by sample cleanup along
with phospholipids. Both of these nonprotein-based ionic components
can interfere with IEF (see Section 3.2). Short bursts with a tip sonicator,
on ice, are suggested.
4. At all steps, it is important to keep the system chilled, especially in the
presence of urea-containing samples that should never be heated. Excessive
heating of urea (above 37  C) can accelerate the formation of isocyanate
(a natural breakdown product of urea), which will in turn carbamylate free
amines. When this occurs on proteins (either at the amino-terminal residue
or on the epsilon-amine group of lysine residues), it prevents these sites
from becoming protonated, causing an acidic shift in the isoelectric point.
The end result of heavy sample carbamylation in 2D gel electrophoresis is

5
An excellent practical 2D electrophoresis handbook with many tips and tricks with respect to basic
methodology can be downloaded for free at the following site: http://www5.gelifesciences.com/aptrix/
upp01077.nsf/Content/2d_electrophoresis2delectrophoresis_handbook
530 David B. Friedman et al.

beautiful charge trains of proteins that appear to be post-translationally

modified, but are completely artifactual.
5. Measure protein concentration using a variety of standard methods.
Take care to use a method that is compatible with the buffer that the
proteins are extracted in. For example, CHAPS and thiourea (although
perfectly suitable for extraction) will interfere with either the Bradford
or BCA assays, making the data inaccurate and unreliable. In these cases,
aliquots should be precipitated prior to quantification in a suitable buffer,
or the use of a detergent compatible assay should be utilized.
6. For cell culture experiments, protein concentration can be estimated
from the cell number. For example, for protein samples prepared from
the 697 pre-B lymphoma cell line (available from the German Collec-
tion of Microorganisms and Cell Cultures, Deutsche Sammlung von
Mikroorganismen und Zellkulturen GmbH, DSMZ No: ACC 42), 107
cells correspond to approximately 1 mg of extracted proteins. It is
essential to wash the cells thoroughly (at least twice in PBS) in order
to remove components of the growth medium (especially serum pro-
teins) before collecting the cells. For the final cell pellet, all PBS is
carefully removed with a fine pipette tip. Cell pellets (usually 108 cells
per tube) are stored at
80  C.6
7. The extract can now be applied to the dehydrated IPG strips, either by
in-gel rehydration or by cup loading. Appropriate ampholytes or IPG
buffers should be added prior to sample application, at 0.5% (v/v) for
most applications (but up to 2% can be tolerated if necessary). If bro-
mophenol blue was not present in the sample buffer, add a small amount
(either as a few grains of dry solid or a few microliters of a solution in
water) to use as a tracking dye during IEF as well as a visual aid during
cup loading. Alternatively, samples can be stored at
80  C indefinitely.

3.2. Sample cleanup and precipitation

As already mentioned, the presence of nonprotein-based ions (e.g., salts,
phospholipids) can sometimes interfere with IEF by reducing the resistance
in the IPG strip such that it becomes difficult to attain the high voltages
necessary for high-resolution focusing without overheating the strips. Since
most commercial IEF units focus multiple strips together in a parallel circuit,
one strip of significantly different resistance can adversely affect the resolu-
tion of the other strips.
Often a cleanup or precipitation step can remove these interfering ions,
or at least normalize all samples to be of similar resistance. Precipitation not

6
For protein extraction from cell pellets using 2D gel sample buffer, it is useful for the solution to contain 2%
(v/v) Pharmalytes 3–10 to prevent aggregation of DNA. Cells (e.g., 108 cells) are best disrupted by rapid
addition of 1 mL sample buffer (Rabilloud, 1998) to the cell pellet. Nucleases (DNase, RNase, and
Benzonase) are commonly added as well, but their efficacy is questionable in this highly denaturing buffer.
Isoelectric Focusing and Two-Dimensional Gel Electrophoresis 531

only removes several contaminants at the same time, but also disrupts
complexes efficiently and prevents protease activities irreversibly. The
most effective methods for precipitation are using methanol and chloroform
according to Wessel and Flugge (1984) or TCA, deoxycholate, and acetone
according to Arnold and Ulbrich-Hofmann (1999). Before IEF, the pro-
teins must be redissolved in solubilization solution (see Section 3.1). In
addition, several protein precipitation kits for 2D gel analysis are available
from a number of commercial vendors. An adaption of the method of
Wessel and Flugge (1984) is described as follows:
1. Bring up predetermined amount of protein extract to 100 mL with water.
2. Add 300 mL (3-volumes) water.
3. Add 400 mL (4-volumes) methanol.
4. Add 100 mL (1-volume) chloroform.
5. Vortex vigorously and centrifuge; the protein precipitate should appear
at the interface.
6. Remove the water/MeOH mix on top of the interface, being careful not
to disturb the interface. Often the precipitated proteins do not make a
visibly white interface, and care should be taken not to disturb the interface.
7. Add another 400 mL methanol to wash the precipitate.
8. Vortex vigorously and centrifuge; the protein precipitate should now
pellet to the bottom of the tube.
9. Remove the supernatant and briefly dry the pellets in a vacuum centrifuge.
10. Resuspend the pellets in a suitable amount of 2D gel-compatible buffer
(see Section 3.1).
When using precipitation as a cleanup step, it is recommended to have a
starting protein concentration in the range of 1–10 mg/mL. When protein
samples are too diluted, it will be difficult to quantitatively recover proteins
following precipitation cleanup. Freeze/thawing should also be kept to a
minimum; freezing samples in 1 mL aliquots or less will usually suffice. For
DIGE experiments, this precipitation step greatly facilitates the requirement
for sample labeling to be performed in 2D electrophoresis sample buffer that
is devoid of free amines and pH balanced.

3.3. Isoelectric focusing using acidic range IPG gels

Most commercial IEF units can accommodate up to 12 individual IPG strips
per run. It is important to keep any IPG strips that are resolved in the same
run to be as equally matched as possible, because in most configurations the
individual IPG strips form a parallel circuit between the anode and cathode
electrodes. Any IPG strip that is significantly different from the others in
composition (especially with respect to ionic content) can adversely affect
the resolution of all of the strips, sometimes dramatically. It is, therefore,
strongly recommended to also cofocus only IPG strips of the same pH
532 David B. Friedman et al.

gradient and strip length, using samples derived from the same experiment
(e.g., sample types and composition should be a similar as possible).7
The following method is presented for 24 cm IPG strips pH 4–7:
1. IPG strips are rehydrated in the 2D gel sample buffer (see Section 3.1)
with total protein amounts as high as 0.5–2 mg protein, diluted or
resuspended in a final volume of 450 mL solution per 24 cm strip in a
reswelling tray. The strips are overlaid with 2–3 mL paraffin oil to
prevent urea crystallization.8
2. Rehydration of the strips in the presence of sample should take at least
12 h, but overnight rehydration is recommended, at room temperature.
Some proteins, especially high-molecular-weight proteins, need more
time to enter into the strip during the rehydration step.
3. After complete rehydration, the strips are moved to a flatbed electro-
phoresis system of which the temperature of the cooling block is held
constant at 20  C.
4. Humid paper wicks are applied on the ends of the rehydrated IPG strips,
and electrodes are positioned in place and the strips are covered with
paraffin oil.9
5. See Table 30.1 for typical focusing program.10
Rehydration loading is the easiest method for sample introduction for
2D gel electrophoresis, where protein samples in sample buffer are intro-
duced passively while the IPG strip takes up the sample solution, and the

7
When using the DIGE technique where different combination of samples are multiplexed and coresolved
using a series of gels, the potential of the different strips containing different sample amounts is minimized by
properly randomizing the loading of the samples such that each IPG strip contains the same variety of
samples. For example, in a 3-dye experiment of control versus treated samples using four biological replicates
(using Cy3 and Cy5 to randomly label each of the eight individual samples and Cy2 for the mixed-sample
internal standard), each of the four gels would contain one control and one treated sample along with an
aliquot of the Cy2-labeled mixed-sample internal standard.
8
It is very important to make sure that the rehydration solution is evenly distributed across the entire strip
length. The sample is first distributed along the length of a rehydration well and then the dehydrated IPG
strip is carefully overlayed (with the plastic backing facing up) to ensure that the sample is evenly distributed
beneath the overlayed strip. The IPG strips are covered with paraffin oil to avoid crystallization of the highly
concentrated urea (ca. 10 M urea is saturation at room temperature).
9
Placing humid paper wicks (filter paper wetted with water, but not oversaturated) between the electrodes
and both ends of the IPG strips will provide a sink for excess ions as well as proteins with isoelectric points
outside the pH range of the IPG strip. In extreme cases, it is possible to exchange the paper wicks during the
IEF run. The paraffin oil prevents drying out of the strip and crystallization of the urea/thiourea present, and
also serves to prevent carbon dioxide to enter the system. Carbon dioxide from the air can dissolve into the
IPG gel, acting as a buffer with pK 6.3 and thus changing the pH gradient.
10
IEF is mostly performed in a horizontal flatbed type of equipment. The focusing effect requires high field
strength; therefore, a power supply with high voltage is required. IEF must be performed at an exactly
controlled temperature, because the pK values of the proteins as well as the buffers are highly temperature
dependent. Thus a reliable cooling system is necessary for this method to be reproducible. It is important to
start with a low voltage to avoid aggregation of proteins. It is very useful to apply a voltage/time gradient on
the strips in order to minimize differences in conductivities coming from varying salt loads between the IPG
strips. The best results are obtained with high-end voltages of 8,000–10,000 V, with the overall amount of
total volt-hours being ultimately important for high-resolution focusing. Optimal gradients and total Vh
vary with strip length, pH range, and commercial formulations.
Isoelectric Focusing and Two-Dimensional Gel Electrophoresis 533

Table 30.1 Recommended isoelectric focusing programs using IPG gelsa

Voltage program for 24 cm IPG 4–7

Mode Voltage (V) Time (h) Vh
Step and hold 500 1
Gradient 1000 1
Gradient 10,000 3
Step and hold 10,000 5
Total 10 65.0 kV h
Voltage program for 24 cm IPG 3–11NL or IPG 3–10NL
Step and hold 150 2
Step and hold 300 2
Step and hold 500 2
Gradient 1000 4.5
Gradient 10,000 4.5
Step and hold 10,000 1.5
Total 16.5 45.0 kV h
Voltage program for 24 cm IPG 7–11NL
Step and hold 150 2
Step and hold 300 2
Step and hold 500 2
Gradient 1000 3.5
Gradient 10,000 3.5
Step and hold 10,000 5
Total 18.0 73.8 kV h
a
For best results, IEF should start at low voltage, and voltages should then slowly ramp up to their
maximum as indicated above. In some circumstances, it may be beneficial to add additional low-
voltage steps to the 4–7 program. Ultimately, it is the final kV h that becomes the most important
factor for high-resolution IEF.

proteins are evenly distributed across the entire pH gradient. In some

commercial IEF units, the IPG strip rehydration and focusing can be
performed in the same apparatus, eliminating extra manual steps. This
configuration also allows for the process of a so-called active rehydration,
whereby applying a low voltage (30–50 V) across the strip during rehydra-
tion loading can, in some cases, improve the sample application, because this
drives salt ions toward the electrodes, removes proteases from the proteins,
and helps large molecules from entering the gel. However, this configura-
tion of the IPG strip rehydration (‘‘facedown’’) during IEF can also produce
lower resolution in some cases due to the extra mechanical stress of the IEF
gel being sandwiched between the focusing tray and the plastic gel backing,
especially in the case of very abundant proteins. In these cases, using an
apparatus that performs the focusing with the plastic backing directly in
534 David B. Friedman et al.

contact with the focusing tray (the gel is ‘‘sunny-side up’’) can be advanta-
geous for more difficult samples (Fig. 30.3).
6. Optional: active rehydration (available with some instrument configura-
tions). Sample rehydration occurs with the IPG strip face down and in
contact with the electrodes, which supply a low (typically 30–50 V)
electric field across the strip during rehydration. In some cases, active
rehydration can be found to facilitate the entry of high-molecular-
weight proteins into the IPG strip. IEF units with this configuration
typically commence with the focusing with the IPG strips in the face-
down orientation and do not require user intervention at this point.

3.4. Isoelectric focusing using alkaline range IPG gels

The great advantage of rehydration loading is the reduction of handling
steps. However, it can also have some disadvantages especially in alkaline
pH ranges where proteins can aggregate on the strip surface and cause
horizontal streaks in the second-dimension gel.
The conditions for running alkaline IPG strips have to be changed
compared to the acidic range. During IEF at alkaline pH, water transport
and migration of the reducing agent DTT take place. To minimize these
effects, protein samples are always applied by cup loading at the anodic side
and not by in-gel rehydration (Fig. 30.3). The maximum protein load on
alkaline gels is lower than on the acidic gels. It is strongly suggested not to
load more than 0.5 mg protein per cup to prevent streaking.
The following method is presented for 24 cm IPG strips pH 3–11 and
7–11:
1. Ensure that the final sample volume will not exceed the volume of the
cup (typcially between 100 and 120 mL, check manufacturer).11
2. The IPG strip is rehydrated (without protein sample) with the rehydra-
tion buffer containing HED (DeStreak; Olsson et al., 2002) in the
reswelling cassette. IPG buffers can be added at this stage to 0.5% final
concentration for flatbed IEF and vertical SDS–PAGE. Since no proteins
are present during rehydration, the time can be as short as 6 h (to enable
rehydration and focusing to be done on the same day), but rehydration
for >12 h is generally recommended.
3. After complete rehydration, the IPG strips are placed on a flatbed
electrophoresis system of which the temperature of the cooling block
is held constant at 20  C.
4. Humid paper wicks are applied on the ends of the rehydrated IPG strips,
and electrodes are positioned in place. Application cups are then

11
The protein concentration in the sample cup should not be too high (<10 mg/mL) to avoid protein
precipitation at the entry point. If possible, the sample should be diluted with the sample buffer.
Isoelectric Focusing and Two-Dimensional Gel Electrophoresis 535

carefully positioned just below the anode (on the acidic side of the
gradient) such that the bottom of the cup is in contact with the top of
the IPG gel and makes a seal, but not crushing the gel. The strips are then
covered with paraffin oil.9
5. Ensure that oil does not leak into the cups before proceeding. Then add
approximately 10–20 mL paraffin oil to the cup, and carefully pipet the
sample under the oil (check manufacturer for maximum sample volume
per cup).12
6. See Table 30.1 for typical focusing program.10

3.5. Equilibration of IPG gels

After the IEF step is completed, it is necessary to equilibrate the IPG strips in
two steps prior to the second dimension. The equilibration buffer contains 6 M
urea and 30% glycerol to diminish electroendosmotic effects (Sanchez et al.,
1997), which are responsible for poor transfer from first to second dimension.
Further, the equilibration buffer contains 2% (w/v) SDS to make the proteins
are negatively charged for the SDS–PAGE. Loss of proteins during the
equilibration step and subsequent transfer from the first to the second dimen-
sion are mainly due to proteins which are strongly absorbed to the IPG gel
matrix and due to wash-off effects. The majority of these proteins appear to be
lost during the first minutes of equilibration, whereas protein losses during the
second equilibration step are only minimal (Sanchez et al., 1997).
1. After focusing is complete, the IPG strips are removed from the paraffin
oil and incubated in the equilibration buffer. This can be done in a
variety of vessels, sealed tubes, and even in the flatbed focusing IPG strip
holder to minimize strip handling.
2. The strips are first equilibrated in 25–100 mL equilibration buffer supple-
mented with 1–2% (w/v) DTT (15–20 min) to reduce disulfide bonds.
3. The solution is discarded, and replaced with an equal volume of equili-
bration buffer supplemented with iodoacetamide (2.5  w/w more
than what was used for DTT) for an additional 15–20 min to alkylate
the formed sulfhydryl groups with a carbamidomethyl group (net mass
increase of 57 Da). This is to prevent reoxidation of the sulfhydryl
groups during the second-dimension SDS–PAGE in order to prevent
streaking of proteins.13

12
Paper-bridge loading is a modification of the cup-loading technique that uses a piece of clean filter cardboard
containing the protein sample solution and applied at one of the ends of the IPG strip. Both this as well as the
conventional cup-loading technique needs more handling steps and can lead to losses of proteins with their
isoelectric points close to the pH value of the application point. However, with these techniques all proteins
have the same charge sign and will not form complexes during electrophoretic migration into the IPG strip,
and so in some cases it can deliver better results especially for IEF at pH ranges greater than pH 8.
13
Although suggested in several scientific papers, it is generally not recommended to prealkylate proteins prior
to IEF, because it is often not easy to guarantee 100% alkylation.
536 David B. Friedman et al.

4. Equilibrated IPG strips can now be run on the second-dimension SDS–

PAGE gels, or can be double wrapped in plastic wrap and frozen for storage
at
20 or
80  C for up to several months. For samples labeled with DIGE,
storage in this way for at least a week does not seem to affect the performance
of the Cydyes. If necessary, IPG strips can be frozen in this way directly after
focusing, leaving the equilibration and reduction/alkylation steps until the
second-dimensional separation is ready to be performed.

3.6. SDS–PAGE: The second dimension

When equilibration is completed, the strips are ready for the second dimen-
sion (SDS–PAGE). When more than one gel is to be run (which is always
the case in a comparative proteomics study), the most reproducible results
can be obtained when gels are run together. Typically, electrophoresis
chambers holding 10–12 large-format (24  20 cm) SDS–PAGE gels are
used, and are available from a number of commercial vendors. For enhanced
reproducibility, and to reduce depletion of SDS during the run, it is
recommended to utilize a cooling system to keep the buffer at 20  C.
Homemade gels are commonly used and are very economical but also
very labor-intensive to prepare reproducibly. Readymade precast gels are
now available on the market from a number of vendors, including using a
low-fluorescence backing medium for DIGE. A variety of acrylamide
concentrations with varying cross-linker ratios (acrylamide:bis-acrylamide)
can be used. For general purposes, 12% homogeneous pore size gels (12% T;
2.6% C) afford the best compromise between ease of preparation and
resolution. Gradient gels can provide better separation and resolution over
a broader mass range, but are slightly more difficult to prepare reproducibly
on a large scale. Dedicated gel casters for reproducible gradient gels are
commercially available.
Hand-cast gels can either be made individually or in bulk using com-
mercial multigel casters. The total volume of acrylamide solution necessary
per multigel casting should be determined for each individual setup and
equipment ahead of time. The following is an example for multiple gel
casting where the final volume is 2.3 L. For smaller volumes, adjust the
amounts proportionally.
1. Typically, glass cassettes with fixed spacers of 1.0 or 1.5 mm thickness
are separated by plastic sheets to prevent the plates from sticking to each
other after polymerization. If desired, each gel can be given an unique
serial identity number by placing a piece of printed Whatman paper
between the glass plates prior to casting.
2. In a vacuum flask is prepared 920 mL 30% (w/v) acrylamide:bisacry-
lamide solution (37.5:1), 575 mL 1.5 M Tris–HCl, pH 8.8 and 770 mL
water (see Step 4).
Isoelectric Focusing and Two-Dimensional Gel Electrophoresis 537

3. The mixture is degassed for 10 min to remove microair bubbles that

can interfere with polymerization and cause point streaking after pro-
tein staining in some cases.
4. Optional: after degassing, SDS can be added to the acrylamide solution
to a final concentration of 0.01% (add 2.3 g in 35 mL water). However,
many have shown that excellent resolution can be obtained without
SDS at this stage, most likely because the SDS that carries proteins
through the gel is already associated with the protein during the
equilibrium stage and/or carried through by the running buffer (and
any SDS present in the gel at the start of the run would leave the gel
when the electric field is applied, before proteins entered the gel).
Many commercial vendors now offer precast gels made both with
and without SDS. If omitting SDS at this stage, then add 805 mL
water instead of 770 mL in Step 2 (see Step 2).
5. Just prior to pouring, using gentle mixing or a stir bar, add 700 mg
ammonium persulfate (APS, made fresh) and 300 mL TEMED to the
acrylamide mixture. (Use proportionally less APS and TEMED when
using smaller volumes of acrylamide mixtures for different pouring
configurations).
6. Multigel casting chambers are typically filled from the bottom to a
height of about 2 cm below the top of the glass plates. The gels are
carefully overlaid with 1.0–1.5 mL buffer-saturated iso-butanol to
allow for polymerization.14
7. After full polymerization, clean the cassettes of residual polyacrylamide
residues, overlay with water or SDS running buffer, and double seal
with plastic wrap for storage up to several weeks at 4  C.
8. If first-dimensional IPG strips (equilibrated and reduced/alkylated)
were previously frozen, allow a few minutes for the strips to thaw
completely prior to unsealing and handling them. Rinse and remove
the water or SDS running buffer out of the top well space of the
second-dimension cassette.
9. The IPG strip should be dipped in SDS running buffer and then placed
inside the top well space of the SDS–PAGE gel assembly. The plastic
backing of the IPG strip should adhere to the inside surface of one of
the glass plates (typically the longer plate if possible), and can then be
easily tapped into place using a thin card or ruler (making sure that

14
Casting a large number of gels requires some practice. The amount of catalyst that is used is minimal to
prevent the gels from polymerizing too rapidly, which leads to excessive heating of the casting chamber.
Ideally, initial polymerization, which can be seen by the development of a distinct gel surface below the
butanol layer, should take 30–60 min. Subsequently, gels can be washed, covered with gel buffer, and stored
at room temperature. The amounts of catalyst can be increased by 10% if necessary. ‘‘Fast’’ polymeration
should occur within an hour or two, after which the isopropanol can be replaced with water or SDS running
buffer. However, gels should ideally be left to polymerize overnight to enable complete polymerization, and
the top of the gels should be lightly sealed with plastic wrap to prevent evaporation of the overlay (especially
if the butanal is left on overnight).
538 David B. Friedman et al.

pressure is applied to the plastic backing, not the IPG gel). Once the
SDS running buffer is overlayed within the running chamber, good
contact and removal of air bubbles can be easily achieved using the
same tapping procedure. Once placed, the IPG strip should not move
during the second-dimension run.
10. Optional: to ensure good contact between the strip and the gel, an
agarose solution can be used to keep the IPG strip in place. The agarose
solution is kept at 65 C and added first on top of the gel. Immediately
after, the equilibrated strip is placed on the gel. This optional procedure
is favored by some, but is cumbersome and can be difficult if the
solution solidifies before the strip is in place and any air bubbles
removed.15
11. Optional: a molecular weight marker can be run at the side (loaded
through a small paper wick).
12. The current is set to 5–10 mA/gel initially to let the proteins enter the
gel and subsequently set to 20 mA/gel over night at 15  C until the blue
front reach the end of the gel. If available, using constant wattage,
setting less than 1 watt per gel for at least the first hour followed by no
more than 15 watt per gel until completion is preferred because the
voltage and amperage will vary during the run as salts and ions leave
the gel.
After completion, gels can be fixed and stained using a variety of visual
and fluorescent staining techniques. DIGE gels can be imaged on a fluores-
cent imager immediately after the second dimension is completed. Gels can
then be stored in the refrigerator or in a cold room at 4  C for months16. It is
no problem to identify proteins using mass spectrometry from 2D gels that
were stored over a long period of time.

REFERENCES
Anderson, N. G., and Anderson, N. L. (1978a). Analytical techniques for cell fractions. XXI.
Two-dimensional analysis of serum and tissue proteins: Multiple isoelectric focusing.
Anal. Biochem. 85, 331–340.
Anderson, N. L., and Anderson, N. G. (1978b). Analytical techniques for cell fractions.
XXII. Two-dimensional analysis of serum and tissue proteins: Multiple gradient-slab gel
electrophoresis. Anal. Biochem. 85, 341–354.

15
The agarose overlay is prepared in such a way that it remains fluid at relative low temperatures. To realize
this a mixture of low-melting agaroses is used (e.g., 0.4% (w/v) standard low Mr from Bio-Rad Laboratories,
and 0.1% (w/v) type VII-A-low gelling temperature from Sigma), dissolved in 1 SDS running buffer.
A trace of bromophenol blue can be added as tracking dye.
16
For long-term storage, 0.02% (w/v) sodium azide can be added (optional) to prevent bacterial and fungal
growth. For DIGE gels and other applications where the gel is affixed to one of the glass plates, the gels can
be stored with the glass cassette intact to minimize dehydration during storage.
Isoelectric Focusing and Two-Dimensional Gel Electrophoresis 539

Arnold, U., and Ulbrich-Hofmann, R. (1999). Quantitative protein precipitation from

guanidine hydrochloride-containing solutions by sodium deoxycholate/trichloroacetic
acid. Anal. Biochem. 271, 197–199.
Bjellqvist, B., Ek, K., Righetti, P. G., Gianazza, E., Gorg, A., Westermeier, R., and
Postel, W. (1982). Isoelectric focusing in immobilized pH gradients: Principle, method-
ology and some applications. J. Biochem. Biophys. Methods 6, 317–339.
Franco, A. T., Friedman, D. B., Nagy, T. A., Romero-Gallo, J., Krishna, U., Kendall, A.,
Israel, D. A., Tegtmeyer, N., Washington, M. K., and Peek, R. M. (2009). Delineation
of a carcinogenic Helicobacter pylori proteome. Mol. Cell. Proteomics 25, 25.
Friedman, D. B., and Lilley, K. S. (2009). Difference gel electrophoresis (DIGE).
In ‘‘The Protein Protocols Handbook’’ (Walker John, ed.), 3rd edn. pp. 379–408.
Humana Press, Totowa, NJ.
Friedman, D. B., Stauff, D. L., Pishchany, G., Whitwell, C. W., Torres, V. J., and
Skaar, E. P. (2006). Staphylococcus aureus redirects central metabolism to increase iron
availability. PLoS Pathog. 2, e87.
Friedman, D. B., Wang, S. E., Whitwell, C. W., Caprioli, R. M., and Arteaga, C. L. (2007).
Multi-variable difference gel electrophoresis and mass spectrometry: A case study on
TGF-beta and ErbB2 signaling. Mol. Cell. Proteomics 6, 150–169.
Gorg, A., Obermaier, C., Boguth, G., Harder, A., Scheibe, B., Wildgruber, R., and
Weiss, W. (2000). The current state of two-dimensional electrophoresis with immobi-
lized pH gradients. Electrophoresis 21, 1037–1053.
Gorg, A., Weiss, W., and Dunn, M. J. (2004). Current two-dimensional electrophoresis
technology for proteomics. Proteomics 4, 3665–3685.
Hartinger, J., Stenius, K., Hogemann, D., and Jahn, R. (1996). 16-BAC/SDS-PAGE:
A two-dimensional gel electrophoresis system suitable for the separation of integral
membrane proteins. Anal. Biochem. 240, 126–133.
Hatakeyama, H., Kondo, T., Fujii, K., Nakanishi, Y., Kato, H., Fukuda, S., and
Hirohashi, S. (2006). Protein clusters associated with carcinogenesis, histological differ-
entiation and nodal metastasis in esophageal cancer. Proteomics 6, 6300–6316.
Hoving, S., Voshol, H., and van Oostrum, J. (2000). Towards high performance two-
dimensional gel electrophoresis using ultrazoom gels. Electrophoresis 21, 2617–2621.
Ibel, K., May, R. P., Kirschner, K., Szadkowski, H., Mascher, E., and Lundahl, P. (1990).
Protein-decorated micelle structure of sodium-dodecyl-sulfate–protein complexes as
determined by neutron scattering. Eur. J. Biochem. 190, 311–318.
Karp, N. A., and Lilley, K. S. (2005). Maximising sensitivity for detecting changes in protein
expression: Experimental design using minimal CyDyes. Proteomics 5, 3105–3115.
Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of
bacteriophage T4. Nature 227, 680–685.
Lilley, K. S., and Friedman, D. B. (2004). All about DIGE: Quantification technology for
differential-display 2D-gel proteomics. Expert Rev. Proteomics 1, 401–409.
O’Farrell, P. H. (1975). High resolution two-dimensional electrophoresis of proteins. J. Biol.
Chem. 250, 4007–4021.
O’Farrell, P. Z., Goodman, H. M., and O’Farrell, P. H. (1977). High resolution two-
dimensional electrophoresis of basic as well as acidic proteins. Cell 12, 1133–1141.
Olsson, I., Larsson, K., Palmgren, R., and Bjellqvist, B. (2002). Organic disulfides as a means
to generate streak-free two-dimensional maps with narrow range basic immobilized pH
gradient strips as first dimension. Proteomics 2, 1630–1632.
Rabilloud, T. (1998). Use of thiourea to increase the solubility of membrane proteins in
two-dimensional electrophoresis. Electrophoresis 19, 758–760.
Righetti, P. G., Simo, C., Sebastiano, R., and Citterio, A. (2007). Carrier ampholytes for
IEF, on their fortieth anniversary (1967–2007), brought to trial in court: The verdict.
Electrophoresis 28, 3799–3810.
540 David B. Friedman et al.

Sanchez, J. C., Rouge, V., Pisteur, M., Ravier, F., Tonella, L., Moosmayer, M.,
Wilkins, M. R., and Hochstrasser, D. F. (1997). Improved and simplified in-gel sample
application using reswelling of dry immobilized pH gradients. Electrophoresis 18, 324–327.
Schagger, H., and von Jagow, G. (1991). Blue native electrophoresis for isolation of
membrane protein complexes in enzymatically active form. Anal. Biochem. 199,
223–231.
Suehara, Y., Kondo, T., Fujii, K., Hasegawa, T., Kawai, A., Seki, K., Beppu, Y.,
Nishimura, T., Kurosawa, H., and Hirohashi, S. (2006). Proteomic signatures corres-
ponding to histological classification and grading of soft-tissue sarcomas. Proteomics 6,
4402–4409.
Unlu, M., Morgan, M. E., and Minden, J. S. (1997). Difference gel electrophoresis: A single
gel method for detecting changes in protein extracts. Electrophoresis 18, 2071–2077.
Vesterberg, O. (1972). Isoelectric focusing of proteins in polyacrylamide gels. Biochim.
Biophys. Acta 257, 11–19.
Vesterberg, O., and Svensson, H. (1966). Isoelectric fractionation, analysis, and characteri-
zation of ampholytes in natural pH gradients. IV. Further studies on the resolving power
in connection with separation of myoglobins. Acta. Chem. Scand. 20, 820–834.
Wessel, D., and Flugge, U. I. (1984). A method for the quantitative recovery of protein in
dilute solution in the presence of detergents and lipids. Anal. Biochem. 138, 141–143.
Westermeier, R. (2005). Electrophoresis in Practice (4th edn). Method 10 - IEF in immo-
bilized pH gradients, pp. 257–273. Wiley-VCH.
 C H A P T E R T H I R T Y- O N E

Protein Gel Staining Methods:

An Introduction and Overview
Thomas H. Steinberg

Contents
1. Introduction 542
2. General Considerations 543
3. Instrumentation: Detection and Documentation 545
4. Total Protein Detection 545
4.1. Colorimetric total protein stains 545
4.2. Fluorescent total protein stains 549
4.3. Preelectrophoresis sample labeling 556
5. Phosphoprotein Detection 556
5.1. Pro-Q Diamond phosphoprotein gel stain 557
5.2. Phos-tag phosphoprotein stains 557
6. Glycoprotein Detection 557
6.1. General glycoprotein detection 557
6.2. O-GlcNAc detection 559
References 559

Abstract
Laboratory scientists who encounter protein biochemistry in many of its myriad
forms must often ask: is my protein pure? The most frequent response: run a
denaturing SDS polyacrylamide gel. Running this gel raises another series of
considerations regarding detection, quantitation, and characterization and so
the next questions invariably center on suitable protein gel staining and detec-
tion methods. A total protein profile can be determined with the colorimetric
methods embodied in Coomassie Blue and silver staining methods, or increas-
ingly, with fluorescent stains. Protein quantitation can be done following stain-
ing, with fluorescence- and instrumentation-based methods offering the
greatest sensitivity and linear dynamic range. Protein posttranslational mod-
ifications such as phosphorylation and glycosylation can be reliably determined
with several fluorescence-based protocols. Staining and detection with two or
more different stains can be done in series to establish relative profiles of
modified versus total protein or to assess purity at two levels of quantitative

McArdle Laboratory for Cancer Research, University of Wisconsin–Madison, Madison, Wisconsin, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63031-7 All rights reserved.

541
542 Thomas H. Steinberg

sensitivity. The choice of staining method and protocol depends on the required
rigor of detection and quantitation combined with available instrumentation
and documentation capabilities. Other considerations for staining methods
include intended downstream analytical procedures such as mass spectrometry
or peptide sequencing, which preclude some methods. Nonfixative staining
methods allow western blotting after gel staining. Laboratory custom and
budget or intellectual curiosity may be the ultimate determinate of the chosen
gel staining protocol.

1. Introduction
This chapter focuses on electrophoresis-based protein detection with
emphasis on current methods for detection of total protein and detection of
the most frequently encountered protein posttranslational modifications
(PTMs), phosphorylation, and glycosylation. In the 20 years since the
publication of first edition of this volume, denaturing sodium dodecyl
sulfate polyacrylamide gel electrophoresis (SDS–PAGE) followed by pro-
tein staining has remained a fundamental laboratory procedure to determine
the composition and characterization of protein-containing mixtures at all
stages of a purification scheme. It is standard practice to display protein
profiles from a sampling of all purification steps—cell lysate to the final
product—to evaluate polypeptide composition, purity, and quantitation.
Thus a sample from each step should reveal a progressively simpler
polypeptide profile showing a greater proportion of the target protein.
The previously summarized methods and principles of protein electro-
phoresis and detection remain relevant (Dunbar et al., 1990; Garfin, 1990a,b;
Merril, 1990).
The widespread use of SDS–PAGE protein analytical methods has
driven the commercial development and common use of standardized
electrophoresis apparatus and reagents, precast multiwell 1D and 2D slab
gels (minigels; ca. 8
8 cm,
1.0 mm thick), and ready-to-use staining
formulations or kits for simplified, rapid staining. Additionally, proteomics-
oriented electrophoresis techniques have emerged, directed toward analysis
of complex protein mixtures by large format 2-D gel electrophoresis
followed by mass spectrometry (MS) of excised protein spots. The utility
of gel-based methodology in proteome analysis has spurred further com-
mercial development of extant colorimetric stains for total protein and
development of new fluorescent stains and modalities for total protein and
for protein PTMs, driving the coevolution of instrumentation for sensitive
detection, easy documentation, and accurate quantitation of both fluores-
cent and colorimetric signals. A recent compendium of proteomics proto-
cols describes methods and applications for protein detection in 2-D and
Protein Gel Stains 543

1-D gels and subsequent analytical methods (Walker, 2005). Recent

reviews of protein detection in gels for proteomics summarize the history
and scope of gel staining and detection methods (Miller et al., 2006; Patton,
2000, 2002; Smejkal, 2004). Clearly, there is convergence in the considera-
tions and methods for protein identification in gels for proteomic analysis
and protein purification.
Zymography—in-gel enzymatic activity analysis—is a complementary,
gel-based protein characterization methodology beyond the scope of this
chapter. A comprehensive handbook of zymographic protocols has recently
been published (Manchenko, 2003). A related, emerging field—activity-
based proteomics—utilizes small molecule probes to covalently link reporter
moieties to catalytic sites of specific classes of active enzymes in complex
mixtures such as lysates or whole cells. This increasingly popular approach can
be used to in-gel assays to identify enzymes and to track their purification
( Jessani and Cravatt, 2004; Paulick and Bogyo, 2008).
A detection method that accurately evaluates the total protein profile is
usually sufficient for a general purification protocol. Protein detection in
SDS–PAGE can be done by labeling samples—by covalent modification
with a fluorescent dye—prior to running gels or, more commonly, by non-
covalent postelectrophoresis staining with organic dyes or by metal deposition
techniques. Posttranslational modifications can be measured by specialized
staining formulations and protocols that target the modified moieties in a
background of total protein. PTM detection methods always require a control
lane displaying proteins known to contain and others known to be devoid of
the targeted PTM. If selective PTM detection utilizes fluorescence, bear in
mind that intrinsic protein fluorescence may be a confounding factor,
especially if ultraviolet (UV) light-excited, blue or green light-emitting fluor-
ophores are used. Therefore, an additional control—chemical or enzymatic
treatment of a sample—to remove the postulated PTM may be necessary.

2. General Considerations
Common features of all gel staining protocols are grounded in good
laboratory practices: cleanliness, careful manipulations, and attention to
detail. Postelectrophoresis gel staining is generally accomplished in covered
polycarbonate or polypropylene dishes. Accumulation of residual stain may
compromise results, particularly with fluorescent stains. Dishes should be
cleaned with 70–100% ethanol or methanol followed by a water rinse. The
gels are incubated on a rocking platform or an orbital shaker or at a
moderate speed, for example, 60 rpm. The gel should float freely in
solution, neither stuck to the bottom of the dish nor float on top of the
fixing, washing, and staining solutions. If a stacking gel is used, it should be
544 Thomas H. Steinberg

removed; if a commercial, ‘‘prepoured’’ gel is used, the wells and the foot of
the gel should be trimmed away to reduce staining deposition artifacts,
particularly if the SDS bolus from the sample buffer remains in the gel—
this typically travels with the Bromophenol Blue tracking dye. For all
methods and protocols presented below, pure—dionized or double-
distilled—water is used.
Historically, standard gel staining protocols required several sequential
procedures, some requiring more than one step and some requiring
intervening washes with water or buffer; common to all were:
1. The gel is fixed to immobilize protein in the gel and to remove SDS,
buffers, or other interfering components.
2. The gel is stained with organic dyes or silver deposition formulations.
3. Quenching a staining reaction or destaining to reduce background was
required.
Fixation generally has been accomplished with variations of acidic
alcohol mixtures (e.g., 7–10% acetic acid, 30–50% methanol); organic
dyes and destaining solutions were often formulated with the same solvents
as for fixation. Given that literally hundreds of variations of protein stains
and staining protocols have been introduced in the literature over the
years—often with lengthy, painstaking protocols and competing claims of
detection sensitivity (or other features of utility)—the commercialization of
stains, kits, and protocols have been something of a boon to the yeoman
laboratory scientist, although the proprietary considerations that accompany
commercial development may hinder clear description and understanding
of some of these reagents. Trade names that are useful for broad product
recognition can be a bit confusing. For example, ‘‘Gelcode’’ refers to a
product line that encompasses a range of total and PTM-specific stains,
while ‘‘SYPRO’’ refers to a product line of several total protein gel stains
that are chemically diverse. For the curious, material safety data sheets may
give clues to buffer compositions and the patent literature may allow
gleaning of critical physical or chemical information needed to fully under-
stand a staining method or material composition. Many kits have their basis
on the open literature, and vendors that reveal more rather than less about
the intellectual source and composition of the product are preferred.
Gel staining protocols can vary even within the same staining formula-
tions, depending on the time considerations, and desired quantitative rigor of
detection. Not surprisingly, there is a strong market drive for: (1) instant
gratification and reasonable quantitation and (2) staining formulations with
protocols that avoid strong acids, organic solvents, or heavy metals, due to
environmental and disposal concerns. Thus vendors generally offer ‘‘rapid
protocols’’ for total protein detection formulations. These may involve trun-
cation of fixation or staining steps and, frequently, acceleration of staining or
destaining by judicious heating in a microwave oven. In considering the use
Protein Gel Stains 545

of commercialized stains, it is important to consult product manuals and

company web sites, since products are frequently updated and supported by
several alternate, one hopes, detailed, accurate protocols.

3. Instrumentation: Detection
and Documentation
Obviously, detection by visual inspection has been the basis of protein
gel staining; documentation has been accomplished by photography of
stained gels or by drying stained gels and pasting them into a laboratory
notebook. Unless quantitative data are required, most purification detection
and documentation requirements now can be met with relatively low-cost
flatbed scanners or with camera mounted visible or UV light box stations.
In the past two decades, instrumentation for detection and documentation
(often with quantitation software), driven by fluorescence detection
requirements, has become widely available for both fluorescent and colori-
metric protein detection modalities. Fluorescence detection can be as simple
as visual inspection with a handheld light-emitting diode (LED) ‘‘keychain’’
flashlight or as complex as fully automated multiple mode imaging stations
with linear detection capabilities spanning 4 or 5 orders of magnitude and
correspondingly powerful image analysis and quantitation software for
advanced proteomics applications.
The manufactures of advanced instrumentation generally provide good
information regarding the best acquisition modes and data analysis for
protein colorimetric and fluorescent stains, often providing validated pro-
tocols and examples of applications of commercial or open-source staining
formulations. For fluorescent applications, excitation and emission setting
are frequently identified not only by the wavelength settings, but also by the
names of frequently used fluorescent dyes or products for DNA or protein
stains (e.g., Cy dyes, ethidium bromide, SYBR dyes, SYPRO dyes).
In recent years, new varieties of simple transilluminators with LED or
other inexpensive filtered monochromatic light sources, paired with an
orange-filtering emission cover, offer inexpensive, safe modes of detection
of many of the UV-excitable dyes.

4. Total Protein Detection

4.1. Colorimetric total protein stains
Simplicity of detection by visual inspection, relative ease of use, and wide-
spread familiarity among the user base continue to ensure that staining with
Coomassie Brilliant Blue (CBB) is the most commonly used total protein
546 Thomas H. Steinberg

gel stain, and silver staining continues to be the alternative colorimetric

method, for increased detection sensitivity over Coomassie staining. If mass
spectrometry of excised protein is desired, staining methods that are do not
introduce covalent protein modifications are preferred.

4.1.1. Coomassie Brilliant Blue staining

CBB R-250 and the dimethyl derivative CBB G-250 are disulfonated triphe-
nylmethane dyes that stain protein bands bright blue. The dyes bind via
electrostatic interaction with protonated basic amino acids (lysine, arginine,
and histidine) and by hydrophobic associations with aromatic residues. The
dye does not bind to the polyacrylamide with high affinity but does penetrate
the gel matrix and bind with low affinity, necessitating a destaining step, unless
the stain is a colloidal formulation. Coomassie stains are generally treated as
endpoint stains, allowing significant time flexibility for fixation and staining
steps. Coomassie stains are noncovalent and are reversible and do not interfere
with subsequent mass spectrometry of excised protein bands.
Staining with CBB R-250 has been done with 0.025–0.10% (w/v dye)
in an acidic alcohol formulation: 30–50% (v/v) methanol (or, less
frequently, ethanol) with 7–10% (v/v) acetic acid solution. The dye is
dissolved, and the solution is then filtered through Whatman #1 paper to
produce a stable formulation. Fixation and staining are typically done in
 10 gel volumes of solution, for example, 50 ml for a standard minigel. For
the most sensitive and consistent results, the gel is fixed and stained in the
same acidic alcohol solution; under these conditions, the gel shrinks.
Destaining is in 7% acetic acid, minus the alcohol to return the gel to full
size. Destaining generally requires several solution changes and can be
accelerated by addition of dye-adsorbent paper or foam rubber/plastic.
Thorough analyses of Coomassie staining methods have resulted in
colloidal formulations such that the stain does not effectively enter the gel
matrix while binding specifically to protein bands, allowing low-back-
ground staining with reliable quantitation and increased sensitivity relative
to the standard CBB-R 250 staining method, above. For colloidal staining,
CBB G-250 is preferred to CBB R-250; initial formulations contained
0.1% CBB G-250 in 2% (w/v) phosphoric acid, 6% (w/v) ammonium
sulfate (Neuhoff et al., 1985). An improved, widely used formulation is
prepared by dissolving 10% (w/v) ammonium sulfate in 2% (w/v) phos-
phoric acid followed by addition of CBB G-250 to 0.1% (w/v) from a 5%
(w/v) aqueous stock solution; this is combined with 20% (v/v) methanol
prior to staining (Neuhoff et al., 1988). The addition of methanol may
increase the proportion of monodisperse dye, resulting in more rapid stain-
ing and more intense bands, but with some background; increasing ammo-
nium sulfate concentrations drives the dye toward the colloidal state.
Another, more recent formulation recommends greater dye and phosphoric
acid concentrations in a final formulation of 0.12% (w/v) dye, 10% (w/v)
Protein Gel Stains 547

ammonium sulfate, 10% (w/v) phosphoric acid, and 20% (v/v) methanol in
a single stable staining solution, prepared by sequentially combining desired
amounts of phosphoric acid, ammonium sulfate, and powdered dye in an
aqueous solution to 80% of the desired volume, then adding methanol to
the final volume (Candiano et al., 2004). Colloidal staining solutions should
not be filtered.
Most of the commercially available Coomassie staining formulations are
proprietary derivatives of the various formulations and protocols referenced
above.
Recommendations regarding fixation vary, but good results can be
obtained with a simple procedure in which following electrophoresis the
gel is washed in water to remove most of the SDS placed in the staining
solution for several hours to overnight. Protein bands are visible within the
first few minutes, and background is initially clear, and eventually weakly
blue. Following staining the gel is washed with water to reduce back-
ground. A useful, recent succinct summary of Coomassie staining protocols
includes a ‘‘hot’’ Coomassie staining protocol utilizing an acetic acid-based
CBB R-250 formulation at 90  C or a ‘‘fast’’ colloidal TCA-based CBB
G-250 formulation at 50  C (Westermeier, 2006). Several of the commer-
cially available formulations include simple microwave-based protocols to
accelerate staining (and destaining) with CBB G-250 colloidal formulations.

4.1.2. Silver staining

Silver staining, widely regarded as the standard by which all other ‘‘ultrasensi-
tive’’ staining methods are judged, is the most complex and variable protein
gel staining methodology, with many dozens of published protocols, all of
them requiring several stepwise procedures; the basis of protein detection is
reduction of protein-bound silver ions to metallic silver and, to a lesser extent,
in some protocols, localized deposition of silver sulfide. Silver staining may set
the standard of rigor for ultimate detection sensitivity but quantitation is not a
simple matter, due to the complex nature of the color development step and
difference between proteins for silver signal sensitivity. There are commer-
cially available silver staining kits that variously employ validated protocols
and reagent mixtures from the literature that allow reproducible results for
first-time users and thus can be an important teaching and benchmarking tool
for those who will develop and optimize ‘‘home brew’’ silver staining reagents
and protocols for routine in-house use. Silver staining principles, methods,
and protocols are reviewed and summarized by Merril (1990), Rabilloud
(1990), and Poland et al. (2005).
The key steps are, generally:
1. The gel is fixed to immobilize protein bands and to remove from the gel
interfering substances such as SDS, buffers, and salts that bind silver and
would contribute to background staining.
548 Thomas H. Steinberg

2. The gel is incubated with substances that bind proteins and enhance
silver binding or that interfere with background staining by residual
unbound silver. Taken together, these various strategies are referred to
as sensitization.
3. The gel is impregnated with silver ions.
4. The colorimetric silver stain signal is developed by reduction of bound
silver ions to insoluble and visible metallic silver.
5. Development is stopped, to prevent background staining in the gel matrix
that would obscure protein bands. The time dependency of the latter two
steps in particular contribute to the complexity of the technique, as does
the necessity for several time-dependent washes among steps 2–4.
Silver staining protocols are broadly distinguished on the basis of the
silvering agents and corresponding development conditions. Alkaline meth-
ods use a silver diamine complex in an alkaline environment as a silvering
agent with development in an acidic formaldehyde solution, whereas acidic
methods use weakly acidic silver nitrate as a silvering agent with development
in an alkaline formaldehyde solution. Acidic staining methods have recently
become more popular on the basis of reduced cost and the perception that
background staining is more easily controlled, relative to alkaline methods.
There are a wide variety of sensitizing agents. Aromatic sulfonates such
as naphthalene disulfonate or sulfosalicylic acid bind to proteins and also
bind silver; residual protein-bound SDS may also contribute to increased
silver binding. In a similar vein, CBB-stained gels that are poststained with
silver stain also show enhanced silver staining intensity. Glutaraldehyde is
widely used as a sensitizing agent; the basis of utility is binding to free amino
groups in proteins and thus the introduction of protein-bound reducing
aldehyde groups. Sulfiding reagents such as tetrathionates or thiosulfates can
introduce in close proximity to protein a free S2 ion that immediately
reacts with silver ion to form insoluble silver sulfide.
Mass spectrometry of proteins following silver staining may be problem-
atic due to protein modification resulting from use of glutaraldehyde in
fixation and sensitization steps. Protocols that are compatible with mass
spectrometry omit glutaraldehyde, which may reduce relative sensitivity.
The sensitizing steps rely on tetrathionate and thiosulfate, and the develop-
ment step is also correspondingly extended. In that regard, Colloidal
Coomassie Blue staining followed by a rapid glutaraldehyde-free silver stain-
ing protocol may be a convenient two-step colorimetric approach to evaluate
the purity of a protein sample when subsequent mass spectrometry is desired.

4.1.3. Zn2þ reverse staining

There are several staining methods that use metal cations to rapidly stain
proteins after SDS–PAGE without recourse to fixatives or organic dyes.
Zinc ion ‘‘reverse’’ staining utilizes the ability of proteins and protein–SDS
Protein Gel Stains 549

complexes to bind and sequester Zn2þ in a milieu where the precipitation

reaction between imidazole and zinc ion produces zinc imidazolate (ZnIm2)
to create an opaque background, contrasting with transparent protein–
SDS–Zn2þ zones. The technique is useful as a nonfixative procedure
where subsequent microanalyses such as biological or enzymatic assays of
eluted bands are anticipated and is compatible with mass spectrometry or
microsequencing methods. The method is rapid, with detection sensitivity
reported to rival Coomassie Blue staining.
A rapid protocol suitable for 1 mm thick gels is as follows:
1. Following electrophoresis, the gel is incubated in 0.2 M imidazole, 0.1%
SDS for 15 min.
2. The imidazole solution is discarded and the gel is developed by incuba-
tion in 0.3 M zinc sulfate for 30–45 s.
3. The developer is discarded and the gel is washed several times, with
water, 1 min per wash.
4. The gel is stored in 0.5% (w/v) sodium carbonate.
Overdevelopment can be problematic, but can be corrected by incuba-
tion in 100 mM glycine to dissolve excess ZnIm2. Though the transparent
protein bands against the opaque background do not provide the striking
colorimetric contrast of Coomassie or silver staining, imaging can be done
with the gel illuminated against a black background (Fernandez-Patron,
2005; Fernandez-Patron et al., 1998).

4.2. Fluorescent total protein stains

Fluorescent staining methods can combine detection sensitivity that rivals
silver staining with workflow advantages similar to Coomassie or zinc ion
staining, and offer linear quantitation ranges 10–100-fold greater than the
colorimetric methods. Detection is instrumentation dependent, requiring
a monochromatic excitation light source, selective optical filtration to
separate the longer wavelength emitted light from the shorter wavelength
(and much brighter) excitation light, and a detection mode. For many
fluorescent stains, detection can also be by visual inspection, but with
reduced sensitivity in comparison to photographic or instrumentation
methods. With any fluorescent compound, some degree of photobleach-
ing is a consequence of light exposure. Many of the commercially available
fluorescent stains have been developed to be relatively photostable.
Nevertheless, it is prudent to avoid prolonged exposure to intense ambient
light prior to visual inspection and image acquisition. Excitation and
emission maxima of the fluorescent stains and dyes discussed in this
chapter are presented in Table 31.1. Customarily, and for the purpose of
discussion below, light excitation and emission colors are frequently
distinguished in broad categories: ultraviolet (UV), 250—400 nm; blue,
Table 31.1 Fluorescent protein gel stains

Fluorescent protein gel stain Target Excitation peaks (nm) Emission peak (nm) Vendor
Nile red Total protein 270, 550 600 Sigma-Aldrich
SYPRO Orange Total protein 280, 470 569 Life Technologies
SYPRO Red Total protein 280, 547 631 Life Technologies
SYPRO Tangerine Total protein 280, 490 640 Life Technologies
SYPRO Ruby Total protein 280, 450 610 Life Technologies
Deep Purple Total protein 400, 500 610 GE Healthcare
Krypton Total protein 520 580 Thermo Fisher
Krypton Infrared Total protein 690 718 Thermo Fisher
Flamingo Total protein 270, 512 535 Bio-Rad
LUCY 506 Total protein 505 515 Sigma-Aldrich
LUCY 569 Total protein 569 580 Sigma-Aldrich
C16-FL Total protein 470 530 Life Technologies
Pro-Q Diamond Phosphoprotein 555 580 Life Technologies
Phos-tag 300/460 Phosphoprotein 300, 460 630 Perkin Elmer
Phos-tag 540 Phosphoprotein 540 570 Perkin Elmer
Pro-Q Emerald 300 Glycoprotein 288 533 Life Technologies
Pro-Q Emerald 488 Glycoprotein 512 525 Life Technologies
Krypton glycoprotein Glycoprotein 654 673 Thermo Fisher
Glycoprofile III Glycoprotein 430 480 Sigma-Aldrich
TAMRA alkyne O-GlcNAc 545 580 Life Technologies
Dapoxyl alkyne O-GlcNAc 370 580 Life Technologies
Fluorescent protein gel stains discussed in the text are summarized regarding stain specificity, excitation and emission maxima, and vendor. The web sites are as follows, with
parenthetical indications regarding the division specializing in protein stains, if applicable: Bio-Rad Laboratories, http://www.bio-rad.com; GE Healthcare (Amersham), http://
www.gehealthcare.com; Life Technologies (Invitrogen, Molecular Probes), http://www.lifetechnologies.com; Perkin Elmer, http://www.perkinelmer.com; Sigma-Aldrich,
http://www.sigma-aldrich.com; Thermo Fisher Scientific (Pierce), http://www.thermofisher.com.
Protein Gel Stains 551

400–500 nm; green, 500–550 nm; yellow/orange, 550–580 nm; red,

580–650 nm; near infrared, 650–850 nm.
Fluorescent stains fall into two general categories: (a) fluorogenic stains
that show significant fluorescent enhancement corresponding to localiza-
tion with protein bands and (b) intrinsically fluorescent stains that bind
selectively to protein bands and do not bind to the gel matrix. The first
commercialized fluorescent total protein stains, SYPRO Orange and
SYPRO Red gel stains, were developed in the 1990s in the context of
routine fluorescence detection of stained DNA in gels by 300 nm UV
transillumination followed by Polaroid photography, with the intent of
introducing a simple, one-step total protein-specific staining and documen-
tation work flow similar to ethidium bromide or SYBR Green DNA
staining methods, with detection sensitivity exceeding Coomassie Blue
staining methods. Further development of total protein stains has resulted
in several commercial products or formulations with detection sensitivity
limits matching or exceeding silver staining detection limits. Fluorescent gel
stains that do not require preelectrophoresis covalent sample labeling are
compatible with subsequent mass spectrometry of eluted protein bands.

4.2.1. Nile red protein gel stain

Nile red is a phenoxazone dye that shows strong fluorescence enhancement
upon transition from aqueous to hydrophobic environments such as SDS
micelles or protein–SDS complexes. Nile red does not interact significantly
with SDS monomers. This property has been exploited to develop a rapid,
nonfixative total protein staining method for SDS gels. The protocol relies
upon electrophoresis in nonstandard conditions with running buffer SDS at
0.05% (w/v), below the detergent’s critical micelle concentration, rather
than 0.1% (w/v) SDS typically used in 1D and 2D SDS–PAGE. Protein
sample preparation is standard (as in Garfin, 1990a,b) such that, it is argued,
SDS–protein complexes remain stable during electrophoresis and protein
band migration is considered to be normal. The protocol is simple and
rapid: Nile red is diluted from a stock solution (0.4 mg/ml in DMSO) 200-
fold into water to 2 mg/ml, and a 10-fold volume excess (e.g., 50 ml staining
solution for a 5 ml minigel) is added to a gel with immediate and thorough
agitation. The dye precipitates quickly in water such that staining is optimal
in 2–5 min and does not improve thereafter. Following staining, the gels are
briefly washed with water. Excitation can be with UV light or with green
light sources; protein bands appear pale red. Detection sensitivity is similar
to that obtained with Coomassie stains. Because there is no fixation, Nile
red-stained gels can be subsequently electroblotted with good transfer
efficiency (Daban, 2001; Daban et al., 1991). High fluorescent background
in the gel matrix can be problematic due to dye precipitation, and the
affinity of this stain for SDS micelles combined with dye insolubility
552 Thomas H. Steinberg

precludes use of this after dye SDS–PAGE done with running buffer
containing 0.1% SDS. Photostability can also be problematic.

4.2.2. SYPRO Orange, SYPRO Red, and SYPRO Tangerine

protein gel stains
SYPRO Red and SYPRO Orange protein gel stains are described by
Steinberg et al. (1996a,b) and Haugland et al. (1997); SYPRO Tangerine
protein gel stain is described by Steinberg et al. (2000b) and Yue et al. (2003);
all three stains are also discussed by Steinberg et al. (2005). These dyes contain
a hydrophilic functional group, an aromatic fluorophore, and an aliphatic tail,
facilitating both good aqueous solubility and intercalation into protein–SDS
complexes, SDS micelles, or membranes, with strong fluorescent enhance-
ment in nonpolar environments. These features, combined with good chem-
ical and photostability allow simple, versatile staining protocols following
SDS–PAGE under standard buffer conditions (0.1% SDS) with detection
sensitivity surpassing Colloidal Coomassie Blue G-250. Staining sensitivity
can be increased a further fourfold with nonstandard buffer conditions (0.05%
SDS; cf. Nile red) but that is not a required feature for utility with these dyes.
SYPRO Orange and SYPRO Red, proprietary sulfopropylaminostyryl
dyes, are commercially available as 10 mM stock solutions in DMSO.
The staining protocol is simple:
1. Following SDS–PAGE, the gel is placed in staining solution.
2. Prior to image acquisition, the gel is washed briefly with water.
To prepare a staining solution, the dye is diluted 5000-fold to 2 mM in
dilute acetic acid (7%, v/v is standard; 2–10% is equally effective). The
staining solution typically is prepared fresh, but remains stable for many
months. Following electrophoresis the gel is placed in 10–20 gel volumes
(50–100 ml for a 5-ml volume minigel) of staining solution, typically in a
polypropylene or polycarbonate dish with continuous gentle agitation.
Staining can be monitored periodically by placing the dish—with the gel
remaining in the staining solution—on a UV light box or on a blue-light
transilluminator, or staining can be monitored with a handheld UV light or
a blue LED. For a typical 10 mg cell lysate, fluorescent bands of abundant
proteins can be detected with 10 or 15 min against a fluorescent back-
ground. As staining progresses, protein band signal increases while back-
ground decreases, and the less-abundant proteins become more visible. For
a 1-mm-thick gel, staining is complete in 1 h and is stable for many days
with the gel remaining in the staining solution. It appears that in dilute
acetic acid the protein–SDS–dye complexes are stabilized, while the
remaining SDS diffuses out of the gel, resulting in low background.
Destaining thus is minimal. Prior to imaging acquisition, the gel is washed
with two changes of water for 2–3 min per wash to remove from the gel
surface residual dye, SDS, and acetic acid.
Protein Gel Stains 553

SYPRO Tangerine protein gel stain, a proprietary carbazolylvinyl stain,

is distinguished from SYPRO Orange and SYPRO Red protein gel stains
on the basis of recommended staining diluent and intended utility. This
protein stain can be used in a neutral pH buffered salts diluent—50 mM
phosphate, 150 mM NaCl, pH 7.0—such that stained protein bands are not
fixed and can be subsequently subjected to zymography, elution and rena-
tured for in vitro activity assays, or electroblotted. The dye is commercially
available as a 10 mM stock solution; staining solutions are prepared fresh
with 5000-fold dilution to 2 mM. Staining proceeds as with SYPRO
Orange, and is complete within 1 h; the gel is washed briefly with water
before image acquisition. Excitation is with UV light or a blue light source.

4.2.3. SYPRO Ruby protein gel stain and other Ruthenium-based

formulations
Organometallic ruthenium ion-based luminescent stains for protein detec-
tion in gels and on blots (Bhalget et al., 2001), introduced as SYPRO Ruby
protein gel stains, were developed to provide fluorescent detection sensitiv-
ity rivaling or exceeding silver staining methods with simple end-point
staining protocols, and were introduced in ready-to-use formulations for
staining after SDS–PAGE (Berggren et al., 2000) or after isoelectric focusing
electrophoresis (Steinberg et al. 2000a). The development and formulation
of these stains had its basis in colloidal Coomassie staining strategies whereby
the organic component chelates luminescent ruthenium(II) and also
provides the basis for noncovalent protein in much the same manner as
Coomassie staining primarily by ionic interactions with basic amino acids
and secondary by hydrophobic interactions. An explicit description of a
ruthenium II tris (bathophenanthroline disulfonate) preparation and stain-
ing protocol and comparison to SYPRO Ruby protein gel stain (Rabilloud
et al., 2001) has led to the suggestion that the staining principle in SYPRO
Ruby is a very similar compound; mass spectrometry has revealed some
minor differences, a consequence of proprietary chemistry. A reformulated
SYPRO Ruby gel stain, also a ready-to-use solution, with improved
properties over the original formulation, suitable for both SDS–PAGE
and isoelectric focusing gels, was then developed (Berggren et al., 2002).
Thus SYPRO Ruby is distinguished on the basis of the fluorescent chemi-
cal moiety and, importantly, also on the basis of the staining formulation
that allows a stable ready-to-use solution and a relatively simple protocol:
1. The gel is fixed in 50% (v/v) methanol, 10% (v/v) acetic acid; fixation
can be 30 min to overnight.
2. The gel is stained with SYPRO ruby stain. This is an endpoint stain;
staining can be 3 h to overnight, and is stable.
3. The gel is briefly destained with 10% (v/v) methanol, 7% (v/v) acetic
acid.
554 Thomas H. Steinberg

Staining can be accelerated with a microwave-based protocol. SYPRO

Ruby has a relatively high extinction coefficient and quantum yield, so it is
intrinsically ‘‘bright’’ to the eye and is chemically stable and photostable.
Excitation can be with UV light or with blue light sources; the protein
bands appear orange-red to the eye. The basis for fluorescent protein
staining is differential binding to protein bands relative to the gel matrix,
which does not bind the dye. Thus the protein staining is not fluorogenic
per se. The brightness, stability, detection sensitivity, and ease of use of this
product, combined with extensive literature coverage for proteomics
applications and an aggressive marketing campaign has led to SYPRO
Ruby becoming the comparison standard against subsequently developed
protein gel stains, just as silver stain also remains a detection sensitivity
benchmark.

4.2.4. Epicocconone: Deep Purple, Lighting Fast protein gel stains

Epicocconone, a fluorophore from the fungus Epicoccum nigrum, has been
utilized in total protein staining kits under several trade names, including the
Deep Purple and Lightning Fast total protein gel stains, with several
protocol variations. In one protocol:
1. An SDS–PAGE gel is fixed for 1 h in 7.5% (v/v) acetic acid.
2. The gel is washed with water (2
30 min).
3. The gel is stained for 1 h in an aqueous solution into which an aliquot of
the fluorophore stock solution is diluted.
4. The gel is incubated 0.05% in (v/v) ammonia (3
10 min).
5. Image acquisition is immediate; excitation can be with UV or blue or
green visible light.
The stain sensitivity has been described to rival or exceed SYPRO Ruby
(Bell and Karuso, 2003; Mackintosh et al., 2003). The staining is reversible
and is compatible with mass spectrometry. The dull red signal is not
strikingly bright, and the basis of sensitivity is the fluorogenic nature of
staining, such that background is very low. The staining seems to require
trace amounts of SDS that remain complexed with protein after the acetic
acid fixation. The stain is only weakly fluorescent in water (green) but in the
presence of SDS-treated protein fluorescence increases and is redshifted.
Epicocconone has been shown to react with lysine amines, to form a
fluorescent adduct that can then be hydrolyzed by base, accounting for
the fluorogenic, reversible nature of this product (Coghlan et al., 2005). The
concentrated, staining stock solution must be stored frozen and thawed
before use. Image acquisition should follow immediately upon completion
of staining, as stability may be an issue.
Protein Gel Stains 555

4.2.5. Fluorescein derivatives

Fluorescein derivatives containing hydrocarbon tails have been shown to be
effective protein gel stains. A comparison of 5-dodecanoyl amino-(C12-
FL), 5-hexadecanoyl amino-(C16-FL), and 5-octadecanoyl amino-fluores-
cein (C18-FL), in several simple staining protocols analogous to classic CBB
R-250 protocols, demonstrated staining comparable to silver staining sensi-
tivity. The most effective fluorophore was C16-FL. In one example, gels
were stained in two changes of 30% (v/v) ethanol, 7.5% (v/v) acetic acid,
1 mM dye, washed with two changes of water, and destained with 7.5%
acetic acid. The stain is fluorogenic, apparently on the basis of binding to
residual SDS in a complex with protein; mass spectrometry compatibility
has been demonstrated (Kang et al., 2003). Excitation and emission are in
the blue and green, respectively, as for fluorescein.

4.2.6. Krypton protein gel stains

Krypton protein gel stain is a proprietary hydroxyquinolone formulation
with green excitation and orange fluorescence (Wolf et al., 2007). Krypton
Infrared protein gel stain is a proprietary formulation containing coumar-
ine/chinolone and/or Martina-type dyes with red/near infrared excitation
and near infrared emission (Czerney et al., 2008). These products are
available in 10
formulations that are diluted into a working solution and
have been demonstrated to be competitive with SYPRO Ruby and Deep
Purple protein gel stains for detection sensitivity and dynamic linear range.
The staining protocol follows a standard fix–stain–rapid destain work flow
and have been developed for rapid staining such that staining can be done in
1–4 h overall; for maximum sensitivity and signal linearity the longer
protocol is preferred. These stains are compatible with mass spectrometry.

4.2.7. Flamingo protein gel stain

Flamingo protein gel stain is a proprietary coumarin-based cyanine dye formu-
lation (Berkelman, 2006). The staining protocol requires fixation prior to
staining; the dye is available in a 10
stock solution that is diluted into water
just before use. The stain shows fluorescence enhancement upon binding to
protein and is reported to have a low background that can be further reduced
with a short destaining step. Excitation is in the green range and the signal is
orange/red. The stain has been reported to be competitive with silver stain and
with SYPRO Ruby and is mass spectrometry compatible.

4.2.8. LUCY protein gel stains

LUCY protein gel stains are proprietary trimethincyanine dyes that show
fluorescent enhancement in SDS/protein complexes. Staining is most
frequently done in dilute acetic acid, in same manner as with SYPRO
Orange. The stains are reported to be mass spectrometry compatible
(Kovalska et al., 2006).
556 Thomas H. Steinberg

4.3. Preelectrophoresis sample labeling

The development and use of succinimidyl esters of charge-balanced cyanine
dyes (Cy dyes) for covalent amine labeling of protein samples prior to
electrophoresis has been the basis of fluorescent two-dimensional differential
gel electrophoresis (2-D DIGE), a very important proteomics technology
(Minden et al., 2002; Tonge et al. 2001; Waggoner et al., 1993). DIGE is
discussed in more detail in Chapter 30. There are many succinimidyl esters of
florescent compounds, and other reactive florescent covalent protein labels
are also widely used (Haugland, 2005). With the broad instrumentation base
now available, gel-based analysis of the composition of any fluorescently
labeled protein preparation should become a routine procedure.

5. Phosphoprotein Detection
The importance of reversible protein phosphorylation at selected
amino acid residues as a fundamental cell signaling mechanism is beyond
dispute. Current phosphoprotein stains are proprietary formulations with
phosphate-binding moieties covalently attached to fluorophores. The mode
of detection is selective binding to phosphorylated amino acids, but with no
fluorescent enhancement. Many soluble fluorescent compounds can be, to
some degree, total protein stains. A selective phosphoprotein staining solu-
tion combines a phosphate-binding moiety, a fluorophore that shows low
intrinsic total protein staining, a formulation that suppresses nonspecific,
total protein staining, and a destaining protocol that favors dissociation of
residual nonspecific stain. There are only a few phosphates per polypeptide
relative to the number of amino acids targeted by total protein stains, so
phosphoprotein detection sensitivity is generally at the same protein mass
sensitivity level as detection by total proteins as colloidal Coomassie staining
or SYPRO Orange staining, and not as sensitive as with silver or SYPRO
Ruby stain. Protein phosphorylation determination in gels must always be
done with appropriate controls, including gel lanes displaying proteins
of known positive or negative phosphorylation status and, if possible,
control samples treated with an effective phosphatase. After the phospho-
protein signal is detected, the gel should be poststained with a total
protein stain. The commercially available phosphoprotein stains are
compatible with subsequent staining with Coomassie, Silver, or total
fluorescent stains, such as SYPRO Ruby, but not with SDS-dependent
fluorogenic stains such as SYPRO Orange. The phosphoprotein stains are
compatible with subsequent analytical procedures such as mass spectrometry
or peptide sequencing.
Protein Gel Stains 557

5.1. Pro-Q Diamond phosphoprotein gel stain

Pro-Q Diamond phosphoprotein gel stain (Life Technologies), a ready-
to-use formulation, was developed from principles of immobilized metal
affinity chromatography for phosphopeptides, utilizing a fluorophore-con-
jugated metal chelating moiety in a solution containing a metal ion, salts,
and a water-miscible organic solvent, buffered to pH 4 (Agnew et al., 2006).
The staining protocol requires fixation in methanol/acetic acid, washing
with water to remove the fixative, staining with the Pro-Q Diamond
formulation and subsequent destaining in a mixture containing 1,2-propa-
nediol or, more effectively, acetonitrile (e.g., 50 mM sodium acetate, pH 4,
15–20% 1,2-propanediol, or 15–20% acetonitrile). Detection of phospho-
proteins in the minigel format is of as little as 1–2 ng for b-casein, a
pentaphosphorylated protein, or 8 ng for pepsin, a monophosphorylated
protein, with l000-fold linear dynamic detection range. In vitro protein
phosphorylation or dephosphorylation can be demonstrated with Pro-Q
Diamond staining followed by total protein staining (Steinberg et al., 2003).

5.2. Phos-tag phosphoprotein stains

The Phos-tag phosphoprotein stains (Perkin Elmer), available in kit form,
employ the same general staining strategy, albeit with different buffer com-
positions, as Pro-Q Diamond stain. These stains contain the well-described
Phos-tag-binding moiety, an alkoxide-bridged dinuclear zinc(II) complex;
the basis for the molecular design was from the enzymology of reversible
phosphate–Zn(II) coordination. A Phos-tag-phenyl-phosphate complex
with a dissociation constant of 25 nM has been described, in contrast to
micomolar dissociation constant for other phosphate-binding moieties
(Kinoshita et al., 2004; Koike et al., 2007). Phos-tag technology has been
shown to be useful for gel-based phosphoprotein analysis in other contexts
than direct in-gel staining of phosphoproteins (Kinoshita et al., 2006;
Yamada et al., 2007). Phos-tag gel stain kits are available as two fluorophore
choices, identified by the excitation maxima: Phos-tag phosphoprotein gel
stains 300/460 (dual excitation maxima) or 540. Use of Phos-tag 300/460
stain for detection of a bacterial response regulator aspartate phosphorylation
has recently been reported (Barbieri and Stock, 2008).

6. Glycoprotein Detection
6.1. General glycoprotein detection
Glycosylation is the most frequent protein posttranslational modification in
eukaryotes. Oligosaccharides are usually linked to asparagine side chains
(N-linked glycosylation) or to serine and threonine hydroxyl side chains
558 Thomas H. Steinberg

(O-linked glycosylation). Postelectrophoretic glycoprotein staining in gels

generally utilizes periodate oxidation of vicinal glycol residues followed by
hydrazide conjugation by a Schiff’s base mechanism.

6.1.1. Chromogenic: Acid fuchsin dye

A chromogenic method that uses acid fuchsin dye (Zacharius et al., 1969)
has been the basis of several commercially available kits, including the
Gelcode (Thermo Scientific) and Glyco-Pro (Sigma-Aldrich) glycoprotein
stain kits. The gel is fixed in 50% methanol, washed, incubated in 1%
periodate to oxidize vicinal glycols, washed, incubated in fuchsin sulfite
reagent, incubated in a reducing reagent (sodium metabisulfite or sodium
borohydride), washed, and stored in dilute acetic acid. Glycoproteins are
indicated by magenta bands that begin to appear during the staining reaction
and slowly intensify thereafter.

6.1.2. Fluorescent general glycoprotein stains

Pro-Q Emerald 300 and Pro-Q Emerald 488 glycoprotein gel stain kits
(Life Technologies; the numerals refer to commonly used excitation
sources, respectively), Krypton glycoprotein staining kit (Thermo Scientific
Pierce), and Glycoprofile III fluorescent glycoprotein staining kit (Sigma-
Aldrich) all utilize periodate oxidation followed by conjugation with a
fluorescent hydrazide. With these kit reagents, subsequent reduction, for
example, with sodium metabisulfite, is not needed.
Most available fluorescent hydrazides are intrinsically fluorescent. The
brightly green fluorescent Pro-Q Emerald 300 hydrazide reagent is nonflu-
orescent in solution and is weakly fluorogenic upon conjugation to small
molecule aldehydes. The stained glycoprotein bands are bright green; the
conjugated dye may act as nucleation sites for the deposition of free reagent,
which is in solution at near saturating concentration (Haugland et al., 2005).
The utility of Pro-Q Emerald 300 and Pro-Q Emerald 488 glycoprotein
staining kits has been described in some detail (Hart et al., 2003; Steinberg
et al., 2001). In these studies, Pro-Q Emerald 300 glycoprotein staining was
found to be the most sensitive selective glycoprotein stain, but is suited for
use only with a UV light source. Data and observations in both studies
emphasize that many fluorescent hydrazides exhibit nonspecific staining of
total proteins. Moreover, selectivity is also strongly influenced by the
composition of the staining diluent. It is helpful to include a control lane
containing proteins of known glycosylation state, by choosing well-
characterized proteins and/or by glycosidase digestion. A parallel gel that
is not subject to periodate oxidation may also reveal either intrinsic nonspecific
protein staining or staining of endogenous aldehydes or ketones due to
other oxidative events. Finally, it may be noted that acidic fuchsin sulfite-
stained protein bands could be detected by fluorescence scanning with a
Protein Gel Stains 559

532-nm laser and a 675-nm long-pass filter, with a two- to fourfold increase
in sensitivity over the observable colorimetric limit.

6.2. O-GlcNAc detection

6.2.1. Azide–alkyne ‘‘click chemistry’’ reagents
Many proteins are dynamically modified by attachment of O-linked N-acet-
ylglucosamine (O-GlcNAc) in a b-linkage to serine and threonine residues.
This PTM has shown site-specific reciprocity to phosphorylation in a regulated
manner. O-GlcNAcylation thus may have a dynamic ‘‘Yin–Yang’’ role med-
iating phosphorylation-modulated signaling as well as other specific protein
signaling functions (Golks and Guerini, 2008; Hart, 1999). Historically, gel-
based O-GlcNAc detection has been problematic, but sensitive O-GlcNAc-
specific fluorescent labeling for detection in gels now can be accomplished.
The method presented has its basis in the copper(I)-catalyzed Huisgen
[3 þ 2] cycloaddition between azides and alkynes, a ‘‘click chemistry’’ reac-
tion (Agnew et al., 2008; Clark et al., 2008). Proteins (at any stage of purity)
are labeled enzymatically (overnight, 4 C) with the Click-iT O-GlcNAc
Enzymatic Labeling System (Life Technologies) utilizing a permissive
mutant b-1,4-galactosyltransferase (Gal-T1 Y289L) that transfers azido-
modified galactose (GalNAz) from UDP-GalNAz to O-GlcNac residues
on the target proteins (Ramakrishnan and Qasba, 2002). The azide-modified
O-GlcNAcylated proteins are then fluorescently labeled by the Cu(I)-cata-
lyzed click reaction with tetramethylrhodamine–alkyne (TAMRA–alkyne)
or with Dapoxyl alkyne dyes, available with the catalytic reagents in the
Click-iT TAMRA or Click-iT Dapoxyl protein analysis detection kits,
respectively (Life Technologies). The labeled proteins are separated by
SDS–PAGE and image acquisition is done with appropriate instrumentation.
The enzymatic labeling kit contains a positive control protein, a-crystalline.
The technique has been demonstrated to detect 10–40 fmol of O-GlcNAc in
a minigel format with the TAMRA alkyne detection reagent, and is com-
patible with downstream mass spectrometry analyses. Labeled proteins can
also be poststained with a total protein stain or with other PTM-specific
stains such as Pro-Q Diamond phosphoprotein stain or Pro-Q Emerald 300
glycoprotein stain.

REFERENCES
Agnew, B., Beechem, J., Gee, K., Haugland, R., Liu, J., Martin, V., Patton, W., and
Steinberg, T. (2006). Compositions and methods for detection and isolation of phosphorylated
molecules. United States Patent 7102005.
Agnew, B., Buck, S., Nyberg, T., Bradford, J., Clarke, S., and Gee, K. (2008). Click
chemistry for labeling and detection of biomolecules. Proc. SPIE 6867, 686708.
560 Thomas H. Steinberg

Barbieri, C. M., and Stock, A. M. (2008). Universally applicable methods for monitoring
response regulator aspartate phosphorylation both in vitro and in vivo using Phos-tag base
reagents. Anal. Biochem. 376, 73–82.
Bell, P. J. L., and Karuso, P. (2003). Fluorescent compounds. United States Patent Applica-
tion 20030157518.
Berggren, K., Chernokalskaya, E., Steinberg, T. H., Kemper, C., Lopez, M. F., Diwu, Z.,
Haugland, R. P., and Patton, W. F. (2000). Background-free, high sensitivity staining of
proteins in one- and two-dimensional sodium dodecyl sulfate-polyacrylamide gels using
a luminescent ruthenium complex. Electrophoresis 21, 2509–2521.
Berggren, K. N., Schulenberg, B., Lopez, M. F., Steinberg, T. H., Bogdanova, A.,
Smejkal, G., Wang, A., and Patton, W. F. (2002). An improved formulation of
SYPRO Ruby protein gel stain: Comparison with the original formulation and with a
ruthenium II tris (bathophenanthroline disulfonate) formulation. Proteomics 2, 486–498.
Berkelman, T. R. (2006). Coumarin-based cyanine dyes for non-specific protein binding. United
States Patent Application 20060166368.
Bhalget, M. K., Diwu, Z., Haugland, R. P., and Patton, W. F. (2001). Luminescent protein
stains and their method of use. United States Patent 6316267.
Candiano, G., Bruschi, M., Musante, L., Santucci, L., Ghiggeri, G. M., Carnemolla, B.,
Orecchia, P., Zardi, L., and Righetti, P. G. (2004). Blue silver: A very sensitive colloidal
Coomassie G-250 staining for proteome analysis. Electrophoresis 25, 1327–1333.
Clark, P. M., Dweck, J. F., Mason, D. E., Hart, C. R., Buck, S. B., Peters, E. C.,
Agnew, B. J., and Hsieh-Wilson, L. C. (2008). Direct in-gel fluorescence detection
and cellular imaging of O-GlcNAc-modified proteins. J. Am. Chem. Soc. 130,
11576–11577.
Coghlan, D. R., Mackintosh, J. A., and Karuso, P. (2005). Mechanism of reversible staining
of protein with epicocconone. Org. Lett. 7, 2401–2404.
Czerney, P. T., Desai, S., Lehmann, F. G., Murtaza, Z. S., Schweder, B. G., Wenzel, M. S.,
and Wolf, B. D. (2008). Protein probe compounds, compositions, and methods. United States
Patent Application 20080026478.
Daban, J.-R. (2001). Fluorescent labeling of proteins with Nile red and 2-methoxy-2,
4-diphenyl-3(2H)-furanone: Physicochemical basis and application to the rapid staining
of sodium dodecyl sulfate polyacrylamide gels and Western blots. Electrophoresis 22,
874–880.
Daban, J.-R., Bartolome, S., and Samso, M. (1991). Use of the hydrophobic probe Nile red
for the fluorescent staining of protein bands in sodium dodecyl sulfate-polyacrylamide
gels. Anal. Biochem. 199, 169–174.
Dunbar, B. S., Kimura, H., and Timmons, T. M. (1990). Protein analysis using high-
resolution two-dimensional polyacrylamide gel electrophoresis. In ‘‘Meth. Enzmyol.’’
(M. P. Deutscher, ed.) 182, pp. 441–459. Academic Press, San Diego, CA.
Fernandez-Patron, C. (2005). Zn2þ reverse staining technique. In ‘‘The Proteomics Proto-
col Handbook’’ ( J. M. Walker, ed.), pp. 215–222. Humana Press, Totowa, NJ.
Fernandez-Patron, C., Castellanos-Serra, L., Hardy, E., Guerra, M., Estevez, E., Mehl, E.,
and Frank, R. W. (1998). Understanding the mechanism of the zinc-ion stains of
biomacromolecules in electrophoresis gels: generalization of the reverse-staining tech-
nique. Electrophoresis 19, 2398–2406.
Garfin, D. E. (1990a). One-dimensional gel electrophoresis. In ‘‘Meth. Enzmyol.’’
(M. P. Deutscher, ed.) 182, pp. 425–441. Academic Press, San Diego, CA.
Garfin, D. E. (1990b). Isoelectric focusing. In ‘‘Meth. Enzmyol.’’ (M. P. Deutscher, ed.)
182, pp. 459–478. Academic Press, San Diego, CA.
Golks, A., and Guerini, D. (2008). The O-linked N-acetylglucosamine modification in
cellular signaling and the immune system. EMBO Rep. 9, 748–753.
Protein Gel Stains 561

Hart, G. W. (1999). The O-GlcNAc modification. In ‘‘Essentials of Glycobiology’’

(A. Varki, J. Esko, H. Freeze, G. Hart, and J. Marth, eds.), 2nd edn. p. 653. Cold Spring
Harbor Laboratory Press, Cold Spring Harbor, NY.
Hart, C., Schulenberg, B., Steinberg, T. H., Leung, W.-Y., and Patton, W. F. (2003).
Detection of glycoproteins in polyacrylamide gels and on electroblots using Pro-Q
Emerald 488 dye, a fluorescent periodate Schiff-base stain. Electrophoresis 24, 588–598.
Haugland, R. P. (2005). The Handbook: A Guide to Fluorescent Probes and Labeling Technologies.
10th edn. Invitrogen, Carlsbad, CA, 1126pp.
Haugland, R. P., Singer, V. L., Jones, L. J., and Steinberg, T. H. (1997). Non-specific protein
staining using merocyanine dyes. United States Patent 5616502.
Haugland, R. P., Steinberg, T. H., Patton, W. F., and Diwu, Z. (2005). Reagents for labeling
biomolecules having aldehyde or ketone moieties. United States Patent 6967251.
Jessani, N., and Cravatt, B. F. (2004). The development and application of methods for
activity-based protein profiling. Curr. Opin. Chem. Biol. 8, 54–59.
Kang, C., Kim, H. J., Kang, D., Jung, D. Y., and Suh, M. (2003). Highly sensitive and
simple fluorescence staining of proteins in sodium dodecyl sulfate-polyacrylamide-based
gels by using hydrophobic tail-mediated enhancement of fluorescein luminescence.
Electrophoresis 24, 3297–3304.
Kinoshita, E., Takahashi, M., Takeda, H., Shiro, M., and Koike, T. (2004). Recognition of
the phosphate monoester dianion by an alkoxide-bridged dinuclear zinc(II) complex.
Dalton Trans. 1189–1193.
Kinoshita, E., Kinoshita-Kikuta, E., Takiyama, K., and Koike, T. (2006). Phosphate-
binding tag, a new tool to visualize phosphorylated proteins. Mol. Cell. Proteomics 5,
749–757.
Koike, T., Kawasaki, A., Kobashi, T., and Takahagi, M. (2007). Method for labeling phos-
phorylated peptides, method for selectively adsorbing phosphorylated peptides, complex compounds
used in the methods, process for producing the complex compounds, and raw material compounds for
the complex compounds. United States Patent 7202093.
Kovalska, V., Kryvorotenko, D., Losytskyy, M., Nording, P., Rueck, A.,
Schoenenberger, B., Yarmoluk, S., and Wahl, F. (2006). Detection of polyamino acids
using trimethincyanine dyes. United States Patent Application 20060207881.
Mackintosh, J. A., Choi, H.-Y., Bae, S.-H., Veal, D. A., Bell, P. J., Ferrari, B. C., Van
Dyk, D. D., Verrills, N. M., Paik, Y.-K., and Karuso, P. (2003). A fluorescent natural
product for ultrasensitive detection of proteins in one-dimensional and two-dimensional
gel electrophoresis. Proteomics 3, 2273–2288.
Manchenko, G. P. (2003). Handbook of Detection of Enzymes on Electrophoretic Gels. 2nd edn.
CRC Press, LLC, Boca Raton, xivþ554pp.
Merril, C. R. (1990). Gel-staining techniques. In ‘‘Meth. Enzmyol.’’ (M. P. Deutscher, ed.)
182, pp. 477–488. Academic Press, San Diego, CA, xxixþ894pp.
Miller, I., Crawford, J., and Gianazza, E. (2006). Protein stains for proteomic applications:
Which, when, why? Proteomics 6, 5385–5408.
Minden, J., Waggoner, A., and Fowler, S. J. (2002). Difference detection methods using matched
multiple dyes. United States Patent Application 20020177122.
Neuhoff, V., Stamm, R., and Hansjorg, E. (1985). Clear background and highly sensitive
protein staining with Coomassie Blue dyes in polyacrylamide gels: A systematic analysis.
Electrophoresis 6, 427–488.
Neuhoff, V., Arold, N., Taube, D., and Ehrhardt, W. (1988). Improved staining in polyacryl-
amide gels including isoelectric focusing gels with clear background at nanogram sensitivity
using Coomassie Brilliant Blue G-250 and R-250. Electrophoresis 9, 255–262.
Patton, W. F. (2000). A thousand points of light: The application of fluorescence detection
technologies to two-dimensional gel electrophoresis and proteomics. Electrophoresis 21,
1123–1144.
562 Thomas H. Steinberg

Patton, W. F. (2002). Detection technologies in proteome analysis. J. Chromatogr. B 771,

3–31.
Paulick, M. G., and Bogyo, M. (2008). Application of activity-based probes to the study of
enzymes involved in cancer progression. Curr. Opin. Genet. Dev. 18, 97–106.
Poland, J., Rabilloud, T., and Sinha, P. (2005). Silver staining of 2-D gels. In ‘‘The
Proteomics Protocol Handbook’’ ( J. M. Walker, ed.), pp. 215–222. Humana Press,
Totowa, NJ.
Rabilloud, T. (1990). Mechanisms of protein silver staining in polyacrylamide gels:
A 10-year synthesis. Electrophoresis 11, 785–794.
Rabilloud, T., Strub, J. M., Luche, S., van Dorsselaer, A., and Lunardi, J. (2001).
A comparison between Sypro Ruby and ruthenium II tris (bathophenanthroline
disulfonate) as fluorescent stains for protein detection in gels. Proteomics 1, 699–704.
Ramakrishnan, B., and Qasba, P. K. (2002). Structure-based design of beta-1, 4-galactosyl-
transferase-I (beta 4Gal-T1) with equally efficient N-acetylgalactosaminyltransferase
activity: Point mutation broadens beta 4Gal-T1 donor specificity. J. Biol. Chem. 277,
20833–20839.
Smejkal, G. B. (2004). The Coomassie Chronicles: Past, present and future perspectives in
polyacrylamide gel staining. Expert Rev. Proteomics 1, 381–387.
Steinberg, T. H., Jones, L. J., Haugland, R. P., and Singer, V. L. (1996a). SYPRO orange
and SYPRO red protein gel stains: One-step fluorescent staining of denaturing gels for
detection of nanogram levels of protein. Anal. Biochem. 239, 223–237.
Steinberg, T. H., Haugland, R. P., and Singer, V. L. (1996b). Applications of SYPRO
orange and SYPRO red protein gel stains. Anal. Biochem. 239, 238–245.
Steinberg, T. H., Chernokalskaya, E., Berggren, K., Lopez, M. F., Diwu, Z.,
Haugland, R. P., and Patton, W. F. (2000a). Ultrasensitive fluorescence protein detec-
tion in isoelectric focusing gels using a ruthenium metal chelate stain. Electrophoresis 21,
486–496.
Steinberg, T. H., Lauber, W. M., Berggren, K., Kemper, C., Yue, S., and Patton, W. F.
(2000b). Fluorescence detection of proteins in sodium dodecyl sulfate-polyacrylamide
gels using environmentally benign, nonfixative, saline solution. Electrophoresis 21,
497–508.
Steinberg, T. H., Pretty On Top, K., Berggren, K., Kemper, C., Jones, L., Diwu, Z.,
Haugland, R. P., and Patton, W. F. (2001). Rapid and simple single nanogram detection
of glycoproteins in polyacrylamide gels and on electroblots. Proteomics 1, 841–855.
Steinberg, T. H., Agnew, B. J., Gee, K. R., Leung, W.-Y., Goodman, T., Schulenberg, B.,
Hendrickson, J., Beechem, J. M., Haugland, R. P., and Patton, W. F. (2003). Global
quantitative phosphoprotein analysis using Multiplexed Proteomics technology. Proteo-
mics 3, 1128–1144.
Steinberg, T. H., Hart, C. R., and Patton, W. F. (2005). Rapid, sensitive detection of
proteins in minigels with fluorescent dyes: Coomassie Fluor Orange, SYPRO Orange,
SYPRO Red, and SYPRO Tangerine stains. In ‘‘The Proteomics Protocol Handbook’’
( J. M. Walker, ed.), pp. 215–222. Humana Press, Totowa, NJ.
Tonge, R., Shaw, J., Middletown, B., Rowlinson, R., Rayner, S., Young, J., Pognan, F.,
Hawkins, E., Currie, I., and Davison, M. (2001). Validation and development of
fluorescence two-dimensional differential gel electrophoresis proteomics technology.
Proteomics 1, 377–396.
Waggoner, A. S., Ernst, L. A., and Mujumdar, R. B. (1993). Method for labeling and detecting
materials employing arylsulfonate cyanine dyes. United States Patent 5268486.
Walker, J. M. (ed.) (2005). In ‘‘The Proteomics Protocols Handbook’’, Humana Press,
Totowa, NJ, xviiþ988pp.
Westermeier, R. (2006). Sensitive, quantitative, and fast modifications for Coomassie blue
staining of polyacrylamide gels. Pract. Proteomics 6(Suppl. 2), 61–64 (online journal).
Protein Gel Stains 563

Wolf, B.D, Desai, S., Czerney, P. T., Lehmann, F. G., Schweder, B. G., and Wenzel, M. S.
(2007). Protein detection and quantitation using hydroxyquinolone dyes. United States Patent
Application 20070281360.
Yamada, S., Nakamura, H., Kinoshita, E., Kinoshita-Kikuta, E., Koike, T., and Shiro, Y.
(2007). Separation of a phosphorylated histidine protein using phosphate affinity
polyacrylamide gel electrophoresis. Anal. Biochem. 360, 160–162.
Yue, S. T., Steinberg, T. H., Patton, W. F., Cheung, C.-Y., and Haugland, R. P. (2003).
Carbazolylvinyl dye protein stains. United States Patent 6579718.
Zacharius, R. M., Zeli, T. E., Morrison, J. H., and Woodlock, J. J. (1969). Glycoprotein
staining following electrophoresis on acrylamide gels. Anal. Biochem. 30, 148–152.
 C H A P T E R T H I R T Y- T W O

Elution of Proteins from Gels

Richard R. Burgess

Contents
1. Introduction 565
2. Elution of Proteins from Gels by Diffusion 566
2.1. Preparing the gel and protecting the protein 567
2.2. Locating protein in the gel 567
2.3. Elution by diffusion 568
2.4. SDS removal and concentration 568
2.5. Renaturation of enzyme activity 569
2.6. Preliminary tests on your enzyme 569
2.7. Limitations of the method 569
2.8. Many applications 569
3. Replacing the SDS Gel with a Reverse Phase HPLC 570
4. Electrophoretic Elution 570
5. Conclusion 571
References 571

Abstract
New protein biochemical technologies have been developed that allow one to
learn more and more about a protein with less and less material (on the order of
mg). Polyacrylamide gel electrophoresis, which was originally strictly an analyt-
ical method, has in some cases become a high-resolution preparative method.
This chapter will focus on the elution of proteins from SDS gels, with an
emphasis on recovering enzyme or biological activity of the resulting protein.

1. Introduction
SDS polyacrylamide gel electrophoresis has proved to be an incredibly
useful analytical method for determining the number and sizes of polypep-
tides in a sample. When performed skillfully it has the ability to resolve
many individual sized proteins. Many of us in the early days of gel

McArdle Laboratory for Cancer Research, University of Wisconsin-Madison, Madison, Wisconsin, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63032-9 All rights reserved.

565
566 Richard R. Burgess

electrophoresis wished we could take advantage of the high resolution of

the gels to obtain small quantities of pure proteins. Many methods have
been developed that allow us now to do this. They include:
1. Transfer of proteins out of the gel and onto a nitrocellulose or PVDF membrane.
Subsequent detection and quantification of the protein can be achieved
by Western blot analysis with an appropriate antibody (see Chapter 33).
Sometimes one can renature at least some of the protein to regain
enzymatic function or the ability to bind another protein by far-Western
analysis (e.g., see Burgess et al., 2000). Edman degradation of the protein
on the membrane can give limited N-terminal amino acid sequence
information.
2. Enzyme assays of proteins while still in the gel matrix. This has been covered
extensively (Manchenko, 2003).
3. Preparative gel electrophoresis. In this procedure the proteins are electro-
phoresed off the end of a gel and into a chamber where fractions can be
collected. Several commercial products for preparative electrophoresis
are now available.
4. Elution of the protein from the gel and recovery of the protein. This approach has
been the subject of several extensive reviews (Harrington, 1990; Seelert
and Krause, 2008). Proteins eluted from gels have been used for a wide
variety of purposes including: protein chemistry, proteolytic cleavage;
amino acid composition, and sequence determination; identifying a poly-
peptide by trypsin digestion and MALDI-TOF mass spectroscopy; antigen
for antibody production; and identifying a polypeptide corresponding
to an enzyme activity.
This chapter will focus on this fourth approach, that of eluting protein
from SDS gels, particularly on recovering enzymatic or biological activity of
the eluted protein.

2. Elution of Proteins from Gels by Diffusion

The early work on removal of SDS from protein was published by
Weber and Kutter (1971), but when applied to microgram amounts of
protein eluted from gels it proved cumbersome and usually resulted in loss
of most of the protein.
Much of the early work in this area has been published (Hager and
Burgess, 1980), but bears revisiting. The basic procedure involves the gel
electrophoresis itself, locating the protein of interest on the gel, eluting the
protein from the gel, removing SDS, and renaturing the protein for
subsequent study or use.
Elution of Proteins from Gels 567

2.1. Preparing the gel and protecting the protein

To the first approximation, any of a large variety of gel recipes can be used.
There are several important considerations. The polymerization step of
polyacrylamide gels involves the generation of free radicals to initiate the
polymerization reaction. Therefore, if you are pouring your own gels you
should wait at least 12 h to allow complete polymerization. Since the
initiation of the reaction involves creating an oxidizing environment, one
should take precautions to protect the protein against oxidation. This can
readily be accomplished by adding to the sample applied to the gel an
internal carrier protein such as b-lactoglobulin that is inexpensive, small,
and runs ahead of most proteins and that will take the brunt of any oxidizing
chemicals remaining in the gel. You should also add an anionic thiol
compound to the upper reservoir buffer (we use 0.1 mM sodium thiogly-
colate) that will run through the gel, again destroying any oxidizing poten-
tial and residual free radicals within the gel. The presence of the thiol in the
gel may also help reduce the reaction of protein cysteines with free
acrylamide to produce cysteinyl-S-propionamide (Chiari et al., 1992).
You should use high-quality SDS, containing only the dodecyl form
(C12), because impure SDS can contain variable amounts of C14 or C16
that bind much more tightly to proteins and can be very difficult to remove
(Kunitani and Kresin, 1989).

2.2. Locating protein in the gel

Once the gel has been run, one needs to locate the region of the gel that
contains the band of interest or the band of a particular MW. Of course, if
you do not know the size of your enzyme band, you can cut the gel into
sections, elute protein from each section, and assay all for enzyme activity.
Originally, we located our band of interest by soaking the gel for 5 min in
cold 0.25 M KCl and then destaining with cold distilled water for 1 h. Now
we know that you can use almost any convenient way to stain the gel,
including with zinc-imidazole (Castellanos-Serra and Hardy, 2001), or with
new very sensitive fluorescent protein dyes like SYPRO Ruby (see Chapter
31). You can even use Coomassie Blue stain. In one case, we succeeded in
recovering activity from a band that was stained with Coomassie, destained,
dried down, and stored for 10 years! Another method is to run colored MW
markers (available from many companies) in flanking lanes of the gel. That
way you can be guided to cut out particular sections of the gel based on
prior molecular weight information.
568 Richard R. Burgess

2.3. Elution by diffusion

People have tended to ignore this simple procedure that is inexpensive,
effective, and can be done on many samples at the same time. A typical
protocol below is based on and adapted from Hager and Burgess (1980).
1. Locate a band on an SDS gel by rinsing the gel with cold ddH2O and
staining for 5 min with ice-cold 0.25 M KCl and 1 mM DTT. Rinse
with ddH2O and destain for 10–60 min with cold ddH2O and 1 mM
DTT. Alternatively, one can divide a gel lane into many 3–5 mm
sections, put each section of gel into a tube, and wash as above.
2. Crush the gel slice in 1 ml elution buffer. We originally used a 3.5-ml
siliconized Pyrex test tube (10  75 mm) and a small Teflon pestle
(Kontes, K886001 size 19), but now use 0.3 ml of elution buffer and a
1.5-ml polypropylene microfuge tube and a disposable polypropylene
pestle (such as Kontes pellet pestle, K749521-1500). The elution buffer
is 50 mM Tris, pH 7.9, 0.1 mM EDTA, 1 mM DTT, 0.15 M NaCl,
25–100 mg/ml BSA (can be omitted), and 0.1% SDS.
3. Once the gel section has been crushed, one incubates the small gel
fragments for 1–8 h on a rotator to allow passive diffusion. In Hager
and Burgess (1980), the elution kinetics indicate half lives of elution
from a 8.75% polyacrylamide gel matrix to be less than 30 min for a
36-kDa polypeptide and 1–1.5 h for 150 kDa polypeptides (complete
elution in 4 and 16–24 h, respectively).
4. The mixture is centrifuged at maximum speed in a microfuge for 2 min
to pellet the crumbled gel. The supernatant (the protein eluate) is then
transferred into a clean microfuge tube.

2.4. SDS removal and concentration

Since elution is most efficient when 0.1% SDS is present in the elution
buffer, one must remove SDS after elution. This can be most effectively
accomplished by acetone precipitation that not only removes the SDS but
also concentrates the protein. A typical protocol below is based on Hager
and Burgess (1980).
1. Four volumes of cold acetone ( 20  C) are added to the protein eluate
and the sample allowed to precipitate for 30 min in a dry ice–ethanol
bath. Since the samples freeze in the dry ice–ethanol bath, we thaw the
samples by placing them briefly in an ice water bath just before
centrifugation.
2. Centrifuge at maximum speed in a microfuge for 5 min. The supernatant
is removed and discarded. In test experiments with 0.16 mg of radioac-
tive RNA polymerase, we found that our recovery of polymerase was
Elution of Proteins from Gels 569

71%, 90%, and 99% when carrier BSA concentration in the elution
buffer was 0, 15, and 100 mg/ml, respectively (Hager and Burgess, 1980).
3. Tests showed that more than 99.9% of the SDS remained in the acetone
supernatant. The residual SDS can be removed by washing the pellet
briefly with 1 ml of ice-cold 80% acetone and recentrifuging.

2.5. Renaturation of enzyme activity

1. The acetone precipitate is allowed to dry for about 10 min.
2. The precipitate is solubilized and denatured in 20 ml of 6 M guanidine
hydrochloride (GuHCl) in dilution buffer (dilution buffer is 50 mM
Tris, pH 7.9, 20% glycerol, 0.1 mM EDTA, 1 mM DTT, 0.15 M NaCl,
25–100 mg/ml BSA (can be omitted), and 0.1% SDS). Solubilization
takes place within 20 min at room temperature.
3. The solubilized protein is rapidly diluted 50-fold by adding 1.0 ml of
dilution buffer and permitted to renature 1–12 h at room temperature.
4. The renatured protein is assayed by an appropriate assay.

2.6. Preliminary tests on your enzyme

One can test the ability of a given protein to be renatured by this method by
subjecting the enzyme of interest to preliminary testing to see whether it is
able to regain activity after dissolving in 6 M GuHCl and diluting. If recovery
is good, then test for recovery of activity after denaturing in SDS elution
buffer, acetone precipitating, dissolving in 6 M GuHCl, and diluting.

2.7. Limitations of the method

While this method is quite versatile, it cannot be used for all enzymes.
It cannot be used if the enzyme activity is dependent on two or more
different sized polypeptides or requires a dissociable cofactor such as heme
that will be separated from the catalytic protein during the SDS gel electro-
phoresis. It might also be difficult if the protein contains posttranslational
modifications such as glycosylation or proteolytic processing that affect
refolding or if there are essential disulfides that are difficult to reform.

2.8. Many applications

Even given the limitation noted above, many proteins have been success-
fully renatured using this method. This includes DNA topoisomerases,
DNA ligase, bacterial sigma factors (Haldenwang et al., 1981; Wiggs et al.,
1981), eukaryotic transcription factors, H-Ras GTPase, methyl reductases,
centromere binding protein, and the protein component of the pro-
tein/RNA RNases. It has been used with enzymes from bacteria, yeast,
570 Richard R. Burgess

rat, human, Drosophila, and many other organisms. It has been used with
enzyme containing multiple identical subunits and with proteins with
multiple different subunits if the appropriate gel slices have been mixed
together. It has been used with proteins that have essential S–S bridges.

3. Replacing the SDS Gel with a Reverse

Phase HPLC
A clever variation of the method of Hager and Burgess described
above was published by Prokipcak et al. (1994). They replaced the SDS
gel and acetone precipitation with RP-HPLC and lyophilization. The
protein was applied to a C4 RP-HPLC, and eluted with a gradient of
0–50% acetonitrile/0.1% TFA. The fractions were lyophilized, resuspended
in a small volume of 6 M GuHCl, diluted to refold, and assayed for
enzymatic activity.

4. Electrophoretic Elution
While the above protocols describe in detail the elution of protein
from gels by diffusion, many of the operations would be perfectly applicable
to electrophoretic elution. While these approaches may result in slightly
higher elution efficiency in slightly shorter time than elution by diffusion,
they are more expensive and harder to scale up to many gel sections.
The Schleicher and Schuell ElutrapTM and the Bio-Rad Model 422 Electro-
eluterTM are the two most widely used commercial electrophoretic elution
apparati. Both involve the placing of a section of a gel containing a protein
band of interest into the apparatus and eluting the protein by electrophoresis
of the protein out of the gel piece, through a large-pore membrane or frit
and into a chamber with a small-pore membrane that will only pass small
molecules or proteins smaller than about 5 kDa. The protein of interest is
pipetted out of the chamber and used as desired. The detailed methods for
using these commercial products can be found in excellent reviews by
Harrington (1990) and Seelert and Krause (2008) and in company literature.
Another commercial approach is the ProteoPLUSTM electroeluter. A small
screw capped tube is constructed with upstream and downstream protein
retention membranes that permit the passage of electric current when the
tubes are placed in an electrophoresis tank. After the protein has eluted from
the gel piece, it can be dialyzed in the same tube for subsequent use.
An example of the application of this approach is given by Lei et al. (2007).
The Bio-Rad Whole Gel Eluter allows one to electrophoretically transfer a
whole slab gel into 26 fractions. Typically a protein sample is applied into
Elution of Proteins from Gels 571

one gel-wide lane and subjected to SDS gel electrophoresis. The 26 grooves
of the Eluter are aligned in parallel to the protein bands and the protein
transferred out of the gel, into the grooves and into the harvesting box.
Another recent advance includes the ability to elute a whole slab gel
into numerous individual fractions suitable for proteomic applications in
multiwall plates (Antal et al., 2007).

5. Conclusion
Protein bands separated by high-resolution gel electrophoresis can be
recovered from the gel and used for a multitude of applications. The protein
can be eluted out of the gel section by crushing the gel and allowing the
protein to diffuse out of the gel or it can eluted by applying an electric field
to the gel section and trapping the eluted protein in an appropriate
membrane bounded apparatus. Submicrogram to 100 mg amounts of
protein can be obtained and often renatured to regain enzymatic activity.

REFERENCES
Antal, J., Banyasz, B., and Buzas, Z. (2007). Shotgun electrophoresis: A proteomic tool for
simultaneous sample elution from whole SDS-polyacrylamide gels slabs. Electrophoresis
28, 508–511.
Burgess, R. R., Arthur, T. M., and Pietz, B. C. (2000). Mapping protein–protein interaction
domains by order fragment ladder far-Western analysis using His-tagged proteins. Meth.
Enzymol. 328, 141–157.
Castellanos-Serra, L., and Hardy, E. (2001). Review. Detection of biomolecules in electro-
phoresis gels with salts of imidazole and zinc II: A decade of research. Electrophoresis 22,
864–873.
Chiari, M., Righetti, P. G., Negri, A., Ceciliani, F., and Ronchi, S. (1992). Preincubation
with cysteine prevents modification of sulfhydryl groups in proteins by unreacted
acrylamide in a gel. Electrophoresis 13, 882–884.
Hager, D. A., and Burgess, R. R. (1980). Elution of proteins from sodium dodecyl sulfate-
polyacrylamide gels, removal of SDS, and renaturation of enzymatic activity: Results
with sigma subunit of E. coli RNA polymerase, wheat germ DNA topoisomerase, and
other enzymes. Anal. Biochem. 109, 76–86.
Haldenwang, W. G., Lang, N., and Losick, R. (1981). A sporulation-induced sigma-like
regulatory protein from B. subtilis. Cell 23, 615–624.
Harrington, M. G. (1990). Elution of protein from gels. Meth. Enzymol. 182, 488–495.
Kunitani, M. G., and Kresin, L. M. (1989). Analysis of alkyl sulfates in protein solutions by
isocratic and gradient ion chromatography. Anal. Biochem. 182, 103–108.
Lei, Z., Anand, A., Mysore, K. S., and Sumner, L. W. (2007). Electroelution of intact
proteins from SDS-PAGE gels and their subsequent MALDI-TOF MS analysis. Methods
Mol. Biol. 355, 353–363.
Manchenko, G. P. (2003). Handbook of detection of enzymes on electrophoresis gels. 2nd
edn. CRC Press LLC, Boca Raton.
572 Richard R. Burgess

Prokipcak, R. D., Harris, D. J., and Ross, J. (1994). Purification and properties of a protein
that binds to the C-terminal coding region of human c-Myc mRNA. J. Biol. Chem. 269,
9261–9269.
Seelert, H., and Krause, F. (2008). Review: Preparative isolation of protein complexes and
other bioparticles by elution from polyacrylamide gels. Electrophoresis 29, 2617–2636.
Weber, K., and Kutter, K. (1971). Reversible denaturation of enzymes by sodium dodecyl
sulfate. J. Biol. Chem. 246, 4504–4509.
Wiggs, J., Gilman, M., and Chamberlin, M. (1981). Heterogeneity of RNA polymerase in
Bacillus subtilis: Evidence for an additional sigma factor in vegetative cells. Proc. Natl. Acad.
Sci. USA 78, 2762–2766.
 C H A P T E R T H I R T Y- T H R E E

Performing and Optimizing Western

Blots with an Emphasis on
Chemiluminescent Detection
Alice Alegria-Schaffer, Andrew Lodge, and Krishna Vattem

Contents
1. Western Blotting 574
2. Types of Western Blots 575
2.1. Direct and indirect 575
2.2. Far-Western 577
2.3. Semiquantitative Western blotting 577
3. Detection Methods 579
3.1. Enzyme conjugate 579
3.2. Colorimetric detection 581
3.3. Fluorescence detection 581
3.4. Chemiluminescence detection 582
4. The Chemiluminescence Signal 583
4.1. Signal capture 583
4.2. Signal intensity and duration 583
4.3. Optimizing a Western blot 585
5. Common Problems and their Explanations 588
5.1. No signal 588
5.2. Signal fades quickly 588
5.3. High background 589
5.4. New bottle of substrate does not produce a signal 589
5.5. Brown or yellow bands on the membrane 589
5.6. Bands or entire blot glowing in the darkroom 590
5.7. Ghost/hollow bands 590
6. Blotting and Optimization Protocols using
Chemiluminescent Substrates 593
6.1. Western blotting protocol using chemiluminescent substrates 593
6.2. Western blot stripping protocol 595
6.3. Optimizing antigen concentration 596

Thermo Fisher Scientific, Pierce Protein Research, Rockford, Illinois, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63033-0 All rights reserved.

573
574 Alice Alegria-Schaffer et al.

6.4. Optimizing membrane blocking 596

6.5. Optimizing the primary antibody concentration 597
6.6. Optimizing membrane washing 597
6.7. Optimizing enzyme conjugate concentration 597
6.8. Optimizing the detection method 598
References 598

Abstract
Immunodetection refers to any detection method that exploits the interaction
of an antibody and antigen. The choice of detection method, such as enzyme-
linked immunosorbent assays (ELISAs) or Western blotting, depends on the
researcher’s preferences and requirements. If a researcher wants to quantify a
low-abundance target protein then a chemiluminescent ELISA is used. If a
researcher wants to identify a protein that is in high abundance, a colorimetric
Western blot will suffice. If there are multiple targets within an assay, then
multiplex fluorescence is typically used.
This article focuses on Western blotting. Although colorimetric and fluores-
cent detection methods are discussed, chemiluminescent detection is used
most often and is, therefore, discussed in great detail. Included is specific
information about the chemiluminescent signal and factors that affect its inten-
sity and longevity. We also describe types of blotting and present data and
suggestions for obtaining semiquantitative data. Although classical Western
blotting is typically used for qualitative purposes, we present information about
effective quantitative analysis using specific controls. Common occurrences
within the methodology and their possible explanations are also detailed.
One frequent result is the appearance of ghost bands, which, based on our
research, can be caused by high amounts of target or antibody cross-reactivity.
Also included are the basic Western blot protocol and protocols for trouble-
shooting common problems and optimizing many of the specific factors that
influence results.

1. Western Blotting
Western blotting is a powerful and commonly used tool to identify
and quantify a specific protein in a complex mixture (Towbin et al., 1979).
The technique enables indirect detection of protein samples immobilized
on a nitrocellulose or polyvinylidene fluoride (PVDF) membrane. In a
conventional Western blot, protein samples are first resolved by sodium
dodecyl sulfate polyacrylamide gel electrophoresis (SDS–PAGE) and then
electrophoretically transferred to the membrane. Following a blocking step
using a nonrelevant protein, the membrane is probed with a primary
antibody (poly- or monoclonal) that was raised against the target antigen.
The membrane is then washed and incubated with an enzyme-conjugated
Performing and Optimizing Western Blots 575

secondary antibody that is reactive toward the primary antibody. The

membrane is washed again and incubated with an appropriate enzyme
substrate. The signal is either visually evaluated, if a colorimetric substrate
was used, or is detected with X-ray film or imaging instrumentation for
chemiluminescence and fluorescence.
The most significant advances in Western blotting methodology are
highly sensitive-enhanced chemiluminescent substrates, imaging systems,
and, most recently, the availability of a wide variety of photostable fluor-
ophores. The widespread use of extremely sensitive chemiluminescent
substrates (Mattson and Bellehumeur, 1996; Walker et al., 1995) has resulted
in nearly eliminating the use of radioisotope-labeled probes. Protein A or G
labeled with 125I was once commonly used as a secondary detection reagent;
however, the enhanced chemiluminescent substrates can detect proteins
down to the low-femtogram level with high signal-to-noise ratios.
Charged-coupled device (CCD) cameras are commonplace in most
labs. Imagers have a large dynamic range and a high degree of exposure
control, which allows background signal adjustments. Furthermore, their
accompanying analytical software enables densitometry analysis. Although
X-ray film is still widely used and more sensitive, CCD imaging systems
eliminate the hassles of film handling and developing and the associated
chemical wastes.
Recently, there is a rapid trend toward fluorescent dyes. This trend is
attributed to the new generation of fluorophores that offer tremendous
improvements in brightness, photostability, and pH sensitivity compared
to the traditional fluorophores. Western blotting detection via fluorescence
is typically performed when there are two different targets on a single blot.
A fluorophore pair is chosen based on their separate and distinct excitation–
emission spectra. The production of an easily identifiable color difference
enables multiplexing experiments. Most notably, these new fluorescent dyes
coupled with software have allowed detection and quantitation of cell
signaling pathways (Choudhary et al., 2007; Tipsmark et al., 2008).

2. Types of Western Blots

2.1. Direct and indirect
A direct Western blot refers to a detection method that uses a reporter-labeled
primary antibody that directly binds to the target protein. Indirect detection
uses a labeled secondary antibody (Ramlau, 1987) that binds to a nonlabeled
primary antibody (Fig. 33.1). Because incubation with a secondary antibody
is eliminated in direct detection, this strategy is performed in less time than
an indirect Western blot. Additionally, background signal from secondary
antibody cross-reactivity is eliminated (Bergendahl et al., 2003). Direct
 B Substrate
Detectable
product

A Substrate
Detectable Enzyme
product

Enzyme

Figure 33.1 Schematic of direct and indirect Western blotting methods. In the direct detection method, labeled primary antibody binds to
antigen on the membrane and reacts with the substrate, creating a detectable signal. In the indirect detection method, unlabeled primary
antibody binds to the antigen. Then a labeled secondary antibody binds to the primary antibody and reacts with the substrate.
Performing and Optimizing Western Blots 577

detection also enables probing for multiple targets simultaneously. Labeling a

primary antibody, however, sometimes has an adverse effect on its immuno-
reactivity, and even in the best of circumstances, a labeled primary antibody
cannot provide signal amplification. Consequently, the direct method is
generally less sensitive than indirect detection and is best used only when
the target is relatively abundant. An indirect method that amplifies the signal
and eliminates the secondary antibody is primary antibody biotinylation.
Labeling with biotinylation reagents typically results in more than one biotin
moiety per antibody molecule. Each biotin moiety is capable of interacting
with an enzyme-conjugated avidin, streptavidin, or Thermo Scientific Neu-
trAvidin1 Protein. These multiple enzymes catalyze the conversion of appro-
priate substrate to amplify the signal. Essentially, the avidin conjugate replaces
a secondary antibody and its appropriate molar concentration is the same as if
a secondary antibody were used. Be aware, however, if the sample applied to
the gel is naturally biotinylated, there is the potential for signal generation that
could interfere with target protein detection, especially when using highly
sensitive substrates.

2.2. Far-Western
Occasionally, an antibody to a specific antigen is unsuitable for Western blot
analysis or simply unavailable. Blotting is still possible if a binding partner to
the target protein is available for use as a probe. This type of application is
referred to as a far-Western blot and is routinely used for the discovery or
confirmation of a protein–protein interaction. Variations on this theme are
myriad and involve all the previously mentioned strategies. As with primary
antibodies, labeled binding partners are used, which are frequently labeled
in an in vitro translation reaction with 35S. Biotinylating the probe and
detecting with an avidin or avidin-like conjugate is another possibility and
has the added effect of amplifying the signal. Care must be taken to ensure
the probe is not overlabeled, which can compromise its ability to interact
with the target. Alternatively, expressing a recombinant probe in bacteria
with a tag, such as GST, HA, c-Myc, or FLAG, enables detection via a
labeled antibody to the particular tag (Burgess et al., 2000).

2.3. Semiquantitative Western blotting

Although Western blotting is often considered qualitative, it can be a
quantitative method provided that specific controls are included. Semi-
quantitative Western blotting is advantageous when no ELISA is available
for a specific sample or when a component in a biological sample interferes

1
DyLightTM, MemCodeÒ , NeutrAvidinÒ , PierceÒ , RestoreÒ and SuperSignalÒ are trademarks of Thermo
Fisher Scientific.
578 Alice Alegria-Schaffer et al.

with an ELISA. Often antibodies directed against one protein may interact
with equal specificity with another closely related protein. In such situa-
tions, an ELISA may generate false positives or overestimation of the target
protein abundance. Because Western blotting involves resolution of pro-
teins by gel electrophoresis, variations in molecular weight can be exploited
to distinguish and quantify the target protein alone (Sato et al., 2002; Xing
and Imagawa, 1999).
To evaluate the effectiveness and accuracy of quantitative Western
blotting, we used a purified target protein as an internal control and to
create a standard curve. The sample must contain sufficient target protein
such that its amount was within the standard curve range. To convert band
intensity into a quantitative measurement, the Western blot was analyzed
densitometrically using a CCD camera and imaging system. Finally, an
ELISA was used to confirm the trends observed by Western blot
(Mathrubutham and Vattem, 2005).

2.3.1. Assay methods

A431 cells were cultured for analyzing IkBa and p53. For IkBa, cells treated
with epidermal growth factor (EGF; 10 ng/ml). For p53, cells were treated
with either 50 mM cisplatin (Cis) or 5 mg/ml of doxorubicin (Dox) for 1 h.
Cells were collected at time intervals and lysed. Total cellular protein was
estimated using the Pierce Micro BCA Protein Assay (Thermo Fisher
Scientific, Rockford, IL).
Western blots were prepared using pure protein or cell lysates.
A standard curve was included with each Western blot. The linear range
for IkBa and p53 was determined using recombinant IkBa (Upstate
Biotechnology, Lake Placid, NY) from 0.0015 to 50 ng and recombinant
p53 (Active Motif, Carlsbad, CA) from 1.8750 to 60 ng. After blocking,
blots were incubated overnight at 4
C with anti-IkBa (Upstate Biotech-
nology) or anti-p53 (Active Motif) antibodies at 1 mg/ml prepared in
blocking buffer. Blots were washed and then incubated with horseradish
peroxidase (HRP)-conjugated secondary antibodies (1:1000 dilution from
10 mg/ml stock) for 1 h at room temperature and then washed again. Signal
was detected using SuperSignal2 West Dura Substrate (Thermo Fisher
Scientific). Blots were imaged using the Kodak 20000MM Image Station,
and protein band densities were analyzed using the spot density analysis
software provided with the image station.
The quantitative Western blot data were confirmed using p53 and IkBa
ELISA assays (Assay Designs, Ann Arbor, MI). The ELISAs were performed
according to the manufacturer’s instructions.

2
SuperSignalÒ Technology is protected by U.S. Patent # 6,432,662.
Performing and Optimizing Western Blots 579

2.3.2. Results and discussion

Densitometric analysis of the Western blots indicated that the linear range
for IkBa was 0.097–3.12 ng and for p53 was 1.87–60 ng. This range served
as an excellent internal positive control and counteracted variations in
Western blotting efficiency.
In the Western blot analysis, IkBa levels remained constant for the first
5 min after EGF treatment, rapidly dropped by 30% after 30 min and finally
gradually increased during the next 24 h (Fig. 33.2). Western blot densito-
metric analysis revealed that p53 levels changed slightly over the first 8 h
after treatment with Dox and then rapidly increased during the next
20–30 h. Treatment with Cis resulted in a gradual increase in p53 for
30 h (data not shown).
The trend in IkBa (Fig. 33.2) and p53 levels was confirmed by ELISA
and also correlated well with published data (Kwok et al., 1994; Sun and
Carpenter, 1998). With respect to p53, although the Western blot data were
confirmed by ELISA, closer comparison of the data revealed that the
Western blot was more sensitive to quantity changes of p53 between
20 and 30 h. At the 20–30 h time point there was a large amount of p53
present, but the ELISA did not detect the gradual increase of p53 between
these times, whereas the Western blot did. In contrast, changes in p53
during the first 8 h, when levels of p53 are relatively low, were detected
better by ELISA than by quantitative Western blot. Therefore, the ELISA is
more sensitive to changes in p53 at low levels, but the Western blot is more
sensitive to p53 changes at high levels.
Although Western blotting and ELISA are immunodetection methods,
the applications are conducted differently and usually use different immu-
noreagents. Variation in absolute protein amounts might be caused by
differences in the recombinant protein used for controls, primary antibody
specificity, and variations in protein interactions within each technique.

3. Detection Methods
3.1. Enzyme conjugate
Alkaline phosphatase (AP, 140 kDa) used to be the enzyme of choice
and was typically detected with precipitating chromogenic substrates.
Colorimetric reactions proceed at a steady rate, allowing accurate control
of relative sensitivity and reaction development. As protein research pro-
gressed, HRP (40 kDa) became more popular because of its stability and
smaller size, which enables more molecules conjugated per IgG and greater
sensitivity. Furthermore, chemiluminescent substrates for HRP enabled
even higher sensitivity.
 C
1,50,000

1,00,000

Net intensity
1 2 3 4 5 6 7 8
A
50,000

B
0
0 1 2 3 4
IkB (ng)
E
0.014
D
0.040 0.012

IkB (ng/mg lysate)

IkB (ng/mg lysate)

0.035 0.010

0.030 0.008

0.025 0.006

0.020 0.004
d in in d l l l
te 4h te m
g/ in
m
g/ in
m
g/ 4 h
ul
a 5m 30
m
F
2
ul
a m m m m m
tim
F F EG tim 50 5 50 30 50 2
ns EG EG ns F F F
U U EG EG EG

Figure 33.2 Quantitation of IkBa in EGF-treated cells. Panel A: Known amounts of recombinant IkBa (rIkBa) were Western blotted to
generate a standard curve. Lanes 1–6 contain 0.15–0.012 ng of rIkBa. Panel B: Lysates of EGF-treated A431 cells were Western blotted for
IkBa. Treatments were as follows: nontreated (lanes 1, 2), 50 ng for 5 min (lanes 3, 4); 50 ng for 30 min (lanes 5, 6); and 50 ng for 24 h (lanes
7, 8). Panel C: Recombinant IkBa standard curve (r2 ¼ 0.99). Panel D: Variation in the amount of IkBa with EGF exposure time as
determined by quantitative Western blotting. Panel E: Quantitation by ELISA.
Performing and Optimizing Western Blots 581

3.2. Colorimetric detection

Colorimetric or chromogenic substrates are perhaps the simplest and most
cost-effective method of detection. When these substrates come in contact
with the appropriate enzyme, they are converted to insoluble, colored
products that precipitate onto the membrane and require no special
equipment for processing or visualizing. Substrates such as TMB
(3,30 ,5,50 -tetramethylbenzidine), 4-CN (4-chloro-1-naphthol), and DAB
(3,30 -diaminobenzidine tetrahydrochloride) are used with HRP. Substrates
for AP include BCIP (5-bromo-4-chloro-30 -indolylphosphate-p-toluidine
salt) and NBT (nitro-blue tetrazolium chloride), which yields an insoluble
black-purple precipitate, and Fast Red (naphthol AS-MX phosphate þ Fast
Red TR Salt). The performance of a particular substrate may vary dramati-
cally when obtained from different suppliers because results can be affected
by the concentration and purity of the substrate and by additives and buffer
components that are a part of the formulation.

3.3. Fluorescence detection

Western blotting detection via fluorescence is typically performed when
there are two different targets on a single blot and high sensitivity is
required. Fluorescent dyes (often called ‘‘fluorophores’’ or simply ‘‘fluors’’)
are molecules whose chemical bonds excite when absorbing a photon of
light at one wavelength, causing them to emit a photon of light at a longer
wavelength (lower energy) as the molecule relaxes to the ground state.
Small, chemically stable fluorophores that have convenient ranges of effi-
cient (intense) excitation and emission wavelengths are useful as detectable
chemical tags or labels for antibodies and other biomolecular probes.
Some fluorescence-based systems use fluorescent proteins (e.g., phyco-
erythrin) or bioluminescent reporter systems; however, these techniques are
time-consuming, limited in their ability to detect multiple targets and do
not typically provide the level of photostability and sensitivity offered by
synthetic fluorescent dyes. Fluorescence techniques that use specific probes
labeled with carefully selected sets of fluorescent dyes enable the detection
of multiple targets and provide greater compatibility with a wide range of
fluorescence instrumentation.
Although fluorescein, rhodamine, and amino-methyl-coumarin-acetate
(AMCA) dyes are the traditionally used fluorophores, they have numerous
limitations. Most notably, these traditional dyes have relatively low-
fluorescence intensity and tend to photobleach. A new generation of
fluorophores has overcome such limitations (Table 33.1) and offers tremen-
dous improvements in brightness, photostability, and pH sensitivity. These
new fluorophores cover the entire visible spectrum and much of the
infrared spectrum.
582 Alice Alegria-Schaffer et al.

Table 33.1 Spectral properties of commonly used fluorescent dyes*

Fluorescent dye Emission Ex/Ema eb

DyLight Fluor 405, Alexa Fluorc 405, Blue 400/420 30,000
Cascade Blue
DyLight Fluor 488, Alexa Fluor 488, Green 493/518 70,000
fluorescein, FITC
DyLight Fluor 549, Alexa Fluor 546, Yellow 560/574 150,000
Alexa 555, Cy Dye 3d, TRITC
DyLight Fluor 594, Alexa Fluor 594, Red 593/618 80,000
Texas Red
DyLight Fluor 633, Alexa Fluor 633 Red 638/658 170,000
DyLight Fluor 649, Alexa Fluor 647, Red 654/673 250,000
Cy Dye 5
DyLight Fluor 680, Alexa Fluor 680, Near IR 692/712 140,000
Cy Dye 5.5
DyLight Fluor 750, Alexa Fluor 750 Near IR 752/778 220,000
DyLight Fluor 800, IRDye 800 Infrared 777/790 270,000
* The listed extinction coefficients are representative. The extinction coefficient for a given fluorescent
dye and its analogs varies with the dye’s purity, solvent and molecular structure changes, including the
position and makeup of reactive groups and other moieties.
a
Excitation and emission maxima in nanometers.
b
Molar extinction coefficient (M 1 cm 1).
c
Alexa FluorÒ is a trademark of Molecular Probes Inc.
d
CyÒ is a trademark of Amersham Pharmacia Biotech UK Limited Corp.

When performing a fluorescent Western blot, typically low-fluorescence

(treated) membranes are used because membrane polymers autofluoresce in
the visible range of the spectrum, which can interfere with detection. Using
fluors with nonoverlapping excitation–emission spectra is critical for iden-
tifying targets. A typical fluor pair includes Thermo Scientific DyLight3
Fluors 549 and 649 or the DyLight Near-infrared (IR) Fluors, 680 and 800.
The near-IR fluors are especially useful because protein samples and mem-
brane polymers are less likely to autofluoresce within these spectral ranges,
resulting in lower background and higher sensitivity (Patonay and Antoine,
1991; Sowell et al., 2002).

3.4. Chemiluminescence detection

The most popular Western blotting substrates are luminol-based and pro-
duce a chemiluminescent signal. Chemiluminescence is a chemical reaction
that produces energy released in the form of light (Fig. 33.3). In the presence

3
DyLightTM, MemCodeÒ , NeutrAvidinÒ , PierceÒ , RestoreÒ and SuperSignalÒ are trademarks of Thermo
Fisher Scientific.
Performing and Optimizing Western Blots 583

O O * O

NH O− O−
H2O2 + HRP + Light
NH O− O−

NH2 O NH2 O NH2 O

@ 425 nm

Figure 33.3 Chemiluminescence reaction scheme. Luminol is oxidized in the pres-

ence of HRP and hydrogen peroxide to form an excited state product (3-aminophtha-
late) that emits light at 425 as it decays to the ground state.

of HRP and a peroxide buffer, luminol oxidizes and forms an excited state
product that emits light as it decays to the ground state. Light emission occurs
only during the enzyme–substrate reaction and, therefore, once the substrate
in proximity to the enzyme is exhausted, signal output ceases. In contrast,
colorimetric substrates, such as TMB, produce precipitate that remains
visible on the membrane even after the reaction has terminated.

4. The Chemiluminescence Signal

4.1. Signal capture
Although chemiluminescent Western blotting has become commonplace,
attempting to capture that elusive signal can be frustrating. Because a Western
blot is composed of a series of linked techniques that require skill to perform,
failure to capture signal can be caused by many factors. With so many
variables (Table 33.2), troubleshooting a problem blot can be equated to
that proverbial needle in a haystack. The classical protocol is often ineffective
in detecting a particular protein. For example, the primary antibody may not
recognize the immobilized antigen in its denatured state. Although the
protein can be kept in its native state by using nondenaturing conditions,
this makes determination of target molecular weight more challenging.
Furthermore, exceptionally large or hydrophobic proteins often preclude
effective membrane transfer. On the other hand, small proteins (e.g., 10
kDa) can flow through the membrane pores, failing to bind. Towbin’s
original protocol is therefore often modified to ensure target detection.
Such modifications may involve using different reagents for indirect detection
or a labeled primary probe for direct detection. Sometimes transfer is bypassed
altogether and detection is accomplished in-gel (Desai et al., 2001).

4.2. Signal intensity and duration

When all Western blotting factors are optimal, a chemiluminescent signal
can last for 6–24 h, depending on the specific substrate used. How much
light is generated and for how long depends on the specific substrate being
584 Alice Alegria-Schaffer et al.

Table 33.2 Factors that affect Western blotting results

Factor Variable characteristic

Target antigen Conformation, stability, available epitope(s),
polypeptide size
Polyacrylamide gel Manufacturer, percent polyacrylamide, age, lot
Membrane Manufacturer, type, lot
Primary antibody Specificity, titer, affinity, incubation time, and
temperature
HRP conjugate Enzyme activation level and activity, source animal,
concentration
Blocking buffer Protein type, concentration, cross-reactivity
Washes Buffer, volume, duration, frequency
Substrate Sensitivity, manufacturer lot, age
Detection method Film age, imaging instrument manufacturer, exposure
time

120
Too much enzyme
Appropriate enzyme
100

80
Intensity

60

40

20

0
1 2 3 4 5 6
Time (min)

Figure 33.4 Example signal emission curves. When there is too much enzyme present
in a chemiluminescent Western blot system, signal output peaks soon after substrate
application and rapidly exhausts the substrate. In an optimal system, the signal emission
peaks approximately 3 min after applying the substrate and plateaus for several hours.

used and the enzyme-to-substrate ratio present in the system. Although the
amount of substrate on a blot is relatively constant, the amount of enzyme
present depends on how much was added and other factors (Table 33.1).
Too much enzyme conjugate applied to a Western blot system is the single
greatest cause of signal variability, high background, short signal duration,
and low sensitivity.
A signal emission curve that decays slowly (Fig. 33.4) is desirable because
it demonstrates that each component of the system has been optimized and
Performing and Optimizing Western Blots 585

allows reproducible results. A signal that decays too quickly can cause
variability, low sensitivity and lack of signal documentation. A long-lasting
signal minimizes variability with transfer efficiency, different manufacturer
lots of substrate and other factors.
Although HRP continues its activity for as long as substrate is available,
HRP can become inactive with prolonged exposure to substrate. Free
radicals produced during the oxidation reaction can bind to HRP in such
a way that the enzyme can no longer interact with the substrate.
An abundance of HRP in the system in turn produces an abundance of
free radicals that increases the probability of HRP inactivation. Free radicals
also can damage the antigen, antibodies and the membrane, prohibiting
reprobing effectiveness.

4.3. Optimizing a Western blot

Each Western blot system must be optimized to obtain consistent results.
Many factors influence the intensity and longevity of a signal, and each of
these factors can be optimized as discussed in the following sections. The
protocol section at the end of this chapter includes specific instructions and
suggestions about optimization procedures.

4.3.1. Blotting membrane

HRP–luminol interactions and subsequent signal generation are likely
unaffected by membrane composition. Nitrocellulose and PVDF mem-
branes, however, do differ in their protein-binding properties. Generally,
nitrocellulose binds proteins better, often produces crisper bands and some-
times results in greater sensitivity. PVDF is more hydrophobic, is difficult to
wet, and sometimes results in more background signal; however, it has high
tensile strength and excellent handling characteristics. For best results,
empirically determine which membrane type, manufacturer and lot is
optimal for each Western blotting system. Once a specific membrane has
proven effective in a system, it may be beneficial to use the same lot
throughout the course of the study.

4.3.2. Target protein

Transfer efficiency can vary dramatically among proteins. Proteins differ in
their ability to migrate from the gel and their propensity to bind to the
membrane using a particular set of conditions. Transfer efficiency depends
on factors such as gel composition, the gel-membrane contact, position of
the electrodes, transfer duration, protein size and composition, field
strength, and the presence of detergents. Optimal transfer of most proteins
is obtained in low-ionic strength buffers and with low electrical current.
Transfer efficiency can be evaluated by staining the membrane with an
immunoblot-compatible or reversible stain (Sasse and Gallagher, 2008).
586 Alice Alegria-Schaffer et al.

Some common protein stains for membranes include: Ponceau S, a red stain
that fades with time; coomassie dye, a sensitive blue stain that tightly binds
to proteins and can interfere with Western blotting; and Thermo Scientific
MemCode4 Stain (Thermo Fisher Scientific), an easily reversible blue stain
(Antharavally et al., 2004).

4.3.3. Blocking buffer

Many different blocking reagents are available for Western blotting.
Because no blocking reagent is appropriate for all systems, empirical testing
is essential (Spinola and Cannon, 1985). An optimal blocking buffer max-
imizes the signal-to-noise ratio and does not react with the system’s anti-
bodies or target. For example, using 5% nonfat milk as a blocking reagent
when using avidin–biotin systems results in high background because milk
contains variable amounts of endogenous biotin, which binds the avidin.
When switching substrates, antibodies or the target, a diminished signal or
increased background can result simply because the blocking buffer was not
optimal for the new system.
Some systems may benefit from adding a surfactant, such as Tween-205
detergent, to the blocking buffer. Surfactants can minimize background by
preventing the blocking reagent from nonspecifically binding to the target
or by blocking hydrophobic sites on the membrane to which antibodies can
bind. Adding too much detergent, however, can prevent adequate block-
ing. Typically, a final concentration of 0.05% detergent is used; however,
for best results, determine if detergents enhance a specific system and their
optimal concentration. Always use a high-quality detergent that is low in
contaminants.

4.3.4. Antibodies
Not only is the affinity of the primary antibody for the antigen important,
but primary and secondary antibody concentrations also have a profound
effect on signal intensity. Too much HRP on the blot can be caused by
either primary and secondary antibody concentrations or both. Minimal
primary antibody is advantageous, as it promotes target-specific binding and
low background.
If a blot failed to generate an adequate signal, removing all detection
reagents (stripping) from the blot and reprobing with either a different
primary antibody or different concentrations of antibodies often conserves
valuable sample and time; however, insufficient stripping can leave active
HRP on the blot that will produce a signal. Applying substrate on the
stripped blot and subsequent detection will reveal if active HRP remains on

4
DyLightTM, MemCodeÒ , NeutrAvidinÒ , PierceÒ , RestoreÒ and SuperSignalÒ are trademarks of Thermo
Fisher Scientific.
5
TweenÒ is a trademark of ICI Americas.
Performing and Optimizing Western Blots 587

the blot. Also, an abundance of inactive HRP molecules not removed by

stripping can inhibit the primary antibody from binding to the target.
Stripping and reprobing blots is an effective method to gain information
about a specific system, but it is not a definitive way to determine the
optimal system parameters.

4.3.5. Detection method

Traditionally, film has been used to detect a Western blot chemiluminescent
signal. Film requires no expensive equipment and provides excellent
sensitivity, but suffers from a narrow range of linear detection intensity.
Adjusting film exposure time is frequently required to obtain a publication-
quality blot, which can be time-consuming and generate waste. Immersing
the developed film in a set of Thermo Scientific Reagents (Fig. 33.5) will
remove excessive signal caused by overexposure. These reagents evenly
remove silver from the film, which maintains the signal-to-noise ratio
while improving the film’s appearance.
CCD cameras and their accompanying analytical software can adjust
background levels and perform densitometry. The imager has advantages
over film-based detection because the large dynamic range and the high
degree of exposure control, allowing for the best possible image documen-
tation without issues of high background or high-signal intensity obscuring
data. Additionally, optimization of exposure time avoids band signal

A B
Before After
treatment treatment

Figure 33.5 Reduction of background signal by chemical treatment of the X-ray film.
Recombinant human TNFa was separated by SDS–PAGE and transferred to a nitrocel-
lulose membrane. The blot was probed with mouse antihuman TNFa and goat
antimouse-HRP antibodies and detected with SuperSignal West Dura Substrate.
Panel A: The blot was exposed to film for 30 s, resulting in considerable background
speckling. Panel B: The film was then treated with Pierce Background Eliminator for
2 min to remove speckling.
588 Alice Alegria-Schaffer et al.

saturation and allows observation of minor variations in density. In contrast,

film has a small dynamic range, and band signal can quickly reach saturation.
When signal intensity is high, film’s low dynamic range, propensity to
reach saturation quickly and exposure control limitations often result in
overexposed images.

5. Common Problems and their Explanations

5.1. No signal
An initial exposure that fails to capture the chemiluminescent signal indi-
cates a Western blotting system that requires optimization. Frequently, lack
of signal is caused by too much enzyme (i.e., HRP) in the system. It may
seem counterintuitive to use less enzyme conjugate when a signal cannot be
detected; however, for successful signal documentation, the correct balance
of enzyme and substrate must be present. Substrate oxidation by the enzyme
is irreversible and, therefore, once the substrate is oxidized, it can no longer
interact with the enzyme to generate light. Because enzyme activity persists,
the substrate is the limiting factor and once exhausted, signal output ceases.
Rarely, lack of signal is caused by an insufficient amount of active enzyme
present. Too much or not enough enzyme can be caused by any of the
factors involved in the Western blot system.
To produce a signal that can be captured, adjust the system’s parameters.
For reproducible results, prepare a new gel and apply less sample or titrate
the antibodies. When optimizing antibody concentrations, image the blot
twice: once immediately after adding the substrate and a second time at an
interval (e.g., 1 h) after incubating the substrate. The second detection
provides information about the optimal enzyme concentration and helps
optimize parameters.
Also, if the initial exposure did not capture a signal, a second incubation
in substrate may yield a signal if some HRP remains active. Stripping all
detection reagents from the blot and reprobing can save valuable sample
while optimizing parameters. Perform an additional incubation in the
substrate and strip the blot only to recover some information about the
system. If blot-to-blot consistency or comparison is desired, the same
conditions must be used and the same procedure must be followed each
time the experiment is performed.

5.2. Signal fades quickly

When a particular system produces a chemiluminescent signal that fades
quickly, the Western blotting system requires optimization, as described
earlier. The good news is that a signal was obtained, indicating that the blot
Performing and Optimizing Western Blots 589

is almost optimized. Sometimes a particular system produces a signal that

fades more quickly than usual although all parameters are the same. This
type of result is minimized in a fully optimized system. A successful but
suboptimal system is subject to the slight variations inherent to the method,
such as transfer efficiency and changes in sample and antibody activity
during storage and handling.

5.3. High background

High background signal is the result of either insufficient blocking, anti-
bodies cross-reacting with the blocking proteins, or the use of too much
enzyme conjugate. Researchers sometimes believe that a particular substrate
causes background or can increase background. The substrate in itself
typically does not produce signal without the enzyme being present.
When using a substrate with greater sensitivity than what was previously
used, high background often results if the parameters were not altered to
compensate for the substrate’s sensitivity. Using optimal concentrations of
antibodies promotes target-specific binding and low background.

5.4. New bottle of substrate does not produce a signal

Occasionally, a signal cannot be captured when the only variable that has
changed in a particular system is a new bottle or different lot of substrate.
Typically, this result is caused by a Western blot system that has not been
fully optimized. Western blotting substrates are inherently variable. Many
manufacturers simply control for a minimum sensitivity, and it is possible
that the new substrate is more sensitive than the previously used lot. In a
fully optimized blotting system, substrate sensitivity variations, as well as
other variables, are minor or unnoticeable.

5.5. Brown or yellow bands on the membrane

HRP becomes brown when it is oxidized and inactive. Within a given
amount of enzyme conjugate, there always exists a portion that is oxidized.
In an optimized system, the amount of oxidized HRP is miniscule and
cannot be visualized on the blot. The appearance of yellow or brown bands
indicates the presence of a large amount of HRP and, therefore, the
oxidized and inactive portion is visible. A blotting system that results
in yellow bands requires optimizing using much less enzyme conjugate.
Additionally, too much HRP in a localized area produces an abundance of
free radicals during enzymatic activity. Free radicals can inactivate HRP and
damage antibodies, target and the membrane, prohibiting effective
reprobing.
590 Alice Alegria-Schaffer et al.

5.6. Bands or entire blot glowing in the darkroom

If a pattern of bands or the entire blot is glowing after incubation in the
substrate, then there is too much HRP present in the system. This occur-
rence indicates that further dilution of the secondary HRP-conjugated
antibody is required and possibly the primary antibody as well. Too much
enzyme can be caused by many of the factors involved in the Western
blotting system. If the entire blot is glowing, optimization of blocking and
washing also might be necessary.

5.7. Ghost/hollow bands

Protein bands that appear as a halo with no signal in the middle of the band
or an entire band that appears white in a dark background are typically
referred to as ghost bands. This result occurs when the substrate is depleted
within the white area.
The specific cause of ghost bands is not well-understood. Therefore, we
tested several factors that could contribute to this result. Our study, briefly
described in the following sections, revealed that some common causes for
ghost-band effects include applying too much target protein to the gel and
using antibodies that cross-react with component(s) in the blocking solution
(Vattem and Mathrubutham, 2005).

5.7.1. Assay methods

A431 cells (ATCC) were cultured and treated with cisplatin. Cells were
lysed, sonicated, and clarified by centrifugation. Total protein concentration
was estimated using Thermo Scientific Pierce6 BCA Protein Assay (Thermo
Fisher Scientific).
Various amounts of purified NFkB protein (Upstate Biotechnology) or
30 mg of A431 lysates were transferred to nitrocellulose membranes. The
membranes were incubated overnight with either rabbit anti-NFkB anti-
body (1 mg/ml stock; diluted 1:5000; Upstate Biotechnology) or rabbit
anti-p53 (1 mg/ml stock diluted 1:1000; GENEKA). HRP-conjugated
secondary antibodies (10 mg/ml stock diluted 1:1000) were applied to the
blots and incubated for 1 h at room temperature. Either Thermo Scientific
SuperSignal1 West Pico or West Dura Chemiluminescent Substrate
(Thermo Fisher Scientific) was added (0.5 and 5.0 ml) and incubated for
various periods. Membranes were exposed to X-ray film for various periods.

6
DyLightTM, MemCodeÒ , NeutrAvidinÒ , PierceÒ , RestoreÒ and SuperSignalÒ are trademarks of Thermo
Fisher Scientific.
Performing and Optimizing Western Blots 591

500
200
100
50
10
5

NFkB (ng)

1:1,000 1:5,000
(10ng/ml) (2ng/ml)

1:20,000 1:50,000
(0.5ng/ml) (0.2ng/ml)

Figure 33.6 Secondary antibody concentration affects the appearance of ghost bands.
Western blots were performed using various concentrations of secondary antibody.
Ghost bands (oval) are present in lanes containing >200 ng of NFkB. Membranes were
incubated in SuperSignal West Pico Substrate for 5 min and exposed to X-ray film for
30 s.

5.7.2. Results and discussion

Protein bands on membranes treated with 1:1000, 1:5000 or 1:20,000
dilution of a 10 mg/ml stock of secondary antibody were not significantly
different; however, 1:50,000 dilution (0.2 ng/ml) of secondary specifically
enhanced the appearance of ghost bands. This effect was more evident in
lanes that contained >200 ng NFkB (Fig. 33.6). Also, 64 h after completing
the experiment, when signal intensity had slightly reduced, ghost bands
became more evident in all lanes having >200 ng NFkB, irrespective of
secondary antibody dilution (data not shown).
A short substrate incubation of 30 s resulted in ghost bands in lanes
containing >200 ng of protein. Generally, longer substrate incubation
prolongs signal duration. Using either the West Pico or West Dura substrate
enabled detection down to 10 ng of proteins, even after 64 h of completing
the Western blot. West Dura is a more sensitive and longer lasting substrate
than West Pico. Although West Dura produced a significantly stronger
signal when detecting 5 ng of proteins, lanes with greater than 100 ng
resulted in rapid signal loss, which emphasizes the importance of using an
592 Alice Alegria-Schaffer et al.

A 1 2 3 4 1 2 3 4 B

p53 Ghost
bands

Milk: 0.5% 2%
1 2 3 4 1 2 3 4
Protein
of interest

p53

Milk: 5% 10%

A431 lysates

Figure 33.7 Low concentration of nonfat milk enhances the detection of low-abun-
dance proteins, but antibodies cross-reacting with milk proteins result in ghost bands.
Panel A: Lanes 1–4 are lysate samples from different cell harvest times (0, 8, 20, and
30 h) after treatment with 50 mM cisplatin. Different concentrations of nonfat milk were
used for blocking and antibody incubation. The low concentration of milk (0.5%)
enhanced detection of p53. Panel B: Membrane was incubated with 5% nonfat
milk blocking buffer for 48 h. The primary antibody bound to the blocked regions,
producing ghost bands.

appropriate amount of target protein with highly sensitive detection

systems.
Incubating the blot with a nonfat milk solution is commonly used to
minimize background signal. Varying the concentration of milk in the
blocking solution did not result in ghost bands. Surprisingly, using 0.5%
nonfat milk in the blocking solution enhanced detection of low-abundance
proteins (Fig. 33.7, panel A); however, antibody cross-reacting with the
blocking protein did result in ghost bands. If the primary or secondary
antibody binds to milk proteins, these blocked sites chemiluminescence,
producing a dark background with white lanes where sample proteins
populate the membrane surface (Fig. 33.7, panel B).
Rapid signal loss may result from enhanced catalysis of substrate by a
high density of enzyme-conjugated secondary antibody binding to a high
amount of target protein, causing rapid substrate depletion in these regions.
Performing and Optimizing Western Blots 593

A high amount of transferred protein binds high amounts of HRP-

conjugated secondary antibody. Excessive product (i.e., signal) produced
in these concentrated areas of immune complexes can result in a rapid
substrate depletion, resulting in ghost bands. In conclusion, excess loading
of gels, high concentrations of secondary antibody, suboptimal substrate
incubation times, and antibodies cross-reacting with blocked areas can
cause ghost bands. Optimizing Western blot parameters is especially critical
to prevent ghost bands when using highly sensitive detection systems.

6. Blotting and Optimization Protocols using

Chemiluminescent Substrates
6.1. Western blotting protocol using
chemiluminescent substrates
1. Separate the proteins in the sample by gel electrophoresis.
2. Prepare the transfer buffer: Use Tris–glycine transfer buffer dissolved in
400 ml of ultrapure water plus 100 ml methanol. Use and store the
transfer buffer at 4
C. Note: Transfer buffer: 25 mM Tris, 192 mM
glycine, pH 8.0, 20% methanol.
3. Construct a gel ‘‘sandwich’’ (Fig. 33.8) for wet transfer. For semidry
transfer, prepare the sandwich in the same order between the anode
and cathode. Note: Soak the pads and filters in transfer buffer before
sandwich assembly.
4. Transfer proteins from the gel to a membrane. For wet transfer using a
mini-transfer apparatus designed for an 8  10 cm gel, transfer at 40 V
for 90 min keeping the buffer temperature at 4
C. For semidry transfer
use 15 V for 90 min. Note: Determine transfer success by staining the
gel or reversibly staining the membrane.
5. Remove the membrane and block nonspecific binding sites with a
blocking buffer for 20–60 min at room temperature (RT) with shaking.
6. Incubate the blot with the primary antibody solution (see Table 33.3)
containing 10% blocking solution with rocking for 1 h. If desired,
incubate the blot overnight at 2–8
C.
7. Wash the membrane three times for 5 min each with Tris-buffered
saline (TBS), phosphate-buffered saline (PBS) or other physiological
wash buffer containing 0.05% Tween-20 detergent. If using an
enzyme-conjugated primary antibody, proceed to step 10.
8. Incubate the blot with the enzyme conjugate (see Table 33.3) contain-
ing 10% blocking solution for 1 h with rocking at room temperature.
9. Wash the membrane five times for 5 min each in wash buffer to remove
any nonbound conjugate. It is crucial to thoroughly wash the
membrane after incubation with the enzyme conjugate.
594 Alice Alegria-Schaffer et al.

Gel
Transfer membrane
Filter paper
Pads
Support grid

Gel/membrane/filter
sandwich

Buffer tank

Anode (+)
Electrodes
Cathode (−)

Direction of
transfer

Figure 33.8 Electrophoretic transfer unit and blot setup.

Table 33.3 Primary and secondary antibody concentrations to use with Thermo
Scientific Chemiluminescent Substrates

West West West

Substrate ECL Picoa Duraa Femtoa
Primary antibody 0.2–10 0.2–1.0 0.02–1.0 0.01–0.2
concentration
(mg/ml)
Secondary antibody 67–1000 10–50 4–20 2–10
concentration
(ng/ml)
a
Part of the Thermo Scientific SuperSignal Product Family.

10. Prepare the substrate. Use a sufficient volume to ensure that the blot is
completely wetted with substrate and the blot does not become dry
(0.1 ml/cm2).
11. Incubate the blot with substrate for 1 min when using ECL or 5 min
when using SuperSignal Substrates.
Performing and Optimizing Western Blots 595

12. Remove the blot from the substrate and place it in a plastic membrane
protector. A plastic sheet protector works well, although plastic wrap
also may be used. Remove all air bubbles between the blot and the
surface of the membrane protector.
13. Image the blot using film or a cooled CCD camera.

6.2. Western blot stripping protocol

One of the many advantages of using a chemiluminescent substrate is that it
allows the removal of all detection reagents (Kaufmann et al., 1987). These
stripping methods are ineffective for precipitating substrates and produce
variable results with fluorescent dyes.
1. Prepare the stripping buffer. Use one of the following suggested stripping
buffers:
(a) Restore Western Blot Stripping Buffer (Thermo Fisher Scientific)
(b) 0.1 M glycine–HCl (pH 2.5–3.0)
(c) 50 mM Tris–HCl (pH 7), 2% sodium dodecyl sulfate (SDS), 50 mM
dithiothreitol (DTT). Note: Prepare this buffer immediately before
use.
2. Place the blot to be stripped in stripping buffer. Use a sufficient volume
to ensure that the blot is completely wetted (approximately 20 ml for an
8  10 cm blot). Note: Some denaturation and loss of target protein may
occur.
3. Incubate for 5–15 min at room temperature. Optimization of both
incubation time and temperature is essential for best results. Some
interactions require at least 15 min and may require incubation at
37
C. If using buffer C, incubate for 30 min at 70
C.
4. Remove the blot from the stripping buffer and wash using wash buffer
(PBS/TBS or other physiological buffer containing 0.05% Tween-20
detergent).
5. To test for complete removal of the enzyme conjugate and primary anti-
body, perform the tests listed below. If signal is detected in either case, repeat
steps 2–4, stripping for an additional 5–15 min or increasing the temperature
to 37
C. Optimize stripping time and temperature to ensure complete
removal of antibodies while preventing damage to the antigen.
 To test for complete removal of the enzyme conjugate, incubate the
membrane with substrate and image the blot. If no signal is detected
after a 5-min film exposure, the enzyme conjugate has been success-
fully removed.
 To test for complete removal of the primary antibody, incubate the
membrane with enzyme conjugate then wash with wash buffer. Incu-
bate with substrate and image the blot. If no signal is detected after a
5-min film exposure, the primary antibody has been successfully
removed.
596 Alice Alegria-Schaffer et al.

After determining that the membrane is properly stripped, commence

the second probing experiment. Typically, a blot can be stripped and
reprobed several times but may require longer exposure times or a more
sensitive substrate. Subsequent reprobing may result in decreased signal if
the antigen is labile. Analysis of the individual system is required. Note:
Reblocking a membrane is generally not necessary after stripping, but may
be required in some systems.

6.3. Optimizing antigen concentration

1. Prepare different concentrations of the protein sample in SDS–PAGE
sample buffer. Test a wide range of sample concentration, keeping in
mind the detection limit of the substrate being used.
2. Apply an equal volume of each concentration on the gel and separate by
electrophoresis. Transfer the samples to a membrane.
Note: For a rough indication of optimal concentration, perform a dot
blot. Dot antigen dilutions onto dry membrane strips, using the smallest
possible volume. Allow the membranes to dry 10–15 min or until no visible
moisture remains and proceed with step 3.
3. Block the membrane with a standard blocking reagent and probe with
primary antibody followed by enzyme conjugate. If optimized dilutions
have not yet been determined, use a mid-range value according to the
sensitivity of the substrate.
4. Wash membrane and add the substrate. Image the blot as desired.

6.4. Optimizing membrane blocking

1. Separate the protein sample by electrophoresis and transfer to a mem-
brane or dot protein samples onto the membrane as described
previously.
2. Cut strips from the membrane according to the number of conditions
being tested. The following combinations should be tested with each
blocker:
 Blocker þ primary antibody þ enzyme conjugate þ substrate
 Blocker þ enzyme conjugate þ substrate
 Blocker þ substrate
3. Add the strips to various blocking solutions, ensuring the strip is
completely immersed in the solution. Incubate each strip for 1 h at
room temperature with shaking.
4. Add primary antibody and/or enzyme conjugate solutions containing
10% blocking agent to appropriate groups. If optimized dilutions have
not yet been determined, use a mid-range value according to the
sensitivity of the substrate.
5. Wash membrane and add the substrate. Image the blot as desired.
Performing and Optimizing Western Blots 597

Note: To check for endogenous peroxidase activity in the blocker or in

the sample, incubate blocked membranes with substrate. If signal is detected,
use a different blocking buffer or use a peroxidase suppressor after the
blocking step. Large blocking buffer volumes minimize nonspecific signal.

6.5. Optimizing the primary antibody concentration

1. Separate the protein sample by electrophoresis and transfer to the mem-
brane. Alternatively, dot the protein sample onto the membrane as
described in Section 6.3. Block the membrane using an appropriate
blocking reagent. Cut strips from the membrane according to the
number of primary antibody conditions being tested.
2. Prepare dilutions of primary antibody in wash buffer containing 1/10
volume of blocking agent and apply to the membrane strips. Incubate for
1 h at room temperature.
3. Wash the strips and incubate with enzyme conjugate for 1 h at room
temperature. Wash again and develop signal using an appropriate
substrate. Detect signal using film or a CCD camera.
Note: To check for nonspecific binding of primary antibody to blocker,
incubate membrane in primary antibody working solution followed by
enzyme conjugate working solution. If signal is observed, correct the
problem by using a different blocking buffer or use less primary antibody.

6.6. Optimizing membrane washing

1. Use a wash buffer such as PBS or TBS or other physiological buffer
containing 0.05% Tween-20.
2. Wash the membrane by agitating at least three times for 5 min each after
primary antibody incubation, and at least five times for 5 min each after
incubating with enzyme conjugate.
3. If nonspecific background appears upon final detection, use larger
volumes of wash buffer or increase the number and time of each wash.
If no improvement occurs, the problem lies with another variable.

6.7. Optimizing enzyme conjugate concentration

When determining the optimal concentration for a new Western blotting
system, a simple experiment often saves much frustration with signal
variability.
1. Apply the same amount of target in three (or more) wells of the gel.
2. Separate the protein sample by electrophoresis and transfer to the mem-
brane. Block the nonspecific binding sites and probe with primary
antibody.
598 Alice Alegria-Schaffer et al.

3. After washing, cut the blot into strips containing the target.
4. Probe each strip with a different enzyme conjugate concentration. For
example, for SuperSignal West Pico Substrate use 1:40,000, 1:60,000,
and 1:80,000 dilutions (from a 1-mg/ml stock). Incubate strips for 1 h at
room temperature with rocking.
5. Wash the strips and add the substrate. After substrate incubation, image
the strips.
6. Wait 1–2 h and image the strips a second time.
7. Evaluate the results. Use the dilution that produces the strongest signal.

6.8. Optimizing the detection method

1. Separate the protein sample by electrophoresis and transfer to membrane
or dot the protein sample onto the membrane as described in Section 6.3.
2. Block nonspecific binding sites and probe with primary antibody and
enzyme conjugate containing 1/10 volume of blocking agent.
3. If antibody concentrations have not been optimized, choose a mid-range
value. Wash the membrane after each incubation.
4. Cut strips from the membrane according to the number of substrate
exposure conditions being tested. Prepare working solution of substrate
to be tested.
5. Incubate the membrane strips with the substrate for a time period
consistent with the manufacturer’s instructions.
6. Remove the strips from the substrate using forceps, and gently tap edge
onto a paper towel to remove excess substrate.
7. Place the strips in a plastic cover and image blot for varying lengths of
time. Select a time with clear signal and low background. Note: A CCD
camera might require slightly longer exposure times than film.

REFERENCES
Antharavally, B. S., Carter, B., Bell, P. A., and Mallia, K. A. (2004). A high-affinity
reversible protein stain for Western blots. Anal. Biochem. 329(2), 276–280.
Bergendahl, V., Glaser, B. T., and Burgess, R. R. (2003). A fast western blot procedure
improved for quantitative analysis by direct fluorescence labeling of primary antibodies.
J. Immunol. Methods 277, 117–125.
Burgess, R. R., Arthur, T. M., and Pietz, B. C. (2000). Mapping protein–protein interaction
domains using ordered fragment ladder far-Western analysis of hexahistidine-tagged
fusion proteins. Methods Enzymol. 328, 141–157.
Choudhary, S., Lu, M., Cui, R, and Brasier, A. R. (2007). Involvement of a novel Rac/
RhoA guanosine triphosphatase-nuclear factor-kappaB inducing kinase signaling path-
way mediating angiotensin II-induced RelA transactivation. Mol. Endocrinol. 21(9),
2203–2217.
Performing and Optimizing Western Blots 599

Desai, S., Dworecki, B., and Cichon, E. (2001). Direct immunodetection of antigens within
the precast polyacrylamide gel. Anal. Biochem. 297, 94–98.
Kaufmann, S. H., Ewing, C. M., and Shaper, J. H. (1987). The erasable Western blot. Anal.
Biochem. 161, 89–95.
Kwok, T. T., Mok, C. H., and Menton-Brennan, L. (1994). Up-regulation of a mutant
form of p53 by doxorubicin in human squamous carcinoma cells. Cancer Res. 54,
2834–2836.
Mathrubutham, M., and Vattem, K. (2005). Methods and considerations for quantitative
Western blotting using SuperSignalÒ Chemiluminescent Substrates. In ‘‘Thermo Fisher
Scientific Application Note # 12’’. www.thermo.com/pierce.
Mattson, D. L., and Bellehumeur, T. G. (1996). Comparison of three chemiluminescent
horseradish peroxidase substrates for immunoblotting. Anal. Biochem. 240, 306–308.
Patonay, G., and Antoine, M. D. (1991). Near-infrared fluorogenic labels: New approach to
an old problem. Anal. Chem. 63, 321A–327A.
Ramlau, J. (1987). Use of secondary antibodies for visualization of bound primary reagents in
blotting procedures. Electrophoresis 8, 398–402.
Sasse, J., and Gallagher, S. R. (2008). Detection of proteins on blot transfer membranes.
Curr. Protoc. Immunol. Chapter 8, Unit 8.10B.
Sato, M., Nishi, N., Shoji, H., Kumagai, M., Imaizumi, T., Hata, Y., Hirashima, M.,
Suzuki, S., and Nakamura, T. (2002). Quantification of Galectin-7 and its localization
in adult mouse tissue. J. Biochem. 131, 255–260.
Sowell, J., Strekowski, L., and Patonay, G. (2002). DNA and protein applications of near-
infrared dyes. J. Biomed. Opt. 7, 571–575.
Spinola, S. M., and Cannon, J. G. (1985). Different blocking agents cause variation in the
immunologic detection of proteins transferred to nitrocellulose membranes. J. Immunol.
Meth. 81, 161.
Sun, L., and Carpenter, G. (1998). Epidermal growth factor activation of NF-kappaB is
mediated through IkappaBalpha degradation and intracellular free calcium. Oncogene 23,
2095–2102.
Tipsmark, C. K., Strom, C. N., Bailey, S. T., and Borski, R. J. (2008). Leptin stimulates
pituitary prolactin release through an extracellular signal-regulated kinase-dependent
pathway. J. Endocrinol. 196(2), 275–281.
Towbin, H., Staehelin, T., and Gordon, J. (1979). Electrophoretic transfer of proteins from
polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl.
Acad. Sci. USA 76, 4350–4354.
Vattem, K., and Mathrubutham, M. (2005). Factors that cause the appearance of ghost bands
when using chemiluminescent detection systems in a Western blot. In ‘‘Thermo Fisher
Scientific Application Note # 11’’. www.thermo.com/pierce.
Walker, G. R., Feather, K. D., Davis, P. D., and Hines, K. K. (1995). SuperSignalTM
CL-HRP: A new enhanced chemiluminescent substrate for the development of the
horseradish peroxide label in Western blotting applications. J. NIH Res. 7, 76.
Xing, C., and Imagawa, W. (1999). Altered MAP kinase (ERK1, 2) regulation in primary
cultures of mammary tumor cells: Elevated basal activity and sustained response to EGF.
Carcinogenesis 20, 1201–1208.
 C H A P T E R T H I R T Y- F O U R

Detergents: An Overview
Dirk Linke

Contents
1. Introduction 604
2. Detergent Structure 604
3. Properties of Detergents in Solution 605
3.1. Phase diagrams and critical micellization concentration 609
3.2. Effects of temperature: Detergent solubility, krafft point,
cloud point, and phase separation 611
4. Exploiting the Physicochemical Parameters of Detergents for
Membrane Protein Purification 612
5. Detergent Removal and Detergent Exchange 613
6. Choosing the Right Detergent 613
6.1. Optical spectroscopy 614
6.2. Mass spectrometry and nuclear magnetic resonance (NMR)
measurements 615
6.3. Protein crystallization 615
7. Conclusions 615
Acknowledgments 616
References 616

Abstract
Detergents are used in molecular biology laboratories every day. They are
present in cell lysis buffers (e.g., in kits for plasmid isolation), in electrophoresis
and blotting buffers, and, most importantly, they are used for cleaning laboratory
glassware and the hands of the laboratory staff. For these routine applications,
a detailed knowledge of detergent properties is not necessary—they just work.
When it comes to the isolation and purification of membrane proteins, one cannot
rely on routine protocols. Many membrane proteins are only stable in a small
number of different detergent buffer systems, and worst of all, different mem-
brane proteins prefer very different detergents. Unfortunately, detergent proper-
ties are considered the domain of colloid science or physical chemistry, and thus,
while the available amount of physico-chemical data on detergents is astound-
ing, this data is rarely compiled in a way that is useful to biochemists. The aim of

Department I, Protein Evolution, Max Planck Institute for Developmental Biology, Tübingen, Germany

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63034-2 All rights reserved.

603
604 Dirk Linke

this chapter is to provide an overview of the physical and chemical properties of

detergents commonly used in membrane protein science and to explain how
these properties can be exploited for protein purification.

1. Introduction
What is a detergent? The term detergent is used in many different
contexts. A household detergent is a complex mixture of surfactants (to dissolve
grease), abrasives (to scour), chelating agents (to counter the effect of ions that
contribute to water hardness), oxidants (for bleaching), enzymes (to degrade
fats, proteins, or complex carbohydrates), colors, perfumes, optical brighteners,
buffer substances (to stabilize the pH), and a number of other stabilizing
ingredients, for example, to modify the foaming properties or to inhibit
bacterial or fungal growth. The term is sometimes also used to distinguish
soap from other cleaning agents (detergents). In molecular biology laboratories,
the term detergent is typically used as a synonym for surfactant. Surfactants
(‘‘surface acting agents’’) are wetting agents that lower the surface tension of a
liquid, and the interfacial tension between two liquids. The chemical and
physical basis of these properties is discussed in detail in this chapter. Obviously,
surfactant is the more precise term, but in scientific literature, it is only used in
the context of physical chemistry and colloid science; in biochemistry, the term
detergent is widespread, and the author of this chapter is not going to try and
change this—from here on, detergent is used as a synonym for surfactant.
Historically, the first detergents used were saponins and soap (or rather,
the fatty acid salts therein). Soap is made from vegetable oils or animal fats
which are hydrolyzed by lye to yield fatty acid salts and glycerol; this process
was already used in ancient Babylon and Egypt, and was brought to
perfection by chemists during the early middle ages in the Middle East.
Saponins are natural detergents that can be extracted from plants, for
example from Soapwort (Saponaria officinalis) or Soapnut (Sapindus). While
soaps are not used in membrane protein science, the saponin Digitonin from
Purple Foxglove (Digitalis purpurea) is. The first synthetic detergents were
developed and used during World War I in Germany, when soap was
scarce. Synthetic detergents have replaced soap in many household applica-
tions since then. Most detergents used in laboratory applications today were
originally developed as constituents of laundry or industry detergents.

2. Detergent Structure
Detergents (or surfactants) are organic compounds of very diverse
structure. Generally speaking, detergent molecules consist of two parts: an
extended, hydrophobic hydrocarbon moiety, and a polar or charged
Detergents: An Overview 605

headgroup. In the simplest case, the hydrocarbon part of a detergent is an

unbranched, saturated alkane; hydrophobicity is then increased with
increasing chain length. Alternatively, unsaturated or branched-chain
alkanes, aromatic hydrocarbons or steroid moieties are found in the lipo-
philic group of detergents, sometimes in combination. The hydrophilic
headgroup is even more variable.
The most simple classification of detergents is based on the four basic
types of headgroups, namely nonionic, anionic, cationic, or zwitterionic
detergents; representative members of the four classes are shown in
Table 34.1. Various other classification systems have been introduced to
grasp the complexity of detergent structures and properties. One of the
standard classifications is based on hydrophile–lipophile balance number
(HLB). It describes the balance between type and size of the hydrophilic
and lipophilic parts of a detergent molecule and determines the water
solubility of a detergent (Neugebauer, 1990). HLB values of surface-active
substances range from 0 to 40, but detergents typically have an HLB of
12–15. Nonionic detergents with a lower HLB (<10) are not water-soluble
and are used as antifoaming agents or to emulsify water in oil, while those
with a higher HLB (>16) are used as stabilizers. Note that HLB is a
numerically calculated number based on the detergent’s molecular struc-
ture; it is not a measured parameter. The HLB is a useful tool to classify
nonionic detergents, but it does not help to compare ionic with nonionic
detergents. Sodium dodecylsufate (SDS), for example, has a calculated HLB
of 40 and is a good laboratory detergent, while any nonionic substance with
this high HLB would not be a useful detergent.

3. Properties of Detergents in Solution

Detergents in aqueous solutions can form spherical aggregates of a
defined size, called micelles. In these micelles, the hydrophobic tails of the
detergent molecules form the core of the sphere, while the hydrophilic
headgroups occupy the interface to the aqueous solution. The number of
detergent molecules in a micelle is called the aggregation number. The
overall shape of a detergent molecule determines its propensity to form
micelles; this is expressed as the packing parameter P, defined as
v
P¼
al
where v is the volume of the detergent chain, a is the cross-sectional area of
the headgroup, and l is the length of the hydrophobic chain. If P is small
(<1/3), spherical micelles are formed, while for bigger P values (>1/2),
Table 34.1 Common laboratory detergents and their properties

Micellar Cloud point

CMC (mM) Aggregation weight ( C) in low- PubChem
Detergent Headgroup Tail (at 25  C) number MW (Da) (kDa) salt buffersa substance IDsb Comments

Nonionic detergents
Big CHAP Gluconamidopropyl Cholesterol 2.9–3.4 10 878.1 9 – SID: 26758300
moieties (2) derivative
Deoxy-Big CHAP Gluconamidopropyl Cholesterol 1.1–1.4 11 862.1 11 – SID: 26758553
moieties (2) derivative
Digitonin Complex Cholesterol – 60–70 1229.3 – – SID: 168187 Natural compound
polysaccharide derivative with high lot-to-lot
variability
Brij-35 (C12E23) Linear PEG (23) Linear hydrocarbon 0.09 40 1199.6 48 >100 SID: 24898176
alcohol (C12)
C12E8 Linear PEG (8) Linear hydrocarbon 0.11 123 538.8 66 74–79 SID: 24898996
alcohol (C12)
C10E4 Linear PEG (4) Linear hydrocarbon 0.64–0.81 54 334.5 18 20 SID: 3729667
alcohol (C10)
C10E6 Linear PEG (6) Linear hydrocarbon – 74 422.6 31 – SID: 24874132
alcohol (C10)
C8E4 Linear PEG (4) Linear hydrocarbon 6.5–8.5 – 306.4 26 35–40 SID: 24900138
alcohol (C8)
C8POE Linear PEG Linear hydrocarbon 6.6 – 330 (average) – 58 Mixture of molecules
alcohol (C8) with different
headgroup lengths
(E ¼ 2–9)
Triton X-100 Linear PEG p-(2,2,4,4- 0.17–0.3 100–150 630 (average) 80 64–65 SID: 7889640 Mixture of molecules
Tetramethylbutyl) with different
phenol headgroup lengths
(average 9.6),
strong UV
absorption
Triton X-114 Linear PEG p-(2,2,4, 0.2–0.35 – 540 (average) – 20–25 SID: 24902129 Mixture of molecules
4-Tetramethylbutyl) with different
phenol headgroup lengths
(average 7–8)
classic detergent for
phase separation,
strong UV
absorption
NP-40 Linear PEG p-(2,2,4, 0.3 100–150 600 60–90 63–67 SID: 813297 Mixture of molecules
4-Tetramethylbutyl) (average) with different
phenol headgroup lengths
(average 9.0),
strong UV
absorption
Tween-20 Polysorbate Linear fatty 0.059 – 1230 – 76 SID: 47207229 Mixture of molecules
acid (C12) (average) with different
headgroup lengths
Tween-80 Polysorbate Linear fatty acid, 0.012 58 1310 76 93 SID: 7848130 Mixture of molecules
unsaturated (C18:1) (average) with different
headgroup lengths
b-Dodecylmaltoside b-Glycosidic maltose Linear hydrocarbon 0.15 98 510.6 70 <0 SID: 691815
alcohol (C12)
b-Decylmaltoside b-Glycosidic maltose Linear hydrocarbon 1.6 – 482.6 – <0 SID: 24894148
alcohol (C10)
b-Octylglucoside b-Glycosidic glucose Linear hydrocarbon 20–25 84 292.4 25 <0 SID: 205023
alcohol (C8)
b-Octylthioglucoside b-Glycosidic glucose Linear hydrocarbon 9 – 308.4 – 7 SID: 698369
thiol (C8)
MEGA-8 N-Methylglucamine Linear fatty acid (C8) 58 – 321.5 – – SID: 737553
MEGA-9 N-Methylglucamine Linear fatty acid (C9) 19–25 – 335.5 – – SID: 737552
MEGA-10 N-Methylglucamine Linear fatty acid (C10) 6–7 – 349.5 – – SID: 751715
LDAO N-Oxide Linear hydrocarbon 2.1–8.3 74 229.4 17 – SID: 158752 Can also be considered
amine as a zwitterionic
detergent

(continued)
Table 34.1 (continued)

Micellar Cloud point

CMC (mM) Aggregation weight ( C) in low- PubChem
Detergent Headgroup Tail (at 25  C) number MW (Da) (kDa) salt buffersa substance IDsb Comments

Zwitterionic detergents
CHAPS Dimethylammonium- Cholesterol derivative 6.5 4–14 614.9 6 SID: 684915
1-propanesulfonate
CHAPSO Dimethylammonium- Cholesterol derivative 8 11 630.9 7 SID: 698662
1-propanesulfonate
SB-10 Dimethylammonium- Linear hydrocarbon 25–40 41 307.6 12.5 SID: 738617 Zwittergent 3-10
1-propanesulfonate alcohol (C10)
SB-12 Dimethylammonium- Linear hydrocarbon 2–4 55 335.6 18.5 SID: 661588 Zwittergent 3-12
1-propanesulfonate alcohol (C12)
SB-14 Dimethylammonium- Linear hydrocarbon 0.1–0.4 83 363.6 30 SID: 661589 Zwittergent 3-14
1-propanesulfonate alcohol (C14)
Anioinic detergents
Cholate, sodium salt Carboxylic acid Cholesterol derivative 9–15 2 430.6 0.9 SID: 152893 Typically as sodium
salt
Desoxycholate, Carboxylic acid Cholesterol derivative 4–8 4–10 414.6 1.7–4.2 SID: 668198 Typically as sodium
sodium salt salt
Lauroylsarcosine Methylaminoacetate Linear fatty acid (C12) 2 293.4 0.6 SID: 151872 Typically as sodium
(‘‘Sarkosyl’’), salt
sodium salt
Sodium dodecylsulfate Sulfate Linear hydrocarbon 7–10 62 288.4 18 SID: 152210 Also available as
(SDS) alcohol (C12) lithium salt, with
better solubility at
cold temperatures
Cationic detergents
Cetyltrimethyla- Quarternary amine Linear hydrocarbon 0.9 170 364.5 62 SID: 148809 Typically as bromide
mmonium bromide (C16) salt or chloride salt
(CTAB)
Dodecyltrimethy- Quarternary amine Linear hydrocarbon (C12) 15 70 280.3 20 SID: 159639 Typically as bromide
lammonium salt or chloride salt
bromide (DTAB)
a
Note that cloud points are influenced by the buffer composition, see text.
b
Link to PubChem structures and annotations (http://www.ncbi.nlm.nih.gov/sites/entrez?db¼pcsubstance).
Data were compiled from several reviews (Arnold and Linke, 2007, 2008; Helenius et al., 1979; Hinze and Pramauro, 1993; Neugebauer, 1990), and from datasheets of commercial suppliers of detergents.MW, molecular weight;
CMC, critical mizellization concentration.
Detergents: An Overview 609

P > 1/2

P < 1/3

Figure 34.1 The formation of micelles depends on the molecular shape of the deter-
gent. It is best described by the packing parameter P that is calculated by comparing the
headgroup volume with the hydrophobic chain volume and length.

detergent molecules arrange preferentially into lamellar aggregates

(Fig. 34.1) (Neugebauer, 1990). Intermediate forms, for example cylindrical
micelles, exist. It has been shown that the packing parameter P generally
correlates to the HLB parameter described above, with smaller HLB values
resulting in bigger P values.
It is the formation of micelles that is the basis for membrane protein
solubilization. The hydrophobic parts of the membrane proteins are
covered by detergent molecules, such that the protein sticks out of the
surrounding micelle only by its hydrophilic extensions. It is thus important
to understand all physical parameters that influence the formation and
destruction of micelles. In addition to the packing parameter P mentioned
above, the concentration of the detergent, as well as temperature, pH,
ionic strength, and presence of other additives in the buffer play an
important role.

3.1. Phase diagrams and critical micellization concentration

Detergents only form micelles in a defined concentration and temperature
range. This is best described by a graph, called the phase diagram (Fig. 34.2).
At a given temperature, the minimal detergent concentration at which
micelles are observed is called the critical micellization concentration
(CMC). Below this concentration, only detergent monomers exist in the
solution; above the CMC, detergent monomers are in equilibrium with the
detergent micelles. At high detergent concentrations, other, nonmicellar
phases exist, which are typically immiscible with water. These phases
are usually liquid-crystalline in nature; they can be hexagonal, reverse
hexagonal, or lamellar in structure (Fig. 34.2 shows a lamellar structure).
610 Dirk Linke

A
Phase separation

Temperature

Micellar

Monomeric Liquid crystalline

Concentration
B
Temperature

Micellar

Liquid
Monomeric Phase separation crystalline

Concentration

Figure 34.2 Simplified phase diagrams. Panel A shows a detergent with a lower
consolute boundary. Most nonionic detergents fall into this group. Panel B shows a
detergent with an upper consolute boundary. A number of glycosidic and zwitterionic
detergents fall into this group. Note that the liquid-crystalline phase can also consist of
hexagonally packed cylindrical micelles.

An alternative phase type which usually occurs at higher temperatures is

referred to as phase separation, and describes disordered aggregates of
micelles.
Detergents: An Overview 611

3.2. Effects of temperature: Detergent solubility, krafft point,

cloud point, and phase separation
Temperature has drastic effects on the solubility and phase behavior of
detergents, and so has the detergent concentration in aqueous solution. As
membrane proteins are typically solubilized with the help of micelles, it is
important to know the temperature and concentration ranges in which a
given detergent will form micelles at all. The Krafft point describes the
temperature above which detergent solubility increases to above the CMC
(i.e., at temperatures below the Krafft point, no micelles are formed). The
Krafft point is a temperature value that is specific for each detergent (Gu and
Sjoblom, 1992) (Fig. 34.3). The same is true for the cloud point, which
describes the temperature at which micelles start to aggregate and become
immiscible with water.
The line in the phase diagram that forms the border between the micellar
solution and the region of phase separation is called the consolute boundary
(Arnold and Linke, 2007). While for most detergents, the cloud point is at a
relatively high temperature that has to be exceeded for phase separation
(crossing the so-called lower consolute boundary of the detergent), some
detergents have a very low cloud point, and phase separation has to fall
below that temperature value for phase separation to occur (crossing the
so-called higher consolute boundary of the detergent) (Fig. 34.2). Note that
the properties of detergents (phase diagram, Krafft point, cloud point,
consolute boundaries) change dramatically with increasing ionic strength,
Concentration

y
lit
bi
lu
So

CMC

Krafft point

Temperature

Figure 34.3 Definition of the Krafft point. The gray area describes the possible
concentrations of detergent soluble at a given temperature. Above the Krafft point,
detergent solubility increases dramatically as micelles can be formed.
612 Dirk Linke

and thus in many cases have to be determined experimentally for a given

buffer system. The same is true for the presence of nonionic solutes like
polyethylene glycols, sugars, or glycerol, among others.

4. Exploiting the Physicochemical Parameters

of Detergents for Membrane Protein
Purification
As mentioned above, the presence of micelles is essential for keeping
membrane proteins in solution. Thus, a membrane protein solution needs
to contain detergent at a concentration above the CMC; additionally,
enough detergent needs to be available to solubilize all membrane proteins
in the solution (simply speaking, there should be at least one micelle per
protein molecule). This can be a problem at high protein concentrations,
or when working with detergents that have a very low CMC (e.g.,
b-dodecylmaltoside). The molar concentration of micelles in the solution
can be calculated from two specific detergent properties, the aggregation
number (the number of detergent molecules in a micelle which depends
on the packing parameter), and the molecular weight of the detergent
monomer. Table 34.1 lists common laboratory detergents with all relevant
parameters, including the molecular weight of the micelles.
The first step in membrane protein purification is typically the solubili-
zation of a native membrane. Again, enough detergent needs to be available
in the solution to accommodate all proteins in micelles; moreover, the
membrane lipids form mixed micelles with the detergent used, changing
the properties of the micelles. For each solubilization process, the optimal
detergent:protein and detergent:lipid ratio needs to be determined, which
unfortunately is a trial-and-error process. Which detergent to use strongly
depends on the downstream processing planned for the membrane proteins.
Note that some detergents selectively solubilize certain types of membranes.
Examples are Sarkosyl and Triton X-100, which solubilize bacterial inner,
but not outer membranes (Arnold and Linke, 2008), and Digitonin, which
can be used to selectively solubilize eukaryotic plasma membranes
(Mooney, 1988).
Phase separation effects can be exploited for membrane protein separa-
tion (Arnold and Linke, 2007). In principle, this works for all detergents; the
basic idea is to shift the cloud point of the detergent used to ambient
temperature by adding salts or other buffer additives to the system, or to
shift the temperature to cross the consolute boundary. Care has to be taken
that the high salt concentrations or high temperatures involved do not harm
the membrane protein which is to be purified. After phase separation has
occurred, the detergent-rich phase that contains the membrane proteins can
Detergents: An Overview 613

be collected, while soluble proteins and other hydrophilic substances remain

in the aqueous phase. As the volume of the detergent-rich phase is only a
fraction of the total volume, this method can also be used to concentrate
membrane proteins. After the procedure, excess detergent needs to be
removed, for example, by dialysis.

5. Detergent Removal and Detergent Exchange

While the solubilization of membranes requires high amounts of
detergent, such high concentrations are usually not wanted in downstream
applications. Thus, the detergent concentration has to be lowered, which
can of course be easily achieved by dilution. In cases where low concentra-
tions of detergent are wanted, several methods to specifically reduce the
detergent amount are available. Size exclusion chromatography using a
low-detergent buffer is efficient for detergent removal or exchange if the
micellar size is significantly different from the size of the membrane protein
in question (Furth et al., 1984). Dialysis can be used to remove detergent or
exchange it with a different one, provided that the detergents used are
dialyzable. Detergents can only be dialyzed if their micelles are sufficiently
small to pass the dialysis tubing as a whole, or if the detergents have a high
CMC so that a relatively high amount of detergent is present in a mono-
meric, not micellar, form. Other efficient methods to remove detergents are
the use of BioBeads or similar hydrophobic resins that will bind detergents
(Rigaud et al., 1998), or the additions of cyclodextrins that can form
inclusion complexes with monomeric detergent molecules (DeGrip et al.,
1998). Alternatively, affinity or ion-exchange chromatography can be used
to bind the protein to a column, wash away excess detergent, and elute with
a suitable low-detergent buffer (Arnold and Linke, 2008).

6. Choosing the Right Detergent

Unfortunately, there are no detergents that are suitable for all mem-
brane proteins. While this needs to be tested for every membrane protein/
detergent combination, some general rules apply. Typically, ionic deter-
gents are harsher on proteins than nonionic or zwitterionic ones. Complete
denaturation of proteins by detergents like SDS or CTAB may be a
welcome effect in polyacrylamide electrophoresis or in DNA isolation
techniques, but for the isolation and purification of intact membrane pro-
teins, these detergents are rarely used. As a rule of thumb, detergents with
614 Dirk Linke

bigger headgroups and longer hydrocarbon chains have milder properties,

while short-chain detergents have a higher propensity to denature proteins
or to disrupt membrane protein complexes (Privé, 2007).
Even if a detergent is suitable for a membrane protein in question
(i.e., does solubilize it without denaturing it), it might be incompatible
with downstream applications. Even simple biochemical assays are influ-
enced by the presence of detergents. A number of known incompatibilities
are listed in this section. This list is far from complete, and thus, whenever
using detergent buffers, a possible detrimental effect on the experiment has
to be taken into account. Note also that the purity of the detergents used
can pose significant problems. Many detergents contain trace amounts of
peroxides, or residual organic solvents from the synthesis procedure (Ashani
and Catravas, 1980). For many sensitive applications, it is highly recom-
mended to use only detergents of the highest available purity, and to always
prepare detergent solutions freshly to prevent oxidation or hydrolysis.

6.1. Optical spectroscopy

Protein concentration measurements using UV absorption are not possible
in the presence of Triton X-100 and related detergents that contain aromatic
rings in their structure (NP-40, e.g., is a derivative of Triton X-100), because
they strongly absorb UV light at 280 nm wavelength. Many colorimetric
assays used for protein concentration measurements are not compatible with
detergents. Generally speaking, the BCA assay is more resistant to detergents
than either Bradford or Lowry assays, but in many cases, detergents have to
be removed, for example, by precipitation. Commercial suppliers have
developed numerous variations of the known colorimetric assays to over-
come these problems at least partially, and it is strongly recommended to
refer to the instruction booklets of these assays when samples containing
detergents are to be measured. It is noteworthy that recently, a new assay
technology based on a transition metal complex for colorimetric readout at
660 nm wavelength (patent pending) has been developed, which is supposed
to be fully compatible with most detergents even in the presence of reducing
and chelating agents (Antharavally et al., 2009).
Many other spectroscopic techniques are influenced by detergents,
including linear and circular dichroism spectroscopy, where detergents
with chiral headgroups can produce a significant signal, or infrared spec-
troscopy, where the hydrophobic as well as the hydrophilic groups absorb at
wavelengths relevant to protein biochemistry. Detergents strongly bind to
hydrophobic dyes, influencing, for example, their fluorescence spectra, or
their behavior in other colorimetric assays.
Detergents: An Overview 615

6.2. Mass spectrometry and nuclear magnetic resonance

(NMR) measurements
Detergents are generally detrimental to mass spectrometry applications, as
they themselves can be ionized, leading to strong signals in different detec-
tor systems. As a rule of thumb, MALDI-MS is more resistant to detergents
than ESI-MS, but in general, detergents need to be removed. Some spe-
cialty detergents have been developed to overcome these problems, includ-
ing cleavable detergents for MALDI-MS (Norris et al., 2003) or
perfluorinated detergents for ESI-MS (Ishihama et al., 2000). Special deter-
gents have also become available for NMR spectroscopy, where the deter-
gent micelle adds to the apparent molecular weight of the protein to be
analyzed, leading to slower motion in solution, and thus, to more isotropic
signals. Small, lipid-like detergents like dihexanoylphosphatidylcholine have
been successfully employed in NMR studies of membrane proteins
(Fernandez and Wuthrich, 2003).

6.3. Protein crystallization

Detergents produce a number of problems when using them in protein
crystallization assays. The most basic problem is that detergents are simply
another parameter that needs to be checked in empirical screens. Thus, they
simply add another level of complexity; the optimal detergent, and its optimal
concentration for crystallization of a membrane protein has to be found.
Moreover, detergents tend to phase-separate under the high-salt or high-
PEG conditions typically used in protein crystallization screens. And last but
not least, detergent purity can be an issue; as an example, it has been shown
that even trace amounts of a-dodecylmaltoside in b-dodecylmaltoside buffers
can effectively ruin crystal quality (Fromme and Witt, 1998).

7. Conclusions
The most important conclusion, when talking about detergents in
membrane protein research, is that one always has to consider or even test
whether a certain membrane protein–detergent combination is suitable for
the isolation and purification procedure, but also for the downstream
applications planned. To even increase the complexity of the problem,
some membrane proteins are not stable without at least a little amount of
native membrane lipids; this implies that solubilization and purification
should in some cases not be complete. It also implies that sometimes
mixtures of detergents (or lipids) might be more suitable than a pure
substance. This is, in part, realized by commercially available detergent
616 Dirk Linke

mixtures, that are indeed sometimes useful in stabilizing membrane

proteins. Prominent examples are Digitonin, which is never pure as it is
isolated from a plant, or Triton X-100, which actually consists of several
different species with different polyethyleneglycol headgroup lengths.
As biochemists have a tendency to keep their buffer recipes as simple as
possible (and the author of this chapter is no exception from that rule),
deliberate mixtures of, for example, ionic and nonionic detergents are rarely
found in literature. One reason for this surely is that parameters like CMC,
cloud point, or micellar size are impossible to predict for mixtures of
detergents. Thus, finding the right detergents to use is still, at least to
some extent, a trial-and-error process.

ACKNOWLEDGMENTS
The author thanks Andrei Lupas for continuing support. Funding by the Max Planck
Society, the German Science Foundation (SFB766/B4), and the Bill & Melinda Gates
Foundation (Grand Challenges Explorations Program) is gratefully acknowledged.

REFERENCES
Antharavally, B. S., Mallia, K. A., Rangaraj, P., Haney, P., and Bell, P. A. (2009).
Quantitation of proteins using a dye-metal based colorimetric protein assay. Anal.
Biochem. 385, 342–345.
Arnold, T., and Linke, D. (2007). Phase separation in the isolation and purification of
membrane proteins. Biotechniques 43, 427–440.
Arnold, T., and Linke, D. (2008). The use of detergents to purify membrane proteins. Curr.
Protoc. Protein Sci. Chapter 4, 4.8.1–4.8.30.
Ashani, Y., and Catravas, G. N. (1980). Highly reactive impurities in Triton x-100 and
Brij 35: Partial characterization and removal. Anal. Biochem. 109, 55–62.
DeGrip, W. J., van Oostrum, J., and Bovee-Geurts, P. H. M. (1998). Selective detergent-
extraction from mixed detergent/lipid/protein micelles, using cyclodextrin inclusion
compounds: A novel generic approach for the preparation of proteoliposomes. Biochem.
J. 330, 667–674.
Fernandez, C., and Wüthrich, K. (2003). NMR solution structure determination of mem-
brane proteins reconstituted in detergent micelles. FEBS Lett. 555, 144–150.
Fromme, P., and Witt, H. T. (1998). Improved isolation and crystallization of photosystem I
for structural analysis. Biochim. Biophys. Acta Bioenerg. 1365, 175–184.
Furth, A. J., Bolton, H., Potter, J., and Priddle, J. D. (1984). Separating detergent from
proteins. Methods Enzymol. 104, 318–328.
Gu, T. R., and Sjoblom, J. (1992). Surfactant structure and its relation to the Krafft point,
cloud point and micellization—Some empirical relationships. Colloids Surf. 64, 39–46.
Helenius, A., McCaslin, D. R., Fries, E., and Tanford, C. (1979). Properties of detergents.
Methods Enzymol. 56, 734–749.
Hinze, W. L., and Pramauro, E. (1993). A critical review of surfactant-mediated phase
separations (cloud-point extractions)—Theory and applications. Crit. Rev. Anal. Chem.
24, 133–177.
Detergents: An Overview 617

Ishihama, Y., Katayama, H., and Asakawa, N. (2000). Surfactants usable for electrospray
ionization mass spectrometry. Anal. Biochem. 287, 45–54.
Mooney, R. A. (1988). Use of digitonin-permeabilized adipocytes for cAMP studies.
Methods Enzymol. 159, 193–202.
Neugebauer, J. M. (1990). Detergents: An overview. Methods Enzymol. 182, 239–253.
Norris, J. L., Porter, N. A., and Caprioli, R. M. (2003). Mass spectrometry of intracellular
and membrane proteins using cleavable detergents. Anal. Chem. 75, 6642–6647.
Privé, G. G. (2007). Detergents for the stabilization and crystallization of membrane
proteins. Methods 41, 388–397.
Rigaud, J. L., Levy, D., Mosser, G., and Lambert, O. (1998). Detergent removal by
non-polar polystyrene beads—Applications to membrane protein reconstitution and
two-dimensional crystallization. Eur. Biophys. J. Biophys. Lett. 27, 305–319.
 C H A P T E R T H I R T Y- F I V E

Purification of Membrane Proteins

Sue-Hwa Lin* and Guido Guidotti†

Contents
1. Introduction 619
2. Preparation of Membranes 620
3. Solubilization of Native Membrane Proteins 622
4. Purification of Membrane Proteins 625
4.1. Lectin-affinity chromatography 626
4.2. Ligand-affinity chromatography 627
4.3. Antibody-affinity chromatography 627
5. Detergent Removal and Detergent Exchange 628
6. Expression and Purification of Recombinant Integral
Membrane Proteins 628
References 629

Abstract
Membrane proteins are pivotal players in biological processes. In order to
understand how a membrane protein works, it is important to purify the protein
to fully characterize it. Membrane proteins are difficult to purify because they
are present in low levels and they require detergents to become soluble in an
aqueous solution. The selection of detergents suitable for the solubilization and
purification of a specific membrane protein is critical in the purification of
membrane proteins. The aim of this chapter is to provide an overview for the
isolation of plasma membranes, selection of detergents for solubilization of
membrane proteins, and how the choice of detergents may affect membrane
protein purification.

1. Introduction
Membrane proteins are pivotal players in biological processes. They
are responsible for connecting cells to each other and to the cell matrix, for
organizing the shape of the organelles and the cells, and for the transport of

* Department of Molecular Pathology, University of Texas, Houston, Texas, USA

{
Department of Molecular and Cellular Biology, Harvard University, Cambridge, Massachusetts, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63035-4 All rights reserved.

619
620 Sue-Hwa Lin and Guido Guidotti

ions, metabolites, and proteins across plasma membranes, and RNA trans-
port across nuclear membrane. Given the importance of membrane proteins
for a plethora of cellular functions, it is not surprising that around 50% of the
current drugs for a variety of diseases target membrane proteins.
In order to understand how a membrane protein works and to generate
drugs that target specific sites within the protein, it is important to purify
the protein to fully characterize it. However, of all known proteins that
were purified and identified biochemically, only very few were membrane
proteins. This is in a large part due to the fact that membrane proteins are
more difficult to purify for the following reasons. First, most membrane
proteins are present at low levels; and second, they are embedded in the
lipid bilayer and require detergents to become soluble in aqueous systems.
The selection of detergents suitable for the solubilization and purification of
the specific membrane protein is critical in the purification of membrane
proteins. In general, the basic principle for the purification of membrane
proteins are the same as those used for soluble proteins, but modifications in
the purification scheme is required in order to accommodate the require-
ment of detergents that are essential in maintaining the solubility and
function of these proteins.
Unlike DNA or RNA, it is impossible to present a single set of methods
for the purification of all proteins, especially membrane proteins. Each
membrane protein possesses a unique set of physical characteristics, prefer-
ence for detergent for solubilization, and conditions for purification. An
overview of the properties of detergents is described in Chapter 41. This
chapter describes methods for the isolation of plasma membranes, and for
solubilization and purification of membrane proteins.

2. Preparation of Membranes
Isolation of plasma membranes from cells or tissues is the first step in
the purification of a plasma membrane protein. Because of the limited
biochemical fractionation methods available for effective separation of
detergent solubilized membrane proteins, the time invested in purifying
plasma membrane fractions will improve the results at the later stages.
Due to the low levels of most membrane proteins, it is important to
select a tissue or cell line that is easy to obtain a large amount of starting
material for purification and that also expresses high specific activity of the
protein of interest. Recently, there has been increased interest in identifying
cell surface proteins as markers for different cell lineages or stem cells. As a
result, obtaining a sufficient amount of plasma membranes from a limited
amount of cells with relatively good enrichment of plasma membrane
marker enzyme activity has become a new focus of membrane isolation.
Purification of Membrane Proteins 621

The most commonly used method to break up tissues or cells for

preparation of the membrane fraction is to homogenize them in buffered
isotonic sucrose (0.25 M, pH 7–8) with a Dounce homogenizer. Membrane
proteins are relatively stable while they are embedded in the membrane.
The major potential concern for losing activity during the membrane
purification step is proteolysis. During cell disruption, proteases may be
released. A cocktail of protease inhibitors in a convenient tablet form is
commercially available, for example, Complete Protease Inhibitor Cocktail
(Roche, Indianapolis, IN). Extracellular domains of plasma membrane
proteins are in direct contact with an oxidative environment and most of
the sulfhydryl groups are in disulfide bonds. These disulfide bonds are
formed during the synthesis of these proteins in the endoplasmic reticulum.
Thus, reducing agents may not be needed in this step. In fact, reducing
agents, when present in high concentration, may change the existing
disulfide bonded conformation, resulting in inactivation of the enzymatic
or ligand-binding activity of membrane proteins. Due to the disulfide
bonded structure and glycosylation, the extracellular domain of membrane
protein is relatively resistant to proteolysis. However, there are exceptions
to this. For example, the Ca2þ-dependent cell adhesion molecules, cadher-
ins, are susceptible to proteolysis when Ca2þ ions are removed by EDTA.
In contrast, the cytoplasmic domains, which mediate the signal transduction
of membrane proteins, are usually susceptible to proteolysis. For example,
the receptor tyrosine kinase activity of insulin receptor is largely lost if
protease inhibitors are not present in sufficient quantities during homoge-
nization and the subsequent membrane fractionation steps.
The most frequently used membrane fractionation methods employ a
combination of differential centrifugation and sucrose density gradient
centrifugation steps. Due to the differences in lipid and protein composi-
tions, membranes have different densities that allow them to be separated
from other organelles. Differential centrifugation can remove soluble
proteins, the majority of mitochondria and nuclei from the cell homo-
genates. Sucrose density gradients can further separate membranes with
different densities. However, these multiple steps of centrifugation are
lengthy and also yield only a small percentage of the plasma membranes.
Often, much of the plasma membrane is lost in the early steps of centrifu-
gation. Thus, this method is more suitable for isolating membranes from
tissues that are relatively easy to obtain. Rat liver is one of the tissues that
are most commonly used in the isolation of membranes for biochemical
studies. Many methods have been developed for isolating various mem-
branes from rat livers. A method developed by Neville (1968) that homo-
genizes livers in hypotonic solution followed by a discontinuous sucrose
gradient centrifugation gave a very good recovery of liver plasma mem-
branes and has been commonly used in the preparation of liver plasma
membranes.
622 Sue-Hwa Lin and Guido Guidotti

In the situations, when only a limited amount of tissue culture cells are
used, an increase in the recovery of plasma membrane proteins without
sacrificing the purity of the membrane will be needed. Affinity matrices
provide a simple and quick method to use in the purification of membranes.
The conventional agarose- or acrylamide-affinity matrices cannot be used to
isolate membranes because they sediment with the organelles (e.g., nuclei)
that have relatively high densities. Chemically derived magnetic beads can
be coupled with various proteins, and they have become a new form of
affinity matrix. In contrast to conventional agarose- or acrylamide-based
matrices, magnetic beads can be conveniently separated from the mixture by
using a magnet and thus can be used to separate organelles independent of
their densities. By simply holding the tubes near the magnet, the magnetic
beads can be recovered at the sides of the tubes, allowing easy recovery of
the beads from the mixture. Thus, magnetic beads can be used as a substitute
for centrifugation. This property is likely to have great advantages in
separating plasma membranes from other organelles. We have recently
used magnetic beads with immobilized lectin for the purification of plasma
membrane proteins from cultured epithelial cells (Lee et al., 2008).
This procedure takes advantage of the fact that some of the membrane
proteins are glycosylated and can bind to lectins, the carbohydrate-binding
proteins. In this procedure, the lectin Concanavalin A (ConA) is immobi-
lized onto magnetic beads by binding biotinylated ConA to streptavidin
magnetic beads. The ConA-magnetic beads were mixed with homogenized
cell lysates, and the beads were recovered at the sides of the tubes, allowing
removal of other organelles that are not associated with the magnetic beads.
The membranes were found to be associated with the ConA-magnetic
beads as there was an enrichment in the activity of 50 -nucleotidase, a
membrane protein, from a total cell lysate of prostate PC-3 cells or cervical
HeLa cells. One caveat of this lectin magnetic bead method is that we
could not elute membranes from the ConA-magnetic beads by using the
competitive sugar alpha-methyl mannoside, possibly because the competing
sugar cannot access the binding sites between the plasma membranes and
ConA-magnetic beads. As a result, we used detergents to solubilize the
membrane proteins from the ConA-magnetic beads. The use of detergents
for solubilization of membrane proteins is described in Section 3.

3. Solubilization of Native Membrane Proteins

Membrane proteins are embedded in the lipid bilayer. Integral mem-
brane proteins have at least one stretch of protein sequence embedded in the
membrane while peripheral proteins are associated with the membrane
through electrostatic and in some cases hydrophobic interaction. Peripheral
membrane proteins can be dissociated from membranes by using high salt or
Purification of Membrane Proteins 623

high pH solutions (Schindler et al., 2006), for example 0.5 M NaCl. Because
detergents are not used in the procedure, the peripheral membrane proteins
can be purified by methods similar to those applied to soluble proteins.
Integral membrane proteins need to be solubilized from the lipid bilayer
to become individual proteins before purification. Detergents that possess
amphipathic properties are commonly used to solubilize integral membrane
proteins from membranes. Detergents may be grouped into three classes,
ionic, nonionic, and zwitterionic. A discussion of the different types of
detergents is described in Chapter 41 of this volume. Thus, we will only
discuss their use in terms of purification of membrane proteins.
Proteins are considered ‘‘solubilized’’ from the membrane if the proteins
appear in the supernatant fraction after the detergent treated membranes are
centrifuged at 105,000g for 1 h at 4  C. The process of solubilization of
membrane proteins by detergents can be divided into several phases. In the
first phase, the detergent binds to membranes. As the amount of detergent is
increased, the detergent starts to lyse membranes. Further increase in deter-
gent will lead to the formation of lipid/protein/detergent complexes.
At this stage, the membrane proteins are ‘‘solubilized.’’ Additional amount
of detergent will be needed to ‘‘delipidate’’ the complexes to protein/
detergent and lipid/detergent complexes. Usually, a detergent-to-protein
ratio of around 1–2 is sufficient to solubilize the membrane proteins into
lipid/protein/detergent complex. A detergent-to-protein ratio of around
10 or higher will lead to delipidation (Hjelmeland, 1990). The optimal
detergent to protein ratio that is required to solubilize a specific membrane
protein needs to be empirically determined.
The choice of detergents can be simply stated as the detergent that works
for your protein of interest. Nondenaturing detergents are those detergents
that solubilize the membrane proteins without significantly inactivating the
activity or function of the proteins. Among detergents that are available,
Triton X-100, sodium cholate, CHAPS, octylglucoside are ‘‘nondenatur-
ing’’ most of the time, although a loss of certain fraction of activity during
solubilization is expected.
The presence of detergents will affect several aspects of protein purifica-
tion. The detergents may affect the assay for the activity, for example
transport activity for membrane transporter and ligand-binding assay for a
receptor. As the proteins will no longer be associated with the membranes,
reconstitution of solubilized membrane proteins into phospholipid vesicles
will be needed for measuring transporter activity, and methods that separate
unbound ligand from the ligand–receptor complex will be needed to be
developed for the specific receptor. These requirements may limit the type
of detergents to be used for solubilization. For example, detergents with
high critical micelle concentration (sodium cholate, CHAPS, octylgluco-
side), which are easier to be removed by dialysis, should be used if a
subsequent reconstitution of proteins into phospholipid vesicles is planned.
624 Sue-Hwa Lin and Guido Guidotti

Some types of detergents may interfere with certain protein assays.

The polyoxyethylene derivatives, for example Triton X-100, C12E9, and
Tween series, give false positive in Commassie blue-G250 dye-binding
assays (also known as Bradford assay) (Bradford, 1976). Sodium cholate or
sodium deoxycholate forms precipitate in the Bradford assay. Triton X-100
and NP-40 absorb at 280 nm and interfere with the use of ultraviolet
absorbance method to monitor the chromatographic elution of proteins.
This will require either a change of the protein assay method or choice of
detergents that are compatible with the protein assay. The Bicinchoninic
acid (BCA) protein assay method (Smith et al., 1985) is compatible with
certain amounts of detergents. A modified method based on the Lowry
protein assay (Peterson, 1977) avoids the interference of detergents by first
precipitating the proteins from the solution with deoxycholate and trichlor-
oacetic acid (TCA), and can be used to measure protein concentration
throughout the purification steps. However, the modified Lowry method
takes a longer time compared with the Bradford method to obtain protein
concentrations.
In some cases, whether the detergent is chemically pure is critical. For
example, if the protein is to be crystallized for X-ray structural analysis or
analyzed by mass spectrometry, the impurity in the detergents may interfere
with these analyses. Polyoxyethylene derivatives, for example Triton
X-100, Lubrol PX, the Tween, and Brij series, contain varied polymer
lengths and are not chemically homogeneous. Octylglucoside, dodecyl
maltoside, and CHAPS have well-defined chemical composition and can
be obtained in relatively high purity. Noninonic detergents, including the
polyoxyethylene derivatives, for example Triton X-100 and the Tween
series, are less effective at dissociating protein complexes, but many proteins
are more stable in nonionic detergents than in ionic detergents.
In screening of detergents for solubilization of integral membrane
proteins, one should first identify the detergents that can solubilize the
protein without inactivating its activity. Among those detergents that fulfill
these requirements, one should then consider other factors, including the
compatibility with the methods used for protein determination, characteri-
zation, and chromatography methods. The detergent interference with
various chromatography methods will be discussed in Section 4.
Detergents are screened by preparing membrane fractions at a specific
protein concentration, for example 1 mg/ml. The solubilization solutions
with a range of detergent concentrations, for example 0.2–20 mg/ml, will
then be added to the membrane preparation. Solubilization is usually
achieved at detergent/protein ratios of 0.1–10 (w/w). Solutions are incu-
bated at 0–4  C for 30–60 min and then centrifuged at 105,000g for 1 h at
4  C. The activities of the protein of interest in both the supernatant and the
pellet fractions are then measured. If the majority of the activity is found to
be present in the supernatant fraction, this detergent is suitable for the
Purification of Membrane Proteins 625

membrane protein. If the activity is not detected in the supernatant fraction

but remains in the pellet fraction, the detergent either cannot solubilize the
protein from the membrane or the amount of the detergent used may not be
sufficient. If the activity is not detected in either the supernatant or the pellet
fractions, the detergent has ‘‘denatured’’ the protein.

4. Purification of Membrane Proteins

Once a suitable detergent has been used to solubilize membrane
proteins from the membranes, fractionation methods can be used to isolate
the specific protein of interest. Conventional chromatographic techniques
including gel filtration, affinity, ion exchange, and chromatofocusing can be
applied in the purification of membrane protein. However, the following
precautions are noted when performing chromatography in the presence
of detergents.
1. Include a level of detergent sufficient to keep the integral membrane
proteins in a soluble form in the buffer and prevent protein aggregation.
2. Due to the hydrophobic nature of most detergents, protein separation
methods that are based on hydrophobicity of proteins, for example
phenyl-Sepharose and reverse-phase, may not be suitable for membrane
protein purification.
3. Solubilization of membrane proteins with ionic detergents, for example
cholate and deoxycholate, are not suitable for ion-exchange chromatog-
raphy. Nonionic or zwitterionic detergents should be used for charge-
based preparative procedures including ion-exchange chromatography
and preparative electrophoresis.
4. Detergents that contain a sugar moiety may interfere with specific lectin
chromatography, for example octylglucoside may interfere with ConA
chromatography.
5. Because the solubilized membrane proteins are present in detergent
micelles, membrane proteins have much bigger apparent molecular
weight in gel filtration. Detergents that form large micellar molecular
weight, for example Triton X-100, may add 60–100 kDa to the size of
solubilized membrane proteins. Thus, most of the proteins will appear in
the high molecular weight fractions, leading to difficulty in separating
proteins based on molecular sizes.
6. Association of membrane proteins with detergent micelles, especially
those with large micellar sizes or that are ionic in nature, obscures the
charge of the membrane proteins. Thus, the ability of ion-exchange
chromatography in separating membrane proteins may not be as good as
for nonmembrane proteins.
626 Sue-Hwa Lin and Guido Guidotti

7. In general, affinity chromatography is by far the most useful and success-

fully applied method for purification of integral membrane proteins and
can be used at various purification steps. Because ion-exchange chroma-
tography is sensitive to the ionic strength of the buffer, and gel filtration
requires a concentrated sample of relatively small volume, affinity
chromatography can be used to purify, concentrate, and perform salt
exchange between different chromatography steps. Several affinity
chromatographies that are commonly used in membrane protein purifi-
cation are described in the following sections.

4.1. Lectin-affinity chromatography

There are three types of affinity chromatography, including the use of a
general ligand (e.g., lectins), specific ligands (e.g., enzyme inhibitors, hor-
mones), and antibodies. Immobilized lectin is a general form of affinity
chromatography. Lectins are carbohydrate-binding proteins, which offer a
rapid and mild method to purify plasma membrane glycoproteins. Lectin–
glycoprotein interactions are reversible and can be inhibited by simple
sugars. Because membrane proteins are frequently glycosylated, lectin-
affinity chromatography is very useful for membrane protein purification.
Numerous lectins have been identified with the most widely used lectins to
be concanavalin A (binds a-D-mannose) and wheat germ agglutinin (binds
sialic acid and b-D-GlcNAc). Several aspects about using lectin-affinity
chromatography need to be mentioned here. First, whether the protein of
interest will bind to a specific lectin needs to be empirically tested. Some-
times, due to different glycosylation enzymes present in different tissues, the
same glycoprotein expressed in different tissues of the same animal may not
have the same lectin-binding specificity. Second, a specific lectin-affinity
chromatography purifies a group of proteins with a specific type of glyco-
sylation, thus this method is unable to achieve the extent of purification as
ligand or antibody chromatography. Third, lectins are sensitive to certain
types of detergents. Although nonionic detergents, for example Triton
X-100 (up to 2.5%, w/v), have negligible effects on the binding of
concanavalin A or wheat germ agglutinin with their ligands, some ionic
detergents, for example SDS, may inactivate lectins.
A protocol for a typical use of a lectin column for the purification of
membrane proteins is described here.
1. Wash the WGA-agarose affinity matrix by transferring 2 ml of wheat
germ agglutinin (WGA)-agarose to a disposable plastic column (Poly-
Prep, Bio-Rad Laboratories). Fill the column with 10 ml wash buffer
(50 mM HEPES, pH 7.4, containing 0.1% detergent). Let wash run
through column to remove free uncoupled WGA and the storage buffer.
Purification of Membrane Proteins 627

2. Add washed WGA-agarose to the solubilized membrane extract. Mix

WGA-agarose and solubilized membrane extract by rotation or on a
rocker for 30 min at room temperature.
3. Spin at 600 rpm for 1 min to pellet WGA-agarose. Remove supernatant
with a pipette. Add 10 volumes (based on the volume of WGA-agarose)
of wash buffer, invert several times.
4. Repeat step 3 two more times.
5. Add 2 ml wash buffer to the WGA-agarose and transfer slurry of WGA-
agarose back to the column. Wash with 5–10 column volumes of wash
buffer. Check periodically with protein assay, for example Coommassie
Blue protein assay, until the protein levels drop to background, that is
similar to that in the wash buffer.
6. Elute the glycoproteins from the WGA-agarose by adding half column
volume of elution buffer, for example 50 mM HEPES, pH 7.4, 0.1%
detergent, 0.25 M N-acetyl glucosamine. Collect eluted fractions.
7. Add half column volume of elution buffer. Collect eluted fractions.
8. Repeat step 7 four more times.
9. Detect protein concentration in the eluted fractions. For example, take
5 ml from each eluted fraction and determine the protein concentration
with Coomassie Blue assay. Pool appropriate fractions.

4.2. Ligand-affinity chromatography

The key to a successful ligand-affinity chromatography is that the affinity
between the ligand and the receptor needs to be sufficiently high as to allow
for binding and washing during the purification steps. The ligand is usually
immobilized onto the affinity support through cross-linking. The affinity
between the ligand (or inhibitor) and receptor may be altered due to
the presence of detergent. As such, choosing a detergent that does not
significantly lower the receptor affinity for its ligand will be critical if
ligand-affinity column is to be used for its purification.

4.3. Antibody-affinity chromatography

If the antibody is available, using immobilized antibody against the specific
membrane protein is the most powerful method of purification. Antibodies
are relatively stable in nonionic detergents and thus are compatible with the
presence of detergents in solubilized membrane preparations. The challenge
may be the elution of the protein from the immunoaffinity column
(see Chapter 28).
628 Sue-Hwa Lin and Guido Guidotti

5. Detergent Removal and Detergent Exchange

Most membrane proteins will aggregate and precipitate if detergents
are completely removed. Thus, complete removal of one detergent without
exchanging to another detergent is seldom done if the function of the
membrane protein is to be maintained. In some case, an excess amount of
a detergent present in the samples, for example during initial solubilization,
may interfere with protein activity or concentration measurement, and
removal of the excess amount of detergent may be desirable. In other
cases, the detergent initially used for solubilization may not be appropriate
for subsequent chromatographic or analytical procedure, and exchange to
another detergent may be necessary. For example, the presence of ionic
detergent is not compatible with ion-exchange chromatography or
performing isoelectric focusing, and it will be desirable to switch to non-
ionic or zwitterionic detergents. If the membrane protein needs to be
reconstituted into phospholipid vesicles, it will be desirable to change to a
detergent with high critical micelle concentration (cmc) value so that the
detergent can be removed by dialysis to allow for the membrane proteins to
reconstitute into phospholipids vesicles. Detergent properties have signifi-
cant effects on the ease of detergent removal. Detergents that have a high
(>1 mM) cmc, for example cholate and octylglucoside, can be easily
removed by methods such as dialysis or through ultrafiltration membranes.
Detergents with low (<1 mM) cmc values, such as the nonionic detergents
Triton X-100, C12E9, Brij, Tween, are difficult to remove by dialysis and
adsorption onto chromatography matrix, gel filtration, equilibrium method
can be used. Methods for detergent removal and detergent exchange are
described in Chapter 41.

6. Expression and Purification of Recombinant

Integral Membrane Proteins
Expression and purification of recombinant integral membrane pro-
teins are commonly used to obtain large amounts of membrane proteins for
structure studies or generation of antibodies against the membrane proteins.
Membrane proteins are usually expressed in mammalian cells or insect cells.
The presence of a signal sequence is essential to allow targeting of the
protein to the endoplasmic reticulum for protein synthesis and posttransla-
tional modifications. Thus, tags are usually placed at the end of the sequence
to avoid disturbing the membrane targeting process. Sometimes, a signal
sequence from another protein may be used to improve membrane target-
ing efficiency. Small tags, for example, 6-histidine or small peptide tags, do
Purification of Membrane Proteins 629

not introduce too much change to the proteins and can increase the
likelihood of expressing the membrane proteins in heterologous cell types.
Metal-affinity chromatography, which has high binding capacity and where
its binding is not affected by the presence of detergents, is a good choice for
expressing and purification of membrane proteins. Other affinity epitope
tags, for example, FLAG, myc, or HA, can also be used. However, purifi-
cation of proteins using immobilized antibody has limited capacity and the
cost is much higher than using metal-affinity chromatography.
Examples of expression and purification of membrane proteins can be
found in several articles. One example is the purification of a cell adhesion
molecule CEACAM1 expressed in insect cells (Phan et al., 2001). In this
study, the integral membrane protein was expressed with a 7-histidine tag at
its C-terminus. The original signal sequence was used. The recombinant
CEACAM1 protein was found to be localized on the plasma membrane.
The cells were lysed and the membranes solubilized by Triton X-100 and
the his-tagged CEACAM1 purified in one-step using metal-affinity
chromatography.

REFERENCES
Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram
quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 72,
248–254.
Hjelmeland, L. M. (1990). Solubilization of native membrane proteins. Methods Enzymol.
182, 253–264.
Lee, Y. C., Block, G., Chen, H., Folch-Puy, E., Foronjy, R., Jalili, R., Jendresen, C. B.,
Kimura, M., Kraft, E., Lindemose, S., Lu, J., McLain, T., Nutt, L., Ramon-Garcia, S.,
Smith, J., Spivak, A., Wang, M. L., Zanic, M., and Lin, S. H. (2008). One-step isolation
of plasma membrane proteins using magnetic beads with immobilized concanavalin A.
Protein Expr. Purif. 62, 223–229.
Neville, D. M. J. (1968). Isolation of an organ specific protein antigen from cell surface
membrane of rat liver. Biochim. Biophys. Acta 154, 540–552.
Peterson, G. L. (1977). A simplication of the protein assay method of Lowry et al. which is
more generally applicable. Anal. Biochem. 83, 346–356.
Phan, D., Han, E., Birrell, G., Bonnal, S., Duggan, L., Esumi, N., Gutstein, H., Li, R.,
Lopato, S., Manogaran, A., Pollak, E. S., Ray, A., et al. (2001). Purification and
characterization of human cell-cell adhesion molecule 1 (C-CAM1) expressed in insect
cells. Protein Expr. Purif. 21, 343–351.
Schindler, J., Jung, S., Niedner-Schatteburg, G., Friauf, E., and Nothwang, H. G. (2006).
Enrichment of integral membrane proteins from small amounts of brain tissue. J. Neural
Transm. 113, 995–1013.
Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H.,
Provenzano, M. D., Fujimoto, E. K., Goeke, N. M., Olson, B. J., and Klenk, D. C.
(1985). Measurement of protein using bicinchoninic acid. Anal. Biochem. 150, 76–85.
 C H A P T E R T H I R T Y- S I X

Purification of Recombinant
G-Protein-Coupled Receptors
Reinhard Grisshammer

Contents
1. Introduction 632
2. Solubilization: General Considerations 633
3. Purification: General Considerations 634
3.1. Stability of GPCRs in detergent solution 634
3.2. General affinity purification 634
3.3. Receptor-specific ligand affinity chromatography 636
3.4. Analysis of detergent-solubilized GPCRs 636
4. Solubilization and Purification of a Recombinant Neurotensin
Receptor NTS1 637
4.1. Solubilization of the NTS1 fusion protein 638
4.2. Purification of the NTS1 fusion protein by immobilized metal
affinity chromatography 640
4.3. Purification of the NTS1 fusion protein
by a neurotensin column 640
5. Analysis of Purified NTS1 641
6. Conclusions 642
Acknowledgments 642
References 642

Abstract
Structural and functional analysis of most G-protein-coupled receptors (GPCRs)
requires their expression and purification in functional form. The produced
amount of recombinant membrane-inserted receptors depends on the optimal
combination of a particular GPCR and production host; optimization of expres-
sion is still a matter of trial-and-error. Prior to purification, receptors must be
extracted from the membranes by use of detergent(s). The choice of an
appropriate detergent for solubilization and purification is crucial to maintain
receptors in their functional state. The initial enrichment can be carried out by
affinity chromatography using a general affinity tag (e.g., poly-histidine tag).

Department of Health and Human Services, National Institutes of Health, National Institute of Neurological
Disorders and Stroke, Rockville, Maryland, USA

Methods in Enzymology, Volume 463

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63036-6

631
632 Reinhard Grisshammer

If the first purification step does not yield pure receptor protein, purification to
homogeneity can often be achieved by use of a subsequent receptor-specific
ligand column. If suitable immobilized ligands are not available, size exclusion
chromatography or other techniques need to be applied. Many GPCRs become
unstable upon detergent extraction from lipid membranes, and measures for
stabilization are discussed. As an example, the purification of a functional
neurotensin receptor to homogeneity in milligram quantities is given below.

1. Introduction
Structure determination and functional analysis of integral membrane
proteins, which are not naturally abundant, require (i) a recombinant
production system and (ii) a purification strategy to allow the isolation of
functional rather than nonfunctional, incorrectly folded membrane protein.
Expression and purification of prokaryotic and eukaryotic membrane pro-
teins has been covered in the literature. The reader is referred, for example,
to Grisshammer and Tate (1995, 2003) and Grisshammer and Buchanan
(2006). In addition, the reader may consult Chapter 35 of this volume. This
chapter focuses exclusively on G-protein-coupled receptors (GPCRs)
which are eukaryotic integral membrane proteins involved in cell-to-cell
communication and sensory signal transduction (see Gether and Kobilka,
1998).
It is beyond the scope of this chapter to discuss in any great detail all the
possible expression strategies for integral membrane proteins such as GPCRs.
However, a few key points are summarized here. (i) A universal strategy for
the high-level recombinant expression of functional receptors is currently
unavailable. Some GPCRs accumulate in the membrane to high levels,
whereas other often closely related receptors are hardly detected. Despite
their assumed similarities, individual GPCRs behave quite differently in a
given expression host and recombinant production is still a matter of trial-and-
error. Comparative expression studies have, for example, been performed
using the methylotrophic yeast Pichia pastoris (Andre et al., 2006), the baculo-
virus insect cell system (Akermoun et al., 2005), and the Semliki Forest virus
system (Hassaine et al., 2006). A survey of GPCR production in commonly
used expression hosts has been summarized (Sarramegna et al., 2003). (ii) The
use of eukaryotic hosts seems generally better for producing functional,
membrane-inserted GPCRs than prokaryotic hosts (Grisshammer, 2006),
although there are exceptions. For example, the bacterium Escherichia coli
was successfully used for the expression of the neurotensin receptor NTS1
(Grisshammer et al., 1993; White et al., 2004), the M1 muscarinic acetylcho-
line receptor (Hulme and Curtis, 1998), the adenosine A2a receptor
(Weiß and Grisshammer, 2002), and the cannabinoid CB2 receptor
Recombinant GPCR Purification 633

(Calandra et al., 1997; Yeliseev et al., 2005). (iii) Numerous descriptive reports
have been published on the recombinant expression of integral membrane
proteins. However, few publications (see Bonander et al., 2005, 2009; Griffith
et al., 2003; Wagner et al., 2006, 2007) are currently available to understand
the underlying mechanisms of how a given host cell responds to membrane
protein overproduction.
The purification of receptors can be conceptually divided into two steps:
extraction from membranes with a suitable detergent (solubilization), and
subsequent purification by use of general affinity tags, receptor-specific ligand
columns, size exclusion chromatography, and other methods. Of utmost
importance, care must be taken to choose experimental conditions to main-
tain the membrane protein in its active state throughout the purification
procedure. This latter aspect cannot be overemphasized because many
GPCRs become unstable upon detergent extraction from lipid membranes.

2. Solubilization: General Considerations

The extraction of membrane-inserted receptors is accomplished with
the use of detergents (see Chapter 34). The correct choice of detergent is
crucial to maintain solubilized receptors in their functional state for a
prolonged period of time. N-Dodecyl-b-D-maltoside (DDM), a mild
nonionic detergent, is commonly used for the solubilization of GPCRs.
Lipid-like cholesteryl hemisuccinate may be added to increase receptor
stability. The use of shorter chain detergents, which are usually harsher
than longer chain detergents, is appropriate as long as these detergents do
not compromise the integrity of the GPCR under investigation.
Solubilized receptors must be considered as detergent–lipid–receptor
complexes rather than just as receptor protein. This implies that the bio-
chemical properties such as size, shape, or isoelectric point of solubilized
receptors differ from those computed solely by the amino acid sequence.
Likewise, detergent–lipid–receptor complexes cannot necessarily be
regarded as homogeneous particles. Because lipids as well as receptors
sequester detergent during the solubilization step, the ratio of detergent to
membrane will determine how much detergent and lipid are bound around
the receptor. The amount of bound lipid and detergent will change during
the course of purification.
The solubilization procedure must be optimized to yield the highest
extraction efficiency while maintaining the best stability of a given GPCR
in solution. The effect of the systematic variation of detergent to membrane
input can be monitored by performing radioligand-binding assays and
determining the total protein content. This allows the calculation of an
experimental Bmax value (nmol functional receptor/mg of protein) and
634 Reinhard Grisshammer

comparison of that value with the theoretical value for specific binding of
pure functional receptor. If no test for functionality is available, then good
biochemical behavior such as a symmetric size exclusion chromatography
profile can be used as an indicator for integrity.
The preparation of membranes prior to solubilization constitutes an
initial purification step because soluble proteins are removed before
receptor extraction and the ratio of target receptor to contaminants is there-
fore higher. However, receptors can also be enriched from total cell lysate
rather than from solubilized membranes during the first purification step.

3. Purification: General Considerations

3.1. Stability of GPCRs in detergent solution
Many GPCRs are not very stable in detergent solution (except the visual
pigment rhodopsin as long as the receptor is kept in its nonsignaling dark
state (De Grip, 1982)). One possible explanation for this instability may be
inherent structural flexibility, that is, the receptor can adopt several/many
conformations in detergent solution, some of which may aggregate.
Removal of lipid during the purification process may also cause destabiliza-
tion. A number of measures have been taken to increase the stability of
GPCRs in detergent solution. For example, the addition of an inverse
agonist/antagonist ligand (Cherezov et al., 2007; Hanson et al., 2008;
Jaakola et al., 2008; Warne et al., 2008) will cause the receptor to assume
its nonsignaling inactive state which is generally considered more stable than
the activated state(s) (see Kobilka and Deupi, 2007). Lipids or lipid-like
substances such as cholesteryl hemisuccinate ( Jaakola et al., 2008; Tucker
and Grisshammer, 1996; Weiß and Grisshammer, 2002) and glycerol
(Tucker and Grisshammer, 1996) have been included throughout the
purification to improve receptor stability. Site-directed mutagenesis has
also been used to generate GPCRs with increased stability (Magnani et al.,
2008; Roth et al., 2008; Sarkar et al., 2008; Serrano-Vega et al., 2008;
Shibata et al., 2009) and hence more tolerant to a wider range of detergents.

3.2. General affinity purification

The biochemical and pharmacological properties of GPCRs in detergent
solution may arbitrarily be grouped into two categories. (i) Receptors which
can be purified to homogeneity in functional form under optimized buffer
conditions as assessed by radioligand-binding assays and G-protein nucleotide
exchange experiments (White et al., 2007) are referred to as being functional.
(ii) In some cases (Kobilka, 1995; White et al., 2004), a detergent-soluble
species has been observed which does not bind receptor-specific ligands.
Recombinant GPCR Purification 635

I refer to this latter species as nonfunctional, incorrectly folded (at least in

respect to ligand recognition) but still detergent soluble.
Recombinant cloning techniques have made it easy to introduce general
affinity tags at either the N-terminus or C-terminus of a given receptor.
Examples are a Flag epitope tag at the receptor N-terminus for use with an
M1 antibody affinity column (Kobilka, 1995) or poly-histidine tails at the
receptor C-terminus for immobilized metal affinity chromatography
(IMAC) (Grisshammer and Tucker, 1997; Hanson et al., 2008; Hulme
and Curtis, 1998; Jaakola et al., 2008; Klaassen et al., 1999; Kobilka, 1995;
Warne et al., 2003; Weiß and Grisshammer, 2002; Yeliseev et al., 2005). An
antibody column (1D4 antibody; Molday and MacKenzie, 1983) recogniz-
ing the extreme C-terminus of bovine rhodopsin has been used for the
single-step purification of rhodopsin (Oprian et al., 1987; Reeves et al.,
1999) and of a b-adrenergic receptor with the 1D4 epitope tag fused to its
C-terminus (Chelikani et al., 2006).
The exact position of an affinity tag (i.e., well removed from the
receptor transmembrane core or close to a transmembrane helix) determines
whether binding of the receptor to the affinity resin must be done in batch
or can be done in column mode. An affinity tag too close to the transmem-
brane core may be partially masked by the detergent belt around the
receptor protein and hence will be less accessible to the affinity resin.
Batch loading for a prolonged time will capture receptors efficiently in
that case (Weiß and Grisshammer, 2002). Receptors with well exposed
affinity tags can be loaded onto resin packed into a column with a short
exposure time of the purification tag to the affinity resin. Batch purification
procedures have to be performed manually, whereas column-loading pro-
cedures can be automated (White et al., 2004). As general guideline,
detergents with low critical micelle concentration (cmc) values form larger
detergent belts around a membrane protein than detergents with high cmc
values (see, e.g., Bamber et al., 2006) although the exact amount of protein-
bound detergent will depend on the properties of the respective detergent
tested. The accessibility of an affinity tag to the resin may therefore be
reduced in low cmc detergents compared to high cmc detergents.
Many laboratories utilize IMAC as a first enrichment step. Several resin
types are commercially available such as Ni2þ-NTA resin (Ni2þ-nitrilotri-
acetate, Qiagen), Talon resin (Co2þ-carboxymethylaspartate, Clontech),
and IDA resin (iminodiacetate, Zn2þ, Ni2þ, Co2þ, GE Healthcare). How-
ever, the properties of the various IMAC resins are slightly different (Weiß
and Grisshammer, 2002). For example, Ni2þ-NTA resin not only binds the
target membrane protein tighter but also binds more contaminants com-
pared to Talon resin. The receptor expression level (i.e., the ratio of the
target receptor to contaminants) determines the efficiency of this first
purification step; the higher the receptor expression level, the more efficient
the first purification step. A purity of >90% has been achieved for a mutant
636 Reinhard Grisshammer

b2-adrenergic receptor using a single IMAC step (Hanson et al., 2008).

In some cases, the binding capacity of IMAC resin has been found to be
lower for receptors in detergent solution than that for soluble proteins
(Weiß and Grisshammer, 2002; White et al., 2004).
The time needed for purification should be kept to a minimum, as
should the number of purification steps because of incurring protein losses
at all stages.

3.3. Receptor-specific ligand affinity chromatography

After the enrichment of receptors by use of a general affinity tag, a second
purification step may be required to isolate pure, functional receptor protein
for a number of reasons. (i) Receptors from the first purification step are not
yet pure and other contaminants are still present. For example, a major
contaminant was removed from the adenosine A2a receptor by use of a
XAC (xanthine amine congener, antagonist) column step (Weiß and
Grisshammer, 2002). Likewise, the neurotensin receptor NTS1 was pur-
ified to homogeneity by use of a NT (neurotensin, agonist) column (White
et al., 2004) and an alprenolol (nonselective b-adrenergic receptor antago-
nist) Sepharose column was used for the purification of a b-adrenergic
receptor (Warne et al., 2003). If receptors, eluted from the general tag
affinity resin, are a mixture of functional and nonfunctional, incorrectly
folded receptors, then this subsequent receptor-specific ligand affinity chro-
matography step will also remove the nonfunctional receptor population.
(ii) The use of one (or two subsequent) general affinity tag step(s) produces
almost pure receptor which are, however, a mixture of functional and
nonfunctional species. The use of general affinity tags will hence not resolve
correctly folded from incorrectly folded but still detergent-soluble protein.
Incorrect folding of receptors may occur within the expression host, and/or
may arise because of the instability of receptors in detergent solution.
Application of a receptor-specific ligand affinity column will resolve func-
tional from incorrectly folded receptor species. For example, an alprenolol
column has been used for the isolation of functional b2-adrenergic receptor
(Kobilka, 1995). (iii) A single-step purification protocol led to a highly
enriched preparation of a pituitary adenylate cyclase-activating polypeptide
(PACAP) receptor by adding biotinylated peptide ligand (PACAP38) and
avidin resin to the crude-solubilized receptor (Ohtaki et al., 1998).

3.4. Analysis of detergent-solubilized GPCRs

The amount of functional, detergent-solubilized receptor can be deter-
mined by radioligand-binding experiments. Ideally, the radioligand should
bind to the receptor with high affinity and should display a slow off-rate to
avoid artifacts possibly arising because of nonequilibrium conditions during
Recombinant GPCR Purification 637

the process of separation of the ligand–receptor–detergent complex from

free ligand. An outline into receptor-binding studies is given by Hulme
(1990). Radioligand-binding analyses work best with labeled ligands which
are hydrophilic and do not incorporate into empty detergent micelles.
In contrast, hydrophobic ligands can insert into empty detergent micelles
leading to high nonspecific background signals.
It is recommended to determine the protein content by the Amido
Black assay (Schaffner and Weissmann, 1973). Many methods for protein
content determination are inaccurate in the presence of detergents and
lipids. The Amido Black assay includes a precipitation step to remove
detergents, lipids, or other buffer components. Note that not all proteins
bind dyes such as Amido Black equally well, meaning the protein concen-
tration determinations may be biased if the dye binding is different for the
receptor compared to the reference protein (BSA).
From ligand-binding data and protein content analysis, one can calculate
an experimental Bmax value (nmol functional receptor/mg of protein) and
compare this value with the theoretical value for specific binding of pure
functional receptor. Note that the theoretical Bmax value is specific for each
GPCR because it depends on its amino acid composition. Purification of
functional receptors is achieved if the experimental value matches the
theoretical Bmax value.

4. Solubilization and Purification of a

Recombinant Neurotensin Receptor NTS1
A more detailed description of the solubilization, purification, and
analysis of NTS1 has been described by White and Grisshammer (2007) and
is presented here in an abbreviated form. The purification to homogeneity
of the NTS1 fusion protein is achieved by IMAC followed by a NTS1-
specific ligand column step.
A maltose-binding protein (MBP) fusion approach is used for
the expression of functional, membrane-inserted NTS1 in E. coli
(Grisshammer et al., 1993). The expression plasmid encodes MBP with its
signal peptide, followed by the receptor. After removal of the signal peptide
by the E. coli leader peptidase, the NTS1 fusion protein MBP-T43NTR-
TrxA-H10 (NTS1-624), used for purification, starts with the mature E. coli
MBP (Lys1 to Thr366), followed by the N-terminally truncated rat neuro-
tensin type I receptor NTS1 (T43NTR, Thr43 to Tyr424) (Tanaka et al.,
1990), the E. coli thioredoxin (TrxA, Ser2 to Ala109) (the presence of TrxA
increases expression, see Tucker and Grisshammer, 1996), and a decahisti-
dine tag (H10) (Grisshammer and Tucker, 1997). Expression is driven from
a weak promoter of the low-copy number expression plasmid at low
638 Reinhard Grisshammer

temperature (22  C). This avoids the possible overloading of the E. coli
translocation and membrane insertion machinery by the nascent receptor
chain. Few protein molecules are produced at any given time but accumu-
lation of correctly folded receptors is observed over 2 days. Purification of
10 mg of the NTS1 fusion protein typically requires 250 g of E. coli wet cells
which is equivalent to 50 l of cell culture (White et al., 2004). The growth of
E. coli at this scale is most easily done by fermentation.
The expression levels of the NTS1 fusion protein in E. coli are moderate
(i.e., the ratio of contaminants to target receptor is high). Hence, this requires
an optimized protocol for IMAC to enrich the receptor fusion protein
efficiently. To accomplish this, NTS1 fusion proteins with C-terminal tails
of 10 histidine residues rather than six histidine residues (Grisshammer and
Tucker, 1997) are used in combination with Ni2þ-NTA resin. The tight
binding to Ni2þ-NTA resin of decahistidine-tagged receptors allows strin-
gent washing steps using imidazole at a concentration of 50 mM. Imidazole at
this concentration eliminates binding to the Ni2þ-NTA resin of most of
the E. coli contaminants, but does not cause the elution of the fusion protein.
This strategy not only results in efficient purification of receptors from crude
membranes, but works well for the purification from total cell lysate
(Grisshammer and Tucker, 1997).
The apparent affinity of NT for NTS1 is reduced in the presence of the
high concentrations of sodium ions and imidazole in the NiB buffer
(Grisshammer et al., 1999) precluding the direct binding of functional
receptors in the Ni2þ-NTA column eluate onto the NT column. The
concentration of NaCl and imidazole must therefore be reduced from 200
to 70 mM by dilution with buffer to allow binding of functional NTS1 to
the NT column. Because binding of NTS1 to NT is NaCl sensitive,
receptors can be efficiently eluted from the NT column with NaCl at
high concentration (Fig. 36.1).

4.1. Solubilization of the NTS1 fusion protein

1. All steps are performed at 4  C or on ice unless stated otherwise.
2. At room temperature, crush 250 g of frozen cell paste between plastic
sheets with a hammer to obtain small pieces. Place cells into a Waring
blender.
3. Add 500 ml of cold 2 solubilization buffer (100 mM Tris–HCl, pH
7.4, 60% (v/v) glycerol, 400 mM NaCl) to the cells. Operate the
Waring blender until all cells are suspended.
4. Transfer the cell suspension into a beaker with a magnetic stir bar. It is
difficult to determine the volume at this stage because of the air
introduced by the Waring blender into the suspension. The final
volume is therefore adjusted in step 11.
Recombinant GPCR Purification 639

M 1 2 3 4 5

100

50

Figure 36.1 Purification of the neurotensin receptor NTS1. The fusion protein
NTS1-624 was purified on a 100-ml Ni2þ-NTA column followed by a 20-ml NT
column, starting from 250 g of E. coli cells. The progress of purification was monitored
by SDS–PAGE (NuPAGE 4–12% Bis–Tris gel, Invitrogen, 1 MES buffer) and
Coomassie R-250 staining. Lane M, Novagen Perfect Protein Marker (15–150 kDa);
lane 1, 10 mg of supernatant; lane 2, 10 mg of Ni2þ-NTA column flow through; lane 3,
5 mg of Ni2þ-NTA column eluate; lane 4, 10 mg of NT column flow through; lane 5,
5 mg of NT column eluate. The incorrectly folded receptors in the NT column flow
through are initially detergent soluble but aggregate over time (R. Grisshammer,
unpublished results). Reprinted from White et al. (2004).

5. While stirring, add 1 ml of protease inhibitor stock solutions (phenyl

methyl sulphonyl fluoride, PMSF, at 70 mg/ml in ethanol; leupeptin at
1 mg/ml in H2O; pepstatin A at 1.4 mg/ml in methanol), 5 ml of 1 M
MgCl2 (final concentration of 5 mM ), 0.6 ml of DNaseI solution
(10 mg/ml, Sigma D-4527), and 50 ml of cold H2O.
6. While stirring, add dropwise 100 ml of a CHAPS/CHS stock solution
(6% (w/v) (3-[(3-cholamidopropyl)dimethylammonio]-1-propanesul-
fonate)/1.2% (w/v) cholesteryl hemisuccinate Tris salt in H2O). CHS
is not water soluble on its own and needs to be dissolved in CHAPS.
7. While stirring, add dropwise 100 ml of a DDM stock solution (10%
(w/v) n-dodecyl-b-D-maltoside in H2O).
8. Continue to stir for 15 min.
9. Sonicate for 33 min (8 s/g of cells), 1 s on, 2 s off, level 4 (Misonix
sonicator 3000, 1/2 in. flat tip). Keep the sample in an ice/water bath to
avoid local heating during sonication. This mild sonication step
enhances the receptor solubilization efficiency.
640 Reinhard Grisshammer

10. Add an additional 1 ml of each protease inhibitor stock.

11. Determine volume and add cold H2O dropwise under stirring to a final
volume of 1 l.
12. Stir sample for an additional 30 min.
13. Ultracentrifuge sample for 1 h (Beckman 45Ti rotors or equivalent, at
45,000 rpm (235,000  g at rmax)).
14. Retrieve the solubilized receptors (supernatant).
15. Add an imidazole stock solution (2 M, adjusted to pH 7.4 with HCl)
dropwise while stirring (final concentration of 50 mM ). Pass sample
through 0.22 mm filter. The supernatant is now ready for purification of
receptors by IMAC followed by a neurotensin affinity column.

4.2. Purification of the NTS1 fusion protein by immobilized

metal affinity chromatography
The purification is carried out in the cold room using an Äkta Purifier (GE
Healthcare) chromatography system in a fully automated two-column mode;
for details see White and Grisshammer (2007) and White et al. (2004). The
Purifier is equipped with air sensors, a sample pump P950, a modified
injection valve, a 100-ml Ni2þ-NTA column, a 20-ml NT column, and a
fraction collector Frac950. However, all purification steps can also be per-
formed in a simpler setting. The IMAC column can be processed first, and the
Ni2þ-NTA column eluate is then diluted and loaded onto the NT column.
1. Equilibrate a 100-ml Ni2þ-NTA Superflow (Qiagen) column (XK50,
GE Healthcare) with NiA buffer (50 mM Tris–HCl, pH 7.4, 30% (v/v)
glycerol, 50 mM imidazole, 200 mM NaCl, 0.5% (w/v) CHAPS/0.1%
(w/v) CHS, 0.1% (w/v) DDM).
2. Load the supernatant at a flow rate of 2 ml/min onto the Ni2þ-NTA
column using the sample pump equipped with an air sensor to detect the
end of loading. Collect the Ni2þ-NTA column flow through for analysis.
3. After loading is completed, wash the Ni2þ-NTA column with 15
column volumes of buffer NiA at 2 ml/min.
4. For elution, pass 4 column volumes of buffer NiB (buffer NiA but with
200 mM imidazole) at 2 ml/min over the Ni2þ-NTA column.
5. The total volume of fractions containing the eluted NTS1 fusion protein
is  120 ml.

4.3. Purification of the NTS1 fusion protein

by a neurotensin column
Neurotensin has nanomolar affinity for its receptor in the presence of
detergents; therefore, this property can be used for an efficient affinity
purification step (Grisshammer et al., 1999; Tucker and Grisshammer,
Recombinant GPCR Purification 641

1996). The NT resin is based on N-terminally biotinylated NT (biotin-b

Ala-bAla-Gln-Leu-Tyr-Glu-Asn-Lys-Pro-Arg-Arg-Pro-Tyr-Ile-Leu-OH),
bound to tetrameric avidin resin. Avidin has a high isoelectric point and must
be succinylated to reduce the binding of E. coli proteins to the affinity matrix
(Tucker and Grisshammer, 1996). Biotinylated NT is made by solid-phase
peptide synthesis.
1. Equilibrate a 20-ml NT column (XK26) with NT70 buffer (50 mM
Tris–HCl, pH 7.4, 30% (v/v) glycerol, 1 mM EDTA, 70 mM NaCl,
0.5% (w/v) CHAPS/0.1% (w/v) CHS, 0.1% (w/v) DDM).
2. Dilute the Ni2þ-NTA column eluate with buffer NT0 (50 mM Tris–
HCl, pH 7.4, 30% (v/v) glycerol, 1 mM EDTA, 0.5% (w/v) CHAPS/
0.1% (w/v) CHS, 0.1% (w/v) DDM) to reduce the imidazole and NaCl
concentrations to 70 mM, respectively. The dilution step is performed
by the Äkta Purifier (White et al., 2004), but can also be done manually.
3. Load the diluted Ni2þ-NTA column eluate onto the NT column at a flow
rate of 0.4 ml/min. Collect the NT column flow through for analysis.
4. Wash the NT column at 0.7 ml/min with 8 column volumes of buffer
NT70.
5. Elute receptors at a flow rate of 0.5 ml/min with buffer NT1K (NT0
buffer with 1 M NaCl).
6. After the purification is completed, wash the NT column extensively
with buffer A3 (50 mM Tris–HCl, pH 7.4, 1 mM EDTA, 1 M NaCl)
followed by buffer B3 (50 mM Tris–HCl, pH 7.4, 1 mM EDTA, 3 mM
NaN3). Buffer B3 is used as storage buffer.

5. Analysis of Purified NTS1

The quality of the NTS1 preparation can be assessed by radioligand-
binding assay using tritium-labeled neurotensin ([3H]NT). [3H]NT is very
hydrophilic and not ‘‘sticky,’’ resulting in low background binding. The
detergent-solubilized NTS1 fusion protein binds the agonist with nanomo-
lar affinity. The off-rates are slow, which allows the separation of the
ligand–receptor–detergent complex from free ligand (which does not
incorporate into free detergent micelles) by centrifugation-assisted gel
filtration (White et al., 2004). The size difference between [3H]NT and
the ligand–receptor–detergent complex is large enough to avoid ‘‘leakage’’
of free [3H]NT. For routine analyses, the [3H]NT concentration is set to
2 nM. With some assumptions, one can correct for fractional occupancy
(the law of mass action predicts that the fractional receptor occupancy at
equilibrium is a function of the ligand concentration: fractional occupancy
¼ [ligand]/([ligand] þ KD)). A theoretical value for specific binding (Bmax)
642 Reinhard Grisshammer

of 10.4 nmol/mg is calculated for the NTS1-624 fusion protein (molecular

mass of 96.5 kDa), assuming one ligand-binding site per receptor molecule.
The protein content is determined by the Amido Black assay (Schaffner and
Weissmann, 1973).
Starting from 250 g of wet E. coli cells, the following results can be
anticipated: supernatant ( 930 ml,  16 mg protein/ml,  0.2 nmol NTS1/
ml,  10–12 pmol NTS1/mg protein); Ni2-NTA column eluate ( 120 ml,
 0.25 mg protein/ml,  1.2 nmol NTS1/ml,  5 nmol NTS1/mg
protein); neurotensin column eluate ( 14 ml,  0.7 mg protein/ml,
 7 nmol NTS1/ml,  10 nmol NTS1/mg protein).

6. Conclusions
The NTS1 receptor is expressed in functional form in E. coli as a fusion
protein. Receptors are solubilized by a detergent mixture from total cell
extract (rather than from a membrane preparation) and are enriched by
immobilized metal affinity chromatography. Contaminants present in the
Ni2þ-NTA column eluate are removed by the use of a subsequent neuro-
tensin column. The above purification protocol is simple and robust and
yields pure, functional NTS1 fusion protein.

ACKNOWLEDGMENTS
The research of R. G. is supported by the Intramural Research Program of the NIH,
National Institute of Neurological Disorders and Stroke.

REFERENCES
Akermoun, M., Koglin, M., Zvalova-Iooss, D., Folschweiller, N., Dowell, S. J., and
Gearing, K. L. (2005). Characterization of 16 human G protein-coupled receptors
expressed in baculovirus-infected insect cells. Protein Expr. Purif. 44, 65–74.
Andre, N., Cherouati, N., Prual, C., Steffan, T., Zeder-Lutz, G., Magnin, T., Pattus, F.,
Michel, H., Wagner, R., and Reinhart, C. (2006). Enhancing functional production of
G protein-coupled receptors in Pichia pastoris to levels required for structural studies via a
single expression screen. Protein Sci. 15, 1115–1126.
Bamber, L., Harding, M., Butler, P. J., and Kunji, E. R. (2006). Yeast mitochondrial ADP/
ATP carriers are monomeric in detergents. Proc. Natl. Acad. Sci. USA 103, 16224–16229.
(Epub ahead of print)Bonander, N., Darby, R. A., Grgic, L., Bora, N., Wen, J., Brogna, S.,
Poyner, D. R., O’Neill, M. A., and Bill, R. M. (2009). Altering the ribosomal subunit
ratio in yeast maximizes recombinant protein yield. Microb. Cell Fact 8, 10.
Bonander, N., Hedfalk, K., Larsson, C., Mostad, P., Chang, C., Gustafsson, L., and
Bill, R. M. (2005). Design of improved membrane protein production experiments:
Quantitation of the host response. Protein Sci. 14, 1729–1740.
Recombinant GPCR Purification 643

Calandra, B., Tucker, J., Shire, D., and Grisshammer, R. (1997). Expression in Escherichia
coli and characterisation of the human central CB1 and peripheral CB2 cannabinoid
receptors. Biotechnol. Lett. 19, 425–428.
Chelikani, P., Reeves, P. J., Rajbhandary, U. L., and Khorana, H. G. (2006). The synthesis
and high-level expression of a beta2-adrenergic receptor gene in a tetracycline-inducible
stable mammalian cell line. Protein Sci. 15, 1433–1440.
Cherezov, V., Rosenbaum, D. M., Hanson, M. A., Rasmussen, S. G., Thian, F. S.,
Kobilka, T. S., Choi, H. J., Kuhn, P., Weis, W. I., Kobilka, B. K., et al. (2007). High-
resolution crystal structure of an engineered human beta2-adrenergic G protein-coupled
receptor. Science 318, 1258–1265.
De Grip, W. J. (1982). Thermal stability of rhodopsin and opsin in some novel detergents.
Methods Enzymol. 81, 256–265.
Gether, U., and Kobilka, B. K. (1998). G Protein-coupled receptors II. Mechanism of
agonist activation. J. Biol. Chem. 273, 17979–17982.
Griffith, D. A., Delipala, C., Leadsham, J., Jarvis, S. M., and Oesterhelt, D. (2003). A novel
yeast expression system for the overproduction of quality-controlled membrane proteins.
FEBS Lett. 553, 45–50.
Grisshammer, R. (2006). Understanding recombinant expression of membrane proteins.
Curr. Opin. Biotechnol. 17, 337–340.
Grisshammer, R., and Buchanan, S. K. (eds.) (2006). ‘‘Structural Biology of Membrane
Proteins’’, The Royal Society of Chemistry, Cambridge, UK.
Grisshammer, R., and Tate, C. G. (1995). Overexpression of integral membrane proteins for
structural studies. Q. Rev. Biophys. 28, 315–422.
Grisshammer, R., and Tate, C. G. (eds.) (2003). ‘‘Special Issue on Overexpression of Integral
Membrane Proteins’’. Biochim. Biophys. Acta 1610, pp. 1–153.
Grisshammer, R., and Tucker, J. (1997). Quantitative evaluation of neurotensin receptor
purification by immobilized metal affinity chromatography. Protein Expr. Purif. 11, 53–60.
Grisshammer, R., Duckworth, R., and Henderson, R. (1993). Expression of a rat neuro-
tensin receptor in Escherichia coli. Biochem. J. 295, 571–576.
Grisshammer, R., Averbeck, P., and Sohal, A. K. (1999). Improved purification of a rat
neurotensin receptor expressed in Escherichia coli. Biochem. Soc. Trans. 27, 899–903.
Hanson, M. A., Cherezov, V., Griffith, M. T., Roth, C. B., Jaakola, V. P., Chien, E. Y.,
Velasquez, J., Kuhn, P., and Stevens, R. C. (2008). A specific cholesterol binding site is
established by the 2.8 Å structure of the human beta2-adrenergic receptor. Structure 16,
897–905.
Hassaine, G., Wagner, R., Kempf, J., Cherouati, N., Hassaine, N., Prual, C., Andre, N.,
Reinhart, C., Pattus, F., and Lundstrom, K. (2006). Semliki Forest virus vectors for
overexpression of 101 G protein-coupled receptors in mammalian host cells. Protein Expr.
Purif. 45, 343–351.
Hulme, E. C. (ed.) (1990). ‘‘Receptor Biochemistry—A Practical Approach’’, IRL Press at
Oxford University Press, Oxford.
Hulme, E. C., and Curtis, C. A. M. (1998). Purification of recombinant M1 muscarinic
acetylcholine receptor. Biochem. Soc. Trans. 26, S361.
Jaakola, V. P., Griffith, M. T., Hanson, M. A., Cherezov, V., Chien, E. Y., Lane, J. R.,
Ijzerman, A. P., and Stevens, R. C. (2008). The 2.6 angstrom crystal structure of a human
A2A adenosine receptor bound to an antagonist. Science 322, 1211–1217.
Klaassen, C. H., Bovee-Geurts, P. H., Decaluwe, G. L., and Degrip, W. J. (1999). Large-
scale production and purification of functional recombinant bovine rhodopsin with the
use of the baculovirus expression system. Biochem. J. 342, 293–300.
Kobilka, B. K. (1995). Amino and carboxyl terminal modifications to facilitate the produc-
tion and purification of a G protein-coupled receptor. Anal. Biochem. 231, 269–271.
644 Reinhard Grisshammer

Kobilka, B. K., and Deupi, X. (2007). Conformational complexity of G-protein-coupled

receptors. Trends Pharmacol. Sci. 28, 397–406.
Magnani, F., Shibata, Y., Serrano-Vega, M. J., and Tate, C. G. (2008). Co-evolving stability
and conformational homogeneity of the human adenosine A2a receptor. Proc. Natl. Acad.
Sci. USA 105, 10744–10749.
Molday, R. S., and Mackenzie, D. (1983). Monoclonal antibodies to rhodopsin: Character-
ization, cross-reactivity, and application as structural probes. Biochemistry 22, 653–660.
Ohtaki, T., Ogi, K., Masuda, Y., Mitsuoka, K., Fujiyoshi, Y., Kitada, C., Sawada, H.,
Onda, H., and Fujino, M. (1998). Expression, purification, and reconstitution of receptor
for pituitary adenylate cyclase-activating polypeptide. Large-scale purification of a
functionally active G protein-coupled receptor produced in Sf9 insect cells. J. Biol.
Chem. 273, 15464–15473.
Oprian, D. D., Molday, R. S., Kaufman, R. J., and Khorana, H. G. (1987). Expression of a
synthetic bovine rhodopsin gene in monkey kidney cells. Proc. Natl. Acad. Sci. USA 84,
8874–8878.
Reeves, P. J., Klein-Seetharaman, J., Getmanova, E. V., Eilers, M., Loewen, M. C.,
Smith, S. O., and Khorana, H. G. (1999). Expression and purification of rhodopsin
and its mutants from stable mammalian cell lines: Application to NMR studies. Biochem.
Soc. Trans. 27, 950–955.
Roth, C. B., Hanson, M. A., and Stevens, R. C. (2008). Stabilization of the human beta2-
adrenergic receptor TM4-TM3-TM5 helix interface by mutagenesis of Glu122(3.41),
a critical residue in GPCR structure. J. Mol. Biol. 376, 1305–1319.
Sarkar, C. A., Dodevski, I., Kenig, M., Dudli, S., Mohr, A., Hermans, E., and Plückthun, A.
(2008). Directed evolution of a G protein-coupled receptor for expression, stability, and
binding selectivity. Proc. Natl. Acad. Sci. USA 105, 14808–14813.
Sarramegna, V., Talmont, F., Demange, P., and Milon, A. (2003). Heterologous expression
of G-protein-coupled receptors: Comparison of expression systems from the standpoint
of large-scale production and purification. Cell Mol. Life Sci. 60, 1529–1546.
Schaffner, W., and Weissmann, C. (1973). A rapid, sensitive, and specific method for the
determination of protein in dilute solution. Anal. Biochem. 56, 502–514.
Serrano-Vega, M. J., Magnani, F., Shibata, Y., and Tate, C. G. (2008). Conformational
thermostabilization of the beta1-adrenergic receptor in a detergent-resistant form. Proc.
Natl. Acad. Sci. USA 105, 877–882.
Shibata, Y., White, J. F., Serrano-Vega, M. J., Magnani, F., Aloia, A. L., Grisshammer, R.,
and Tate, C. G. (2009). Thermostabilization of the neurotensin receptor NTS1. J. Mol.
Biol. 390, 262–277.
Tanaka, K., Masu, M., and Nakanishi, S. (1990). Structure and functional expression of the
cloned rat neurotensin receptor. Neuron 4, 847–854.
Tucker, J., and Grisshammer, R. (1996). Purification of a rat neurotensin receptor expressed
in Escherichia coli. Biochem. J. 317, 891–899.
Wagner, S., Bader, M. L., Drew, D., and De Gier, J. W. (2006). Rationalizing membrane
protein overexpression. Trends Biotechnol. 24, 364–371.
Wagner, S., Baars, L., Ytterberg, A. J., Klussmeier, A., Wagner, C. S., Nord, O.,
Nygren, P. A., Van Wijk, K. J., and De Gier, J. W. (2007). Consequences of membrane
protein overexpression in Escherichia coli. Mol. Cell Proteomics 6, 1527–1550.
Warne, T., Chirnside, J., and Schertler, G. F. (2003). Expression and purification of
truncated, non-glycosylated turkey beta-adrenergic receptors for crystallization. Biochim.
Biophys. Acta 1610, 133–140.
Warne, T., Serrano-Vega, M. J., Baker, J. G., Moukhametzianov, R., Edwards, P. C.,
Henderson, R., Leslie, A. G., Tate, C. G., and Schertler, G. F. (2008). Structure of a
beta1-adrenergic G-protein-coupled receptor. Nature 454, 486–491.
Recombinant GPCR Purification 645

Weiß, H. M., and Grisshammer, R. (2002). Purification and characterization of the human
adenosine A2a receptor functionally expressed in Escherichia coli. Eur. J. Biochem. 269,
82–92.
White, J. F., and Grisshammer, R. (2007). Automated large-scale purification of a recombi-
nant G-protein-coupled neurotensin receptor. Curr. Protoc. Protein Sci. 6.8.1–6.8.31.
White, J. F., Trinh, L. B., Shiloach, J., and Grisshammer, R. (2004). Automated large-scale
purification of a G-protein-coupled receptor for neurotensin. FEBS Lett. 564, 289–293.
White, J. F., Grodnitzky, J., Louis, J. M., Trinh, L. B., Shiloach, J., Gutierrez, J.,
Northup, J. K., and Grisshammer, R. (2007). Dimerization of the class A G protein-
coupled neurotensin receptor NTS1 alters G protein interaction. Proc. Natl. Acad. Sci.
USA 104, 12199–12204.
Yeliseev, A. A., Wong, K. K., Soubias, O., and Gawrisch, K. (2005). Expression of human
peripheral cannabinoid receptor for structural studies. Protein Sci. 14, 2638–2653.
 C H A P T E R T H I R T Y- S E V E N

Cell-Free Translation of Integral

Membrane Proteins into Unilamelar
Liposomes
Michael A. Goren, Akira Nozawa, Shin-ichi Makino,
Russell L. Wrobel, and Brian G. Fox

Contents
1. Introduction 648
2. Overview of Cell-Free Translation 650
3. Expression Vectors 651
4. Gene Cloning 652
4.1. Materials and reagents 654
5. PCR Product Cleanup 655
5.1. Materials and reagents 655
6. Flexi Vector and PCR Product Digestion Reaction 656
6.1. Reagents 656
7. Ligation Reaction 657
7.1. Reagents 657
8. Transformation Reaction 658
8.1. Materials and reagents 658
9. Purification of Plasmid DNA 659
9.1. Reagents 659
10. Preparation of mRNA 660
10.1. Reagents 660
11. Preparation of Liposomes 661
11.1. Materials and reagents 661
12. Wheat Germ Translation Reaction 662
12.1. Materials and reagents 663
13. Purification by Density Gradient Ultracentrifugation 665
13.1. Materials and reagents 665
14. Characterization of Proteoliposomes 667
15. Considerations for Scale-Up 669
16. Isotopic Labeling for Structural Studies 669

Department of Biochemistry, Dr. Brian Fox Lab, University of Wisconsin-Madison, Madison,

Wisconsin, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63037-8 All rights reserved.

647
648 Michael A. Goren et al.

17. Conclusions 670

Acknowledgments 670
References 670

Abstract
Wheat germ cell-free translation is shown to be an effective method to produce
integral membrane proteins in the presence of unilamelar liposomes. In this
chapter, we describe the expression vectors, preparation of mRNA, two types of
cell-free translation reactions performed in the presence of liposomes, a simple
and highly efficient purification of intact proteoliposomes using density gradi-
ent ultracentrifugation, and some of the types of characterization studies that
are facilitated by this facile preparative approach. The in vitro transfer of newly
translated, membrane proteins into liposomes compatible with direct measure-
ments of their catalytic function is contrasted with existing approaches to
extract membrane proteins from biological membranes using detergents and
subsequently transfer them back to liposomes for functional studies.

1. Introduction
Membrane proteins provide the molecular mechanisms through
which useful molecules gain controlled entry into a cell, and likewise
provide the portals through which cellular products are exported from the
cell. Membrane enzymes synthesize the molecules that make up cellular
membranes (e.g., saturated and unsaturated fatty acids, phospholipids, gly-
cerolipids, sphingolipids, sterols, polyisoprenes). Membrane enzyme com-
plexes are responsible for electron transport and generate electrochemical
gradients, and an exquisite membrane motor ATPase uses these gradients to
generate ATP. They harvest light, provide allosteric receptors that trans-
duce the information from binding of external molecules into cellular
responses via signaling cascades, provide essential surface contacts as differ-
entiating embryonic stem cells begin to assemble into more complex tissues,
and help to elicit the antigenic responses observed in response to pathogen
infection. These few examples do not do justice to the incredible diversity
of functions and the essential roles that membrane proteins and enzymes
contribute to cellular function.
Achieving control of integral membrane protein expression, transfer into
the lipid bilayer, and cofactor incorporation are significant experimental
challenges, and the ability to manipulate these events would be of great
scientific utility. Furthermore, identification of novel ways to address
increasing structural complexity, leading to the expression and facile purifi-
cation of fully folded, functional membrane proteins or complexes
Cell-Free Translation 649

embedded in easily handled and manipulated lipid environments or in

functionally compatible detergent mixtures would also be of great utility.
Recent studies arising from structural genomics efforts suggest that cell-
free translation may have unique possibilities for addressing these challenges.
Although known for a long time, cell-free protein translation is undergoing
a renaissance (Endo and Sawasaki, 2006; Klammt et al., 2004; Schwarz et al.,
2007; Spirin and Swartz, 2008; Vinarov et al., 2006; Yokoyama, 2003). This
approach to protein synthesis, using extracts from either prokaryotic (Boyer
et al., 2008; Kigawa et al., 2004; Klammt et al., 2006; Schwarz et al., 2007;
Wuu and Swartz, 2008; Yokoyama, 2003) or eukaryotic (Endo and
Sawasaki, 2006; Madin et al., 2000) sources, offers an alternative to E. coli
or other living cell-based platforms that are the mainstay of most recombi-
nant protein expression efforts. Because it decouples the production of
proteins and enzymes from cellular homeostasis, cell-free translation
removes variability associated with the use of living expression hosts (Hall
et al., 2005). Cell-free systems permit the production of proteins that form
complexes in the cell (Matsumoto et al., 2008), inclusion bodies (Klammt
et al., 2004; Schwarz et al., 2007), that undergo proteolysis in cells (Goren
and Fox, 2008; Yokoyama, 2003), or that are toxic to cells (Klammt et al.,
2004; Madin et al., 2000; Schwarz et al., 2007). Cell-free translation can
offer dramatic increases in the speed of evaluation of different conditions for
expression, considerable simplification of the steps needed to purify a
recombinant protein, and direct possibilities for automation of all steps
after sequence-verified cloning to catalytic assay of an expressed and purified
protein.
Because it is an open system, cell-free translation allows the addition of
many types of reagents to the protein synthesis reaction in order to address
biological complexity. Some examples of added reagents include detergents
to solubilize proteins (Klammt et al., 2007), liposomes to capture functional
forms of membrane proteins (Goren and Fox, 2008; Nozawa et al., 2007),
cofactors, coenzymes, and metal ions to reconstitute catalytic function (Abe
et al., 2004; Boyer et al., 2008; Goren and Fox, 2008), multiple mRNAs to
yield cotranslation of heteromeric proteins and protein–protein complexes
(Matsumoto et al., 2008), or other enzymes to effect posttranslational
modifications (Goerke and Swartz, 2008; Kanno et al., 2007). Moreover,
many different types of amino acid (AA) labeling strategies can be accom-
plished in cell-free translation by a simple substitution of unlabeled AAs
with AAs containing isotopic substitutions such as 2H, 13C, or 15N for
NMR or selenomethionine (SeMet) for determination of X-ray diffraction
phasing (Klammt et al., 2004; Matsuda et al., 2007; Vinarov and Markley,
2005). Residue type selective labeling can be carried out without the need
for auxotrophs or specialized feeding, and unnatural AAs can be
incorporated in a straightforward manner (Kiga et al., 2002). Another
advantage is that wheat germ cell-free translation has been successfully
650 Michael A. Goren et al.

automated (Vinarov and Markley, 2005; Vinarov et al., 2006) in 24-, 96-,
and 384-well formats for screening in volumes as low as 50 mL, and 6- and
24-well formats for protein production in volumes up to 6 mL.

2. Overview of Cell-Free Translation

This chapter describes the use of wheat germ extract, a highly stable and
productive eukaryotic system for cell-free translation (Kanno et al., 2007;
Madin et al., 2000). Since wheat germ extract is prepared from a eukaryote, it
provides differences in the mechanisms for translation and folding of nascent
proteins as compared to bacterial translation (Endo and Sawasaki, 2006).
Moreover, the eukaryotic system may provide unique chaperones and trans-
location factors that may aid in the expression of some proteins (Goren and
Fox, 2008), possibly including the examples described here.
Figure 37.1 provides a schematic summary of the steps used for cell-free
translation and shows time-course photographs of an expression of green
fluorescent protein (GFP, photo images provided courtesy of our collabo-
rator and mentor, Prof. Yaeta Endo, Cell-Free Science and Technology
Research Center, Ehime University, Japan). For this work, the gene of
interest is cloned into a specialized expression vector that adds 50 -translation
enhancer and 30 -UTR sequences to the transcribed mRNA. Genes are
transcribed using SP6 RNA polymerase, and the highly purified mRNA
is used in the translation reaction. The quality and quantity of the expression

Prepare
plasmid DNA, Mix mRNA with
other reagents mRNA synthesis wheat germ extract

Overlay amino acids, creatine phosphate, creatine kinase, nucleotides, liposomes

Protein
Translation production

Overlay 3h 6h 16 h

Figure 37.1 A schematic representation of the steps used for cell-free translation in a
bilayer mode reaction. A customized plasmid is used to prepare reagent mRNA, which
is added to the translation mixture, overlaid with amino acids, other substrates, and
desired additives, and then the reaction can begin. The photos show an experimental
demonstration of GFP translation over a 16 h period.
Cell-Free Translation 651

plasmid and mRNA preparations are critical, and often underappreciated

inputs to the success of high-yield cell-free translation (Tyler et al., 2005;
Vinarov et al., 2006). Along with the mRNA template, the cell-free
translation utilizes exogenously added AAs as substrates for the protein
synthesis.

3. Expression Vectors
Table 37.1 summarizes specialized vectors used for this work. pEU,
the original cell-free translation vector optimized for wheat germ cell-free
translation (Madin et al., 2000), was modified for high-throughput Flexi
Vector cloning (Promega) to contain 50 -SgfI and 30 -PmeI restriction sites
and a toxic selection cassette in the multicloning site (Blommel et al., 2006).
The vector named pEU-FV produces a protein with no purification tag,
while pEU-His-FV and pEU-HSBC produces a protein with an N-terminal
His6 purification tag. Vectors pEU-NGFP and pEU-GFPC produce
fusions of GFP to either the N- or C-termini of the target protein. These
vectors are useful for control studies and as vehicles for simplified detection
of the translated fusion proteins during purification (Drew et al., 2001).

Table 37.1 Flexi Vector compatible vectors for cell-free translation

Name Producta Utility

pEU Target Expression screening, native
protein
pEU-FV Target Expression screening
pEU-His-FV His6-Target Expression screening; His6
purification
pEU-HSBC His6-Target Expression screening; His6
purification
pEU-NGFP GFP-target Expression screening; detection
pEU-GFPC Target-GFP Expression screening; detection
pEU-Nb5R Target-(cyt b5 reductase Spontaneous association of
N-terminal anchor N-terminal anchor peptide
peptide)b to liposomes
pEU-Cb5 Target-(cyt b5 Spontaneous association of
C-terminal anchor C-terminal anchor peptide
peptide)c to liposomes
a
The sequence of targets, domains, or tags found in the translated protein. For example, pEU-GFPC
produces a target protein with GFP fused to the C-terminus, that is, target-GFP.
b
The N-terminal anchor peptide sequence is GAQLSTLGHMVLFPVWFLYSLLM.
c
The C-terminal anchor peptide sequence is TLITTIDSSSSWWTNWVIPAISAVAVALMYRLYMAED.
652 Michael A. Goren et al.

With proper design of the primer sequences, the GFP tag can be removed
by treatment with tobacco etch virus (TEV) protease (Blommel and Fox,
2007; Sobrado et al., 2008). The vectors pEU-Nb5R and pEU-Cb5 create
fusion proteins with the membrane anchor signals from human cytochrome
b5 reductase and human cytochrome b5, respectively. Fusion proteins con-
taining these tags spontaneously associate with liposomes upon translation
(Nomura et al., 2008), and thus become amenable to purification by density
gradient ultracentrifugation (Goren and Fox, 2008). Other posttranslational
modifications may also be used to target proteins to liposomes (Nosjean and
Roux, 2003). The vectors described herein are available from the NIH
Protein Structure Initiative Material Repository (http://psimr.asu.edu).

4. Gene Cloning
This approach requires a two-step PCR procedure (Blommel et al.,
2006; Thao et al., 2004). Figure 37.2 provides a vector map for pEU-HSCB
and an example of primers designed for cloning of His6-bacteriorhodopsin
(Blommel and Fox, 2007). By using this vector and primer design, the
sequence MGHHHHHHAIAENLYFQ- can be liberated from the trans-
lated protein upon treatment with TEV protease to leave Ser as the first
residue of the mature protein. The tobacco mosaic virus omega sequence is a
translation enhancement sequence (Sawasaki et al., 2000). Incorporation of
the SgfI and PmeI restriction sites in the indicated positions in the respective
primers allows the cloned gene to be transferred by Flexi Vector cloning
(Blommel et al., 2006) into any (or all) of the vectors described in Table 37.1,
along with a number of other commercially available vectors (see http://
www.promega.com for other examples). The sacB-CAT cassette shown in
Fig. 37.2A provides toxic selection when transformants containing this
construct are grown on plates containing 5% sucrose. In this manner, positive
selection for cloning the desired gene can be enforced. The insert also confers
resistance to chloramphenicol. The pF1K homology region enhances the
efficiency of transfer of cloned genes between different Flexi Vectors
(Blommel et al., 2006).
In the example shown in Fig. 37.2B, the first-step PCR uses a 50 forward
primer containing gene-specific nucleotides. An invariant sequence is added
to the 50 end of the first-step PCR forward primer to encode a portion of
the TEV protease site. The first-step PCR also uses a 30 reverse complement
primer as shown in Fig. 37.2C, which contains gene-specific nucleotides
and the PmeI site. Other genes may be cloned by substitution of the
gene-specific sequence in the designated primer sequences.
For the second-step PCR, a universal forward primer (Fig. 37.2B) is
used to add the nucleotides required to complete the TEV site and the SgfI
 A
TMV omega Sgfl

200 400 600 800 1000 1200 1400 1600 1800 2000
sacB-CAT cassette

Pmel pF1K homology

2200 2400 2600 2800 3000 3200 3400 3600 3800 4000 4200
sacB-CAT cassette

4400 4600 4800 5000 5200 5400 5600 5800 6000 6200
Amp resistance

B Sgfl C Pmel

GGTTgcgatgcgCGAAAACCTGTACTTCCAGTCCttggagttattgcc gcgaccagcgactgaTAGTTTAAACGAATTCGAGCTCACAC

CCAAcgctacgcGCTTTTGGACATGAAGGTCAGGaacctcaataacgg cgctggtcgctgactATCAAATTTGCTTAAGCTCGAGTGTG
1st PCR: 3⬘ TEV primer + gene-specific 1st PCR: gene specific + Pmel site

2nd PCR: 5⬘ Sgfl + 5⬘ TEV primer 2nd PCR: reverse primer

TEV protease site N-terminus of target C-terminal of target

V A I A E N L Y F Q S L E L L P A T S D

Figure 37.2 (A) Linear map of pEU-HSCB showing important features of the vector construction. The TMV omega sequence (blue box)
enhances translation efficiency. The pF1K homology sequence (yellow box) prevents ligation of two plasmid backbones. (B) Example of the
50 primer designed for the cloning of Halobacterium salinarium bacteriorhodopsin (GenBank M11720.1). The first-step PCR forward primer is
654 Michael A. Goren et al.

site. A universal reverse PCR primer is used to duplicate the PmeI site and
add additional nucleotides (Fig. 37.2C). For the second-step PCR, a por-
tion of the first-step PCR reaction is added into a new PCR reaction
containing the universal forward primer and reverse primers to obtain the
PCR product properly functionalized for Flexi Vector cloning.

4.1. Materials and reagents

Gene-specific primers (25 nmol synthesis with standard desalting) can be
obtained from IDT (Coralville, IA).
The dNTP mix (10 mM of each nucleotide) is from Promega (Madison, WI).
Pfu Ultra II fusion hotstart DNA polymerase is from Stratagene (La Jolla, CA).
A 10 Flexi Enzyme Blend containing restriction endonucleases SgfI and
PmeI is from Promega.
PCR plates (T-3069-B) are from ISC Bioexpress (Kaysville, UT).
Adhesive covers for PCR plates (4306311) are from Applied Biosystems
(Foster City, CA).
PCR plates are centrifuged in an Allegra 6R centrifuge with a GH3.8 rotor
(Beckman Coulter, Fullerton, CA).
An MJ DNA Engine, DYAD, Peltier Thermal Cycler (MJ Research,
Waltham, MA) can be used for thermocycling reactions.

4.1.1. Protocol
The following steps are used to PCR amplify the desired genes and append
the sequences required for cloning. A sequence-verified gene in a plasmid is
used for the PCR template. These may be obtained from a variety of other
sources or research projects. This PCR protocol can also be used on
genomic DNA, if the gene of interest has no introns, or on reverse
transcribed mRNA (Thao et al., 2004).

50 -ACCTGTACTTCCAGTCCttggagttattgcc, with the upper case nucleotides

corresponding to the 30 TEV primer and the lower case nucleotides corresponding to
the gene-specific sequence. The second-step universal forward primer is
50 -GGTTgcgatcgcCGAAAACCTGTACTTCCAGTCC, with the SgfI site in lower
case. There is an 18 bp overlap between the first-step forward primer and universal
forward primer. The TEV protease recognition sequence is ENLYFQ/S, with proteo-
lysis between Q and S. (C) Example of the 30 primer designed for the cloning of
Halobacterium salinarium bacteriorhodopsin (GenBank M11720.1). The first-step PCR
reverse complement primer is 50 -GCTCGAATTCGTTTAAACTAtcagtcgctggtcgc,
with the upper case, italic nucleotides corresponding to the PmeI site and the lower
case nucleotides corresponding to the gene-specific sequence. The second-step univer-
sal reverse primer is 50 -GTGTGAGCTCGAATTCGTTTAAAC. There is an 18 bp
overlap between the first-step reverse primer and universal reverse primer.
Cell-Free Translation 655

If the plasmid being used for the template and the Flexi Vector that the
PCR product will be cloned into have the same antibiotic resistance, DpnI
can be added to the PCR reaction followed by incubation for 1 h at 37  C.
The DpnI will digest the template plasmid while leaving the PCR product
fully intact.
1. Create a PCR Primers plate by combining forward and reverse primers
for each target gene to 10 mM each in a well of an ISC PCR plate.
2. Create a PCR Master Mix consisting of 2.195 mL of water, 250 mL of
10 Pfu Ultra II Buffer, 25 mL of dNTPs (10 mM each), and 50 mL of Pfu
Ultra II Hotstart DNA polymerase This Master Mix will provide up to
96 reactions, and can be scaled as appropriate. Discard the unused mix.
3. Aliquot 23.0 mL of PCR Master Mix to each well of an ISC PCR plate
that will be used. This is the PCR Reaction plate.
4. Add 1 mL of the mixture from each used well of the PCR primers plate
to each used well of the PCR Reaction plate.
5. Add 1 mL ( 100 ng) of purified plasmid DNA for each gene to be
cloned to a separate well of the PCR Reaction plate.
6. Centrifuge the PCR Reaction plate briefly in an Allegra 6R centrifuge
and 6H3.B rotor to get liquid to the bottom of the wells and then cover
the plate with sealing tape.
7. Put the PCR Reaction plate in the thermocycler and cycle using the
following parameters: (1) 95  C for 2.00 min; (2) 94  C for 20 s; (3) 50  C
for 20 s; (4) 72  C, 15 s/kb; (5) repeat steps 1–4 for 19 more times.
8. For the second-step PCR, 20% of the volume of the first-step reaction is
added into a new PCR reaction along with 0.2 mM of each of the
universal forward and reverse primers. The second-step reaction is
then completed using the following cycling parameters: (1) 95  C for
2.00 min; (2) 94  C for 20 s; (3) 50  C for 20 s; (4) 72  C, 15 s/kb;
(5) repeat steps 2–4 for four more times; (6) 94  C for 20 s; (7) 55  C for
20 s; (8) 72  C for 15 s/kb; (9) repeat steps 6–8 for 24 more times; (10) 72

C for 3.00 min; and (11) 4  C and hold.
9. Analyze the completed PCR reactions for the appropriately sized
products using agarose gel electrophoresis.

5. PCR Product Cleanup

The PCR products are purified prior to SgfI/PmeI digestion.

5.1. Materials and reagents

Quickstep 2 PCR purification kit (EdgeBio, Gaithersburg, MD), containing
SOPE resin, Secure Seal Sterile tape, Performa Ultra 96-well Plate and
V-Bottom 96-well Plate.
656 Michael A. Goren et al.

5.1.1. Protocol
1. Add 4 mL of well-mixed SOPE resin to 20 mL of PCR product from the
FLEXI-TEV PCR protocol.
2. Cover the plate with Secure Seal Sterile tape and vortex. Let the
suspension stand at room temperature while preparing the Performa
Ultra 96-Well Plate.
3. Remove adhesive plate sealers from the top and bottom of the Performa
Ultra 96-Well Plate. Cover with a lid.
4. Stack the Performa Ultra 96-Well Plate on top of the 96-well
flat-bottom microplate.
5. Place the assembly in a cushioned centrifuge plate carrier.
6. Centrifuge for 5 min at 850g.
7. Briefly spin the SOPE/PCR mix to the bottom of the wells. Transfer
the SOPE/PCR reaction mixture by slowly pipetting directly into the
wells of the Performa Ultra 96-Well Plate. Be sure the fluid runs into the
gel matrix. Cover with a lid.
8. Stack the Performa Ultra 96-Well Plate on top of the 96-well V-bottom
microplate.
9. Place the assembly in the centrifuge plate carrier and centrifuge for 5 min
at 850g. Retain the eluates, which contain the purified PCR products.
PCR products can be stored at 20  C until ready for use.

6. Flexi Vector and PCR Product Digestion

Reaction
The following steps are used to digest the pEU vector variant and the
purified PCR products with SgfI and PmeI prior to the ligation. Additional
descriptions of cloning using the Flexi Vector system are presented else-
where (Blommel and Fox, 2007; Blommel et al., 2006). For success in Flexi
Vector cloning, it is important to avoid overdigesting either the PCR
product or the vector. SgfI has star activity so overdigestion removes the
overhanging nucleotide sequence leaving bunt ends on the digested vector
that efficiently ligate and yield a high background of clones with no insert
after transformation.

6.1. Reagents
5 Flexi Digest Buffer (Promega)
10 SgfI/PmeI Enzyme Blend (Promega)
pEU vector variant (see Table 37.1)
Purified PCR products from PCR Product Cleanup protocol, step 9
Cell-Free Translation 657

6.1.1. Protocol
1. Create a pEU Vector Digest Master Mix consisting of 158.3 mL of
sterile, deionized water, 44.0 mL of 5 Flexi Digest Buffer, 2.20 mL of
10 SgfI/PmeI Enzyme Blend, and 13.5 mL of the pEU vector variant
desired (e.g., purified pEU-His-FV at a concentration of 150 ng/mL).
The enzyme blend is dense and tends to settle, so must be mixed
thoroughly into the remainder of the mix.
2. Place the pEU Vector Digest Master Mix in the thermocycler and cycle
as follows: (1) 37  C for 40.00 min; (2) 65  C for 20.00 min; and (3) hold
at 4  C until needed.
3. Create the PCR Product Digest Master Mix consisting of 638 mL of
sterile, deionized water, 220 mL 5 of Flexi Digest Buffer, and 22 mL of
10 SgfI/PmeI Enzyme Blend. This Master Mix will provide up to 96
reactions, and can be scaled as appropriate. Discard the unused mix.
4. Add 8.0 mL of the PCR Product Digest Master Mix to each well of an
ISC PCR plate. This is the PCR Digest plate.
5. Add 2.0 mL of the purified PCR products in the PCR Reaction plate from
Gene Cloning protocol, step 8 to each used well of the PCR Digest plate.
6. Place the PCR Digest plate in the thermocycler and cycle as follows:
(1) 37  C for 40.00 min; (2) 65  C for 20.00 min; and (3) hold at 4  C
until needed. If transformation yields no colonies, or colonies containing
vector without the desired insert, decrease the 37  C incubation time to
minimize star activity.

7. Ligation Reaction
Digested and purified PCR products and pEU vector variants are
ligated in this step. For the best efficiency in this ligation reaction, it is
important to use a high concentration of ligase.

7.1. Reagents
10 T4 DNA ligase buffer (Promega)
High concentration T4 DNA ligase (Promega)
pEU vector variant from Flexi Vector Digestion Reaction, step 2
Purified PCR product in PCR Cleanup plate from PCR Product Digestion
Reaction Cleanup, step 6

7.1.1. Protocol
1. Create a Ligation Master Mix containing 225 mL of sterile, deionized
water, 110 mL of 10 T4 ligase buffer, and 50 mL of high-concentration
T4 DNA ligase.
658 Michael A. Goren et al.

2. Add 5.0 mL of purified PCR product from the PCR Cleanup plate,
2.0 mL of the pEU vector digest, and 3.5 mL of Ligation Master Mix to
each well of a new PCR plate. This is the Ligation plate.
3. Incubate the Ligation plate in a thermocycler at 25  C for 3 h.
4. Immediately proceed to the transformation step or store the Ligation
plate overnight at 4  C.

8. Transformation Reaction
The material from the ligation reaction is used to transform competent
cells by the following steps. Competent cells from Invitrogen and Promega
have also been used successfully by following the manufacturers protocols.

8.1. Materials and reagents

10G cells, chemically competent for transformation (Lucigen, Middleton, WI)
10G cells recovery medium (Lucigen).
Ligation Plate from Ligation Reaction, step 4.
Fisherbrand sterile disposable Petri dishes (60  15 mm) (Fisher Scientific,
Pittsburgh, PA).
YT agar plates containing 0.5% (w/v) glucose, 50 mg/mL of kanamycin,
and 5% sucrose for selection with pEU-HSCB vector (Fig. 37.2).
ColiRoller Plating Beads (Novagen, Gibbstown, NJ).

8.1.1. Protocol
1. Remove 10G chemically competent cells from 80  C freezer and
thaw on ice.
2. Aliquot 10 mL of cells into a PCR plate or strip tubes that have been
chilled on ice.
3. Add 1.0 mL of the ligation reaction and stir with the pipet tip.
4. Incubate on ice for at least 5 min. Set a thermocycler block to 34  C.
5. Heat shock at 34  C for 30 s (per the manufacturer’s directions for small
reaction volumes).
6. Return the transformation reactions to ice and incubate for 2 min.
7. Remove from ice and add 80 mL of room temp recovery medium.
8. Incubate at 37  C for 1 h without shaking.
9. While transformations are incubating, label the bottoms of 96 YT
plates, containing the appropriate antibiotic, A1–H12 and add 5–10
sterile ColiRoller glass beads to each plate.
10. Apply the entire transformation reaction on each of the correspond-
ingly labeled plates. Shake the plates in a circular motion to spread the
Cell-Free Translation 659

culture. Carefully remove ColiRoller beads by tipping the plate over an

appropriate waste container, or, if they will be reused, 100% ethanol.
11. Incubate the plates overnight at 37  C.

9. Purification of Plasmid DNA

All reagents used for in vitro transcription and translation must be
RNase-free. Therefore, all glassware used to prepare reagents for cell-free
translation reactions must be baked for 3 h at 180  C to eliminate contam-
inating RNases. Furthermore, wear gloves, keep one’s mouth closed, and
avoid sneezing while handling reagents in order to prevent RNase contam-
ination from hands or saliva. All buffers must be sterilized by passage
through a 0.2-mm filter, and stored at 20  C unless otherwise indicated.
Also, unless otherwise stated, 18 MO water (Milli-Q water, Millipore,
Billerica, MA) is used to prepare all reagents. Diethylpyrocarbonate-treated
water should not be used in these protocols as the degradation products of
DEPC can inhibit in vitro transcription and translation reactions.

9.1. Reagents
The 10 buffer used with proteinase K is 100 mM Tris–HCl, pH 8.0,
containing 50 mM EDTA and 1% (w/v) SDS.
Proteinase K is from Sigma/Aldrich (St. Louis, MO). Prepare a 10
proteinase K solution by addition of 0.5 mg of proteinase K to 1.0 mL
of 10 proteinase K buffer. Aliquot the proteinase K solution into 10 mL
samples and store them at 80  C.

9.1.1. Protocol
1. Inoculate 150 mL of 2  YT medium with a single colony from a pEU-
His-FV transformation and grow the culture in a shaking incubator
overnight at 37  C. Recover the cells by centrifugation.
2. Purify the plasmid using a Marligen high-purity plasmid Maxiprep kit
(Marligen Biosciences, Ijamsville, MD) and the manufacturer’s instruc-
tions. Resuspend each separate DNA pellet in 500 mL of Milli-Q water
and measure the absorbance at 260 nm to determine the plasmid DNA
concentration. A typical yield is 600–900 mg of plasmid DNA.
3. Plasmid DNA prepared by commercially available kits often contains a
trace contamination of RNase. This contaminant must be removed for
successful transcription and translation. In order to remove trace RNase
contamination, treat the purified plasmid with 50 ng/mL of proteinase K
for at least 60 min at 37  C in 1 proteinase K buffer.
660 Michael A. Goren et al.

4. Extract the remaining protein by adding an equal volume of a 1:1

phenol/chloroform solution to the plasmid preparation, vortex it vigor-
ously, and then centrifuge the preparation at 14,000 rpm (18,000g) and
4  C in an F2402H rotor and Allegra 21R centrifuge or comparable
instrument. Remove the upper aqueous phase to a separate new tube
and repeat the extraction, centrifugation of the phenol/chloroform
solution, and transfer of the aqueous phase to a new tube.
5. Add 0.1 volume of 3 M sodium acetate, pH 5.2, and 2.5 volumes
of ethanol to the aqueous phase obtained from step 4 and mix well.
Chill each tube at 20  C for 10 min to facilitate precipitation of the
plasmid DNA.
6. Centrifuge each tube at 14,000 rpm (18,000g) and 4  C in an F2402H
rotor and Allegra 21R centrifuge. Wash each pellet with 500 mL of
ice-cold 70% ethanol.
7. Centrifuge each tube again for 5 min as above, discard the supernatant,
and thoroughly air-dry the pellet, which contains the desired
plasmid DNA.
8. Dissolve the plasmid DNA pellets in 400 mL of Milli-Q water and
measure the absorbance of each at 260 nm to determine the concentra-
tion of each separate plasmid DNA preparation. Adjust the volume of
each preparation with Milli-Q water to obtain a DNA concentration for
the solution of 1 mg/mL based on the fact that 1 mg/mL of DNA gives an
A260 ¼ 20 (40-fold dilution ¼ 0.5).

10. Preparation of mRNA

In cell-free translation, mRNA is an essential reagent for the protein
synthesis reaction. In order to obtain maximal translation efficiency,
the mRNA preparations must be added in sufficient quantity to saturate
ribosomes present in the translation reaction.

10.1. Reagents
Transcription Buffer plus Mg (5) is 400 mM HEPES-KOH, pH 7.8,
containing 100 mM magnesium acetate, 10 mM spermidine hydrochlo-
ride, and 50 mM DTT. Store this buffer at 20  C.
An NTP solution containing 25 mM each of ATP, GTP, CTP, and UTP is
prepared from 0.2 mm filter-sterilized, 100 mM solutions of each NTP
prepared in Milli-Q water. NTP solutions are stored at 80  C.
SP6 RNA polymerase and RNase inhibitor (RNasin) are from Promega.
Cell-Free Translation 661

10.1.1. Protocol
1. Immediately before use, prepare a transcription mixture containing
2 Transcription Buffer plus Mg, 8 mM NTPs, 3.2 unit/mL of SP6
RNA polymerase, and 1.6 unit/mL of RNasin.
2. Dispense 2.5 mL of each separate plasmid DNA from Purification of
Plasmid DNA, step 8, to a separate well of a PCR plate. To the PCR
plate, dispense 2.5 mL of the transcription mixture into each well and
mix. This is called the Transcription plate.
3. Tightly cap the wells of the Transcription plate to avoid concentrating
the samples by evaporation. Incubate the Transcription plate at 37  C for
4 h. If the transcription reaction is proceeding correctly, a white precip-
itate of magnesium pyrophosphate will form and make the transcription
solution turbid.
4. Remove the white precipitate from the transcription reaction by centri-
fugation in the C0650 rotor and Allegra X-22R centrifuge for 5 min at
6230 rpm (4000g) and 26  C. To avoid coprecipitation of mRNA, the
reaction should not be chilled. Transfer the supernatant to a new tube.
This clarified solution will be used in the translation reactions as the
mRNA solution.

11. Preparation of Liposomes

Unilamelar liposomes are added to the cell-free translation reaction to
capture membrane proteins as they are translated. This step provides an
alternative to solubilization of membrane proteins by detergents, which
often does not allow direct determination of function.

11.1. Materials and reagents

Soybean total extract (20% lecithin) is from Avanti Polar Lipids (Alabaster, AL).
The lipid rehydration buffer is 25 mM HEPES, pH 7.5, containing
100 mM NaCl.
Track-etch polycarbonate membranes, 0.4 and 0.1 mm, are from Nucleopore
(Pleasanton, CA).
A mini-extruder for preparation of unilamelar liposomes is from Avanti
Polar Lipids.

11.1.1. Protocol
1. Dissolve 1 g of soybean total extract (20% lecithin) in 3 mL of chloroform.
2. Flush the lipid solution with a stream of N2 gas in order to remove the
bulk of the organic solvent. Dry the remaining lipid further under
vacuum for 30 min.
662 Michael A. Goren et al.

3. Resuspend the dried lipid in 67 mL of lipid rehydration buffer (final

concentration of 15 mg/mL). Vortex the lipid solution until homo-
geneous. Hydration of the lipids can be aided by 3–5 freeze/thaw cycles.
4. Form unilamelar liposomes by extrusion through a mini-extruder. Pass
the liposome solution 11 times through a 0.4-mm track-etch polycar-
bonate membrane and then 11 times through a 0.1-mm membrane.
5. Aliquot the liposomes and flash-freeze them. Liposomes prepared in this
way can be stored at 80  C.

12. Wheat Germ Translation Reaction

Figure 37.3 compares the set up of the bilayer reaction with that of a
dialysis reaction. The bilayer reaction separates the extract and mRNA from
other reagents by the difference in density of the extract and upper buffer
solutions. In this case, diffusion of substrates and products occurs through
the entire buffer interface. Because of this simplicity in set up, the bilayer
cell-free translation reaction is amenable to automation (Sawasaki et al.,
2002a; Tyler et al., 2005; Vinarov et al., 2004, 2006). In our hands, the
bilayer reaction has yielded  0.2 mg/mL of various membrane proteins.
However, as diffusion dilutes the reaction, the yield is limited.
The dialysis reaction is another method for performing cell-free transla-
tion (Fig. 37.3). A 12-kDa cutoff membrane at the bottom of the reactant cup
retains the concentrated wheat germ extract, mRNA, and expressed protein.
Diffusion replenishes ATP, AAs, cofactors, and other additives required for
continued translation and removes inhibitory products. The dialysis method
of cell-free translation can yield a 5- to 10-fold higher level of expression than

Bilayer method Dialysis method

Buffer: ATP, AA, cofactors Extract,

Buffer:
mRNA
ATP, AA,
cofactors
Extract, mRNA

Figure 37.3 Schematic representation of two methods of cell-free translation. In the

bilayer method, the extract and mRNA are separated from ATP, amino acids (AA),
cofactors, and other buffer additives by the difference in density of the two solutions
(also see Fig. 37.1). In the dialysis method, the extract and mRNA are contained within
a dialysis membrane. In both cases, diffusion gives transfer of substrates, products, and
additives between the extract and buffer.
Cell-Free Translation 663

the bilayer method because the constituents of the extract are not diluted by
diffusion, which occurs during the course of the translation in the bilayer
reaction. Thus the dialysis method can yield purified membrane proteins in
the range of 0.2 mg/mL to greater than 2 mg/mL in the standard reactions
(Goren and Fox, 2008). Although not as easily adaptable to robotics, the
dialysis reaction can be performed on the benchtop and is scalable from 50 mL
to 10 mL or greater with no overall changes in volumetric productivity. In
the smaller volumes, this method has utility for simple screening of a few
proteins, while in the larger volumes it also has utility for scale-up of the
production of proteins whose properties have been investigated at small scale
and found to be favorable for more extensive studies. Protocols for cell-free
translation in automated batch, bilayer, and dialysis reactions have been
published (Endo and Sawasaki, 2006; Sawasaki et al., 2002a; Vinarov et al.,
2006). For translation of integral membrane proteins, we have found that
translation reactions are productively modified by the addition of unilamellar
liposomes (Goren and Fox, 2008; Nozawa et al., 2007).

12.1. Materials and reagents

WEPRO2240 wheat germ extract is from CellFree Sciences, Ltd.
(Yokohama, Japan). This preparation has  240 OD260/mL and does
not contain AAs. Store the extract at 80  C, thaw on the bench, and
flash-freeze the unused lysate before storing. For addition to translation
reactions, the extract is diluted from the concentrated commercial
preparation to a final OD260 of 60 with 1 Reaction Buffer.
Unlabeled AAs are from Advanced ChemTech (Louisville, KY). A mixture
of the 20 unlabeled AAs is prepared in Milli-Q water with each AA
present at 2 mM. Do not filter these preparations because some AAs will
not dissolve in these solutions.
5 Reaction Buffer is prepared from 150 mM HEPES–KOH, pH 7.8, and
contains 500 mM potassium acetate, 12.5 mM magnesium acetate, 2 mM
spermidine hydrochloride, 20 mM DTT, 6 mM ATP, 1.25 mM GTP,
80 mM creatine phosphate, and 0.025% (w/v) sodium azide. Store this
buffer at 80  C.
1 Overlay Buffer is prepared by fivefold dilution of 5 Reaction Buffer.
The overlay buffer is then amended with AAs solution to give a final
concentration of 0.3 mM for each AA.
Creatine kinase is from Roche Applied Sciences (Indianapolis, IN). Dissolve
in Milli-Q water to make a 50 mg/mL solution and store at 80  C.
Dilute the stock solution to 1 mg/mL prior to use. Avoid multiple cycles
of freezing and thawing of the concentrated solution. Discard the diluted
solution.
664 Michael A. Goren et al.

12 kDa molecular weight cutoff (MWCO) dialysis cups are from Cosmo
Bio (Tokyo, Japan). Prior to use, check the integrity of the membrane by
adding 500 mL of Milli-Q water and monitor for any leakage. If there is
no leakage, shake out the water prior to use.
The purified mRNA preparation is from Preparation of mRNA, step 4.
The liposome solution from Preparation of Liposomes, step 6.
24-Deep well pyramid-bottom plate, maximum volume of 10 mL (Artic-
white, Bethlehem, PA).
96-Well U-Bottom Plate (Greiner Bio-One, Monroe, NC).

12.1.1. Protocol for a bilayer reaction

1. Prepare a 20 mL translation mixture by mixing 2 mL of a 15 mg/mL
liposome solution, 4.25 mL of Milli-Q water, 2.75 mL of 5 Reaction
Buffer, 3.75 mL of 2 mM unlabeled AAs, 1 mL of 1 mg/mL creatine
kinase, and 6.25 mL of WEPRO2240 wheat germ extract. Depending
on the number of separate reactions desired, scale these volumes and
include  10% extra volume to account for handling losses.
2. Transfer 5 mL of an mRNA preparation to an individual well of a
96-well U-bottom plate.
3. Add 20 mL of translation mixture to the mRNA sample in each well,
and mix.
4. Form a bilayer by carefully adding 125 mL of 1 Overlay Buffer.
Be careful not to disrupt the layers as that can dilute the extract and
reduce protein production.
5. Incubate the reaction for 20 h at 26  C. Do not disturb the bilayer during
this time.
6. Protein translation levels can be determined by denaturing electropho-
resis SDS–PAGE with creatine kinase serving as an internal intensity
standard.

12.1.2. Protocol for a dialysis reaction

1. Dissolve the purified mRNA pellet in 50 mL of the translation mixture
prepared as described in Protocol for Bilayer Reaction, step 1.
2. Place the translation mixture into a 12-kDa MWCO dialysis cup.
3. Prepare the reservoir dialysis buffer by mixing 6.5 mL of Milli-Q water,
2.0 mL of 5 Reaction Buffer, and 1.5 mL of 2 mM unlabeled AAs.
Sonicate the mixture for 5 min, and then pass the solution through a
0.2-mm filter. Add 2.5 mL of reservoir dialysis buffer to each well of the
24-deep well plate.
4. Suspend the dialysis cup in a reservoir dialysis buffer. Be careful not
to trap air bubbles underneath the dialysis cup, which will disrupt
replenishment of additives.
Cell-Free Translation 665

5. Cover the 24-deep well plate with Saran Wrap to prevent evaporation
of the reservoir dialysis buffer. Incubate the translation reaction at 26  C
for 16 h.
6. Protein levels are determined by denaturing electrophoresis with creatine
kinase serving as an internal intensity standard.

13. Purification by Density Gradient

Ultracentrifugation
Figure 37.4 shows a schematic representation of the purification of
proteoliposomes containing membrane proteins produced by wheat germ
cell-free translation. After assembly of the density gradient, the proteolipo-
somes are separated from the cell-free extract proteins by a centrifugation
step. In most cases, the proteoliposomes are sharply concentrated at the
interface between the 30% Accudenz solution and the upper buffer.
Figure 37.5 shows a denaturing PAGE analysis of the separation of pro-
teoliposomes obtained by cotranslation of human stearoyl-CoA desaturase
(hSCD1) and cytochrome b5 (cytb5) in the wheat germ extract. hSCD1
accounted for  4% of the total protein present after translation in the presence
of liposomes. The density gradient purification yielded 24 mg of hSCD1 from a
50-mL translation reaction with greater than 80% purity. The density gradient
purification also gave a 25-fold increase in the specific activity of the enzyme,
and provided near complete recovery of the enzyme activity from the extract.

13.1. Materials and reagents

Accudenz is from Accurate Chemical and Scientific (Westbury, NY).
Prepare 80% (w/v) and 30% (w/v) solutions in 25 mM HEPES, pH 7.5,
containing 100 mM NaCl and 10% (w/v) glycerol. Store these solutions at
room temperature.
Buffer Liposomes
+
189,000 g membrane
240 min protein
30%
-Membrane
Accudenz
protein
expressed in
lysate
40% Lysate
-Liposome
Accudenz proteins

Figure 37.4 Schematic representation of the purification of proteoliposomes containing

membrane proteins produced by cell-free translation.
666 Michael A. Goren et al.

Density gradient top to bottom

250
1 2 3 4 5 6 7 8 9 10
150
100
75
50
37

25
20

15

10

Proteoliposomes Extract proteins

Figure 37.5 Denaturing PAGE analysis of the density gradient obtained after cell-free
translation of hSCD1. The order of fractions from the density gradient is indicated.
hSCD1 (black star) and a plant Hsp70-like protein (white star) are noted in lane 2,
which represents the interface between the upper buffer (lane 1) and the remainder of
the 30% Accudenz layer. The majority of wheat germ extract proteins remain in the
40% Accudenz layer. The rightmost lane contains molecular weight markers (kDa,
noted alongside image). Adapted from Goren and Fox (2008).

Ultraclear centrifuge tubes (5 mm id  41 mm h) are from Beckman

Coulter (Fullerton, CA).
The lipid rehydration buffer is 25 mM HEPES, pH 7.5, containing
100 mM NaCl.

13.1.1. Protocol
1. Carefully mix 75 mL of 80% (w/v) Accudenz solution with up to 75 mL
of the translation reaction, and place the mixed sample in the bottom of
an ultraclear centrifuge tube. This creates the 40% Accudenz layer.
2. Carefully layer 350 mL of 30% (w/v) Accudenz solution on top of the
mixture in the centrifuge tube. Minimize mixing of the gradient by
using a gel loading pipet tip to add the successive buffer layers.
3. Carefully layer 100 mL of lipid rehydration buffer on top of two other
layers in the centrifuge tube. This is the density gradient tube.
4. Spin the density gradient tube in an SW 50.1 rotor and L-60 ultracen-
trifuge for at least 4 h at 45,000 rpm (189,000g) and 4  C. The time
required for floatation is dependent on the properties of the proteolipo-
somes created, so optimization of density gradient and the centrifugation
time may be required to obtain the best separation for other proteins.
5. Carefully remove 60 mL fractions from the top to the bottom of the
gradient in order to fractionate the density gradient. Store the individual
Cell-Free Translation 667

fractions in separate 1.5-mL microfuge tubes. Typically, proteolipo-

somes will migrate to the interface between the 30% (w/v) Accudenz
solution and the liposome rehydration buffer.
6. Use denaturing SDS–PAGE to analyze the individual fractions for
protein content.

14. Characterization of Proteoliposomes

Another chapter has elegantly described the process of transfer of
solubilized membrane proteins from detergent into liposomes for functional
assays (Rigaud and Levy, 2003). In this chapter, the liposome is an integral
part of the enzyme production and purification process, so functional assays
are immediately feasible without further manipulations of the preparation.
The facile recovery of proteoliposomes from the cell-free translation reac-
tion by density gradient centrifugation permits functional assays on highly
enriched preparations with a minimal investment of time and effort. This
provides some obvious advantages in the discovery and characterization of
the membrane proteome (Sawasaki et al., 2002b; Wu et al., 2003), a large
and still poorly understood fraction of all proteins.
Figure 37.6 shows results of assays for conversion of 14C-labeled stear-
oyl-CoA (18:0) to oleoyl-CoA (18:1) obtained from different combinations
of proteoliposomes of hSCD1 and cyt5b produced by cotranslation or
independently with wheat germ extract (Goren and Fox, 2008). There is
no desaturase activity in the cell-free extract only control (lane 9). In order
to obtain hSCD activity, both cytb5 and Fe2þ must be added (lanes 1–3
versus lanes 4–8). Furthermore, iron and hemin must be optimized to obtain
maximal activity (lanes 4–8). Furthermore, in this case, cotranslation of
hSCD1 and cytb5 versus separate translation and subsequent mixing of
the proteoliposomes gave equivalent catalytic activity (lane 6 versus lane 8).
In many cases, the cell-free extract will not exhibit competing enzymatic
reactions that are present in whole cells, microsomes, and other impure
natural membrane preparations. This possibility can be experimentally
verified by use of the cell-free extract as a control. If verified, the low
background reactivity of the cell-free extract can have important implica-
tions for discovery of the function of unknown membrane proteins and
complexes. Moreover, the absence of competing background reactions may
facilitate additional characterizations using site-directed mutagenesis, inhi-
bitors, and other catalytic methods that have been more routinely carried
out with purified soluble enzymes.
The availability of functional proteoliposomes also facilitates investiga-
tions of the topology of cell-free translated membrane proteins by the
combination of proteolytic digestion and mass spectrometry (Speers and
668 Michael A. Goren et al.

Lane 1 2 3 4 5 6 7 8 9

18:0

18:1

Lane 1 2 3 4 5 6 7 8 9
hSCD1 (mM) 2.9 – 2.9 2.9 2.9 2.9 2.9 2.9 –
cytb5 (mM) – 5.8 5.8 5.8 5.8 5.8 5.8 5.8 –
Fe2+ (mM) 29 – – 29 29 29 29 29 29
Hemin (mM) – 5.8 5.8 – 2.9 5.8 11.6 5.8 5.8
% Conversion 3 – 2 33 54 43 17 42 –

Figure 37.6 Catalytic activity observed from various reconstitutions of the complex of
human stearoyl-CoA desaturase and human cytochrome b5 produced by cell-free transla-
tion and purified by density gradient centrifugation. The soluble domain of human
cytochrome b5 reductase was expressed in E. coli and added to these assays. The influence
of adding Fe2þ (required for hSCD1 activity) and hemin (required for cytb5 activity) is also
demonstrated. Lane 1, hSCD1 plus iron. Lane 2, cytb5 plus hemin. Lane 3, cotranslation of
hSCD1 and cytb5 supplemented with hemin. Lane 4, cotranslation of hSCD1 and cytb5
supplemented with Fe2þ. Lane 5, cotranslation of hSCD1 and cytb5 supplemented with
Fe2þ and 0.5 equiv. of hemin. Lane 6, cotranslation of hSCD1 and cytb5 supplemented
with Fe2þ and 1 equiv. of hemin. Lane 7, cotranslation of hSCD1 and cytb5 supplemented
with Fe2þ and 2 equiv. of hemin. Lane 8, mixture of separately translated and purified
hSCD1 with separately translated and purified cytb5 supplemented with Fe2þ and 1 equiv.
of hemin. Lane 5, cotranslation of hSCD1 and cytb5 supplemented with Fe2þ and 0.5
equiv. of hemin. Adapted from Goren and Fox (2008).

Wu, 2007; Wu et al., 2003). In some cases, the ability to perform protein
synthesis using radiolabeled AAs, unnatural AAs, or to specifically target one
or more AAs with diagnostic mass tags can advance these studies.
Further purification likely requires that the proteoliposomes be dissolved
with detergents, much as would be required for purification of a membrane
protein starting with a microsomal preparation from a living tissue. How-
ever, the combination of cell-free translation, liposomes, and density gradi-
ent centrifugation allows these additional purification efforts to start at a
much higher level of purity. Monitoring the decrease in light scattering
from the proteoliposome as detergent is added provides a simple diagnostic
for detergent optimization (Seddon et al., 2004; Womack et al., 1983). New
approaches using NMR, analytical ultracentrifugation, and gel filtration can
provide insight into the ability of a given detergent to yield monodisperse
protein-detergent micelles (Maslennikov et al., 2007; Slotboom et al.,
2008), which is generally considered to be an indicator of the compatibility
of protein and detergent. It is also possible to perform cell-free translation of
membrane proteins in the presence of detergents (Klammt et al., 2007;
Cell-Free Translation 669

Nozawa et al., 2007), with the constraints that the detergents used are
compatible with the cell-free translation reaction and with solubilization
of the nascent membrane protein.

15. Considerations for Scale-Up

With demonstration of suitable behavior using the bilayer or small-
volume dialysis reactions, scale-up of protein production is most effectively
carried out by use of the dialysis approach. Methods for performing the dialysis
reaction in large scale have been published (Madin et al., 2000; Spirin and
Swartz, 2008). More recently, a robotic dialysis device has become available to
address the need to produce large quantities of proteins using cell-free transla-
tion. Wheat germ cell-free translation has been used to solve the NMR
structures of numerous soluble proteins (Phillips et al., 2007; Tyler et al.,
2005), and more recently an X-ray structure has been solved (Makino et al.,
2009). Correspondingly, E. coli cell-free translation has been effectively used to
produce membrane proteins for refolding and, in the case of EmrE, preparation
of a SeMet-labeled sample for structure determination (Klammt et al., 2007;
Chen et al., 2007). To prepare sufficient mRNA for a large-scale protein
production, carry out a transcription reaction in a 50 mL conical tube with a
total volume of 4 mL of 1 Transcription Buffer plus Mg containing 4 mM
NTPs, 0.05 mg/mL of plasmid DNA, 0.5 unit/mL of SP6 RNA polymerase,
and 0.25 unit/mL of RNasin. Incubate the reaction at 37  C for 3–5 h.

16. Isotopic Labeling for Structural Studies

Because AAs are added to the wheat germ extract as substrates for the
protein synthesis, SeMet, 2H-, 13C-, 15N- or other isotopically labeled AAs
can be easily substituted. There is no significant metabolism of AAs in wheat
germ extract except for alanine transaminase (beta-chloro-L-alanine), aspar-
tate transaminase (aminooxyacetate), and glutamine synthetase (L-methio-
nine sulfoximine), which can be inhibited by the compounds indicated
(Endo and Sawasaki, 2006; Morita et al., 2004). Otherwise, the level of
isotopic enrichment in translated proteins consistently matches that of the
AA precursors used in the translation reactions. 15N and 13C, 15N-labeled
AAs are from Cambridge Isotope Laboratories (Andover, MA) or other
sources. For NMR studies, the 15N-labeled AAs are prepared at 8 mM, and
the 13C, 15N-labeled AAs are prepared at 5 mM, also in Milli-Q water.
These preparations are substituted into the standard protocols to give the
same final concentrations as described above. Likewise SeMet (Acros
Organics, Morris Plains, NJ) is prepared as a 2-mM stock solution and
substituted for methionine as described earlier.
670 Michael A. Goren et al.

17. Conclusions
The process simplifications afforded by the cell-free translation
approaches described herein provide significant labor and time advantages,
and this conclusion is emphasized when cell-free translation can produce
integral membrane proteins and enzymes that cannot be reasonably
obtained in a catalytically active state by any other method. By comparison,
other methods for the preparation of membrane proteins beginning with
expression in living hosts are time-, labor-, and material-intensive. Through
application of these methods, uniform samples of membrane proteins can be
easily obtained as the starting point for optimization of catalytic assays,
purification procedures, antibody production, structure determination,
and many other types of studies. This work provides an attractive alternative
to the sequence of expressing in a living host, extracting membrane proteins
with detergents, and then transferring them back into liposomes or other
lipid bilayers for functional studies. It is reasonable to assume that continued
application of this approach could open many new avenues to investigations
of membrane protein structure and function.

ACKNOWLEDGMENTS
This work was supported by NIGMS grant GM50853 to BGF, NIGMS Protein Structure
Initiative grant 1U54 GM074901 ( J. L. Markley, PI, G. N. Phillips, Jr., and B. G. Fox,
Co-Investigators) and NSF East Asia and Pacific Summer Institutes for U.S. Graduate
Students to MAG.

REFERENCES
Abe, M., Hori, H., Nakanishi, T., Arisaka, F., Ogasawara, T., Sawasaki, T., Kitamura, M.,
and Endo, Y. (2004). Application of cell-free translation systems to studies of cofactor
binding proteins. Nucleic Acids Symp. Ser. (Oxf.) 48, 143–144.
Blommel, P. G., and Fox, B. G. (2007). A combined approach to improving large-scale
production of tobacco etch virus protease. Protein Expr. Purif. 55, 53–68.
Blommel, P. G., Martin, P. A., Wrobel, R. L., Steffen, E., and Fox, B. G. (2006). High
efficiency single step production of expression plasmids from cDNA clones using the
Flexi Vector cloning system. Protein Expr. Purif. 47, 562–570.
Boyer, M. E., Stapleton, J. A., Kuchenreuther, J. M., Wang, C. W., and Swartz, J. R. (2008).
Cell-free synthesis and maturation of [FeFe] hydrogenases. Biotechnol. Bioeng. 99, 59–67.
CellFree Sciences, Ltd. Yokohama, Japan. http://www.cfsciences.com.
Chen, Y. J., Pornillos, O., Lieu, S., Ma, C., Chen, A. P., and Chang, G. (2007). X-ray structure
of EmrE supports dual topology model. Proc. Natl. Acad. Sci. USA 104, 18999–19004.
Drew, D. E., von Heijne, G., Nordlund, P., and de Gier, J. W. (2001). Green fluorescent
protein as an indicator to monitor membrane protein overexpression in Escherichia coli.
FEBS Lett. 507, 220–224.
Cell-Free Translation 671

Endo, Y., and Sawasaki, T. (2006). Cell-free expression systems for eukaryotic protein
production. Curr. Opin. Biotechnol. 17, 373–380.
Goerke, A. R., and Swartz, J. R. (2008). Development of cell-free protein synthesis plat-
forms for disulfide bonded proteins. Biotechnol. Bioeng. 99, 351–367.
Goren, M. A., and Fox, B. G. (2008). Wheat germ cell-free translation, purification, and
assembly of a functional human stearoyl-CoA desaturase complex. Protein Expr. Purif. 62,
171–178.
Hall, J. F., Ellis, M. J., Kigawa, T., Yabuki, T., Matsuda, T., Seki, E., Hasnain, S. S., and
Yokoyama, S. (2005). Towards the high-throughput expression of metalloproteins from
the Mycobacterium tuberculosis genome. J. Synchrotron Radiat. 12, 4–7.
Kanno, T., Kitano, M., Kato, R., Omori, A., Endo, Y., and Tozawa, Y. (2007). Sequence
specificity and efficiency of protein N-terminal methionine elimination in wheat-
embryo cell-free system. Protein Expr. Purif. 52, 59–65.
Kiga, D., Sakamoto, K., Kodama, K., Kigawa, T., Matsuda, T., Yabuki, T., Shirouzu, M.,
Harada, Y., Nakayama, H., Takio, K., Hasegawa, Y., Endo, Y., et al. (2002). An
engineered Escherichia coli tyrosyl-tRNA synthetase for site-specific incorporation of an
unnatural amino acid into proteins in eukaryotic translation and its application in a wheat
germ cell-free system. Proc. Natl. Acad. Sci. USA 99, 9715–9720.
Kigawa, T., Yabuki, T., Matsuda, N., Matsuda, T., Nakajima, R., Tanaka, A., and
Yokoyama, S. (2004). Preparation of Escherichia coli cell extract for highly productive
cell-free protein expression. J. Struct. Funct. Genomics 5, 63–68.
Klammt, C., Lohr, F., Schafer, B., Haase, W., Dotsch, V., Ruterjans, H., Glaubitz, C., and
Bernhard, F. (2004). High level cell-free expression and specific labeling of integral
membrane proteins. Eur. J. Biochem. 271, 568–580.
Klammt, C., Schwarz, D., Lohr, F., Schneider, B., Dotsch, V., and Bernhard, F. (2006).
Cell-free expression as an emerging technique for the large scale production of integral
membrane protein. FEBS J. 273, 4141–4153.
Klammt, C., Schwarz, D., Dotsch, V., and Bernhard, F. (2007). Cell-free production of
integral membrane proteins on a preparative scale. Methods Mol. Biol. 375, 57–78.
Madin, K., Sawasaki, T., Ogasawara, T., and Endo, Y. (2000). A highly efficient and robust
cell-free protein synthesis system prepared from wheat embryos: Plants apparently contain
a suicide system directed at ribosomes. Proc. Natl. Acad. Sci. USA 97, 559–564.
Makino, S.-I., Bingman, C. A., Berge, S., Larkin, A., Fox, B. G., Phillips, G. N. Jr, and
Markley, J. L. (2009). Crystal structure of agmatine iminohydrolase produced by wheat germ cell-
free translation (in preparation).
Maslennikov, I., Kefala, G., Johnson, C., Riek, R., Choe, S., and Kwiatkowski, W. (2007).
NMR spectroscopic and analytical ultracentrifuge analysis of membrane protein detergent
complexes. BMC Struct. Biol. 7, 74.
Matsuda, T., Koshiba, S., Tochio, N., Seki, E., Iwasaki, N., Yabuki, T., Inoue, M.,
Yokoyama, S., and Kigawa, T. (2007). Improving cell-free protein synthesis for stable-
isotope labeling. J. Biomol. NMR 37, 225–229.
Matsumoto, K., Tomikawa, C., Toyooka, T., Ochi, A., Takano, Y., Takayanagi, N.,
Abe, M., Endo, Y., and Hori, H. (2008). Production of yeast tRNA (m(7)G46) methyl-
transferase (Trm8-Trm82 complex) in a wheat germ cell-free translation system.
J. Biotechnol. 133, 453–460.
Morita, E. H., Shimizu, M., Ogasawara, T., Endo, Y., Tanaka, R., and Kohno, T. (2004).
A novel way of amino acid-specific assignment in 1H–15N HSQC spectra with a wheat
germ cell-free protein synthesis system. J. Biomol. NMR 30, 37–45.
Nomura, S. M., Kondoh, S., Asayama, W., Asada, A., Nishikawa, S., and Akiyoshi, K.
(2008). Direct preparation of giant proteo-liposomes by in vitro membrane protein
synthesis. J. Biotechnol. 133, 190–195.
672 Michael A. Goren et al.

Nosjean, O., and Roux, B. (2003). Anchoring of glycosylphosphatidylinositol-proteins to

liposomes. Methods Enzymol. 372, 216–232.
Nozawa, A., Nanamiya, H., Miyata, T., Linka, N., Endo, Y., Weber, A. P., and Tozawa, Y.
(2007). A cell-free translation and proteoliposome reconstitution system for functional
analysis of plant solute transporters. Plant Cell Physiol. 48, 1815–1820.
Phillips, G. N. Jr., Fox, B. G., Markley, J. L., Volkman, B. F., Bae, E., Bitto, E.,
Bingman, C. A., Frederick, R. O., McCoy, J. G., Lytle, B. L., Pierce, B. S., Song, J.,
et al. (2007). Structures of proteins of biomedical interest from the Center for Eukaryotic
Structural Genomics. J. Struct. Funct. Genomics 8, 73–84.
Rigaud, J. L., and Levy, D. (2003). Reconstitution of membrane proteins into liposomes.
Methods Enzymol. 372, 65–86.
Sawasaki, T., Hasegawa, Y., Tsuchimochi, M., Kasahara, Y., and Endo, Y. (2000). Con-
struction of an efficient expression vector for coupled transcription/translation in a wheat
germ cell-free system. Nucleic Acids Symp. Ser. 44, 9–10.
Sawasaki, T., Hasegawa, Y., Tsuchimochi, M., Kamura, N., Ogasawara, T., Kuroita, T.,
and Endo, Y. (2002a). A bilayer cell-free protein synthesis system for high-throughput
screening of gene products. FEBS Lett. 514, 102–105.
Sawasaki, T., Ogasawara, T., Morishita, R., and Endo, Y. (2002b). A cell-free protein
synthesis system for high-throughput proteomics. Proc. Natl. Acad. Sci. USA 99,
14652–14657.
Schwarz, D., Klammt, C., Koglin, A., Lohr, F., Schneider, B., Dotsch, V., and Bernhard, F.
(2007). Preparative scale cell-free expression systems: New tools for the large scale
preparation of integral membrane proteins for functional and structural studies. Methods
41, 355–369.
Seddon, A. M., Curnow, P., and Booth, P. J. (2004). Membrane proteins, lipids and
detergents: Not just a soap opera. Biochim. Biophys. Acta 1666, 105–117.
Slotboom, D. J., Duurkens, R. H., Olieman, K., and Erkens, G. B. (2008). Static light
scattering to characterize membrane proteins in detergent solution. Methods 46, 73–82.
Sobrado, P., Goren, M. A., James, D., Amundson, C. K., and Fox, B. G. (2008). A protein
structure initiative approach to expression, purification, and in situ delivery of human
cytochrome b5 to membrane vesicles. Protein Expr. Purif. 58, 229–241.
Speers, A. E., and Wu, C. C. (2007). Proteomics of integral membrane proteins—Theory
and application. Chem. Rev. 107, 3687–3714.
Spirin, A. S., and Swartz, J. R. (eds.) (2008). In ‘‘Cell-Free Protein Synthesis-Methods and
Protocols’’, WILEY-VCH Verlag Gmbh & Co., Weinheim.
Thao, S., Zhao, Q., Kimball, T., Steffen, E., Blommel, P. G., Riters, M., Newman, C. S.,
Fox, B. G., and Wrobel, R. L. (2004). Results from high-throughput DNA cloning of
Arabidopsis thaliana target genes using site-specific recombination. J. Struct. Funct. Geno-
mics 5, 267–276.
Tyler, R. C., Aceti, D. J., Bingman, C. A., Cornilescu, C. C., Fox, B. G., Frederick, R. O.,
Jeon, W. B., Lee, M. S., Newman, C. S., Peterson, F. C., Phillips, G. N. Jr.,
Shahan, M. N., et al. (2005). Comparison of cell-based and cell-free protocols for
producing target proteins from the Arabidopsis thaliana genome for structural studies.
Proteins 59, 633–643.
Vinarov, D. A., and Markley, J. L. (2005). High-throughput automated platform for nuclear
magnetic resonance-based structural proteomics. Expert Rev. Proteomics 2, 49–55.
Vinarov, D. A., Lytle, B. L., Peterson, F. C., Tyler, E. M., Volkman, B. F., and
Markley, J. L. (2004). Cell-free protein production and labeling protocol for
NMR-based structural proteomics. Nat. Methods 1, 149–153.
Vinarov, D. A., Loushin Newman, C. L., and Markley, J. L. (2006). Wheat germ cell-free
platform for eukaryotic protein production. FEBS J. 273, 4160–4169.
Cell-Free Translation 673

Womack, M. D., Kendall, D. A., and Macdonald, R. C. (1983). Detergent effects on

enzyme activity and solubilization of lipid bilayer membranes. Biochim. Biophys. Acta
733, 210–215.
Wu, C. C., MacCoss, M. J., Howell, K. E., and Yates, J. R. 3rd. (2003). A method for the
comprehensive proteomic analysis of membrane proteins. Nat. Biotechnol. 21, 532–538.
Wuu, J. J., and Swartz, J. R. (2008). High yield cell-free production of integral membrane
proteins without refolding or detergents. Biochim. Biophys. Acta 1778, 1237–1250.
Yokoyama, S. (2003). Protein expression systems for structural genomics and proteomics.
Curr. Opin. Chem. Biol. 7, 39–43.
 C H A P T E R T H I R T Y- E I G H T

Determination of Protein Purity

David G. Rhodes* and Thomas M. Laue†

Contents
1. Composition-Based and Activity-Based Analyses 680
1.1. Problems and pitfalls 681
2. Electrophoretic Methods 681
2.1. Methods 682
2.2. Pitfalls and problems 684
3. Chromatographic Methods 685
3.1. Gel filtration chromatography 685
3.2. Reversed phase HPLC 686
4. Sedimentation Velocity Methods 687
4.1. Method 687
5. Mass Spectrometry Methods 687
6. Light Scattering Methods 688
References 689

Abstract
There is no means for directly quantifying the purity of a protein sample.
Demonstration of purity of a protein preparation always involves an assessment
of the quantity of particular types of impurities, or simply demonstrating their
absence. Whether the goal is interpretation of analytical data, demonstration of
the quality of a process, or assuring safety of a biopharmaceutical product,
determination of sample purity is of central importance. This chapter will focus
primarily on methods for detecting protein impurities in protein samples and
will describe methods, their capabilities, and their limitations.

To assess the purity of a sample, one must first identify the type of
impurity that is to be measured (i.e., nucleic acid, carbohydrate, lipid,
unrelated proteins, isozymes, inactive enzyme), and then identify a
characteristic property (chemical assay or physical feature) which can distin-
guish the putative contaminant(s) in the presence of the protein of interest
in a specified solution condition. Purity is then the demonstration that the

* Lipophilia Consulting, Storrs, Connecticut, USA

{
Department of Biochemistry and Molecular Biology, University of New Hampshire, Durham,
New Hampshire, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63038-X All rights reserved.

677
678 David G. Rhodes and Thomas M. Laue

contaminant in question is below a specified level. Note that this specifica-

tion need not characterize the contaminant. A chromatographic peak may
remain an unknown even after a purification procedure has reduced its
concentration to levels below detection limits. Clearly, the apparent level of
purity will depend on which assay methods are chosen and their sensitivity.
Since most isolation procedures are quite good at removing nonprotein
contaminants, this chapter will focus primarily on methods for detecting
contamination of a protein sample by other proteins. Selective assays for the
presence of nucleic acids, lipids, and carbohydrates can be found in other
volumes of this series.
Purity of proteins has become a significant issue in regulation of
pharmaceutical products as protein-based products become more preva-
lent. The International Conference on Harmonisation (ICH) quality
guidelines for biological products (ICH Q6B, 1998) consider impurities
in drug substance (bulk material) and drug product (dosage form or
finished product) and recognize the inherent heterogeneity in biological
molecules like proteins. The guidelines state: ‘‘In addition to evaluating
the purity of the drug substance and drug product, which may be com-
posed of the desired product and multiple product-related substances, the
manufacturer should also assess impurities which may be present. Impu-
rities may be either process or product-related. They can be of known
structure, partially characterized, or unidentified.’’ The guidelines go on to
distinguish process-related impurities (derived from the manufacturing
process, and including host cell proteins, host cell DNA, cell culture
inducers, antibiotics, media components, or the effects of downstream
processing) and product-related impurities, including ‘‘molecular variants
arising during manufacture and/or storage, which do not have properties
comparable to those of the desired product with respect to activity,
efficacy, and safety.’’ In the guidelines, there is recognition of the analytical
capabilities described above. For example, with regard to drug substance,
‘‘The absolute purity of biotechnological and biological products is
difficult to determine and the results are method-dependent.. . .
Consequently, the purity of the drug substance is usually estimated by a
combination of methods. The choice and optimization of analytical pro-
cedures should focus on the separation of the desired product from
product-related substances and from impurities.’’ The discussion for drug
product is similar, but also includes testing for degradation products and
product-related impurities due to interaction with excipients.
There are several sensitive methods that may be used to detect the
presence of impurities in a sample (Table 38.1). Each method tests for a
specific physical property of the molecules. The method of choice depends
on the following criteria: (1) the quantity of protein available, (2) the
nature of the impurity being evaluated, (3) the accuracy of the estimate
needed, (4) the sensitivity of the test needed, and (5) any peculiarities of
Table 38.1 Methods for detecting impurities

Method Fractionationa Sample recoveryb Quantity Easec Expensed Property assessed

Electrophoresis
Denaturing gel þ  ng–mg þ þ Chain length
Native gel þ þ ng–mg þ þ Charge/size
Isoelectric focusing þ þ ng–mg þ  Isoelectric pH
Chromatography
Gel permeation þ þ mg þ þ Size, rg
Ion exchange þ þ mg þ þ Net charge
Affinity þ þ mg þ  Specific binding
Reverse phase þ þ mg þ  Polarity
Sedimentation
Velocity þ þ mg þ  Mass/size
Equilibrium  þ mg   Mass, association
Light scattering
Static  þ mg þ  Mass/size
Dynamic  þ mg þ  Mass/size
Mass spectrometry þ  ng–mg   Mass/charge ratio
Composition   Variable þ  Component content
Activity   Variable þ þ Presence of active site
a
þ, Sample fractionated during analysis; separated components may be isolated; , incomplete fractionation, no components isolable; , no sample fractionation.
b
þ, Good recovery of sample following assay; , sample may be recovered in favorable cases; , sample destroyed during analysis.
c
þ, Simple method; , somewhat more difficult; , special equipment or expertise may be required.
d
þ, Equipment and supplies relatively inexpensive; , some equipment or supplies may be costly; , expensive equipment or supplies required.
680 David G. Rhodes and Thomas M. Laue

the protein or its solvent which might interfere with the use of the
techniques. It is easiest and most common to carry out one or more
fractionation procedures and demonstrate that only one component is
detectable. A wide variety of fractionation procedures may be used to
detect impurities, with the criterion for purity being the presence of a
single detectable component. Orthogonal methods are advisable to guard
against situations in which a chosen detection method is incapable of
resolving an unknown impurity from the primary analyte. Finally, it is
important to handle samples carefully, so that the process of preparing a
sample for analysis does not alter the impurity profile. Cleanliness, proper
temperature, and containers made of appropriate materials will all
contribute to informative analysis.
Many appropriate methods are detailed elsewhere in this volume and
will not be reviewed here. Several analytical methods for determining
molecular weight or molecular size are outlined elsewhere in this volume,
so where appropriate, this discussion will refer to details in that chapter
(Rhodes et al., 2009). In this chapter, emphasis will be on the variations of
those methods which are better suited to detection of contaminant than to
quantitative determination of size, mass, or other molecular parameters.
This chapter will also outline criteria for selecting among these methods,
some of their limitations and advantages, and some description of assays for
purity that do not involve fractionation.

1. Composition-Based and
Activity-Based Analyses
Methods that provide molar quantification of amino acids, specific
prosthetic groups, or of active sites may be used to assess the purity of a
sample. In cases where the specific activity of the pure material is known it
may be appropriate to determine the unit activity as a measure of relative
purity. These methods are secondary in nature, since reference is always
made to the quantity of the analyte that would be expected for a sample that
is free of contaminants. Moreover, these methods are appropriate primarily
with heterogeneous systems in early phases of purification or molecules like
membrane proteins which may require a specific environment to retain
function. Typically, two measurements are required: the mass of the protein
(or material) in the sample that will be used in the analysis, and quantifica-
tion of the known activity or of the particular analyte. A calculation is then
made of the expected quantity of analyte based on the mass of the material
used in the analysis. The purity is then expressed as ratio of the amount
measured to the amount expected. Good quantification has been achieved
using end-group analysis (Chang, 1983), and quantitative analysis for
Determination of Protein Purity 681

specific prosthetic groups, and for enzymatic activity (Biggs, 1976). The
details for this method are the same as those described in this volume
Rhodes et al. (2009).

1.1. Problems and pitfalls

Composition-based or activity-based estimates are sensitive only to those
impurities being assessed, and typically provide little specific information
(e.g., size, charge, etc.) concerning the nature of the contaminants. Activity
measurements may provide no information about the contaminant unless
the contaminant interferes with the assay or the active protein. Thus, few
clues are provided to help the experimenter in removing the contaminant.

2. Electrophoretic Methods
Because the electrophoretic methods provide some of the simplest,
least expensive, and often most sensitive approaches to determine the
number of protein components in a sample, they are the most commonly
used methods. Because of the exceptionally low cost and simplicity, these
methods are often the first screen for sample purity, even in early-stage,
highly heterogeneous samples. The methods of SDS gel electrophoresis
(see Friedman, 2009; Garfin, 2009) and electrophoretic determinations of
molecular weight and size are described elsewhere in this volume (Rhodes
et al., 2009). Any of these approaches could be used independently to assess
the purity of a sample, depending on which characteristic of the protein is to
be tested (Table 38.1). If the expected contaminants differ in molecular
weight from that of the desired protein, SDS gel electrophoresis would
resolve the impurity. Molecules of similar molecular weight but different
amino acid composition would generally not be distinguishable using SDS
gel electrophoresis, but could have different electrophoretic mobilities in
nondenaturing gel electrophoresis. Finally, proteins of almost any molecular
weight might be separable using isoelectric focusing techniques. The latter
method is detailed elsewhere in this volume (Friedman, 2009), and will not
be treated here.
Depending on the type of detection method available, nanogram to
microgram quantities of sample are required. Since a contaminant consti-
tutes a fixed weight fraction of a given sample, while the detection of a
contaminant depends on the total amount of contaminant, an overloaded
gel may provide a better chance of detecting a contaminant. However, an
upper limit to the quantity of sample that can be applied to a gel is set by
sample solubility and by considerations of resolution (Lunney et al., 1971).
The latter limit is usually more restrictive, with band broadening and
682 David G. Rhodes and Thomas M. Laue

distortion occurring at higher sample concentrations and with large sample

volumes. Bands broadened due to overloading could obscure an impurity
with similar electrophoretic mobility. Unfortunately, the most sensitive
band detection methods (requiring the smallest amount of sample) do not
allow sample recovery. If the protein sample is denatured, or otherwise
solubilized under harsh conditions, it is usually difficult to recover
functional protein. However, if nondenaturing electrophoresis is used, it
is possible to recover the sample from the gel by electrophoretic extraction
(Friedman, 2009; Garfin, 2009). If denaturing electrophoresis is performed,
it may not be possible to recover the native protein (but see Burgess, 2009).
The success or failure of renaturation depends largely on the structure of the
protein; a large and or multisubunit protein will be less likely to return to
the native conformation than will a small monomeric protein.

2.1. Methods
The details of sample preparation depend on the nature of the electropho-
retic method chosen. The reader is referred to the other chapters in this
volume for discussion of native gel electrophoresis and isoelectric focusing
gels. Most frequently, SDS gels provide the ‘‘front-line’’ method for asses-
sing sample purity. Because purity assessment often is not quantitative, one
need not (necessarily) be as concerned about such problems as nonlinearity
in mobility at protein size extremes, aberrations in mobility due to protein
modifications, or nonuniformities in SDS binding that are discussed in
Rhodes et al. (2009). On the other hand, one does need to be concerned
with consistency and uniformity in sample handling and preparation, and
the conditions to which the sample is exposed during fractionation. For
example, a homogeneous protein preparation in which the reduction of
disulfide is not carried out carefully could appear to be heterogeneous.
Likewise, misinterpretation is possible if nonuniformity in a gel cross linking
results in mobility differences across the gel, but suitable replicates and
controls should minimize such issues.
Because one does not normally know the sizes of all protein components
within a sample, it is difficult to predict the gel concentration which will
give the optimal fractionation of a set of protein components. Gradient gels,
or gels of graded porosity, can cover a very wide range of molecular sizes.
Although the conditions of a gradient gel are unlikely to be optimal for a
particular fractionation, they will cover the widest possible range of
conditions, thus maximizing the probability of identifying a contaminant.
Ability to resolve a contaminant will depend largely on the range of gel
concentrations chosen for the gradient. The gradient should be designed
so that the protein of interest bands at an intermediate gel concentration.
The concentration of acrylamide at the top of the gel (low concentration)
Determination of Protein Purity 683

should be sufficiently low that large molecular weight contaminants may

enter the gel matrix. The bottom of the gel (high concentration) should be
sufficiently concentrated that small protein is retained in the gel matrix. The
lowest concentration commonly available in precast gradient gels is 4%
acrylamide and the standard high concentration limit is 20%. If these limits
are not adequate, lower concentrations (2%) or higher concentrations (as
high as 40%) can be made. Proteins with molecular weights on the order of
1 MDa should be able to penetrate a 2% gel matrix. The high concentration
for most gradient gels generally does not exceed 30% acrylamide. Very few
polypeptides will penetrate a 30% gel, even after prolonged electrophoresis.
The concentration range of a gradient gel determines the resolution of the
gel. Thus, a gel with a gradient from 2% to 30% will be able to examine the
widest range of molecular weights, but will not provide as much discrimi-
nation between two proteins of nearly identical molecular weights as a gel
with an 8–16% gradient. Choice of gradient will depend largely on the
researcher’s knowledge about the system under investigation. If it is known,
for example, that no very small or very large protein is present, one may
wish to use a narrower range of concentrations in the gradient or a gel of
constant concentration.
If a specific gradient is required, the procedure for making gradient gels
differs only slightly from that for making conventional polyacrylamide gels
or that for making sucrose gradients for sedimentation. The procedure is
slightly more difficult than that for conventional polyacrylamide gels only
because two gels at different acrylamide concentrations must be made up in
parallel. The procedure is more difficult than forming sucrose gradients for
sedimentation only because one must work within the time constraints of
the acrylamide polymerization. Several commercial gradient-forming
devices are available and many variations on the technique may be found
by searching the literature.
Electrophoresis procedures for a gradient gel are identical to those used
in the analytical methods with conventional gels (Rhodes et al., 2009, and
elsewhere in this volume). The staining strategy for analysis of the gel will be
determined by the expected contaminant. The potential sensitivity of this
approach and the information gained through its use can be enhanced by
combining it with other techniques to generate a two-dimensional electro-
phoretic analysis. While the second dimension would ordinarily be a gradi-
ent SDS–polyacrylamide gel electrophoresis, the first dimension could be a
nondenaturing gel electrophoresis, an isoelectric focusing gel, or a denaturing
gel under different conditions (e.g., without disulfide reduction).
Isoelectric focusing is another technique which can successfully span a
wide range of potential contaminants and, since it is orthogonal to
electrophoresis, can be combined with these methods. Commercially
available precast gels cover the range of pI from 3 to 10 in the same gel.
The procedures for carrying out isoelectric focusing have been described
684 David G. Rhodes and Thomas M. Laue

elsewhere in this volume and will not be repeated here. The wide range of
pI can be a considerable advantage, but the sensitivity of the technique to
small differences in isoelectric point can be enhanced by covering a
narrower range of pI values (e.g., 4–5, 5–8). Aside from the obvious
capability of differentiating molecules which differ in pI, this method
may be able to resolve contaminants which differ by properties other
than isoelectric point.

2.2. Pitfalls and problems

Because gel electrophoresis is such a simple and inexpensive method, it is
usually the method of first choice for purity assessment. There are, how-
ever, a few potential problems that must be borne in mind when using one
of these techniques. For denaturing gels, both false-negative and false-
positive artifacts are possible. False negatives can result if a comigrating
contaminant is present or if a contaminant is unable to penetrate the gel.
For this reason, it is important to stain the entire gel, stacking and running
portions, and to examine both for protein-staining material. A band of
material left behind in the loading well or at the interface between
the stacking gel and the running gel is an indication of either a high-
molecular-weight contaminant or a contaminant of limited solubility.
Likewise, commonly used protein stains such as Coomassie Brilliant Blue
will bind weakly to fibrous proteins or glycoproteins, resulting in an
underestimate of these contaminants. False positives may result when
covalent modifications are made during preparation of the sample for
electrophoresis or when the gel is nonuniform or contains residual oxi-
dant. Similar problems may arise in nondenaturing gels, with the addi-
tional problems caused by uncertainties as to the net charge on the protein
of interest or the contaminants (see discussion in Rhodes et al., 2009). If
the net charge on the contaminants is zero or of opposite sign from that on
the protein of interest, the contaminant will not appear in the gel unless it
is a center-loaded horizontal gel. It is also helpful to run nondenaturing
gels over as wide a range of pH as possible.
Artifacts associated with isoelectric focusing arise mostly from interac-
tions between the proteins and the polyampholyte beads used to create the
pH gradient. The results are bands (often smeared) appearing in regions of
the gel that may have nothing whatsoever to do with the isoelectric point of
the protein of interest. One way to test for this possibility is to isolate protein
from one of these bands and to analyze the isolate using the same isoelectric
focusing protocol. If the original pattern reflects actual heterogeneity in the
sample, the isolate will form only its original spot. However, if the bands are
artifacts of the method, the original pattern, including the artifact bands will
be regenerated by the isolate.
Determination of Protein Purity 685

3. Chromatographic Methods
3.1. Gel filtration chromatography
Gel filtration chromatography is one of the simplest methods for the
detection of impurities that differ in size from the molecule of interest.
The method is nondestructive and rapid. Samples are diluted as they pass
through the column because this is a zonal method, so starting concentra-
tions well above the minimum detectable limit must be used. The exact
quantity of material needed will depend on the sensitivity of the assay used
to detect for contaminants.

3.1.1. Method
The method outlined in this volume for sample and column preparation for
the purposes of protein size determination should be used for assessing
impurities (Rhodes et al., 2009). The only difference is that the method of
detection must be sensitive to contaminants as well as to the protein of
interest. Although it is not necessary to calibrate the column to assess the
purity of a sample, it is useful to do so for two reasons. First, by using a
calibrated column, the experimenter gains some information concerning
the size of the impurity. Second, by using a calibrated column, both the test
for impurities and the molecular weight determination can be made simul-
taneously. To do this, one frequently uses two assays. The first is a nonspe-
cific assay for protein (e.g., absorbance at 280 nm and/or 220 nm) and the
second is a specific assay (enzymatic, immunological, mass-spectroscopy,
multiangle light scattering, etc.) to confirm the identity of the analyte
protein. Impurities are detected as either separate peaks in the chromato-
gram or as a broadening (e.g., a shoulder) of the elution profile. In principle,
the elution profile should be nearly Gaussian (with a very slight skewing of
the trailing edge, see Problems and Pitfalls). If the analyte protein peak is
strongly skewed, if the peak has a shoulder, or if the specific and nonspecific
assay results to not superimpose the sample has impurity present.

3.1.2. Problems and pitfalls

Gel-permeation chromatography is not as sensitive method as gel electro-
phoresis for the detection of size heterogeneity. Typically, a gel permeation
experiment requires larger quantities of material than a gel electrophoresis
experiment. Since gel-permeation chromatography usually is conducted
under native conditions, it is possible that association behavior will mimic
sample heterogeneity. A protein that exists in a variety of stable oligomers
(e.g., fatty acid synthase) will appear to be heterogeneous by this method.
Likewise, a protein that undergoes rapid, reversible self-association can
provide anomolous, concentration-dependent elution profiles. In each
686 David G. Rhodes and Thomas M. Laue

case, it is possible to distinguish between heterogeneity and molecular

association by further testing. For example, the gel permeation chromatog-
raphy could be repeated with fractions from different portions of the skewed
or broadened peak. Similarly, other methods described in this chapter or
elsewhere in this volume could be applied to these fractions. In either case,
the experimenter obtains useful information about the molecule of interest.

3.2. Reversed phase HPLC

Another commonly used chromatographic separation method is reversed
phase HPLC. In this method, a nonpolar stationary phase such as modified
silica is used with an elution gradient of decreasing polarity. Thus, for
example, a protein in a buffer could be applied to a C-18 modified silica
column equilibrated with aqueous 0.05% trifluoroacetic acid (TFA). The
protein will generally bind to the stationary phase by way of interaction
between the C-18 and hydrophobic amino acids on the protein. The
protein would then be eluted using a gradient from 0.05% aqueous TFA
to some level of compatible organic, such as acetonitrile, methanol, or
ethanol, typically with the same quantity of TFA. The presence of the
organic solvent in the mobile phase decreases the affinity of the protein
for the stationary phase and the protein elutes. As with most gradient
methods, a shallow gradient will provide optimal resolution, but a fast
gradient is a good first step to identify approximate solvent conditions for
elution of the primary analyte and to screen for possible impurities with very
different affinities for the stationary phase. Detection of the protein is usually
based on UV absorbance at 215–220 nm, which will identify peptides as
opposed to measurements based on aromatic amino acids at 280 nm. Aside
from setup time, a gradient can be completed in well under an hour and
most optimized elutions run in 15–20 min.
The simplicity and rapidity of this method, the ready availability of
suitable instrumentation, and the flexibility of the method in accommodat-
ing proteins with widely differing properties, make this a common method
of choice for screening protein samples for purity. With a variety of mobile
phases, a sample can be analyzed under a variety of gradients and conditions,
if necessary, to confirm that the primary analyte is pure. The method can be
combined with other methods such as MS (see below) to confirm the
identity of the protein and to obtain information about possible impurities.
It should be noted that because of the presence of organic solvent, the
proteins that elute cannot be assumed to be native; some degree of denatur-
ation is expected. Denaturation can be minimized by using a stationary phase
with C-4 or C-8; because these will have lower affinity for the protein, it will
elute at lower concentrations of organic mobile phase. Numerous protocols
have been published for specific proteins and manufacturers of HPLC
equipment have extensive product-specific information.
Determination of Protein Purity 687

4. Sedimentation Velocity Methods

Sedimentation velocity is a simple, rapid, nondestructive technique for
evaluating the purity of a protein. The technique is sensitive to the ratio of
the molecular weight to the molecular size. An overview of the topic of
sedimentation velocity is provided in Rhodes et al. (2009). Typically, when
using sedimentation velocity as a test of purity, an experimenter is looking
for the presence of more than one sedimentable component. The virtue of
the technique is that it can be used to test for a wide variety of materials
(especially when using one of the refractive optical systems). Its major
limitation is that it is not nearly as sensitive to small differences between
molecules as the electrophoretic techniques. This is true even when band
sedimentation is used.

4.1. Method
Except for band, or zonal, sedimentation, details describing sample prepa-
ration, optical systems available, and the interpretation of the experimental
results are provided in Rhodes et al. (2009) of this volume and reviewed in
Cole et al. (2008). Consult the manufacturer’s instructions for details on
how to set up a preparative centrifuge to conduct a rate-zonal sedimentation
experiment. For zonal methods in the analytical ultracentrifuge, consult
Eason (1984) or Ralston (1993) for details. The analysis of the data is
analogous to that above describing gel-permeation chromatography.
Analysis using the differential sedimentation coefficient distribution, g(s),
(Cole et al., 2008) is a particularly useful approach to detection of impurity.
Because the method is model independent, the presence of multiple com-
ponents may appear as peaks at different values of s or as broadening of the
peak representing the primary analyte. Again, as is the case with gel filtration
chromatography, self-associating proteins may exhibit sedimentation
behavior with broad peaks in g(s) or multiple peaks, even though they are
pure. Analysis using equilibrium ultracentrifugation or another method can
help to resolve the problem.

5. Mass Spectrometry Methods

Because mass spectrometry provides a direct measure of the covalent
mass distribution in a sample, it is a straightforward and sensitive approach to
test for impurities. Not only is the presence of the impurity detected, but the
impurity is characterized by its mass and therefore, it is often possible to
688 David G. Rhodes and Thomas M. Laue

identify the origin of the impurity. In addition to the direct measurement of

the protein molecular mass, it is also possible to determine covalent
modification identities and sites using the tandem mass spectrometric
(MS/MS or MS2) methods of Collision Induced Dissociation (CID), Elec-
tron Transfer Dissociation (ETD), Electron Capture Dissociation (ECD),
and Infrared Multiphoton Dissociation (IRMPD). CID and IRMPD can be
used on relatively small proteins (<15 kDa) but ETD and ECD can be used
with larger proteins ( 50 kDa). Beyond that, covalent modifications can be
found and site determined by any commonly used proteolytic method such
as trypsin or CNBr (Link, 2009). CID can be performed on instruments
using a quadrupole collision cell before a Time-Of-Flight mass analyzer
(Q-TOF configuration). The ETD, ECD, and IRMPD are usually found
on ion-trap analyzers such as linear ion traps, Orbitraps, and Fourier
Transform Ion Cyclotron Resonance mass spectrometers.
The MS methods, like the light scattering methods, are often most
powerful when used in combination with another method like gel
exclusion chromatography. It is possible that an impurity could have the
same m/z ratio as the primary analyte, so coupling the MS method to an
orthogonal fractionation provides an additional level of confidence in the
purity of the sample. If an asymmetric peak is observed in the elution
profile from a gel exclusion column, mass detection would provide a
means for distinguishing molecules of different sizes from molecules
undergoing self-association. Because MS provides a measure of the mass
of the protein, as opposed to light scattering measurements, which provide a
mass based on a measurement of the radius of gyration, rg, there is less
ambiguity in the result.

6. Light Scattering Methods

As described in Rhodes et al. (2009), current instrumentation is
capable of static and dynamic light scattering measurements of individual
samples or to monitor HPLC or FFF separations. This capability of contin-
uously monitoring size or apparent molecular weight provides enhanced
capability to identify the eluting proteins as being the analyte, a form of the
analyte such as a multimer, or an impurity. Light scattering methods are
simple and nondestructive, and the current instrumentation provides results
as a plot of molecular weight as a function of elution time. Thus, MALS
(multiple angle light scattering, also known as MALLS, multiple angle laser
light scattering) analysis of a peak, which appears to be asymmetric by UV
detection may exhibit apparent molecular weight of M in the main portion
of the peak and 2M at the leading edge, suggesting the presence of a dimer.
It is important to note that the presence of multiples of a molecular weight is
Determination of Protein Purity 689

not proof of self-association; checking these results with an orthogonal

method is advised. Although the results from MALS are not as accurate as
those from MS, the equipment is far less expensive and easier to use.

REFERENCES
Biggs, R. (1976). Tests for fibrinolysis, thrombin clotting time test. In ‘‘Human blood
coagulation, hemostasis and thrombosis’’ (R. Biggs, ed.), 2nd edn. p. 722. Blackwell,
Oxford.
Chang, J. Y. (1983). Amino acid analysis in the picomole range by precolumn derivatization
and high-performance liquid chromatography. Methods Enzymol. 91, 41–48, this series.
Cole, J. L., Lary, J. W., Moody, T. P., and Laue, T. M. (2008). Analytical ultracentrifuga-
tion: Sedimentation velocity and sedimentation equilibrium. Method Cell Biol. 84,
143–179.
Eason, R. (1984). In ‘‘Centrifugation: A practical approach’’. (D. Richard, ed.), 2nd edn.
p. 251. IRL Press, Washington, DC.
Lunney, J., Chrambach, A., and Rodbard, D. (1971). Pore gradient electrophoresis. Anal.
Biochem. 40, 158–173.
Q6B (1998). ICH Q6B Specifications: Test procedures and acceptance criteria for biotech-
nological/biological products.
Ralston, G. (1993). Introduction to analytical ultracentrifugation Beckman Instruments,
Fullerton, CA.
Rhodes, D. G., Bossio, R. E., and Laue, T. M. (2009). Determination of size, molecular
weight, and presence of subunits. This volume.
 C H A P T E R T H I R T Y- N I N E

Determination of Size, Molecular

Weight, and Presence of Subunits
David G. Rhodes,* Robert E. Bossio,† and Thomas M. Laue‡

Contents
1. Introduction 692
2. Chemical Methods 695
2.1. Composition 695
2.2. Colligative properties 697
3. Transport Methods 698
3.1. Sedimentation equilibrium 698
3.2. Sedimentation velocity 701
3.3. Gel-filtration (size-exclusion) chromatography 706
3.4. Electrophoresis 708
3.5. Viscosity 711
3.6. Field-flow fractionation 711
3.7. Mass spectrometry 712
4. Scattering Methods 716
4.1. Electron microscopy 718
5. Presence of Subunits 719
References 721

Abstract
The size or apparent molecular weight of a given protein may be the most cited
distinguishing characteristic of the molecule. In addition to being the basis of
many separation methods, the molecular weight, or simply molecular size,
immediately provides the investigator with an idea of the complexity of the
molecule, whether it is likely to be difficult to produce in quantity, and whether
certain analytical methods are likely to be productive. Knowing whether the
polypeptide of interest can self assemble or exists in a heterogeneous complex
with other polypeptides may provide valuable information regarding biosynthesis

* Lipophilia Consulting, Storrs, Connecticut, USA

{
Department of Chemistry, University of Michigan–Dearborn, Dearborn, Michigan, USA
{
Department of Biochemistry and Molecular Biology, University of New Hampshire, Durham,
New Hampshire, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63039-1 All rights reserved.

691
692 David G. Rhodes et al.

or mechanism. This chapter outlines key methods used to determine the size of
proteins, their molecular weight, and whether subunits are present, with a focus
on the basis of the determinations, their strengths, and their limitations.

1. Introduction
The size of a protein molecule is the property at the basis of many
fractionation methods and is an easy-to-use descriptor for known molecular
entities or unknown impurities (e.g., ‘‘the 20-kDa product’’). Nevertheless,
in spite of significant advances in methodology and data-handling
techniques, care is required if one requires accurate estimates of molecular
size. In this chapter, ‘‘size’’ will refer to the physical dimensions of the
protein, as opposed to the molecular weight of the protein, which is related
to the mass of the protein. This chapter also considers the asymmetry
(or axial ratio) of proteins, as this property normally affects determinations
of the apparent size of the molecule.
Many proteins assemble into larger aggregates, with each constituent
chain being considered a subunit (Timasheff and Fasman, 1971). The
concept of a subunit, however, must be defined in context by the individual
investigator, taking into account the system under consideration and the
objectives of the work with the system. In this discussion, independent
subunits will be defined as those protein moieties which do not have a
contiguous polypeptide backbone. Thus, disulfide-linked peptide chains as
well as segments associated through noncovalent interactions are considered
subunits, even if those subunits may have arisen from proteolytic action on a
parent polypeptide.
The methods used to determine the size or molecular weight of a
protein can be categorized into three broadly defined areas: chemical analysis,
such as composition analysis or the effect of the molecule on the properties
(e.g., vapor pressure, freezing point, or boiling points) of a solvent; transport
methods, based on the movement of the molecule in response to some
applied force (e.g., electrical, centrifugal, mechanical); and scattering methods,
which are based upon the interaction of incident radiation (e.g., light,
X-rays, neutrons) with the molecule. As these methods are sensitive to
different features of the molecule, selection of appropriate methods depends
on what one needs to know about the system and with what accuracy. The
capabilities, advantages, and limitations of a number of techniques are out-
lined in Table 39.1. In addition to these criteria, the protein and the method
must be compatible with regard to the quantity of protein available, the
attainable level of purity (Rhodes and Laue, 2009), solvent requirements,
and any peculiarities of the protein that may interfere with a specific
technique. The various approaches are useful probes of other molecular
Table 39.1 Comparison of techniques for characterization of proteinsa

Primary Contaminant Ease Sample

Method Size Shape Mr Subunits method sensitivityb Precision Accuracy of use size References
Chemical
Composition


Y L, H, N    mg–mg
Colligative
properties
Vapor pressure


Y L, N   þ mg
Osmotic pressure

þ
Y L, N þ þ þ mg Kupke
(1960)
Transport
Sedimentation
Equilibrium

þ þ Y L, N þ þ þ mg
Velocity þ þ  þ Y þ  þ mg
Gel-filtration þ   þ N   þ mg–mg
chromatography
Electrophoresis
SDS denaturing

þ þ N   þ mg
Native þ    N   þ mg
Viscosity þ þ   N L   þ mg Yang (1961)
Mass spectrometry   þ þ Y þ þ 
Field-flow þ    N    mg
fractionation
(continued)
 Table 39.1 (continued)

Primary Contaminant Ease Sample

Method Size Shape Mr Subunits method sensitivityb Precision Accuracy of use size References
Scattering
X-ray diffraction
Crystallography þ þ þ  Y þ þ  mg Knox (1972)
Low-angle scattering þ þ   Y L, H    mg Glatter and
Kratky
(1982)
Rayleigh light þ þ þ  Y H    mg Timasheff
scattering and
Townend
(1970)
Electron microscopy þ þ
 Y    ng–mg Topf et al.
(2008)
a

, not usable; , not good; , acceptable; þ, good.
b
L, sensitive to low M contaminant; H, sensitive to high M contaminant; N, sensitive to nonprotein contaminant.
Size, Molecular Weight, and Subunits 695

properties as well, and these will be brought out in discussions of individual

techniques. Many techniques are best used in combination with other
techniques (e.g., viscosity and sedimentation, composition and SDS gel
electrophoresis, HPLC and mass spectrometry). This is not a complete
list; many methods not included (such as radiation inactivation and cDNA
analysis) may be useful in specific situations.
We have selected for detailed discussion those techniques which are
most useful and/or most commonly used. Suggested readings for a more
complete treatment of other techniques are listed in Table 39.1.

2. Chemical Methods
2.1. Composition
2.1.1. Overview
Although such methods are used only in unusual circumstances in current
practice, measurements that provide molar quantitation of either amino
acids and/or specific prosthetic groups may be used to estimate the mini-
mum molecular weight of a protein as Mmin ¼ m/n, where Mmin is the
molecular weight estimate (g/mol) of the protein, m is the mass of the
protein used in the analysis, and n is the number of moles of the protein-
related component measured in the analysis. Good quantitation has been
achieved using amino acids analyses, end-group analysis, and quantitative
analysis for specific prosthetic groups (Noble and Bailey, 2009). The quan-
tity of material required for such composition-based molecular weight
analyses is most dependent on the sensitivity of the analytical method
being used. Since many newer analytical techniques are sensitive in the
nanomole to picomole range, only microgram or even smaller quantities of
material may be required.
The error in the minimum molecular weight estimate will depend on the
errors incurred in both the mass estimate and in the analysis for the particular
constituent and, therefore, will depend on the methods chosen for these two
measurements. By combining data from analyses for several different consti-
tuents, the accuracy can be improved. For this reason, independent estimates
based on amino acid analyses, quantitative analysis of the end group, or
quantitation of nonprotein cofactors are recommended.
Because composition analysis provides only a minimum molecular
weight, it is very useful to combine these data with results from other
approaches to obtain the mass of the protein. An accurate composition
analysis can be combined with techniques that provide only low-accuracy
total molecular weights to yield high-accuracy results. Perhaps, the most
often used example of this approach is the combination of amino acid analysis
696 David G. Rhodes et al.

(sequence-based methods as described in Burgess, 2009, are more common

now, but see Bartolomeo and Maisano 2006) with SDS gel electrophoresis
(Garfin, 2009). In principle, one should be able to estimate M by determin-
ing the smallest integers that could account for the full amino acid composi-
tion. In practice, uncertainty in the concentrations of all 20 individual amino
acids makes this approach quite unreliable by itself. However, an approxi-
mate molecular weight from SDS gel electrophoresis can guide in the
selection of the appropriate absolute concentrations from the relative con-
centrations, indicating to the investigator whether to ‘‘round up’’ or ‘‘round
down’’ individual values.

2.1.2. Method
An estimate of protein mass is made most accurately from a measurement of
dry weight when the protein is present in a volatile buffer (e.g., ammonium
bicarbonate). Compensation for nonvolatile buffer components can be
made from the difference in mass between dried samples of the protein
containing solution and the buffer alone, but is difficult to do with great
accuracy and should not be attempted with partially volatile buffers. Alter-
natively, protein concentration measurements are often used in place of dry
weight estimates (the mass value used for analysis simply being the product
of mass concentration and the volume). Concentration estimates can be
made refractometrically, spectrophotometrically, or by chemical assay. The
accuracy of such measurement often is limited, even though the precision
may be reasonable, because standardization is a major problem (Nozaki,
1986). This is especially true for glycoproteins, lipoproteins, or other
proteins with unusual compositions where the method of analysis may be
sensitive only to certain constituents of the protein (e.g., peptide bonds). Of
the methods listed, refractometry (or differential refractometry) is the most
accurate means of estimating concentration. Amino acid analysis also pro-
vides a suitable means of quantitation.

2.1.3. Problems and pitfalls

Composition-based minimum molecular weight analyses are sensitive to
sample purity. Any contaminant that influences either the mass estimation
or the quantitative analysis will affect the accuracy of the determination.
Since no fractionation of the starting material is afforded by these analyses,
contributions from contamination or heterogeneity will be averaged into
the results. Therefore, samples must be purified prior to analysis. The
degree of purity required depends on the nature of the contaminant and
the sensitivity of the mass estimate or the analytical method to that
contaminant. In all of these methods, it must be assumed that there is
only one, or an integral number, of moles of the analyzed component per
mole of protein. Since no information concerning quaternary structure is
available, only a minimum molecular weight is obtained. For example,
Size, Molecular Weight, and Subunits 697

quantitative analysis of heme iron in either hemoglobin or myoglobin

would yield nearly identical apparent molecular weights, despite the fact
that the molecular weight of intact hemoglobin is fourfold greater than
that of myoglobin.

2.2. Colligative properties

2.2.1. Overview
Dissolution of a solute in a solvent reduces the chemical potential of that
solvent and results in a number of observable phenomena known collec-
tively as colligative properties (Tinoco et al., 2002). The relationship at the
basis of these phenomena is
ma  m∘a ¼ RT ln ðga Xa Þ; ð39:1Þ

where ma is the chemical potential of solute ‘‘a’’ (cal/mole), m∘a is ma in the

standard state, R is the gas constant (8.3144
107 cal/mol  K), T is the
temperature (K), and g is the activity coefficient of ‘‘a,’’ and Xa is the mole
fraction of ‘‘a.’’ When Xa is less than 1, ma is less than m∘a . The molecular
weight of any solute can, in principle, be determined by the extent to which
the solution activity is changed by the presence of a known weight of solute.
Thus, the molecular weight of any solute, including proteins, could be
determined by measuring freezing point depression, boiling point elevation,
osmotic pressure, or vapor pressure. Because changes in freezing tempera-
ture or boiling temperature are generally quite small for protein-sized
molecules, and occur at extremes of temperature at which the molecules
may not be stable, these methods are generally not applicable for study of
protein solutions. On the other hand, both vapor pressure and osmotic
pressure have been used for measurement of protein molecular weights, but
this will not be discussed in detail here.
Vapor pressure, freezing point, and membrane osmometers are all avail-
able commercially, but the vapor pressure and freezing point osmometers
are most commonly used in clinical and pharmaceutical applications. Vapor
pressure and freezing point osmometers reliably measure molecular weights
in water to approximately 25,000, whereas membrane osmometers are
used for proteins above 20,000. Protein concentrations on the order of
0.1–1 mg/ml and sample volumes of 10–200 ml are required. As with
composition methods, accurate determinations of sample mass are necessary
for any of these methods. It should also be noted that the molecular
weight measured with any colligative property is a number average
and can be severely affected by contaminating low-molecular-weight
species. Additional information can be found in Tinoco et al. (2002) and
in the manufacturer’s literature.
698 David G. Rhodes et al.

3. Transport Methods
3.1. Sedimentation equilibrium
3.1.1. Overview
Sedimentation equilibrium provides an accurate and powerful method for
the determination of the native molecular weight of a protein (Cole et al.,
2008). It is a simple, nondestructive, relatively rapid method. All parameters
that describe sedimentation equilibrium are either readily measured or easily
estimated. It has several unique capabilities, and can provide quantitative
estimates of molecular weights, stoichiometries, and association constants
for a wide range of chemical systems. For example, by sedimenting in
neutrally buoyant (nonsedimenting) detergents, the native molecular
weight of detergent-solubilized proteins may be measured (Fish, 1978).
Many systems not amenable to analysis by any other technique can be
evaluated successfully by sedimentation equilibrium. However, with the
advent of powerful sedimentation velocity analysis, because sedimentation
equilibrium generally requires sophisticated, expensive equipment, and
because other methods are available which provide data that are adequate
for many purposes, this procedure is used much less frequently than in the
past. Nevertheless, equilibrium sedimentation provides one of the most
powerful methods for the quantitative examination of protein–protein
associations, and is irreplaceable for the study of protein systems with
weak to moderate association constants (Cole et al., 2008).

3.1.2. Method
In an equilibrium sedimentation experiment, the purpose is to produce a
measurable protein concentration gradient along the radial axis of the
centrifuge cell. The length of time needed to achieve sedimentation equi-
librium depends on the length of the solution column, the value of the
reduced molecular weight, s, and the diffusion constant. Because the time
to reach equilibrium depends on the square of the column length, many
cells and techniques have been developed to examine short columns (0.75
or 3 mm long) (Ralston, 1993). Using these cells, equilibrium can be
reached in a matter of minutes to a few hours, depending on the protein.
At each rotor speed, the concentration distribution is measured at intervals
of 30–60 min and invariance of the distribution over time indicates that
equilibrium has been reached. Software (e.g., HeteroAnalysis) helps auto-
mate this process (Cole et al., 2008).
The reduced molecular weight, s, is the measured quantity in a
sedimentation equilibrium experiment (Yphantis, 1964). It is defined as:
Size, Molecular Weight, and Subunits 699

Mð1 
vrÞo2
s¼ ; ð39:2Þ
RT
where
v (ml/g) is the partial specific volume of the protein, r (g/ml) is the
solution density, and o2 (s 2) is the square of the angular velocity (o ¼
rpm  p/30) of the rotor. (Methods for
v and r are described below,
following the description of velocity sedimentation). It can be shown that
dln½cr  s
¼ ; ð39:3Þ
dr 2 2
where cr is the solute concentration at radial position in the rotor, ‘‘r.’’ Thus,
s can be determined as the slope of a graph of the natural logarithm of the
concentration as a function of r2/2. The slope may be calculated at each of
several radial positions (or, correspondingly, at each of several concentra-
tions). Molecular weights determined in this fashion are weight-average
values.
Alternatively, s may be obtained from nonlinear least-squares fitting to
equations that describe the concentration distribution (i.e., c as a function of
r2/2) that are beyond the scope of this chapter ( Johnson et al., 1981). These
equations have been derived from thermodynamic first principles for a
variety of models that include association and nonideality of the proteins.
A typical equilibrium centrifugation experiment will require approxi-
mately 100–200 ml of solution at each of three or four concentrations,
which typically vary from 0.1 mg/ml to approximately 3 mg/ml. These
solutions must be matched with appropriate buffers of the same composi-
tion. The equilibrium centrifugation experiment is normally run at different
speeds, from a low speed where the concentration ratio from the top of the
cell to the bottom of the cell is about 3–5, to a high speed where the
meniscus is depleted of solute.
The concentration distribution at equilibrium can be determined by any
number of means, without an analytical centrifuge. If a preparative centri-
fuge (Attri and Minton, 1986) or an ‘‘airfuge’’ (Bock and Halvorson, 1983)
is being used for the measurement, any assay that is proportional to the
concentration of protein may be used following fractionation of the content
of the cell. Because any method for concentration determination may be
used (e.g., enzyme activity), this sort of analysis may be used to determine
molecular weights in complex mixtures and concentrated solutions
(Howlett et al., 2006). However, it is then necessary to perform several
experiments for different lengths of time to ensure that equilibrium has been
reached. These techniques suffer a loss of precision due to the collapse of the
concentration gradient that occurs as the rotor decelerates and as the cell
contents are fractionated. The analytical ultracentrifuge alleviates these
problems by permitting the solution contents to be examined optically as
700 David G. Rhodes et al.

the centrifuge is operating and the contents of the cell remain at

equilibrium.
Currently, there are three optical systems available for the standard
analytical centrifuge, the UV/visible absorbance scanner, the Rayleigh
interferometer, and a fluorescence detector. The three systems provide
complementary information, with the selectivity and sensitivity of the
fluorescence and absorbance optics being useful for some situations, while
the precision and accuracy of the Rayleigh system are needed in others. The
fluorescence system provides exceptional sensitivity (as low as 100 pM
fluorescein) but requires a fluorescent molecule or a fluorescently labeled
molecule (Kroe and Laue, 2009). The precision of the fluorescence optical
system is similar to that of the absorbance optical system, but the accuracy is
limited because the quantum yield of the fluorophore is sensitive to the
environment of the molecule.
The radius must be calculated with accuracy. The optical systems on the
analytical ultracentrifuge provide this information (Ralston, 1993). When
using a preparative centrifuge, the radius must be calculated from geometric
considerations. The original articles describe how this is done for various
rotors and cells (Attri and Minton, 1986); Bock and Halvorson, 1983).

3.1.3. Problems and pitfalls

The principal problem in sedimentation equilibrium analysis is obtaining a
sufficient quantity of highly purified protein for analysis. Even for highly
purified proteins, it is informative to perform a sedimentation velocity
analysis just prior to equilibrium analysis (Cole et al., 2008). A molecular
weight may be measured using the Rayleigh interferometer optical system
with just 20 ml of solution at a concentration of 1 mg/ml and much lower
concentrations are possible for some fluorescent solutes. Although this is no
more material than is often used for gel electrophoresis, the presence of
minor contaminants can make interpretation of the data more difficult,
especially if the contaminant is unknown or unexpected. In cases where a
preparative centrifuge is being used and a sensitive assay is available, less
material may be required, but lower precision can be expected. It should
also be realized that equilibrium sedimentation does not afford the extent of
fractionation that is seen with sedimentation velocity, gel filtration, or gel
electrophoresis. Even so, the gravitational field does fractionate the solution
to some extent, such that very large particles (e.g., dust, flocculant, aggre-
gate) are removed from observation. This means that although proper
technique is required, the scrupulous cleanliness of samples necessary for
scattering techniques is not needed for analytical centrifugation. Note that
buffer components will also be fractionated in the gravitational field. For
this reason, buffers that have a
vr ¼ 1) are
v that is near that of the solvent (

preferred. Thus, buffers such as Tris are good choices for sedimentation,
while phosphate buffers can pose problems under some circumstances
Size, Molecular Weight, and Subunits 701

(Yphantis, 1964). Problems of buffer sedimentation are minimized by

extensive dialysis and careful filling of the cell.
Tests are available for detecting impurities in a sample being analyzed by
sedimentation equilibrium. One is to generate graphs of the apparent
molecular weight (determined as the local slope from the graph of lnc vs.
r2/2) as a function of concentration. If the M(c) graphs are independent of
rotor speed, one can be quite certain that the sample is pure. More sensitive
tests for heterogeneity are outlined by Rhodes and Laue (2009).

3.2. Sedimentation velocity

3.2.1. Overview
Sedimentation velocity is a simple, nondestructive technique for the char-
acterization of the hydrodynamic behavior of a protein. It has the advantage
over gel chromatography of being a primary technique (not requiring
standards) for the determination of hydrodynamic parameters. Because it
is based on first principles, sedimentation analysis can be applied to systems
that cannot be analyzed by any other means. The principal result from a
sedimentation velocity experiment is the sedimentation coefficient (s),
which is a measure of the ratio of the buoyant mass of the protein to its
frictional coefficient:
Mð1 
vrÞ
s¼ ; ð39:4Þ
N0 f
Where s is the sedimentation coefficient in Svedberg units
(
10 13 s 1), N0 is Avogadro’s number, and f is the protein frictional
coefficient. The diffusion coefficient, D (cm2/s), defined in Eq. (39.5) also is
measurable from boundary spreading in measurements of velocity centrifu-
gation experiments:
RT
D¼ : ð39:5Þ
N0 f
The ratio of s/D [Eq. (39.6)] provides a transport-based measure of the
protein molecular weight that is free of any requirements concerning size,
shape, or similarity to protein standards:
s Mð1 
vrÞ
¼ : ð39:6Þ
D RT
Alternatively, if an accurate molecular weight is known, measurement of
the sedimentation coefficient allows the frictional coefficient, f, to be
determined. First, f  , the frictional coefficient expected for an anhydrous
sphere of molecular weight and density equal to those of the protein, is
calculated by writing the Stokes equation:
702 David G. Rhodes et al.

6p
f  ¼ 6prsphere ¼ ; ð39:7Þ
v3Þ=ð4pN0 Þ1=3
½ðM

where  is the viscosity of the solution and rsphere is the radius of the
equivalent sphere. Using Eqs. (39.5)–(39.7), the ratio of f/f  may be
calculated and used to estimate the shape of the protein. This ratio, f/f  , is
compared to standard and theoretical values to estimate the asymmetry and
overall shape of the protein.
With the advent of powerful computer programs to analyze the data,
sedimentation velocity has become the more widely used than equilibrium
sedimentation (Cole et al., 2008). It is recommended that velocity analysis
be performed on any sample prior to sedimentation equilibrium analysis.

3.2.2. Method
Sedimentation velocity experiments are best performed in an analytical
ultracentrifuge, although methods of reasonable accuracy have been devised
for preparative centrifuges. The advantage of the analytical machine is that
its accuracy (1–3%) is nearly an order of magnitude better than can be
expected from techniques using preparative instruments. Moreover, data
are acquired throughout the experiment so that any unusual behavior can be
observed and better understood. The advantage of the preparative machines
is that sensitive or specific assays, such as ELISA, may be used so that in cases
where the purity of the sample is in doubt, a specific assay can identify the
component of interest. Detailed methods for using the analytical ultracen-
trifuge (Ralston, 1993) and for using preparative centrifuges (Freifelder,
1973; Martin and Ames, 1961) are available.
As noted before, the principal measurement in any sedimentation exper-
iment is the concentration as a function of radial position. The sedimenta-
tion coefficient may be determined from the slope of a graph of the natural
logarithm of the distance that the molecules have sedimented as a function
of time (d ln r/dt):
ðd ln r=dtÞ
s¼ ; ð39:8Þ
o2
where r is the distance of the boundary of the molecules from the center of
the rotor and t is the time from the start of the experiment, usually taken to
be the time at which the rotor reaches two-thirds of the final speed. For
experiments in which the sample being examined is a thin zone of mole-
cules, r is taken to be the point of maximum concentration. For broad zone
or boundary experiments, r is taken to be the midpoint between zero
concentration and the plateau concentration. For proper interpretation
and comparison, the sedimentation coefficient must be corrected to
Size, Molecular Weight, and Subunits 703

standard conditions of water at 20  C and zero protein concentration (s∘20;w )

using
! ! !
1 

v 20;w r  ;w  ;b
s 20;w ¼ sobserved
20;w T T
; ð39:9Þ
1
v20;w rT;b 20;w T ;w
where v, r, and  are the partial specific volume, density, and viscosity,
respectively, at the experimental temperature (T) or 20  C, and in buffer (b)
or water (w). Tabulated values for the viscosity of water and measured
values for the viscosity of the buffer at T and 20  C are used to correct for
the effect of temperature on the sedimentation coefficient. The density is
either measured or calculated from tabulated values.
The diffusion coefficient is determined from the spreading of the bound-
ary as it progresses down the analytical cell. A graph may be made of
[c/(dc/dr)max]2 as a function of time, where (dc/dr)max is the concentration
gradient at the point of maximum gradient, for most cases the midpoint of
the boundary (Ralston, 1993). The slope of this line is used to estimate D,
the diffusion coefficient:
  !
1 d½c=ðdc=drÞmax 2
D¼ : ð39:10Þ
4p dt
Although s and D may be determined, as described here, this approach is
provided primarily to illustrate the relationships between these parameters,
since the advent of powerful analysis programs makes graphing data unnec-
essary. Instead, sedimentation velocity data are fit directly to solutions of the
Lamm equation (Cole et al., 2008). These fitting programs can provide good
quantitative analyses of molecular weights and association constants for a
variety of models and using a variety of analysis protocols.
The ratio of s/D (Eq. 39.6), or the estimates of M from data analysis
programs, are the buoyant, or reduced molecular weight, Mð1 
vrÞ=RT.
Determination of M requires other experimental measurements. The sol-
vent density, r, is readily measured or calculable from tabulated data. The
partial specific volume of the protein, v
, can be measured, but more
frequently is calculated from the protein composition using the method of
Cohn and Edsall and tabulated values of
v for the amino acids (McMeekin
and Marshall, 1952). Values of v
for carbohydrates may be found in Gibbons
(1972) and Durschlag (1986). Values for other common protein-associated
components can be found in Steele et al. (1978) or Reynolds and McCaslin
(1985). For more accurate work
v is adjusted for temperature (over the
range 4–40  C) using a coefficient of 4.3
10 4 cm3/g  K, as described by
Durschlag (1986). The effects of pH, buffer composition, and preferential
hydration on
vr are typically neglected, except in cases where a high con-
centration of denaturant is used. Special provision may be made for
704 David G. Rhodes et al.

estimating the isopotential apparent partial specific volume (P) of simple

proteins in solvents containing 8 M urea and 6 M guanidinium chloride, using
the method given by Prakash and Timasheff (1985). Likewise, P may be
estimated for simple proteins in solvents containing varying amounts of
NaCl, Na2SO4, MgSO4, glycine, b-alanine, a-alanine, betaine, and
CH3COO  Na using the method of Arakawa and Timasheff (1985). Nei-
ther estimate is very accurate, however, if the protein contains significant
levels (>10%, w/w) of nonamino acid constituents (e.g., glycoproteins,
lipoproteins) or if the axial ratio is greater than 10.
Hydrodynamic interpretation of sedimentation data can be made, once
an accurate molecular weight and
v estimate are available. First, one
calculates f  using Eq. (39.7) above. The ratio of the measured frictional
coefficient to f  ( f/f  ) will be a value greater than one. Note that ƒ must be
calculated using s∘20;w :
Mð1 
vrÞ
f ¼ 
; ð39:11Þ
N0 s 20;w
where the various terms are used as described above, and s∘20;w is determined
as described in Eq. (39.9).
There are two reasons that the frictional coefficient ratio, f/f  , will be
greater than one. First, proteins have an associated layer of water molecules
that move with the protein as it sediments. While these water molecules
freely exchange with those in the bulk solvent, the net result of the layer is
to increase the effective radius of the protein. As can be seen from
Eq. (39.7), any increase in the effective radius will increase f. Second, the
frictional coefficient depends on the surface area of the protein accessible to
the solvent. For a molecule of given mass and density, a sphere would be the
shape that exposes the minimum surface area, so any molecular asymmetry
will increase f. The degree of molecular asymmetry is estimated by deter-
mining the axial ratio (a/b) of the ellipsoids of revolution, prolate (elon-
gated) or oblate (flattened), that would result in an identical value for f/f  . It
should be clear that the asymmetry is being modeled using these ellipsoids of
revolution as a mathematical construct, and that the calculated result may
not resemble the molecule. However, if one can measure the asymmetry for
the isolated subunits of a multimeric protein, as well as for the intact
oligomer, it is possible to use changes in a/b to distinguish between possible
arrangements of the subunits (Cantor and Schimmel, 1980).

3.2.3. Problems and pitfalls

Sedimentation velocity provides the best and the only primary method for
the determination of hydrodynamic parameters available to molecular biol-
ogists. Electrophoresis and gel filtration require that molecular weight
standards be used, which places restrictions on the data interpretation.
Size, Molecular Weight, and Subunits 705

Of all the techniques described in this chapter, the problems and pitfalls of
sedimentation analysis are the best documented and most easily overcome.
The most demanding aspect of sedimentation is the availability of
sufficient material for analysis. If the standard optical systems on the current
(Beckman XL-A) analytical ultracentrifuge are used, 0.5 ml of solution with
protein concentrations in the range of 0.1–1 mg/ml is needed to obtain
good data.
For proteins that bind significant levels of buffer components
(e.g., detergent-solubilized proteins), the interpretation of sedimentation
coefficients is made difficult by the fact that the bound components will
contribute to the measured s in four terms: M, v
, r, and f. Of these terms,
M,
v, and f are usually the most affected by bound components, and unlike
sedimentation equilibrium, there is no way to ‘‘blank out’’ the contribution
of such components. Thus, the measured sedimentation coefficient is for
the complex of the protein with the bound component, making it difficult
to extract any useful information concerning the protein alone.

3.2.4. Variations
The effect of protein concentration on s is assessed by determining s at each
of several protein concentrations. Typically, there is a slight linear decrease
in s as the protein concentration is increased.
Another widely used method of estimating molecular weights of pro-
teins is the sucrose gradient sedimentation method introduced by Martin
and Ames (1961). In this method, one creates a linear gradient of sucrose in
buffer in a swinging bucket centrifuge tube. The sucrose concentration
typically ranges from approximately 5% at the top of the tube to 20% at the
bottom; the actual range is less important than the linearity and
reproducibility of the gradient. The unknown sample is layered onto a
gradient and a set of standard proteins of known molecular weight layered
onto an equivalent gradient. The centrifugation proceeds for an appropriate
interval (typically 12–24 h) and the material on the gradients collected as
fractions. The protein concentration in these fractions is determined by
spectrophotometric, enzymatic, or other assays.
The basis of the method is the fact that in a linear sucrose gradient, the
distance traveled by a molecule should be a linear function of the time of
centrifugation at a specified speed. In addition, the distance will depend
linearly on s. The ratio of the distance traveled by an unknown protein to
that of a standard will be equal to the ratio of their sedimentation
coefficients, which will, in turn be approximated by the ratio of the
molecular weights to the 2/3 power. This method yields only approximate
values of s and M, but is simple, requires no specialized equipment, and can
be used to estimate s and M for very small amounts of material if a suitable
assay is available.
706 David G. Rhodes et al.

3.3. Gel-filtration (size-exclusion) chromatography

3.3.1. Overview
Gel-filtration chromatography is one of the most powerful and simplest
methods for the estimates of the molecular weight of proteins. Because of
the fractionation afforded by the method, and because assays specific for the
protein of interest may be used (e.g., enzymatic, immunological), sample
purity does not have to be very high. The method is nondestructive, can be
fairly rapid, and has moderate accuracy as long as the protein of interest is
roughly the same shape as the protein standards used to calibrate the column
(Ackers, 1970).
The determination of a molecular weight by gel chromatography relies
on the comparison of the elution volume of the unknown with those of
several protein standards for which molecular weights are known. The
molecular weight of the unknown is estimated from a graph of the loga-
rithm of the molecular weight as a function of elution volume (or Kav, as
described below) made using the data from the protein standards. (It is
worth noting that the actual dependence is on the logarithm of the effective
hydrated radius, or ‘‘Stokes radius’’ of the protein, and that the fit of
standard proteins to this variable is better.) The elution volume (Ve) for
the standards should cover the range from V0 (the void volume  the
volume outside of the stationary phase) to Vi (where Vi is the included
volume  the accessible volume within the pores of the stationary phase).
Although this discussion refers to elution volumes, most reports of SEC data
refer to elution times, which will be proportional to elution volume for a
given pumping rate. A column-independent measure of the protein
behavior, Kav is more useful for comparison of results than simply the
elution volume:
Ve  V0
Kav ¼ ; ð39:12Þ
Vt  V0
where Kav is the fraction of the stationary gel volume which is accessible to
the protein, and Vt is the total volume of the gel bed. Use of Kav is preferred
over Ve since, for a given gel type, values of Kav will vary only slightly from
column to column. The methods described below can be used for both
native and denatured proteins and, therefore, provide a means for establish-
ing the presence of subunits. However, gel filtration in denaturing solvents
typically requires more material than denaturing gel electrophoresis and is
thus not used as often.

3.3.2. Method
The principle of operation and selection of gel media and gel porosity is
described in detail elsewhere (Stellwagen, 2009). SEC stationary phase
materials are available which will fractionate a very wide range of molecular
Size, Molecular Weight, and Subunits 707

sizes; these may not provide adequate resolution for specific needs. In some
cases, a stationary phase with a narrower range may provide superior
resolution, but one must choose a gel in which the protein to be examined
is partially included. Normally, choice of the stationary phase is arbitrary, as
long as the protein does not bind to the matrix. When there is a choice of
bead sizes for a given porosity, the smallest bead size should be used, as this
improves the column resolution. Check the manufacturer’s recommenda-
tions for any limitations on solvents, but in general, nearly any free-flowing
aqueous buffer system may be used. It is recommended that buffers of
moderate ionic strength be used so that electrostatic interactions between
the protein and immobile charges on the matrix are minimized. For best
results and highest resolution, a long, narrow column should be used.
Preparation of the column and equilibration of the column by washing
with buffer should be done according to the manufacturer’s specifications.
Likewise, flow rates should be chosen in accordance with the manufac-
turer’s specifications. In general, lower flow rates afford better resolution
because the solute can fully equilibrate with the gel matrix at all times, but
excessive diffusion can limit resolution if a column is run too slowly. All
samples should be in the same buffer as that used to equilibrate the column.
Sample volumes applied to the column should generally be less than 2% of
the column’s bed volume. In addition, the flow rate of the column should
be kept constant throughout all of the analyses, since flow rate dependence
of the elution volume can be expected (Ackers, 1970).
Elution volume (Ve) is the volume eluted from the column, starting
once one-half of the sample has penetrated the top of the gel bed and
continuing until the maximum (peak) of the protein of interest has eluted.
The void volume (V0) of the column usually can be measured using
commercially available, size-graded Blue Dextran (M ¼ 2,000,000), and
monitoring the effluent spectrophotometrically at 540 or 280 nm. The
included volume can be measured using as a sample some buffer of different
pH or conductivity (avoiding extremes that could affect the column) or one
that contains a small dye (e.g., bromphenol blue). The included volume is
the difference between the elution volume for the small molecule and the
void volume measured with a large, excluded molecule. Care should be
exercised in the choice of a dye, as many aromatic compounds have an
affinity for the stationary phase which results in anomalously high Kav and
Vi values.
Protein standards should be used that span the full range of sizes that can
be analyzed by the stationary phase chosen to assure that the standards
bracket the analyte protein. For best results, a minimum of four different
standards should be used. Kits containing prestained proteins are commer-
cially available. Any convenient assay for detecting the presence of the
protein being analyzed may be used. If, for any reason, the column must
be repacked, the calibration must be performed again.
708 David G. Rhodes et al.

Use of column chromatography to estimate the molecular weight of

denatured proteins poses special problems (Fish, 1978). This method
assumes that the shape of the unknown protein is identical to that of the
standards. This means that the protein must be totally denatured, including
reduction of disulfide bonds. Buffers should contain a reducing agent or
sulfhydryl groups should be alkylated to prevent reformation of any disulfide
bonds. Both the standards and the unknown must be analyzed after the same
treatment and the same buffer conditions. Calibration proceeds as described
above. This is not a commonly used method.

3.3.3. Problems and pitfalls

The greatest source of error in the use of gel-exclusion chromatography for
size determination comes from the requirement that the unknown be
similar in shape and density to the protein standards. Since the protein
standards used are almost universally compact, globular proteins, this
means that fibrous proteins, or protein having fibrous regions, can behave
anomalously on gel columns. One indication of such molecular asymmetry
is if Kav for the unknown increases when analyzed at decreased flow rates,
while those for the standards remain unchanged.
Since it is the size of the protein and not the molecular weight that is
being assessed by this technique, molecular weight estimates for proteins
that are complexed with other molecules (e.g., detergent-solubilized pro-
teins, extensively glycosylated proteins) will be unreliable. Finally, if the
protein interacts with the stationary phase through either affinity (protein
will appear smaller) or repulsion (protein will appear larger), inaccuracies
will result. One can test for column interaction by determining the molec-
ular weight using two chemically different types of stationary phase (e.g.,
Sepharose and acrylamide) or by changing the buffer conditions. There are
extremes of interaction that will be obvious from unacceptable values of
Kav. If Kav > 1, then the protein is binding to the column and a different
stationary phase should be used. Conversely, if Kav < 0, the column is
‘‘channeling’’ and must be repoured and recalibrated or replaced.
While the estimation of molecular weights using gel filtration often is
inaccurate, the combination of size-exclusion chromatography coupled
with light scattering is extremely powerful and widely used (below).
In this case, gel filtration is used to fractionate the sample by size
(Folta-Stogniew, 2000).

3.4. Electrophoresis
3.4.1. Overview
The most widely used method of evaluating the size of a protein molecule is
electrophoresis. The method is simple, inexpensive, rapid, and for a very
wide range of proteins, is sufficiently accurate for many needs. For these
Size, Molecular Weight, and Subunits 709

reasons, it is the method of choice for most protein systems, and almost
always included in characterization studies. SDS gel electrophoresis is the
most widely used method for determining apparent molecular weights of
denatured proteins (Garfin, 2009), but electrophoretic methods for obtain-
ing size, shape, and molecular weight information are not limited to this
approach. Despite its popularity, it is not necessary to include SDS in the gel
formation; native gel electrophoresis of protein samples may be carried out
under almost any buffer condition required. In addition, the sensitivity of
current staining procedures allows these approaches to be applied to very
small amounts of protein (Garfin, 2009).

3.4.2. Method
The basic procedures are identical to those described earlier (Rhodes and
Laue, 2009; Garfin, 2009) for electrophoresis of denatured proteins, except
that the buffer composition is determined by the needs of the investigator.
There are some restrictions; as with gels under denaturing conditions, the
buffer in which the gel is cast cannot contain reducing agents. In addition,
because the conditions to which the protein is exposed may affect charge,
association, or shape, the composition of the buffer in the gel must be
controlled carefully. One must, therefore, be especially cautious about
relative proportions of catalyst, so that excess oxidant is not left in the gel.
Precast gels are inexpensive and generally have good quality control, but the
probability is low that the buffer in the precast gel will be precisely the
condition that is needed. It is often easiest to prerun gels to remove any
possible by-products of polymerization. Alternatively, if the geometry of
the gel apparatus allows the slab to be exposed (e.g., a horizontal slab or a
vertical slab in which one of two glass plates can be removed and later
replaced), the gel can be dialyzed against the buffer of choice prior to
running. Because slabs are generally quite thin, dialysis for several hours is
generally sufficient. This dialysis procedure can also be used to introduce
reducing reagents postpolymerization, or to use a set of gels cast together
(and thus, presumably, uniform in porosity) with a variety of buffer
conditions.
The basis of electrophoretic protein size analysis is based on a simple
principle: that a charged particle in an electric field is forced through the
surrounding medium by a force proportional to the charge on the particle
and the strength of the field, and is subject to a frictional force proportional
to the velocity, the radius of the particle, and the viscosity of the medium.
As with SDS gel electrophoresis, the investigator may control the frictional
coefficient by controlling the porosity of the gel matrix. In addition, under
nondenaturing conditions, the mobility can be significantly affected by
alterations of the intrinsic charge on the protein due to changes of pH at
which the electrophoresis is carried out. This distinction is important. In
SDS gel electrophoresis, the charge is dominated by the negatively charged
710 David G. Rhodes et al.

SDS associated with the protein so the sample is applied to the cathode end
of the gel and the sample always moves toward the anode. In nondenaturing
electrophoresis, the direction of migration will depend on the buffer pH
relative to the isoelectric point (pI) of the protein. If the pI of the protein is
unknown (and quantities are too limited to allow experimental determina-
tion with isoelectric focusing), a horizontal electrophoresis apparatus may
be used, and the starting wells placed in the center of the gel. Otherwise, a
‘‘best guess’’ might be made by using the pI of related proteins.
The mobility will also be affected by the chosen buffer condition.
Beyond association of the protein or conformational changes associated
with changing buffer conditions, the relative mobilities of various proteins
at a given pH will usually be fairly constant. The absolute mobility, how-
ever, will be strongly affected by the concentrations of counterions.
The analysis of native gel electrophoresis mobility data is analogous to
that of SDS gels. The only significant difference is that, in the native system,
one cannot assume that the proteins have equal charge density and are all in
the shape of long rods. As is necessary in gel-exclusion chromatography,
one must compare the mobility of the sample to that of a set of standards
(Rodbard and Chrambach, 1971). Ferguson analysis can also be used to
identify the feature(s) (size and/or charge) that distinguish two components
and to extrapolate to the mobility expected in the absence of sieving effects
(Andrews, 1986). The slope of the Ferguson plot (log of the relative
mobility, Rf, vs. gel concentration) is proportional to rs, the Stokes radius.
Interpolative estimation of an unknown rs is more reliable using this relation
than using, for example, the slope of the Ferguson plot with molecular
weights of the standards. The analysis simply involves running the unknown
and several standards in a set (at least five) of gels of different total gel
concentrations, and plotting log Rf versus the gel concentration for all
standards and for the unknown. The slopes for the standards are then plotted
as a function of the (known) rs, and the rs of the unknown is derived
from this plot and the measured slope. The range of gel concentrations
used will depend on the size of the protein under study, but at some upper
limit of concentration, the protein will be excluded from the gel. A more
closely spaced group of gel concentrations covering a lower range should
then be used.

3.4.3. Problems and pitfalls

Although electrophoresis under nondenaturing conditions can provide
useful information about the physical characteristics of the protein under a
number of different conditions, it is important to be aware that this is a zonal
method, and that concentration effects (or dilution effects) can be signifi-
cant. It is not generally advisable, for example, to use this approach to study
association behavior quantitatively.
Size, Molecular Weight, and Subunits 711

As with gel-exclusion chromatography, this is not a primary technique;

it depends on selection of appropriate standards. These are generally globu-
lar, unmodified, soluble proteins, so that highly asymmetric proteins or
proteins with unusual nonprotein modifications may yield erroneous
results. The Ferguson analysis helps to account for differences in mobility
due to charge differences, but exaggerated charge densities may yield
anomalous results.
Because the buffer conditions in native gel electrophoresis are selected
by the individual investigator, and thus may vary widely, one must be aware
of the solution components responsible for carrying current. High ionic
strength buffers may result in unacceptably slow protein mobilities, and
proteins in low ionic strength buffers may run properly at surprisingly low
currents. Because of this variability and because protein integrity is more
important, one must be particularly aware of power dissipation and cautious
about efficient removal of heat generated in the gel.

3.5. Viscosity
Among the variables upon which the viscosity of a solution depends (T, P,
etc.) are the amount and nature of any solute that is present. The response of
a particle in a fluid under shear will depend on the frictional coefficient of
the particle, which represents the mechanical interaction of the particle with
the moving solution (Yang, 1961). The response of the particle will also
depend on the mass of the particle, which determines the energy required to
attain a given movement ( Johnson et al., 1977). Because the intrinsic
viscosity of a protein depends very strongly on the asymmetry of the
molecule, viscosity measurements are sensitive indicators of protein shape.
In addition, it is possible to combine viscosity data with sedimentation data
to calculate the molecular weight of a protein. Apparatus for rheological
measurements vary widely; their use is generally quite simple, but must be
performed under rigorously controlled conditions.

3.6. Field-flow fractionation

Field-flow fractionation (FFF) is a method based on subjecting particles to a
force field orthogonal to the direction of laminar flow in a narrow channel
(100–500 mm) (Liu et al., 2006). Because the laminar flow in the channel is
parabolic, with the fastest flow in the center and slowest at the edge,
molecules that are most strongly affected by the force imposed on them
will be closer to the edge of the channel and will move more slowly, while
molecules that do not move from the center will move faster. The force
imposed on the molecules could be centrifugal, electrophoretic, magnetic,
or even mechanical flow. The cross-flow approach is the most commonly
used method, and is referred to as flow-FFF (fFFF). In this method, a
712 David G. Rhodes et al.

semipermeable membrane allows solvent, but not protein, to flow out the
bottom of the channel, thus creating a concentration gradient with higher
concentration at the bottom of the channel than at the center or top.
Diffusion works against this concentration gradient and allows the more
rapidly diffusing small molecules to reach the center of the channel where
flow is faster. The larger molecules cannot diffuse as rapidly and move
slowly at the edge of the channel.
To its benefit, this method is nondestructive and can be used with any
buffer system. By appropriate adjustment of the flow rates, this method is
capable of resolving a very wide range of molecular sizes. Current instru-
mentation utilizes many HPLC components and highly automated systems
make the process accessible; a typical experiment runs in approximately
30 min. Nevertheless, the method is technically challenging and will require
time and effort for method development. Finally, the method is very good
for resolving molecules of different sizes (e.g., monomer, dimer, etc.), but
the resulting size distribution is relative and not easily converted to an
absolute size. Consequently, FFF often is used as a fractionation method
for light scattering to provide absolute molecular weights (Liu et al., 2006).

3.7. Mass spectrometry

Mass spectrometry measures an intrinsic property of the protein, its mass, by
measuring the mass-to-charge ratio (m/z) of ions generated from the
protein sample. This property is temperature, density, and concentration
independent. Advances in electrospray ionization mass spectrometry (ESI-
MS) have added significant capabilities to the biochemist’s toolbox (Faull
et al., 2008). ESI-MS and other MS methods can be combined with liquid
chromatographic (LC) methods (i.e., capillary electrophoresis, size-exclu-
sion chromatography, normal and reverse phase LC) to aid in identifying
chromatographic peaks and to provide additional information with rela-
tively little sample (1–5 mg). The protein molecular mass obtained by MS
analysis can be used not only for protein identification; but also for deter-
mining purity (Baldwin et al., 2001) and for analysis of covalent modifica-
tions and degradation products, both qualitatively and quantitatively. The
ESI time-of-flight (TOF) methods are sufficiently gentle that there is little if
any degradation of the protein during analysis. Typical data are obtained as a
family of peaks as a function of m/z that represent individual ionization
states of the protein. For instance, a 30-kDa protein can acquire upwards of
40 Hþ ions, and is observed as a peak in the mass spectrum at (30,000m/
40z ¼ 750 m/z) and is labeled as [M þ 40H]40þ. The charge states form a
Gaussian-like distribution with many charge states ([M þ 39H]39þ,
[M þ 38H]38þ, [M þ 37H]37þ, etc. Deconvolution software (such as
that publicly available at http://ncrr.pnl.gov/software/) is used to determine
the molecular weight of the parent molecule, and minor peaks reveal the
Size, Molecular Weight, and Subunits 713

presence of contaminants or degradation products. The protein molecular

mass obtained is highly accurate (0.01%) and can be used for modified
(e.g., glycosylated, alkylated, or PEGylated) proteins. It is important to note,
however, that any heterogeneity in the modification, such as variable
glycosylation or variable lengths of PEG, will be reflected in the mass
spectrum. This is useful in assessing purity (Baldwin et al., 2001) but can
complicate the calculation of the molecular mass.
This method has a significant advantage over many other methods in
that the sample need not be highly purified. While purity will simplify the
analysis, the method is capable of identifying multiple components in a
single experiment. On the other hand, if the difficulty in purification arises
from affinity of the contaminant for the primary analyte, an analyte-con-
taminant peak may also be observed, since noncovalent interactions can be
preserved in ESI-MS. By diluting the sample, it may be possible to dissociate
these noncovalent adducts and the contaminant can be observed separately.
Selection of a suitable buffer requires care, since the buffer must be
volatile; traditional buffers like Tris and phosphate are not suitable. Typical
buffers include formic acid, trifluoroacetic acid (TFA), acetic acid, ammo-
nium formate, ammonium acetate, ammonium bicarbonate, and hepta-
fluorobutyric acid. Not all of these buffers are suitable for all situations.
TFA has been known to suppress ion formation, so only low concentrations
(0.1% or less) are acceptable. These are not common buffer systems and
changing buffers can affect the protein, so it is important to execute proper
controls to assure that results are being interpreted correctly.
Because the analyte protein is delivered to the ESI-MS instrument in a
buffer, it is relatively straightforward to combine a chromatographic separa-
tion with the mass measurement. The complementarity of a mass measure-
ment and a size measurement that one would obtain with SEC
chromatography is particularly valuable in analyzing associating proteins.
Although the MS can be used as a primary detector, it is most common to
place the MS in combination with a conventional detector such as a UV
absorbance system. With this arrangement, an elution profile from the
SEC–HPLC can be interpreted by mass-based identification of peaks. For
example, a SEC peak skewed to the leading edge may result from an
impurity or from association to multimer. An impurity could present as a
distinct mass peak, whereas an association might result in peaks with mass
values at integral multiples of the mass of the parent peak.

3.7.1. Method
The specific ESI-MS techniques will vary with the individual instrumenta-
tion, but some general guidelines are provided here. As mentioned above,
samples must be in a volatile buffer if a direct presentation to the mass
spectrometer is going to be used. Phosphate- and TRIS-related buffers
cause ion suppression and mass shifts due to noncovalent adduct formation
714 David G. Rhodes et al.

between sodium, potassium, counter ions, and the protein. These salts can
be dialyzed against a volatile buffer or removed chromatographically. If the
sample is analyzed by LC/MS, the small buffer constituents will be fractio-
nated from the larger protein and dialysis is not necessary. In cases where a
protein is added to the column in a salt-based buffer, a diverter valve is used
to divert the salt peak to waste before the salts arrive at the ESI source.
Samples available as a lyophilized powder can be dissolved in a mixture of
methanol or acetonitrile and water with formic or acetic acid, usually at a
ratio of 1:1 aqueous:organic, with the acid present at 5% or less (typically
0.1%). Higher organic concentrations can be used without difficulty for
hydrophobic proteins.
In addition to using the volatile buffers suggested above, it is important
to assure that the protein is fully equilibrated to this condition and that any
titration of the buffer is with suitable acid or base (ammonia, acetic, formic,
or trifluoroacetic). Moreover, for some of these buffers, the concentration
limits are relatively low. Flow rates will be determined by the specific
instrument limits, but in some cases, a splitter will be required so that only
a small portion of the column flow enters the MS. Most LC/MS instru-
ments are capable of operating without a splitter at flow rates as high as
1 ml/min. ESI sources are also available with very small ( 25 micrometer
inner diameter) capillaries, called nano-ESI. These ESI sources operate at
much lower flow rates ( 10–100 nl/min) and are frequently coupled to
capillary LC columns for Capillary LC/nano-ESI-MS. Although the sensi-
tivity is improved, these systems can become clogged without scrupulous
handling of solvents and samples to avoid dust.
The speed of a protein mass spectral analysis is very rapid when analyzing
a single sample (<1 min), and can range up to 1 h for an LC/MS analysis.
New ultrahigh-pressure liquid chromatography analysis (UPLC) is showing
promise to reduce these analysis times and improve the chromatographic
resolution.
Four recently developed ionization modes have allowed for alternate
modes of ionization for large proteins. Desorption electrospray ionization
(DESI), Direct Analysis in Real Time (DART), Laser Ablation Electrospray
Ionization (LAESI) and matrix assisted laser desorption electrospray ioniza-
tion (MALDESI) have shown the ability to ionize proteins, and may prove
useful in cases where ESI and MALDI (matrix-assisted laser desorption
ionization) are ineffective (Faull et al., 2008). MALDESI combines the
advantages of MALDI, such as the ability to handle high-molecular-weight
proteins, with the multiple charging capability of ESI. These ionization
methods may be able to overcome some of the limitations encountered in
the individual methods.
On the mass spectrometer side, an alternate platform is finding new life
in protein analysis, ion mobility spectrometry (IMS). IMS separates ions
according to their mobility through an electric field as they collide with a
Size, Molecular Weight, and Subunits 715

gas (usually He). Protein ions with a large shape/collision cross section
travel slower in the electric field than smaller ions with lower collisional
cross section. From the cross section information, aggregation information
can be determined as well as purity. In some aspects, IMS resembles TOF
but reveals information about protein purity through different measures (the
gas phase ion shape as opposed to its mass). IMS can also be coupled to mass
spectrometers, allowing for interesting multidimensional separations
followed mass analysis. While this technique is still in its starting phases, it
definitely bears watching as the method matures and finds its place in
protein purity analysis.

3.7.2. Problems and pitfalls

The most important factor in interpreting MS data is understanding the
origin of the results. Mass shifts of 22, 44, 66, or 88 are due to sodium
adduction, as Naþ ions exchange for Hþ ions, but retain the overall charge
state. Similarly, mass shifts of 39 are signals of potassium adduction. Ammo-
nium adduction (mass shift of 17 units) can also be observed. These adducts
can be minimized by using ultrapure solvents, but are difficult to completely
remove.
Because the method reports a mass-to-charge ratio, in principle, there is
no difference between a monomer and a doubly charged dimer. While
charge deconvolution algorithms cannot resolve this issue, mass spectro-
meters other than TOF’s can. In instances like these, the ultrahigh mass
resolving power of fourier transform ion cyclotron resonance (FTICRMS)
or Orbitrap mass spectrometers can resolve these uncertainties by using the
isotope peaks of the ions to determine the charge state. These spectrometers
unambiguously differentiate singly charged monomers from doubly charged
dimers. Dimer formation can also be minimized by diluting the sample to
the point where dimers are no longer observed. In cases where tightly
bound ligands cannot be separated by dilution, using higher concentrations
of organic solvent, acid, or both may dissociate the noncovalent complexes.
Acid-labile detergents can also be used to initially dissociate noncovalent
complexes, but once the pH is lowered these detergents dissociate into
noninterfering volatile monomers.
Similarly, without additional information, a single peak could result
from a wide range of charge–mass combinations. Fortunately, for proteins,
the multiple charge states that are normally observed can be deconvoluted
reliably to yield unambiguous results. The presence of tightly bound ligands
can be problematic but can usually be dealt with by changing the pH,
increasing the organic content of the solution, or diluting the solution
until the complex dissociates.
For some proteins the electrospray process does not work well enough
to get usable signal. Very large proteins (>150 kDa) can be difficult to ionize
by ESI. ESI is also matrix sensitive; solutions with high salt concentrations
716 David G. Rhodes et al.

can result in reduced signals. In such cases, alternative ionization modes can
be used. MALDI is useful in these instances, as this mode can handle very
large (>1 MDa) proteins and has a higher tolerance for salts. Chro-
matographic separation followed by lyophilization of the individual chro-
matographic peaks allows for LC/MALDI analysis. MALDI also tends to
form ions with a single charge state, so that protein ions are largely observed
as their [M þ 1H]þ peaks. This can simplify spectral interpretation, as
dimer peaks will show up at twice the m/z, avoiding the doubly charged
dimer issue altogether.
Mass spectrometry is limited due to the fact that proteins are generally
not isotopically pure. The presence of 13C, 15N, 34S, and other isotopes
occurring in proteins causes broadening of mass spectral peaks in TOF as the
mass of the protein increases. This broadening can eventually make m/z
envelope of a protein ion so broad that the molecular weight becomes
indeterminant. In this instance, peaks due to adduct formation, ligand
binding, and di- or multimer formation may become indistinct. Although
this issue is generally less significant below about 6–7 MDa, it does provide
an upper bound to the method. This can be circumvented in some studies
by producing the protein using amino acids or other starting blocks that are
isotopically depleted, these have a lower than natural abundance amount of
13C, 15N, 34S or similar isotopes.

It should also be noted that the power of mass spectrometry-based

methods comes at a significant cost. As of this writing, TOF instruments
range in cost from $250,000 to over $500,000 and FTMS instruments can
cost $2 million.

4. Scattering Methods
Light scattering methods are generally used to obtain molecular
weights and radii of gyration, but specific scattering methods can provide
diffusion coefficients, molecular weight, and thermodynamic parameters
(Chu, 1974; Cottingham, 2005; Timasheff and Townend, 1970).
Instrumentation for key light scattering methods has been developed
which integrates with HPLC and FFF instrumentation and incorporates
nearly all of the data analysis functions. This ease of use and accessibility has
resulted in an resurgence in the application of these powerful methodologies
to protein characterization (Harding et al., 1991).
The basis of light scattering is the interaction of the oscillating electro-
magnetic field of the incident light with the electrons in the protein to
induce oscillating dipoles. These oscillating dipoles radiate light, and the
intensity and directionality of the light depends on many factors, including
the size and structure of the molecule.
Size, Molecular Weight, and Subunits 717

The Zimm relationship forms the basis of what is now referred to as

‘‘static’’ light scattering:
Kc2 1
¼ þ 2A2 c2 ; ð39:13Þ
RðyÞ Mw PðyÞ

where R(y) is the scattering angle- and concentration-dependent Rayleigh

ratio, which is proportional to the intensity of light scattered by the solution
in excess of that scattered by pure solvent, c2 is the protein concentration,
Mw is the weight-average molecular weight of the protein and any other
material in the sample, A2 is the second virial coefficient, K is a constant
(4p2(dn/dc2)2n0/N0l4), dn/dc2 is the refractive increment of the solute, N0
is Avogadro’s number, n0 is the refractive index of the solvent, l is the
wavelength of the incident light, and P(y) is a parameter related to the
angular dependence of the scattered light intensity. P(y) can be approxi-
mated by:

16p2 n0 2 2 2 y
PðyÞ ¼ 1 
r sin þ . . . ; ð39:14Þ
3l2 g 2

where
r g is the RMS radius of gyration of the molecule. By plotting Kc2/R(y)

as a function of sin2(y/2) þ Kc2 for several values of y, one can calculate
r g .
This is an approximation obtained at c2, but is acceptable for many applica-
tions and is the data used in many instruments incorporated in HPLC
instrumentation. By making measurements at several values of c2, one can
extrapolate to c2 ¼ 0 and to y ¼ 0, and the intersection of these lines at the
ordinate (Kc2/R(y)) yields Mw.
The basis for ‘‘dynamic’’ light scattering (also known as photon
correlation spectroscopy or quasielastic light scattering) is the time-depen-
dent constructive or destructive interference between scattering centers
(proteins) in motion in the solution. The normalized autocorrelation
function for the intensity variation with time, g(t) can be shown to be
gðtÞ ¼ expðGtÞ; ð39:15Þ
where G is [(4pn0/l) sin(y/2)]2D. The diffusion coefficient can be related to
size and molecular weight by Eqs. (39.5) and (39.7). Data from dynamic
light scattering measurements can be presented as distributions in rg, from
which the presence of multiple scattering species can be determined.
With laser illumination and a set of fixed angle detectors, one can
simultaneously measure dynamic and static light scattering. From the static
light scattering, one can estimate radius of gyration and molecular weight
and from the dynamic light scattering, diffusion coefficient, hydrodynamic
radius, size distribution, and molecular weight. It is important to be aware of
718 David G. Rhodes et al.

assumptions involved in calculations performed by user-friendly systems;

most calculations, for example, assume that all particles are spherical.
As with MS measurements, the power of these light scattering methods
is greatly enhanced when the measurements are coupled with a
chromatographic separation method. The ability to characterize a protein
by orthogonal methods such as SEC or FFF and light scattering provides a
ready comparison through which the validity of any assumptions can be
tested. The power, reproducibility and ease of use has made SEC–multi-
ple-angle light scattering (SEC-MALS), the principle method for
characterizing protein aggregates in the pharmaceutical industry (Wen
et al., 1996).
When light scattering methods are not coupled with chromatographic
separation, which inherently provides very efficient filtration, it is important
to maintain scrupulously conditions to avoid the presence of ‘‘dust’’ parti-
cles in the solution. The presence of very small amounts of large particles
can easily dominate the scattering signal and compromise the results. On the
other hand, this dependence points out the power of this methodology in
detecting large assemblies or aggregates of proteins.

4.1. Electron microscopy

Electron microscopy has always been an appealing approach to determin-
ing molecular size and shape for large protein molecules or associated
complexes of subunits because it is a direct imaging method. Initial work
in this area was used to identify the shape of very large complexes like the
hemocyanin aggregates (van Bruggen and Wiebenga, 1962), but studies
advanced to imaging of crystalline arrays of membrane proteins (Sletyr
et al., 1988) and more recently to cryo-EM analysis of protein structures
(Topf and Sali, 2005, Topf et al., 2008). This method provides very
detailed information about the native size and shape of the molecule and
is particularly valuable for analysis of membrane-associated proteins that
are incompatible with many other analysis methods. The resolution of
electron microscopy is inherently limited, but with cryopreparation and
signal averaging from large numbers of images, high-resolution electron
density maps can be constructed. These maps can then be combined with
structure filling algorithms to obtain detailed structure maps of the protein
molecule. The modeling requires knowledge of the protein sequence, so
this is not a method by which one would experimentally determine the
molecular weight of the protein, but this is an elegant approach to
determining the size and shape of the molecule to very good resolution.
Although the cryo-EM method is not a common laboratory method for
protein size determination, this approach has made great advances in areas
of protein science which previously were very difficult to study.
Size, Molecular Weight, and Subunits 719

5. Presence of Subunits
To determine whether subunits are present generally involves charac-
terization of one or more properties of the system under conditions which
favor association and under conditions which are likely to favor dissociation.
Thus, the investigator should have prior knowledge of the conditions under
which association and dissociation should be expected, or will need to
investigate a variety of conditions which have been shown in other cases
to result in dissociation. The investigation to identify these conditions may
require significant effort, but if properly planned, will also yield additional
information about the nature of the system. Before initiating an extensive
search for associating or dissociating conditions, it is worthwhile using
extreme conditions as a preliminary evaluation. One normally carries out
one fractionation procedure under some (often physiological) buffer
condition and a second under strongly denaturing conditions.
One very simple, rapid, and low-cost approach for searching for subunits
is to carry out electrophoresis under nondenaturing conditions, using the
buffer in which the protein was isolated, followed by a second dimension
under dissociating or denaturing conditions. The nature of the denaturing
conditions may vary depending on the type of subunit and the association
one expects to find. For example, one could run the second electrophoresis
at extremes of pH, in the presence of urea or SDS, or in the presence or
absence of other buffered components like calcium. Changes in mobility
due to changes of buffer condition should be accounted for and understood
by analyzing the second dimension. That is, analysis should be based on the
apparent size of the molecule(s) under associating and possibly dissociating
buffer conditions. If one is investigating the possibility that disulfide-linked
subunits are present, SDS gel electrophoresis in the absence of reducing
agents may be carried out as the first dimension, and SDS gel electrophoresis
in the presence of reducing agents may be carried out in the second
dimension. Because in-gel alkylation of disulfides is difficult, it is recom-
mended that mercaptoacetate be included in the second gel running buffer
to avoid reoxidation. Differences in the apparent molecular weight or the
appearance of multiple components in the second dimension will indicate
that disulfide-linked subunits were present. Differences in apparent molec-
ular weight deduced from nondenaturing electrophoresis compared with
the apparent molecular weight based on electrophoresis under completely
denaturing conditions (SDS–PAGE) could indicate the presence of subu-
nits, but one must consider the possibility that the electrophoretic behavior
of the molecule is due to extremes of asymmetry or intrinsic charge.
In the case of subunits which are in equilibrium between associated and
dissociated states, the definitive approach to determining the presence of
720 David G. Rhodes et al.

subunits requires experimental determination of the apparent size of the

molecule under conditions where the equilibrium is shifted to either
associating or dissociating conditions. Because the dissociation implies
that the samples being studied will be very dilute, very sensitive methods
must be used.
Sedimentation equilibrium is very useful for determining the molecular
weight of a native protein, and is therefore useful in determining the
stoichiometry of the subunits in the final assembly. This is done by compar-
ison of the native molecular weight with that obtained in separate experi-
ments under denaturing conditions, such as by denaturing gel
electrophoresis. If the protein is composed of subunits of a single molecular
weight, division of the native molecular weight by the denatured molecular
weight will provide the subunit stoichiometry. Likewise, for proteins that
contain more than one chain, comparison of the native molecular weight to
the sum of the monomer molecular weights often will allow the stoichio-
metry of the different subunits in the native complex to be determined. In
cases where there is a wide discrepancy between the subunit molecular
weights or when the native structure contains a large number of monomers,
these estimate are imprecise.
While the subject requires too much detail to be presented in full here, it
is important to note that equilibrium sedimentation provides one of the
most powerful means of determining association constants of mass action-
driven macromolecular associations ( Johnson et al., 1981). The method of
experimentation is essentially as outlined above, except that experiments are
done at concentrations that encompass the range where significant mass
changes occur due to association. The association constant may be estimated
using graphical means, or more accurately using nonlinear least-squares
analyses ( Johnson et al., 1981).
Velocity centrifugation can also be a simple screen for multiple compo-
nents in a solution. The integral sedimentation coefficient distribution,
G(s), (Demeler et al. 1997) is based on an extrapolation of the sedimentation
boundaries to eliminate the effect of diffusion during sedimentation. The
method is model independent and is a useful diagnostic for heterogeneous
distributions in a sample, association behavior, and any nonideality. Similar
and related methods have also been developed (Schuck, 2000; Stafford,
1992) from which one may easily compare the apparent sizes of monomers,
multimers, or multiple species of dissociated subunits.
Gel-permeation chromatography can also be useful in determining the
stoichiometry of the subunits in the final assembly. Again, this is done by
comparison of the native molecular weight determined by gel filtration with
that obtained under denaturing conditions, such as by SDS gel electro-
phoresis. One can test for association and, in principle, determine
stoichiometry using the ratio of the native molecular weight and the
denatured molecular weight.
Size, Molecular Weight, and Subunits 721

Gel chromatography is also useful for diagnosing interacting protein

systems. If, for example, the protein of interest is a multisubunit molecule
undergoing rapidly equilibrating assembly–disassembly, the peak shape is
skewed, with the leading edge being hypersharp and the trailing edge being
diffuse. Likewise, increase of Kav with protein loading concentration usually
indicates that such an interaction is occurring, and that more detailed
analyses will be required (Ackers, 1970).

REFERENCES
Ackers, G. K. (1970). Analytical gel chromatography of proteins. Adv. Protein Chem. 24,
343–446.
Andrews, A. T. (1986). Electrophoresis: Theory, Techniques, and Biochemical and Clinical Appli-
cations. Clarendon Press, Oxford.
Arakawa, T., and Timasheff, S. N. (1985). Calculation of the partial specific volume of
proteins in concentrated salt and amino acid solutions. Methods Enzymol. 117, 60.
Attri, A. K., and Minton, A. P. (1986). Technique and apparatus for automated fractionation
of the contents of small centrifuge tubes: Application to analytical ultracentrifugation.
Anal. Biochem. 152, 319–328.
Baldwin, M. A., Medzihradszky, K. F., Lock, C. M., Fisher, B., Settineri, T. A., and
Burlingame, A. L. (2001). Matrix-assisted laser desorption/ionization coupled with
quadrupole/orthogonal acceleration time-of-flight mass spectrometry for protein discovery,
identification, and structural analysis. Anal. Chem. 73, 1707–1720.
Bartolomeo, M., and Maisano, F. (2006). Validation of a reversed-phase HPLC method for
quantitative amino acid analysis. J. Biomol Tech. 17, 131–137.
Bock, P., and Halvorson, H. (1983). Molecular weight of human high-molecular-weight
kininogen light chain by equilibrium sedimentation in an air-driven ultracentrifuge.
Anal. Biochem. 137, 172–179.
Burgess, R. (2009). Use of bioinformatics in planning a protein purification. Methods
Enzymol. 463, 21–28.
Cantor, C. R., and Schimmel, P. R. (1980). In ‘‘Biophysical Chemistry’’, Part II p. 565.
Freeman, San Francisco, CA.
Chu, B. (1974). Laser Light Scattering. Academic Press, New York.
Cole, J. L., Lary, J. W., Moody, T. P., and Laue, T. M. (2008). Analytical ultracentrifuga-
tion: Sedimentation velocity and sedimentation equilibrium. Methods Cell Biol. 84,
143–179.
Cottingham, K. (2005). Versatile TOFMS. Anal. Chem. 77, 227–231.
Demeler, B., Saber, H., and Hansen, J. C. (1997). Identification and interpretation of
complexity in sedimentation velocity boundaries. Biophys. J. 72, 397–407.
Durschlag, J. (1986). In ‘‘Thermodynamic Data for Biochemistry and Biotechnology,’’
(H. Hinz, ed.), p. 46. Springer-Verlag, Berlin.
Faull, K. F., Dooley, A. N., Halgand, F., Shoemaker, L. D., Norris, A. J., Ryan, C. M.,
Laganowsky, A., Johnson, J. V., and Katz, J. E. (2008). An introduction to the basic
principles and concepts of mass spectrometry. Compr. Anal. Chem. 52, 1–46.
Fish, W. W. (1978). In ‘‘Methods in Membrane Biology,’’ (E. E. Korn, ed.), Vol. 4, p. 289.
Plenum Press, New York.
Folta-Stogniew, E. (2000). Oligomeric states of proteins determined by size-exclusion
chromatography coupled with light scattering, absorbance, and refractive index detec-
tors. Methods Mol. Biol. 328, 97–112.
722 David G. Rhodes et al.

Freifelder, D. (1973). Zonal centrifugation. Methods Enzymol. 27, 140–150.

Garfin, D. (2009). One-dimensional gel electrophoresis. Methods Enzymol. 463, 495–514.
Gibbons, R. A. (1972). In ‘‘Glycoproteins,’’ Part A, (A. Gottschalk, ed.), p. 31. Elseveier,
Amsterdam.
Glatter, O., and Kratky, O. (1982). Small Angle X-Ray Scattering. Academic Press,
New York.
Harding, S. E., Sattelle, D. B., and Bloomfield, V. A. (1991). Laser light scattering in
biochemistry. Biochem. Soc. Trans. 19, 477–516.
Howlett, G. J., Minton, A. P., and Rivas, G. (2006). Analytical ultracentrifugation for the
study of protein association and assembly. Curr. Opin. Chem. Biol. 10, 430–436.
Johnson, J. F., Martin, J. R., and Porter, R. S. (1977). In ‘‘Physical Methods of Chemistry,’’
(A. Weissberger and B. W. Rossiter, eds.),Vol. 1, Part VI, p. 63. Wiley, New York.
Johnson, M. L., Correia, J. J., Yphantis, D. A., and Halvorson, H. R. (1981). Analysis of data
from the analytical ultracentrifuge by nonlinear least-squares techniques. Biophys. J. 36,
575–588.
Knox, J. R. (1972). Protein molecular weight by x-ray diffraction. J. Chem. Educ. 49,
476–479.
Kroe, R. R., and Laue, T. M. (2009). NUTS and BOLTS: Applications of fluorescence-
detected sedimentation. Anal. Biochem. 390, 1–13.
Kupke, D. W. (1960). Osmotic pressure. Adv. Protein Chem. 15, 57–130.
Liu, J., Andya, J. D., and Shire, S. J. (2006). A critical review of analytical ultracentrifugation
and field flow fractionation methods for measuring protein aggregation. AAPS J. 8,
E580–E589.
Martin, R., and Ames, B. (1961). Method for determining the sedimentation behavior of
enzymes: Application to protein mixtures. J. Biol. Chem. 236, 1372–1379.
McMeekin, T. L., and Marshall, K. (1952). Specific volumes of proteins and the relationship
to their amino acid contents. Science 116, 142–143.
Noble, J. E., and Bailey, M. J. A. (2009). Quantitation of protein. Methods Enzymol. 463,
73–96.
Nozaki, Y. (1986). Determination of the concentration of protein by dry weight—
A comparison with spectrophotometric methods. Arch. Biochem. Biophys. 249, 437–446.
Prakash, V., and Timasheff, S. N. (1985). Calculation of partial specific volumes of proteins
in 8 M urea solution. Methods Enzymol. 117, 53.
Ralston, G. (1993). Introduction to Analytical Ultracentrifugation. Beckman Instruments, Full-
erton, CA.
Reynolds, J. A., and McCaslin, D. R. (1985). Determination of protein molecular weight in
complexes with detergent without knowledge of binding. Methods Enzymol. 117, 41–53.
Rhodes, D. G., and Laue, T. M. (2009). Determination of protein purity. Methods Enzymol.
463, 691–724.
Rodbard, D., and Chrambach, A. (1971). Estimation of molecular radius, free mobility, and
valence using polyacylamide gel electrophoresis. Anal. Biochem. 40, 95–134.
Schuck, P. (2000). Size-distribution analysis of macromolecules by sedimentation velocity
ultracentrifugation and Lamm equation modeling. Biophys. J. 78, 1606–1619.
Sletyr, U. B., Messner, P., Pum, D., and Sara, M. (1988). In ‘‘Crystalline Bacterial Cell
Surface Layers.’’ Springer-Verlag, Berlin.
Stafford, W. F. (1992). Boundary analysis in sedimentation transport experiments: A proce-
dure for obtaining sedimentation coefficient distributions using the time derivative of the
concentration profile. Anal. Biochem. 203, 295–301.
Steele, J. H. C., Tanford, C., and Reynolds, J. A. (1978). Determination of partial specific
volumes for lipid-associated proteins. Methods Enzymol. 38, 11–23.
Stellwagen, E. (2009). Gel filtration. Methods Enzymol. 463, 373–386.
Size, Molecular Weight, and Subunits 723

Timasheff, S. N., and Fasman, G. R. (eds.) (1971). In ‘‘Subunits in Biological Systems’’

Part A. Dekker, New York.
Timasheff, S. N., and Townend, R. (1970). In ‘‘Physical Principles and Techniques of
Protein Chemistry,’’ (S. J. Leach, ed.), Part B, p. 147. Academic Press, New York.
Tinoco, I., Sauer, K., Wang, J., and Puglisi, J. (2002). Physical Chemistry: Principles and
Applications in Biological Sciences. 4th edn. Prentice-Hall.
Topf, M., and Sali, A. (2005). Combining electron microscopy and comparative protein
structure modeling. Curr. Opin. Struct. Biol. 15, 578–585.
Topf, M., Lasker, K., Webb, B., Wolfson, H., Chiu, W., and Sali, A. (2008). Protein
structure fitting and refinement guided by cryo-EM density. Structure 16, 295–307.
van Bruggen, E. J. F., and Wiebenga, E. H. (1962). Structure and properties of hemocyanins.
I. Electron micrographs of hemocyanin and apohemocyanin from Helix pomatia at
different pH values. J. Mol. Biol. 4, 1–7.
Wen, J., Arakawa, T., and Philo, J. S. (1996). Size-exclusion chromatography with on-line
light-scattering, absorbance, and refractive index detectors for studying proteins and their
interactions. Anal. Biochem. 240, 155–166.
Yang, J. T. (1961). The viscosity of macromolecules in relation to molecular conformation.
Adv. Protein Chem. 16, 323–400.
Yphantis, D. A. (1964). Equilibrium Ultracentrifugation of Dilute Solutions. Biochemistry 3,
297–317.
 C H A P T E R F O R T Y

Identification and Quantification

of Protein Posttranslational
Modifications
Adam R. Farley* and Andrew J. Link†

Contents
1. Introduction 726
2. Enrichment Techniques for Identifying PTMs 731
2.1. Phosphorylation 731
2.2. Glycosylation 737
2.3. Ubiquitination and sumoylation 738
3. Nitrosative Protein Modifications 740
4. Methylation and Acetylation 741
5. Mass Spectrometry Analysis 744
6. CID versus ECD versus ETD 747
7. Quantifying PTMs 750
8. Future Directions 756
Acknowledgments 758
References 758

Abstract
Posttranslational modifications (PTMs) of proteins perform crucial roles in
regulating the biology of the cell. PTMs are enzymatic, covalent chemical
modifications of proteins that typically occur after the translation of mRNAs.
These modifications are relevant because they can potentially change a
protein’s physical or chemical properties, activity, localization, or stability.
Some PTMs can be added and removed dynamically as a mechanism for rever-
sibly controlling protein function and cell signaling. Extensive investigations have
aimed to identify PTMs and characterize their biological functions. This chapter
will discuss the existing and emerging techniques in the field of mass spectrom-
etry and proteomics that are available to identify and quantify PTMs. We will

* Department of Biochemistry, Vanderbilt University School of Medicine, Nashville, Tennessee, USA

{
Department of Microbiology and Immunology, Vanderbilt University School of Medicine, Nashville,
Tennessee, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63040-8 All rights reserved.

725
726 Adam R. Farley and Andrew J. Link

focus on the most frequently studied modifications. In addition, we will include

an overview of the available tools and technologies in tandem mass spectrometry
instrumentation that affect the ability to identify specific PTMs.

1. Introduction
Protein posttranslational modifications (PTMs) perform essential roles
in the biological regulation of a cell. PTMs are enzymatic, covalent chemi-
cal modifications of proteins that typically occur after translation from
mRNAs. Chemical modifications of proteins are extremely important
because they potentially change a protein’s physical or chemical properties,
conformation, activity, cellular location, or stability. In fact, most proteins
are altered by the addition or removal of a chemical moiety on either an
amino acid or the protein’s N- or C-terminus. Some PTMs can be added
and removed dynamically as a mechanism for reversibly controlling protein
function. Over 400 specific protein modifications have been identified, and
more are likely to be identified (Creasy and Cottrell, 2004). The most
commonly identified PTMs to date include phosphorylation, sumoylation,
ubiquitination, nitrosylation, methylation, acetylation, sulfation, glycosyla-
tion, and acylation (Table 40.1).
Vast amounts of scientific effort have gone toward identifying PTMs and
elucidating their biological functions. Perhaps the best studied cases involve
eukaryotic histones and the myriad of PTMs that are associated with
transcriptional regulation (Berger, 2007; Fuchs et al., 2009; Issad and Kuo,
2008; Olsson et al., 2007; Reid et al., 2009). Histones associate with
chromosomal DNA to form nucleosomes that bundle together to form
chromatin fibers. The histone code hypothesizes that chromatin–DNA
interactions are regulated by combinations of histone PTMs ( Jenuwein
and Allis, 2001; Strahl and Allis, 2000). The acetylation of lysine was first
discovered in histones and is correlated with actively transcribed genes
(Roth et al., 2001). Combinations of methylation, acetylation, ADP-
ribosylation, ubiquitination, and phosphorylation of histone tails function
to regulate specific gene expression programs (Godde and Ura, 2008).
One of the most ubiquitous PTMs, phosphorylation, is the focus of
many biochemical investigations. It has been estimated that 30% of the
human proteome is phosphorylated (Cohen, 2001; Hubbard and Cohen,
1993). For example, tyrosine kinases and phosphatases transduce signals
from ligand-bound receptors on the cell surface to downstream targets in
the insulin/IGF-1 signaling pathway (Taniguchi et al., 2006).
Classic approaches to detecting PTMs on proteins involved Edman
degradation and thin-layer chromatography (TLC). However, these meth-
ods are hampered by the requirement of significant amounts of starting
Protein Posttranslational Modifications 727

Table 40.1 Common PTMs encountered in mass spectrometry

Nominal mass Proposed biological

PTM shift (Da) Stability function
Phosphorylation
pSer, pThr þ 80 Very labile Cellular signaling
processes, enzyme
activity, intermolecular
interactions
pTyr þ 80 Moderately labile
Glycosylation
O-linked 203, >800 Moderately labile Regulatory elements,
O-GlcNAc
N-linked >800 Moderately labile Protein secretion,
signaling
Proteinaceous
Ubiquitination >1000 Stable Protein degradation signal
Sumoylation >1000 Stable Protein stability
Nitrosative
Nitration, nTyr þ 45 Stable Oxidative damage
Nitrosylation, þ 29 Stable Cell signaling
nSer, nCys
Methylation þ 14 Stable Gene expression
Acetylation þ 42 Stable Histone regulation,
protein stability
Sulfation, sTyr þ 80 Very labile Intermolecular
interactions
Deamidation þ1 Stable Intermolecular
interactions, sample
handling artifact
Acylation
Farnesyl þ 204 Stable Membrane tethering,
intermolecular
interactions, cell
localization signals
Myristoyl þ 210 Stable
Palmitoyl þ 238 Moderately labile
Disulfide bond  2 Moderately labile Protein structure and
stability
Alkylation, þ 57 Stable Sample handling
aCys
Oxidation, þ 16 Stable Sample handling
oMet
The associated mass shifts, predicted MS stability, and proposed biological functions are included.
728 Adam R. Farley and Andrew J. Link

material and an inability to identify rare or substoichiometric PTMs.

Because most PTMs result in a concomitant change in the mass of the
modified protein, methods that can detect changes in molecular mass,
namely mass spectrometry-based proteomics, are now routinely utilized to
identify PTMs. Some PTMs, such as phosphorylation or methylation,
increase the mass of a protein, while other PTMs, like signal peptide
cleavage or disulfide bond formation, decrease the mass. Depending on
the mass spectrometry instrument used, the proteomic approaches described
in this chapter have the advantage of high sensitivity and can determine
molecular masses to an accuracy of less than 1 part per million (<1 ppm)
(Makarov et al., 2006; Lu et al., 2008; Scigelova and Makarov, 2006).
Instrument manufacturers and academic researchers are constantly improv-
ing instrument technologies and developing new methods to identify PTMs
(Garcia et al., 2005, 2007). For example, recent advances in the field include
such developments as electron transfer dissociation (ETD) for the fragmen-
tation of peptides and dynamic detectors capable of accurately measuring
masses of undigested proteins (Coon et al., 2004; Mikesh et al., 2006; Pitteri
et al., 2005; Syka et al., 2004).
Enrichment strategies to selectively isolate proteins or peptides with a
desired PTM can be coupled with mass spectrometry-based proteomic
techniques. These enrichment techniques can alleviate the problems created
by rare modifications. Among the variety of affinity enrichment strategies
available, there are two major categories. First are approaches that use
antibodies to recognize a specific PTM or uniquely modified peptide. For
example, antiphosphotyrosine antibodies are used to enrich for peptides
with phosphotyrosine residues (Blagoev et al., 2004; Rush et al., 2005;
Zhang et al., 2005). Second, there are existing and emerging technologies
to enrich for PTMs based on the chemical affinity of a modification for an
immobilized resin. Such techniques include immobilized metal affinity
chromatography (IMAC) for phosphorylations and lectin chromatography
for glycosylations (Ito et al., 2009).
Another widely used proteomic technique useful in the identification of
PTMs is two-dimensional gel electrophoresis (2D-GE). PTMs can alter the
isoelectric point and electrophoretic mobility of a protein in a 2D-gel
experiment. When such a change is detected between different cell types
or growth conditions, the protein can be isolated and sequenced to identify
its PTMs. For example, the differential phosphorylation of a protein will
alter its isoelectric point and may cause charge heterogeneity. The differen-
tially phosphorylated protein may appear as a train of 2D spots with different
isoelectric points but with similar molecular weights. This pattern is some-
times referred to as ‘‘pearls-on-a-string.’’ Specific protein staining methods
for revealing protein PTMs have been devised over the years (Ge et al.,
2004; Patton, 2002). These include fluorescent methods for the direct
detection of phosphoproteins and glycoproteins in gels (Ge et al., 2004;
Protein Posttranslational Modifications 729

Steinberg et al., 2001; Schulenberg et al., 2003a,b, 2004). For these 2D

spots, the proteins can be in-gel digested, and the recovered peptides can be
analyzed by mass spectrometry to identify, validate, and map the expected
PTMs (Hayduk et al., 2004; Steinberg et al., 2003). However, the 2D-GE
approaches to identifying PTMs are not trivial since the recovery and
analysis of the modified peptides are often problematic.
Multiple tools now exist in proteomics to quantify the absolute or
relative abundance of proteins and their specific PTMs (Paoletti and
Washburn, 2006). In vivo and in vitro labeling methods have been developed
to use mass spectrometry for quantifying PTMs and precisely measuring
their changes during cellular events (Goodlett et al., 2001; Gygi et al., 1999;
Oda et al., 1999; Ong et al., 2002; Ross et al., 2004; Zhang et al., 2005).
Quantitation can be crucial in determining the biological significance of a
given PTM, since simply identifying the presence of a modification may not
provide sufficient biological information to model its importance. A study
by Matthies Mann and coworkers performed in HeLa cells found that 14%
of the identified phosphorylation sites were modulated at least two fold by
epidermal growth factor stimulation, demonstating the importance of quan-
tifying PTMs (Olsen et al., 2006).
There are, however, several limitations that must be considered when
using mass spectrometry-based proteomics to identify PTMs. Some PTMs,
such as phosphorylation of serine, tyrosine and threonine, and either
O-linked or N-linked glycosylation, are labile, and maintaining the modifi-
cation during sample preparation can be difficult. If ineffective separation
techniques are used, unmodified peptides or proteins can mask the modified
form during the mass spectrometry analysis. The detection of substoichio-
metric PTMs is especially difficult when the majority of the protein mole-
cules may be unmodified. For mass spectrometry analysis, there is a
multitude of instruments available from different manufacturers that all
have benefits as well as disadvantages. Proteins are typically digested into
peptides, and the peptides are directly analyzed. Modification of a given
peptide can reduce its ionization efficiency, thereby compromising the mass
spectrum and preventing sequence determination. Alternatively, there can
be ambiguity as to which amino acid is modified, since many peptides
contain multiple potential amino acid targets for the PTM. In addition,
when using collision induced dissociation (CID) to fragment peptides
containing labile PTMs such as phosphorylation and glycosylation, the
labile group undergoes a b-elimination reaction to generate a neutral loss
fragment. This pathway is more energetically favorable than fragmentation
at the amide bonds. The resulting mass spectrum often fails to generate
enough sequencing ions to unambiguously identify the peptide or modified
amino acid. Finally, the detection of rare PTMs can be challenging, and
identification often requires highly sensitive mass spectrometry methods or
PTM enrichment strategies.
730 Adam R. Farley and Andrew J. Link

An important concern for proteomic data generated with any mass

spectrometry technique is the volume and complexity of the information
obtained. A typical experiment on a heterologous mixture of digested
proteins analyzed by liquid chromatography coupled with tandem mass
spectrometry can generate tens of thousands of spectra. Within the spectra
obtained, there is the possibility of a nearly endless combination of PTMs
and peptide adducts responsible for the peaks observed by the mass
spectrometer. The large amounts of data necessitate an in silico approach
involving protein database search algorithms to match theoretical spectra
generated from genomic databases to the experimentally observed spectra.
Searching the experimental spectra to identify PTMs requires some a priori
knowledge of the nature of the potential modifications. This arises as a
result of a limit to the number of modifications that can be taken into
consideration when searching data due to computational resource limita-
tions. For example, when considering only phosphorylation of serine (S),
threonine (T), and tyrosine (Y), which adds a nominal mass of 80 Da, the
computational resources required are exponentially higher than when
searching the same dataset without the mass shift. This is due to the
algorithm not only having to consider peptides without any modification
but also having to consider all the permutations when the S, T, or Y
residues are unmodified or modified in the protein sequences. To address
this issue, several groups have been developing software algorithms to
identify unanticipated PTMs from mass spectrometry data (Chalkley et al.,
2008; Hansen et al., 2001, 2005; Tang et al., 2005). The algorithms
calculate mass differences between predicted peptide sequences and exper-
imental MS/MS precursors and localize the mass shift to a sequence
position in the peptide (Hansen et al., 2005).
This chapter will highlight the existing and emerging techniques in the
field of mass spectrometry and proteomics that are available to identify
PTMs. We will focus on some of the most frequently studied modifications,
although many of these techniques can be extended well beyond the scope
of this chapter. As well as discussing the modifications, we will also include
an overview of the available technologies in tandem mass spectrometry
instrumentation that affect the ability of the investigator to identify specific
PTMs. Previous global analysis studies of PTMs will be considered to
emphasize the multitude of difficulties that arise when attempting to discern
useful information from the vast array of data generated in these large-scale
screens. Example workflows will be presented to assist researchers in devel-
oping a method that is most applicable to their current interests and available
resources. In addition, we will include an overview of the various tools
available that allow further quantification of PTMs. It will become evident
that there are a growing number of resources and technologies at the
disposal of the investigator, many of which have overlapping applicability
to identifying PTMs.
Protein Posttranslational Modifications 731

2. Enrichment Techniques for Identifying PTMs

2.1. Phosphorylation
The reversible phosphorylation of serine, threonine, and tyrosine residues is
probably the most heavily studied PTM. Protein phosphorylation signaling
networks mediate cellular responses to a variety of stressors, growth factors,
cytokines, and cellular interactions. Phosphorylation also influences a mul-
titude of cellular processes such as proliferation, apoptosis, migration, tran-
scription, and protein translation (White, 2008). Aberrant regulation of
protein kinases and phosphatases has been implicated in cancer, autoim-
mune diseases, metabolic disorders, and infectious diseases (Blume-Jensen
and Hunter, 2001; Gatzka and Walsh, 2007; Sirard et al., 2007; Taniguchi
et al., 2006). Single phosphoryation events can have dramatic implications
on cellular processes. For example, the dephosphorylation of eIF4E-binding
protein is triggered by nutrient deprivation, environmental stress, or infec-
tion with some picornaviruses and results in a marked increase in eIF4E-
binding activity and an inhibition of translation (Gingras et al., 2001).
The traditional method of identifying phosphorylation sites involves
growing cells with 32P-labeled ATP or phosphoric acid. The phosphopro-
teome incorporates the radioactive phosphate from the growth medium.
The radioactive proteins are detected using such techniques as 2D-GE or
high-performance liquid chromatography (HPLC). The fractionated pro-
teins are then isolated and hydrolyzed. The resulting phosphopeptides are
separated on TLC plates and subsequently sequenced by Edman degrada-
tion. However, these techniques are time consuming, require large amounts
of relatively pure phosphoproteins, and necessitate the use of significant
amounts of radioactive material (McLachlin and Chait, 2001).
Because of these drawbacks, mass spectrometry-based proteomics has
quickly emerged as the preferred technique for identifying phosphorylations.
A recent proteomic study comparing Escherichia coli, Lactococcus lactis, and
Bacillus subtilis found that nearly every enzyme in the glycolytic pathway,
which mediates carbon metabolism, were phosphorylated on various serine,
threonine, and tyrosine residues (Soufi et al., 2008). Interestingly, they also
found that the detected sites of phosphorylation appear far more evolution-
arily conserved than the primary sequence (Soufi et al., 2008). A separate
study investigating the mouse liver phosphoproteome identified many of the
phosphorylation sites previously annotated in the Swiss-Prot mouse database
(Pan et al., 2008). However, more than half of the identified sites were novel,
suggesting that there are still many more phosphorylation sites yet to be
identified in the mouse phosphoproteome (Pan et al., 2008).
Mass spectrometers have the ability to identify specific sites of phosphor-
ylation within complex protein mixtures. This approach does present some
732 Adam R. Farley and Andrew J. Link

problems that are often faced in other proteomic applications, including

limited sample amounts, complex samples, and a wide dynamic range of
protein concentrations within the sample (White, 2008). To compound the
difficulty of these experiments, phosphorylation is often substoichiometric,
and therefore the phosphopeptide is at a lower concentration than the other
peptides from the same protein. The large population of unmodified pep-
tides suppresses the mass spectrometric response to the phosphopeptides.
This phenomenon can lead to a lack of detectable signal for the phospho-
peptide of interest (McLachlin and Chait, 2001). Suppression artifacts can
be minimized by reducing the fraction of unmodified peptides relative to
the PTM variants using any of a variety of enrichment strategies.
One such enrichment strategy applicable to identifying phosphorylations
is the use of strong cation exchange (SCX) resin to selectively isolate phos-
phopeptides. It has been found that, with a pH of less than 2.6, tryptic
phosphopeptides are not retained by the SCX particles (Beausoleil et al.,
2004). This is likely due to their more anionic nature afforded by
the phosphate group. The flow through of the SCX column can be subse-
quently fractionated with a reverse phase (RP) column and subjected to mass
spectrometry analysis (Lim and Kassel, 2006). The separation of the phos-
phorylated peptides from their unmodified counterparts helps alleviate the
problem of suppression during mass spectrometry as discussed above. This
enrichment strategy has the advantage of being applicable to both off-line and
on-line fractionation approaches. One negative aspect of this approach is that
it relies on the inability of the SCX resin to bind the phosphopeptides, as
opposed to other strategies that bind the phosphorylated residues and thereby
positively enrich for them selectively.
Immobilized Metal Affinity Chromatography (IMAC) is the most widely
utilized method for selectively isolating phosphopeptides from complex
mixtures of digested proteins. This method typically utilizes a metal chelating
agent to bind trivalent metal cations, such as Fe3þ or Ga3þ (Sykora et al.,
2007). The charged resin is subsequently used to bind the phosphorylated
peptides. The resin with bound phosphopeptides is washed with a rinse
solution, typically an acetic acid solution, to remove unmodified peptides.
The phosphopeptides can then be eluted with phosphate buffer or a high pH
solution either directly onto a RP column for desalting and subsequent
mass spectrometric analysis or off-line for further enrichment or mani-
pulation. The IMAC approach does present some complications, however.
If multiply phosphorylated peptides are present in abundance, they can
overwhelm the IMAC resin and result in the retention of fewer singly and
doubly phosphorylated species. Additionally, other acidic peptides will have
an affinity for the IMAC column and will be enriched along with the
phosphorylated peptides. This problem is exacerbated by exceedingly com-
plex peptide mixtures such as whole cell extracts. In an attempt to avoid
this problem, many investigators methyl esterify the acidic residues and the
Protein Posttranslational Modifications 733

C-terminus using a chemical reduction procedure (Ficarro et al., 2002).

However, studies have found that the methyl esterification reaction is not
quantitative (Cirulli et al., 2008). Thus, the peptide of interest is present as
modified, partially modified and unmodified versions. The complexity of the
mixture is increased, and the relative amount of a given peptide is actually
reduced (Cirulli et al., 2008).
An approach analogous to IMAC is the use of titanium dioxide (TiO2) as
a substitute for the metal chelating resin (Larsen et al., 2005; Pinkse et al.,
2004; Thingholm et al., 2006). Just as the phosphate groups of the modified
peptides have affinity for the trivalent metal used in IMAC, they also have
an affinity for the titanium dioxide molecules under acidic conditions
(Larsen et al., 2005; Pinkse et al., 2004; Thingholm et al., 2006). By shifting
to a high pH buffer, the phosphopeptides can be eluted from the TiO2
material. This approach has the advantage of requiring less column
preparation time and fewer rinse cycles than the IMAC resin. Studies have
shown that, although these two methods use the same principle of affinity,
they are complementary in that distinct phosphopeptides are detected by
each of the two approaches (Cantin et al., 2007). IMAC is characterized by a
higher affinity for multiply phosphorylated peptides while titanium dioxide
preferentially binds singly phosphorylated species (Bodenmiller, et al.,
2007). This highlights the utility of using both approaches in a single
investigation to obtain maximal coverage of the phosphoproteome.
Another method for phosphopeptide enrichment employs chemically
introduced affinity tags at the sites of phosphorylation (Goshe et al., 2001;
Oda et al., 2001). One such method uses an alkaline environment to remove
H3PO4 from phosphoserine or phosphothreonine residues in a b-elimination
reaction (McLachlin and Chait, 2003; Oda et al., 2001). A dithiol group is then
added to the double bond in a Michael-like addition reaction. The thiol is then
treated with an alkylating reagent linked to a biotin group. The resulting
biotinylated peptides can then be enriched using avidin column chromatogra-
phy (Fig. 40.1) (McLachlin and Chait, 2003). One difficulty of this approach is
the recovery of the tagged species can be problematic due to the high affinity of
biotin for avidin. Also, this method enriches for phosphoserine and
phosphothreonine, not phosphotyrosine. O-linked b-N-acetylglucosamine
(O-GlcNAc) modifications are also succeptible to b-elimination followed by
Michael addition and can therefore compete with phosphorylations for
enrichment (Wells et al., 2002). Finally, under the alkaline environment of
the b-elimination reaction, unwanted side reactions can occur that have
unanticipated affects on peptide mass. These side reactions can make it difficult
to identify the peptides downstream during mass spectrometry.
During mass spectrometry, specific chemical properties of phosphopro-
teins must be taken into account. Phosphorylation is a labile addition to S and
T amino acids. Traditional fragmentation methods result in the preferential
neutral loss of H3PO4 or HPO3 from phosphoserine and phosphothreonine
734 Adam R. Farley and Andrew J. Link

OH

O P OH

N
H
O
OH−
H3PO4 1. b -elimination

N
H
O

SH
2. Michael-like
SH addition

S
HS

N
H
O
S N
S
3. Binding to
affinity resin

H
S N
NH
S
S S +

Figure 40.1 The chemical modification and affinity purification of proteins bearing a
phosphoserine. Reprinted with permission from McLachlin and Chait (2003).
Protein Posttranslational Modifications 735

residues prior to fragmentation along the peptide backbone. The vast majority
of the population undergoes b-elimination reaction, leaving only a fraction
remaining to fragment along the backbone (Bennett et al., 2002; Lee et al.,
2001; Ma et al., 2001). The resulting mass spectrum contains fewer sequenc-
ing ions, and these ions are generally lower in intensity. Therefore, there is a
low signal-to-noise ratio (Fig. 40.2), which hampers interpreting the peptide
sequence. Although capable of losing the phosphate group, phosphotyrosine
seldom generates the intense neutral loss ions that are seen for peptides
containing phosphoserine or phosphothreonine residues. Some mass spectro-
meters that use ETD avoid this neutral loss problem and can yield spectra
suitable for both sequence analysis and phosphoresidue determination. This
alternative fragmentation method will be discussed further when we consider
the various instrument types used in the proteomic identification of PTMs.
On the other hand, neutral loss artifacts can be used to the advantage
of the investigator. The neutral loss signature is a conspicuous indicator
that the peptide contains a site of phosphorylation. The neutral loss peak
observed in the mass spectrum is generally of high intensity due to its
preferential formation. Therefore, if a loss of 98 m/z (H3PO4) or 80 m/z
(HPO3) for a þ1 charged peptide is observed from the precursor ion,
one can infer the presence of a phosphoserine or phosphothreonine
reside within the sequence of the peptide. Modern mass spectrometers
can monitor for this mass loss. An ion trap mass spectrometer can isolate
the neutral loss species and initiate another round of fragmentation (MS3
scan) (Beausoleil et al., 2004; Gruhler et al., 2005; Olsen and Mann,
2004; Ulintz et al., 2009). This additional analysis step can result in
additional sequence coverage to aid in identifying the phosphopeptide
and mapping the PTM to a specific amino acid. An alternative approach
to an MS3 scan is ‘‘Pseudo MSn’’ or multistage activation (MSA) for
phosphopeptide fragmentation in an ion trap mass spectrometer
(Schroeder et al., 2004). The method induces CID of the neutral loss
product from the MS/MS scan of the phosphopeptide. The product ions
from the neutral loss ions along with the initial MS/MS product ions are
trapped together resulting in a composite spectrum. In MSA, the neutral
loss product ions are converted into a variety of structurally informative
fragment ions, which show improved scores in database search algorithms
(Schroeder et al., 2004; Ulintz et al., 2009).
Phosphorylation introduces a negative charge onto the peptide, and this
also affects analysis by mass spectrometry. The negative charge can reduce
the mass spectrometer response when operated in positive-ion mode
(McLachlin and Chait, 2001). This diminished response results in poor
quality spectra and makes identifying the phosphopeptide more difficult.
Furthermore, this increased negative charge can reduce proteolysis during
preparation of protein mixtures prior to mass spectrometry. As a caveat,
Hanno Steen and coworkers found no evidence for decreased ionization/
736 Adam R. Farley and Andrew J. Link

A
100 +2
(M + 2H − H3PO4 )
95
90
85
80
75
70
65
60
55
50
45
40
35
30
25
20
15
10
5
0
400 600 800 1000 1200 1400 1600 1800 2000
m/z

B
10X
9.5
9.0 N M A I N P S* K E N L C S T F C K
8.5
8.0 y8
7.5
7.0 b12
6.5 y12
6.0
5.5
5.0 y6
4.5
4.0 b16
3.5 y10
3.0 y14
2.5 b12
b5
2.0 y5
b4 y4 y9 y13
1.5 b3
b15
1.0
0.5
0.0
400 600 800 1000 1200 1400 1600 1800 2000
M/Z

Figure 40.2 Phosphopeptide MS/MS spectrum from a bovine b-casein tryptic digest.
Panel A demonstrates the intense neutral loss peak observed at 999.2 m/z from the
precursor 1048 at a þ 2 charge state ( 49 Da). This is magnified in Panel B by a factor of
10 to reveal the sequencing ion peaks.

detection efficiencies when selectively investigating phosphopeptides using

electrospray ionization (ESI) (Steen et al., 2006).
As an alternative, a precursor ion scanning technique pioneered by Steve
Carr and coworkers can be employed to avoid some of the problems
generated by performing MS in the positive-ion mode (Carr et al.,
Protein Posttranslational Modifications 737

1996). Tandem mass spectrometry is performed in negative-ion mode,

and phosphorylated residues generate characteristic fragments of  79 Da
for PO3 or 63 Da for PO2. When one of these events is detected in
the precursor ion scan, MS/MS is preformed on the precursor ion
(Zappacosta et al., 2006). This approach can be more sensitive than
positive mode mass spectrometry for detecting phosphopeptides (Carr
et al., 1996). However, negative-mode spectra are difficult to interpret
and not widely studied (Witze et al., 2007). One interesting approach
combines both positive-ion mode and negative-ion mode in the same
experiment. Precursor ion scanning is performed in the usual negative-
ion mode, the instrument’s polarity is switched, and MS/MS fragmenta-
tion spectra are obtained in the positive-ion mode (Carr et al., 1996).
Although identifying sites of phosphorylation is challenging, many pro-
teomic techniques are available that can be customized for characterizing
these critically influential PTMs.

2.2. Glycosylation
Glycosylation is a common PTM and has been shown to affect enzyme
activity, protein localization, stability, signaling, cell adhesion, and protein
interactions (Spiro, 2002). Changes in glycosylation patterns on the surface of
cells and on proteins within biological fluids have been discovered in malig-
nant transformation and tumorogenesis (Durand and Seta, 2000). A variety of
glycosylations can be present within a proteome. O-glycosylated proteins are
modified at serine and threonine, while C-glycosylation targets tryptophan
and N-glycosylation targets asparagine. Proteins with N-glycosylation
share several commonalities such as core structures, prevalence on the
extracellular membrane, and enzymes that remove the N-linked structures
(Medzihradszky, 2005). N-glycosylation sites have has an established
consensus sequence of N-X-S/T where X is any amino acid except proline
(Pless and Lennarz, 1977). Another modification that targets serine and threo-
nine residues is O-linked b-N-acetylglucosamine (O-GlcNAc). Gerald Hart
and coworkers have pioneered the study of this dynamic nucleocytoplasmic
PTM (Holt and Hart, 1986). This modification targets many of the same
proteins selected for phosphorylation. It is postulated that this modification
may serve an antagonistic role to protein phosphorylation (Ahmad et al., 2006).
Some of the same problems to identify phosphorylations with mass
spectrometry also apply to the study of glycosylations. These modifications
tend to be substoichiometric in their abundance relative to unmodified
proteins. Glycosylation and modification by O-GlcNAc are both labile in
the mass spectrometer and often yield poor quality spectra for sequencing.
Even though there is a well characterized consensus sequence known for
the N-glycosylation of proteins, only a subset of the proteins containing
this sequence are glycosylated in vivo.
738 Adam R. Farley and Andrew J. Link

In addition to the alternative peptide fragmentation options mentioned

for phosphorylation, which also apply to glycosylation, enrichment meth-
ods exist for overcoming the issue of low-abundance glycosylations
(Mechref et al., 2008). Glycoproteins can be selectively enriched using
lectin-affinity chromatography. Lectins are highly specific sugar-binding
proteins that can be linked to a solid-phase support. The resin is then
exposed to the complex mixture of digested peptides, washed and eluted
to isolate an enriched fraction of glycopeptides suitable for proteomic
analysis. O-GlcNAc can be enriched using b-elimination followed by
Michael addition of dithiothreitol or biotin pentylamine in a manner
analogous to that previously highlighted for phosphorylations (Wells et al.,
2002). Another method is solid-phase extraction of N-linked glycopeptides
(SPEG) that utilizes hydrazine chemistry to attach the glycopeptides directly
to the solid-phase support (Tian et al., 2007). Washes are performed and the
glycopeptides are eluted by treatment with peptide-N-glycosidase. The
lectin-affinity chromatography approach has the advantage of binding a
wide array of glycoproteins and can effectively isolate different classes of
glycopeptides by eluting with different glycosidases (Yang and Hancock,
2005). There are also specialized columns that contain lectins that uniquely
bind specific classes of glycolytic linkages, affording another degree
of selectivity in the enrichment (Tian et al., 2007). One caveat of using
glycosidases to elute the glycopeptides is that it cleaves the oligosaccharide
from the peptide so that any structural information that could be gained by
mass spectrometry analysis is lost. However, this does simplify the data
analysis since glycosylations can be complex in the nature of their structure
and exist in varied polyglycosylated forms.

2.3. Ubiquitination and sumoylation

Ubiquitin is a 76-amino acid protein that is conjugated to an alpha-amino
group of a substrate lysine via its C-terminal glycine residue. This process
requires the action of E1, E2, and E3 enzymes (Fang and Weissman, 2004).
The sequence of ubiquitin contains seven lysines that can also be selected for
ubiquitination, leading to the formation of ubiquitin chains (Fang and
Weissman, 2004). The internal lysines within ubiquitin that are targeted
for further ubiquitination can determine the biological significance of this
modification. For example, K48 linked ubiquitination chains signal the
attached protein for degradation by the 26S proteosome, and K63 linkages
are involved in DNA damage responses, protein trafficking, and signal
transduction in NF-kB signaling (Ikeda and Dikic, 2008; Pickart and
Fushman, 2004). The turnover of ubiquitinated proteins is very rapid
which results in low steady-state levels of this PTM (Peng et al., 2003).
To further complicate the prediction of ubiquitination sites, there are little
data to support a common sequence motif for this PTM (Sliter et al., 2008).
Protein Posttranslational Modifications 739

Another small protein molecule that is similar to ubiquitin is the small

ubiquitin-like modifier (SUMO). Although the two proteins are only 18%
identical, they show similarity in their 3D structures (Bayer et al., 1998).
Sumoylation and ubiquitination share the same E1, E2, and E3 enzymes,
which conjugate the SUMO molecule to an internal lysine residue within
a conserved motif. Sumoylation occurs preferentially at c–K–X–E
sequences, where c denotes a hydrophobic amino acid (Rodriguez et al.,
2001). Sumoylation targets a multitude of proteins that are involved in such
diverse cellular processes as transcriptional regulation, nuclear body forma-
tion, nuclear pore complexes, and DNA repair (Melchior et al., 2003).
Sumoylation can compete with ubiquitination, thereby preventing protein
degradation and stabilizing key enzymes needed by the cell. This modifica-
tion also has the ability to target proteins to specific cellular locations such as
the nuclear pore complex (Manza et al., 2004).
Effective enrichment strategies have been developed to isolate proteins
modified by ubiquitination and other ubiquitin-like modifications. The
most successful of these approaches have used amino-terminal epitope tags
incorporated into the genetic sequence of the ubiquitin and ubiquitin-like
molecules (Peng et al., 2003; Tagwerker et al., 2006). Polyhistidine tags
have been used, as well as tandem-affinity tags composed of both a poly-
histidine tag and a FLAG tag (Chang, 2006). The tandem-affinity tag
purification method reduces but does not eliminate the rate of false positive
discovery. While it is possible to use antibodies generated against the
ubiquitin class of molecules, their use in immunoprecipitation reactions to
isolate ubiquitinated proteins is still in development. Ubiquitin-binding
proteins may also be used to avoid problems associated with using antibodies
and affinity tags (Kirkpatrick et al., 2005). Epitope tags are convenient in
simple model systems such as Saccharomyces cerevisiae because the multiple
genes that code for untagged ubiquitin can be deleted so that the only
ubiquitin present in the cells contains the epitope tag.
The fact that both ubiquitination and sumoylation are proteinaceous
PTMs enables different approaches to identifying the protein targets with
mass spectrometry. When trypsin digesting ubiquinated proteins, part of the
ubiquitin molecule is cleaved. The remaining tryptic fragment at the site
of ubiquination contains a two residue glycine–glycine remnant from the
C-terminus of ubiquitin linked to the conjugate lysine through a isopeptide
bond (Peng et al., 2003). This peptide causes a nominal mass shift of 114 Da at
the lysine residue as well as a missed tryptic cleavage since trypsin cannot
recognize the modified lysine. For other ubiquitin-like modifications, there
are characteristic mass shifts on lysine associated ubiquitination and ubiquitin-
like modifications as shown in Table 40.2 (Denison et al., 2005). All these
properties can be taken into consideration when searching spectra for
potential PTMs of this type.
740 Adam R. Farley and Andrew J. Link

Table 40.2 Ubiquitin-like modifications and their associated nominal mass remnants
from humans and yeast

Ubiquitin-like Nominal Residual

modifier Organism mass (Da) sequence
SUMO-1 Human 2137 (K)-ELGM. . .QTGG
SUMO-2,3 Human 3552 (R)-FDGQ. . .QTGG
SUMO Yeast 485 (R)-EQIGG
URM1 Yeast 2035 (K)-DYILE. . .LHGG
NEDD8 Yeast 114 (R)-GG
NEDD8 Human 114 (R)-GG
ISG15 Human 114 (R)-GG

One obvious problem with using a mass spectrometry-based proteomics

approach to studying this PTM is the redundancy in mass shifts as indicated
in Table 40.2. From the spectra of an identified peptide, one may not be
able to discern if the modification is ubiquitination or an ubiquitin-
like modification since both types can result in the addition of 114 Da to
the lysine residue. Sumoylation adds a large mass to the peptide that can
affect the ionization of the peptide and reduce the efficiency of ionization.
The peptide fragmentation that occurs on the SUMO modification can
complicate the resulting mass spectrum to the point that it is uninterpret-
able. One laboratory took advantage of the latter problem by developing
sumoylation pattern recognition software (SUMmOn) to search for frag-
ment ion patterns produced by complex PTMs and to identify modified
peptides with their corresponding sites of modification (Pedrioli et al.,
2006). The SUMmOn technology allowed them to both identify a user-
specified fragment ion series and characterize a modified peptide (Pedrioli
et al., 2006).

3. Nitrosative Protein Modifications

Nitration and nitrosylation of tyrosine, tryptophan, methionine, and
cysteine side chains comprise the bulk of nitrosative protein PTMs. These
additions are mediated by reactive nitrogen species produced during devel-
opment, oxidative stress, and aging. Increases in reactive nitrogen species
result from the reaction of excess or deregulated nitric oxide with reactive
oxygen species (Yeo et al., 2008). The reactive nitrogen and oxygen can
target DNA, lipids, and proteins (Barnes et al., 2003). Depending on the
radical generated, the reactive nitrogen species will preferentially react with
Protein Posttranslational Modifications 741

tyrosine residues to form either 3-nitrotyrosine or 3-nitrosotyrosine

(Beckman et al., 1990). Tyrosine nitrosative modifications are perhaps
best characterized by the adducts generated. Increased levels of tyrosine
nitration have been implicated in age-related neurodegenerative diseases
and can serve as a biomarker for oxidative stress (Yeo et al., 2008). Under
normal physiological conditions, the formation of nitrosative tyrosine
adducts is prevented by the presence of reducing agents such as glutathione
(Chen et al., 2004). Tyrosine nitration shifts the pKa of the attached
hydroxyl group from 10.1 to 7.2 and can have consequences in the structure
and function of the modified protein (Sokolovsky et al., 1967).
There are no known selective enrichment strategies for the direct isola-
tion of nitrated peptides or proteins. This is postulated to be a result of the low
chemical activity of the nitric oxide groups of this PTM. These fairly inert
moieties can be converted to amines under reducing conditions with dithio-
nite to generate amines that are reactive to tagging groups (Yeo et al., 2008).
The resulting tagged peptides can then be selectively isolated and analyzed by
proteomic methods. The most widely applied approach to identify protein
nitrosative modifications is use of a combination of 2D-GE and immunoblot
analysis with antibodies specific to a given nitrosative adduct (Zhan and
Desiderio, 2004). Proteins are resolved in the 2D-GE experiment, and
proteins bearing the modification are visualized using the appropriate anti-
body. The immunopositive protein spots are then excised from the gel,
digested and subjected to mass spectrometry identification (Zhan and
Desiderio, 2004). In the case of tyrosine nitrosative modifications, a nominal
mass shift of 45 Da for nitration and a shift of 29 Da for nitrosylation is added as
a differential modification when searching mass spectrometry data to identify
the sites of adduction.

4. Methylation and Acetylation

Lysine and arginine methylation (þ14 Da) and lysine acetylation (þ 42
Da) are two types of PTMs that have come under intense scrutiny, namely for
their role in the histone code ( Jenuwein and Allis, 2001; Strahl and Allis,
2000). The first instance of finding an acetyl group linked to lysine occurred in
a study of histone modifications ( Jenuwein and Allis, 2001). Increased levels of
histone acetylation are generally associated with local activation of gene
transcription (Zhang et al., 2002). Lysine methylation of histone proteins has
been shown to localize to the promoters of repressed genes (Berger, 2007).
The acetylation events have been found to be reversible and dynamic, while
the methylation events are more stable and long-lived (Bernstein and Allis,
2005). Although much scientific effort has gone into studying these PTMs
on histones, acetylation and methylation events occur on many proteins
742 Adam R. Farley and Andrew J. Link

(Glozak et al., 2005; Grewal and Rice, 2004; Sadoul et al., 2008; Yang and
Seto, 2008). In eukaryotes, N-terminal acetylation is one of the most common
PTMs, occurring on approximately 85% of eukaryotic proteins (Polevoda and
Sherman, 2000). The first nonhistone protein found to be regulated by
acetylation and deacetylation was the oncogene p53 (Gu and Roeder, 1997).
As with many of the other PTMs, the classical approach to identifying
methylation and acetylation was the use of modification-specific antibodies.
This technique is limited by cross-reactivity and specificity problems asso-
ciated with an immunological-based method. In the case of histones,
methods exist for their selective isolation from the nuclei of cells (Garcia
et al., 2007). A more general approach to identifying sites of acetylation and
methylation is the use of an in vitro radiolabeling assay. This technique
utilizes acetyltransferases or methyltransferases to catalyze the addition of
radioactive acetyl or methyl donor group. Thin layer chromatography
(TLC) or HPLC separation followed by microsequencing can reveal the
radiolabeled amino acids (Sobel et al., 1994). The radiolabeling approach has
disadvantages in that it requires significant amounts of pure starting material
and utilizes radioactivity to detect modified amino acids.
There are no efficient methods for the selective enrichment of peptides
or proteins with methylation or acetylation modifications. Additionally,
since acetylation and methylation occur on basic residues, the use of trypsin
as a protease can be less effective in cleaving at its usual sites. This can be
more dramatic in the case of methylation since the basic residues can be
methylated at multiple amine sites. The obvious choice to overcome this
problem is selecting an alternative protease. However, the benefit of the
missed tryptic cleavages can work to the investigator’s advantage, serving as
an indicator that the site may be modified. Tandem mass spectral searching
algorithms can take into account the average mass shift caused by acetylation
and methylation of 42 and 14 Da, respectively, as well as the shift of 28 Da
for dimethylation and 42 Da for trimethylation. Some ambiguity can appear
when considering trimethylation since the mass shift between it and an
acylated peptide is only 0.0363 Da. This mass difference can be resolved
with the use of a mass spectrometer with a high enough degree of mass
accuracy and resolution (Zhang et al., 2004).
Another diagnostic tool to identify sites of acetylation and methylation
is analysis of the immonium ions generated in CID type mass spectro-
meters. Peptides that harbor an unmodified lysine generally produce a
characteristic immonium ion peak at 84 m/z (Trelle and Jensen, 2007).
When the lysine is modified, a different immonium is formed at a unique
m/z value (see Figure 40.3). Di- and trimethylated forms of the immonium
do not form due to constraints of the rearrangement reaction that occurs.
However, the trimethylated form produces a unique neutral loss fragment
59 Da less than the precursor due to loss of trimethylamine (Zhang et al.,
2004). Figure 40.3 shows the rearrangement reaction that takes place as
Protein Posttranslational Modifications 743

Mono-methylated Di-methylated Tri-methylated Acetylated

Lysine
lysine lysine lysine lysine
OH
OH OH OH OH
H2N
H2N H2N H2N H2N O
O O O O

O NH
NH2 NH N N+
Immonium ion

Immonium ion

Immonium ion

Immonium ion

Immonium ion
HN HN HN HN HN

NH+ NH+ NH+ N+ O NH2+

3 2

m/z 101 m/z 115 m/z 129 m/z 143 m/z 143
Elimination of ammonia

Elimination of ammonia

Elimination of ammonia

Elimination of ammonia

Elimination of ammonia
and rearrangement

and rearrangement

and rearrangement

and rearrangement

and rearrangement

+
NH+ NH+ N+ NH+ NH+ NH+ N
O

m/z 84 m/z 84 and 98 m/z 84 m/z 84 m/z 84 and 126

Figure 40.3 Proposed structures of the immonium ions generated during CID for
mono-, di-, and trimethylated lysines. Reprinted with permission from Zhang et al. (2004).

well as the associated m/z values of the various immonium ions. These
unique diagnostic characteristics are only generated in mass spectrometers
that are capable of CID.
744 Adam R. Farley and Andrew J. Link

5. Mass Spectrometry Analysis

A variety of mass spectrometry instrument designs and configurations
are available to investigate both the dynamic nature of the proteome and its
associated PTMs. For identifying PTMs, two of the most important features
of a mass spectrometer are its mass accuracy and resolution. Unlike protein
identifications, which are typically based on identifying multiple indepen-
dent peptides from the same protein, PTMs must be identified based on a
single MS/MS spectrum. John Yates and his group have developed
approaches using overlapping peptides produced from independent diges-
tions with various proteases to increase the sequence coverage and reduce
the ambiguity in mapping modifications (MacCoss et al., 2002; Wu et al.,
2003). While the redundant peptide sequences reduce ambiguity, the
approach requires sufficient starting material for multiple mass spectrometry
analyses.
Since the identification of the PTM and its location on the peptide
typically relies on a single MS/MS spectrum, mass spectrometers with high
mass accuracy and resolution have a distinct advantage (Clauser et al., 1999;
Haas et al., 2006; Mann and Kelleher, 2008). Mass spectrometry instrumen-
tation varies in methods for peptide ionization, peptide fragmentation, ion
isolation, and signal detection. Mass spectrometry-based analyses are also
often coupled to chromatographic separation techniques to increase the
level of sensitivity to detect peptides in a given biological sample and to
fractionate complex peptide mixtures. As highlighted previously, many of
these chromatographic approaches can enrich for the desired PTM allowing
for a more effective method to identify the modifications in a complex
protein or peptide mixture. It becomes evident that the combinations of
instrument configurations and separation techniques lead to a dizzying array
of possibilities for the investigator. In this section, we will attempt to discuss
the advantages and possible pitfalls of the mass spectrometers available today.
Perhaps the most widely available and utilized mass spectrometers in
proteomics use electrospray ionization (ESI) to introduce peptides into the
gaseous phase for downstream analysis by the instrumentation. This class of
instruments includes ion traps, triple quadrupoles, time-of-flight (TOF),
quadrupole time-of-flight (q-TOF), orbitrap (Makarov, 2000), and Fourier
transform ion cyclotron resonance (FT-ICR) designs. Ion trap instruments
have the advantage of rapid scan times and high sensitivity. However, they
are limited in their mass accuracy (0.1–1 Da) and resolution. In ion trap
mass spectrometers, the mechanism of CID does not allow for trapping of
fragment masses below 28% of the precursor mass. As a direct result, the lower
1/3 of the MS/MS data is lost. This is referred to as the ‘‘1/3 rule’’ or ‘‘low-
mass cut-off’’ limitation of CID fragmentation. Other mass spectrometers
Protein Posttranslational Modifications 745

such as q-TOF, triple quadrupole, orbitrap, and Fourier transform instru-

ments retain the low-mass-product ions. The low-mass ions may reveal
information about the sequence composition of the peptides that can be
very useful for validating PTMs from the database search.
In an attempt to circumvent the limited accuracy and resolution, the
linear ion trap can be coupled to an orbitrap analyzer to form the hybrid
LTQ-Orbitrap mass spectrometer. The mass accuracy is increased to low
ppm (<5 ppm) and the increased resolution resolves the isotopic envelopes
of the peptides so the charge state of precursor peptides can be determined.
q-TOFs are excellent choices for studying PTMs since they use fragmenta-
tion methods that tend to retain labile peptide modifications. This is
attributed to their higher energies of fragmentation and the shorter time
scale in which the fragmentation occurs. The TOF mass analyzer provides
high mass accuracy and resolution to analyze the precursor and fragment
ions. For tandem mass spectrometry experiments, a TOF mass analyzer can
be separated by a collision cell (TOF–TOF) for routinely generating
MS/MS spectra (Medzihradszky et al., 2000; Vestal and Campbell, 2005).
FT-ICR instruments have the highest accuracy and resolution of the mass
spectrometers available to investigators today. However, their high cost
often makes them impractical for many laboratories. In addition, their low
scan rate decreases the instrument’s sensitivity and necessitates the use of
larger amounts of analyte to obtain a measurable signal.
In general, ESI is best suited to identify small peptides that are acidic in
nature (Covey et al., 1991). ESI typically generates tryptic peptides bearing a
þ2 or þ3 charge state, as opposed to MALDI techniques that typically
produce ions with a þ1 charge state. This complexity can make the
interpretation of spectra more difficult unless an instrument with high
resolution and mass accuracy is used to resolve peptide isotope clusters
and determine the precursor or fragment ions, charge states. The routine
coupling of these instruments to on-line fractionation using low-flow-rate
chromatography allows for a highly sensitive and rapid approach to identi-
fying peptides from a complex protein digest. One significant limitation to
these on-line approaches is that spectra can only be obtained for a peptide as
it is being eluted from a chromatographic column. That is, once the sample
is eluted into the instrumentation, replicate analyses of the peptides are not
available, leading to a temporal limit in data acquisition. This is further
compounded by peptides that coelute, leading to suppression of the ioniza-
tion of the relevant peptide. In addition, unmodified peptides and their
posttranslationally modified counterparts often have similar elution times,
resulting in a diminished signal for the modified peptide.
An alternative to ESI is matrix assisted laser desorption ionization
(MALDI). MALDI instruments typically use TOF analyzers but can also
be used with other mass analyzers. This ionization technique uses a laser to
746 Adam R. Farley and Andrew J. Link

irradiate a sample consisting of an analyte (peptide) and a UV-absorbing

matrix. The matrices are generally aromatic acidic compounds capable of
absorbing the energy from the laser. A portion of the energy absorbed by
the matrix is transferred to the analyte and both desorb into the gaseous
phase where the analyte is ionized and travels downstream in the mass
spectrometer for analysis. MALDI can be applied to either peptides or
whole proteins. When used with the latter, PTMs on undigested proteins
can be identified. This approach does however require some a priori knowl-
edge of the modifications present to account for the mass shift in the
observed protein from its unmodified counterpart. In addition to being
well suited for larger analytes than ESI, MALDI also favors slightly basic
peptides (Covey et al., 1991).
The separation of peptides for analysis with MALDI is typically per-
formed off-line of the mass spectrometry, in contrast to the in-line chro-
matographic separation for ESI. Samples can be fractionated by liquid
chromatography and spotted onto MALDI target plates (Lochnit and
Geyer, 2004; Pflieger et al., 2008; Zhen et al., 2004). Varying methods
exist for spotting the plates, using automated robots specifically designed for
the task. The matrix material can be either mixed with the sample and
spotted, or spotted directly on the plate before or after applying the sample.
The spot size varies according the plate format used and is typically in the
microliter to submicroliter range. The off-line nature of this approach
removes the temporal restrictions that apply to in-line chromatographic
separations. This means that the investigator can return to a sample on a
MALDI plate and perform replicate analyses of the same spot until the
sample is depleted by the instrument’s laser.
The rapid nature of MALDI-TOF instrument and the predominantly
þ1 charge states for ions makes them a convenient option for quickly
measuring the masses of peptides and proteins. A digested protein sample
can be spotted on the MALDI plate and a mass spectrum of the peptides
generated in a short period time without significant training. The investi-
gator may know from previous data what proteins are present and, there-
fore, the expected m/z values for the peptides. These known values can be
compared with what is experimentally detected, and any differences can be
attributed to modified forms of the peptides. This method is simple, fast, and
independent of sequencing algorithms. For suspected phosphoproteins,
alkaline phosphatase treatment of the sample combined with MALDI-
TOF can be used to identify phosphopeptides (Larsen et al., 2001; Zhang
et al., 1998). Differences in the peptides maps (80 Da) generated before and
after phosphatase treatment can help identify phosphopeptides. Mapping
the specific modified amino acid typically requires performing tandem mass
spectrometry on the modified peptides using an instrument inherently
capable of MS/MS (e.g., ion trap, q-TOF, TOF/TOF).
Protein Posttranslational Modifications 747

6. CID versus ECD versus ETD

Mass spectrometry-based proteomic analyses rely on the fragmenta-
tion of peptides in the gas phase at low collision energies to generate peaks
in the mass spectrum. These resulting peaks are then used to determine the
sequence of the peptide, and from this, the associated protein can be
inferred. The primary method for peptide fragmentation is collision induced
dissociation (CID) (Swaney et al., 2008). LTQ, Orbitrap, q-TOF, MALDI-
TOF/TOF, and FT-ICR instruments are all capable of performing CID
fragmentation of peptides. CID is best utilized for small, low charged, and
unmodified peptide cations (Dongre et al., 1996; Good et al., 2007; Huang
et al., 2005). The collision energy transferred to peptides in CID results in the
vibrational excitation of the peptides and is distributed among the covalent
bonds of the peptide chain. When the internal energy transferred to the peptide
in CID exceeds the activation barrier for bond cleavage, the peptide can
fragment. These fragments can be detected by the mass spectrometer if they
occur within the timescale of the instrument. This fragmentation typically
occurs at the amide bonds of the peptide backbone due to the low activation
barrier that exists at these locations. This generates the characteristic b- and
y-product ions seen in CID spectra (Swaney et al., 2008).
Although CID is effective in generating ion series suitable for peptide
sequencing, some caveats exist that diminish its ability to fragment particular
peptides. Internal basic residues and proline amino acids can prevent the
random protonation of the peptide backbone. These residues direct amide
bond dissociation to specific sites and can inhibit the fragmentation of the
peptide resulting in an insufficiently diverse set of sequencing ions (Mikesh
et al., 2006). The common use of trypsin as a protease alleviates some of
these concerns by generating peptides with a basic residue at the C-terminus.
Labile PTMs such as phosphorylation and glycosylation also present alterna-
tive low-energy fragmentation pathways. In these cases, the aforementioned
neutral loss events can occur, generating spectra with insufficient sequenc-
ing ion coverage. Some PTMs can also inhibit the random protonation of
the peptide backbone and subsequently inhibit fragmentation via CID
(Tsaprailis et al., 1999).
An alternative fragmentation method termed electron-capture dissociation
(ECD) has the ability to overcome some of the limitations posed by CID. In
ECD, protonated peptides are held in the Penning trap of an FT-ICR and
exposed to a beam of electrons at thermal or near-thermal energies. The
capture of the thermal electron by the protonated peptide results in backbone
fragmentation without intramolecular vibrational energy redistribution
(Udeshi et al., 2007). One pathway available to account for the fragmentation
of the peptide immersed in thermal electrons is depicted in Fig. 40.4. ECD
748 Adam R. Farley and Andrew J. Link

+ +
H3N H2N
H
O
H O −O O
H
R C
N
N
OH e− R • H N
OH
H O N
+NH3 +NH 3 O

+NH3 +NH3

NH2
H
O
R H O
NH +
HC N
+NH3 • OH
O

C⬘ Z⬘•
+NH3

Figure 40.4 Fragmentation scheme for production of ions of type c and z by reaction
of a low-energy electron with a multiply protonated peptide. Reprinted with permis-
sion from Udeshi et al. (2007).

fragmentation generally produces c- and z-ions series as opposed to the b- and

y-ions generated in CID. ECD is a size and sequence independent process and
can be used to fragment intact proteins. However, for ECD to be efficient, the
sample must be in a dense population of the thermal electrons. This situation is
technically difficult to achieve on instruments using electrostatic radiofre-
quency fields to isolate ions but is easier in FT-ICRs that use static magnetic
fields. In addition, ECD spectra suitable for sequencing require the averaging of
a large number of scans on a time scale of minutes. This necessitates the
introduction of large amounts of analyte into the instrument and precludes
the detection of peptides or proteins in a complex sample mixture (Udeshi
et al., 2007).
The use of a similar fragmentation method electron transfer disassocia-
tion (ETD) overcomes the limitations generated by ECD.
ETD uses ionized radical anions of polyaromatic hydrocarbons to react
with multiply protonated peptides in the gas phase. These radical anions are
stored in the linear quadrupole ion trap of the mass spectrometer. The radical
anions transfer an electron to the ionized peptides and the resulting charge-
reduced peptide fragments in a mechanism believed to be analogous to ECD,
generating characteristic c- and z-sequencing ions (Syka et al., 2004). The
fragmentation patterns generated with ECD and ETD are independent of
peptide length, amino acid composition, and the presence of PTMs (Udeshi
Protein Posttranslational Modifications 749

et al., 2007). This fragmentation process is highly efficient and occurs on a

millisecond time scale. Therefore, ETD is sensitive down to femtomole
amounts of sample and occurs on a time scale corresponding to current
chromatographic separation techniques (Syka et al., 2004). The main benefit
of this fragmentation technique is its ability to preserve labile PTMs while still
allowing sufficient amide bond cleavage to generate spectra from which
sequencing information can be obtained. However, this method does suffer
from a decrease in fragment ion production as the precursor ion charge
decreases (Swaney et al., 2007).
A recent study has utilized a hybrid approach of ETD and CID (ETcaD)
to target the nondissociative electron transfer product ions of ETD with
CID. This method resulted in median sequence coverage of 88.9% for
ETcaD as opposed to 62.5% for ETD and 77.4% for CID alone (Swaney
et al., 2007). Other studies have utilized ETD to identify PTMs, namely
phosphorylation. One group used CID based fragmentation to detect over
1000 phosphopeptides but was only able to define 383 sites of phosphory-
lation, largely due to the neutral loss events on the modified peptides
(Udeshi et al., 2007). Using an LTQ modified with ETD, 1252 sites of
phosphorylation on 629 proteins of a complex mixture were identified
(Udeshi et al., 2007). A separate analysis of the overlap between

A Ac

T A R K S T G G K A P R K Q L
100 [M+4H−2H2O]+4
Normalized intensity: 1.7 x 105
Relative abundance

80

60 y11+2
y13+3 b5
40 b14+3
+2
a5-NH+2
2
y9
20 y10+3
y5 y6
200 400 600 800 1000 Ac 1200 1400 1600
m/z

B T A R K S T G G K A P R K Q L
100
Normalized intensity: 1.0 x 104
Relative abundance

80 .
[M+4H]+3 +2..
[M+4H]
60 [M+4H]+4 C5 C7
C6 C8 Z7
40 C3 C4 Z4 Z6 Z9 C9
Z8 Z10 Z14 C14
20 Z12 ..
Z11
C13 Z13 [M+4H]+1
C2 Z2 Z3 C11 C12

200 400 600 800 1000 1200 1400 1600

m/z

Figure 40.5 A comparison of CID versus ETD spectra for a human histone H3
peptide. Panel A shows the CID spectrum with the corresponding b- and y-ion series.
Panel B depicts the ETD spectrum and its corresponding c- and z-ion series. Reprinted
with permission from Mikesh et al. (2006).
750 Adam R. Farley and Andrew J. Link

phosphopeptides identified with ETD and CID found that only 17.9% of
the sites identified were shared among the two datasets generated (Swaney
et al., 2008). This suggests that the best approach to fragmenting and
identifying sites of PTMs may be a combination of CID and ETD.
Figure 40.5 shows representative spectra generated with both methods
of fragmentation. The characteristic b- and y-ions appear in the CID
spectrum, while c- and z-type ions appear in ETD.

7. Quantifying PTMs
When investigating the biological significance of PTMs, it is often
advantageous to know the relative or absolute abundance of a specific
modification or group of PTMs. This allows for the direct comparison of
the modification of interest across varied biological samples. An example is
comparing the abundance of a PTM in cells or tissues obtained from a normal
versus a disease state. Quantifying these changes can lead to insights into the
role of PTMs in a myriad of processes from cell growth to diseases and
apoptosis. A range of methods are available to quantify PTMs. Traditional
methods utilize 2D-GE and differential staining to identify differences in the
level of protein expression among samples. However, this approach has a low
resolution, and differences in the ability of some proteins to stain can lead to
artifacts (Ong and Mann, 2005). The 2D-GE approach is most amenable to
abundant protein species (Anderson and Anderson, 1998). More recently,
mass spectrometry-based approaches have alleviated some of the problems
posed by gel-based approaches. These methods include protein or peptide
labeling strategies, tagging approaches and differential proteomic methods
using spectral counting (Gygi et al., 1999; Ong et al., 2002; Washburn
et al., 2001). While these methods are more widely applied to quantifying
protein changes among samples, they can also be used to quantify changes in
PTMs from different biological samples.
Quantitative measurements can either be absolute, in terms of concentra-
tion or copy number per cell, or relative, in terms of fold change among
differentially treated samples. Absolute concentrations are more difficult to
obtain, but are more informative and can be used to derive relative measure-
ments. LC-MS/MS experiments allow the plotting of signal intensities as a
peptide elutes from the chromatographic column over time. The area under
this curve for any given species is directly related to the abundance of the
peptide and allows for label-free quantitation. However, the physiochemical
properties of peptides such as hydrophobicity, charge, and size vary widely
and can lead to differences in the mass spectrometer’s detector response. In
addition, co-elution of other peptides and variations in elution conditions can
be problematic for quantitative investigations with this label-free approach.
Protein Posttranslational Modifications 751

Protein abundance estimates using this method can vary by a factor of

3–5-fold from the true abundance (Ong and Mann, 2005).
To circumvent problems associated with ionization efficiencies and
MS-detector responses, one can take advantage of approaches based on
the stable isotope dilution theory. This theory states that a stable, isotopi-
cally labeled peptide is chemically identical to its native counterpart. As a
result, isotopically labeled and unlabeled peptides behave identically in
chromatography and in MS detection. Since the mass spectrometer can
recognize the mass difference between the labeled and unlabeled forms, the
relative signal intensities can be determined (Bantscheff et al., 2007).
There are four main strategies that exploit the stable isotope dilution theory.
First, an isotopically labeled peptide standard corresponding to a PTM of
interest can be introduced into the sample (Gerber et al., 2003). Second,
cells can be metabolically labeled during growth in media consisting pri-
marily of isotopically pure components (Ong and Mann, 2005). Third,
isotopes can be incorporated into the peptide during proteolysis with
enzymes such as trypsin (Miyagi and Rao, 2007). Finally, isotopically
labeled tags can be chemically linked to specific amino acid side chains
(Gygi et al., 1999; Ross et al., 2004). Regardless of which method is chosen,
the mass difference between the ‘‘heavy’’ and ‘‘light’’ forms of the peptides
should be a minimum of 3–4 Da to avoid isotopic overlapping of the peaks
in the mass spectrum. Additionally, deuterated peptides can show different
chromatographic elution times in relation to their ‘‘light’’ counterparts
(Zhang and Regnier, 2002). This makes using the more expensive
13C- and 15N-reagents necessary.

The most basic approach for using stable isotopes is termed absolute
quantitation (AQUA) (Gerber et al., 2003). In this approach, isotopically
labeled peptides are generated synthetically and added as internal standards
to the peptide mixture. Steve Gygi and his colleagues used this approach for
determining the change in the abundance of phosphorylations during the
Xenopus cell cycle (Stemmann et al., 2001). This method can become
cumbersome in that it requires that a labeled synthetic peptide be generated
for every PTM to be quantified. Labeled phosphopeptides standards are
readily available, but this is not the case for other modifications, limiting
broad application of this method. The internal standard is typically added
just before or after proteolysis, in which case it is impossible to normalize for
variations in sample preparation upstream of the digestion step. This tech-
nique requires a prior knowledge of the PTM so the synthetic peptide can
be generated. It is also necessary to know the general abundance of the
modified peptide so the labeled standard can be added in a similar amount.
To minimize sample preparation errors, cell populations can be grown
under different conditions with or without an isotopic label, and the
populations pooled prior to sample preparation and analysis. Any errors in
sample preparation affect both populations equally. Mathias Mann’s
752 Adam R. Farley and Andrew J. Link

laboratory has developed the most popular of these approaches, called stable
isotope labeling by amino acids in cell culture (SILAC) (Ong et al., 2002). In
the SILAC approach, the culture media contains 13C6-arginine and 13C6-
lysine, which label the sites at which trypsin cleaves. Therefore, excluding
the C-terminal fragment, each tryptic product includes one labeled amino
acid. The advantage of the SILAC method over total metabolic labeling is
that the number of incorporated labels is defined and independent of the
amino acid sequence. This simplicity makes data interpretation easier with
SILAC than with other methods that label the peptides in more complicated
patterns. SILAC limits comparison to a maximum of three culture condi-
tions, which include either unlabeled, 13C6-labeled, or 13C615N4-labeled
amino acids. Other limitations of SILAC are that it is only useful with
certain model organisms and cell lines that can be grown in isotopically
defined media and can be prohibitively expensive. However, SILAC’s high
degree of accuracy makes this method well suited for the study of PTMs
(Olsen et al., 2006). A variation of SILAC employs labeling with labeled S-
adenosylmethionine for the identification and relative quantification of
protein methylation (Ong et al., 2004).
An alternative to in vivo metabolic labeling is enzymatic labeling in vitro.
This labeling can be performed either during protein digestion or after
proteolysis with an additional reaction step. Using heavy water (H218O),
trypsin or Glu-C will introduce two 18O atoms. The resulting 4 Da shift is
sufficient to resolve isotopes and acquire relative quantification data (Miyagi
and Rao, 2007). However, some proteases such as endoproteinase Lys-N
only incorporate one 18O atom and yield spectra insufficient to resolve
isotope peak overlap (Rao et al., 2005). Full enzymatic labeling is rarely
achieved and different peptides incorporate the label at different rates
thereby complicating data analysis (Ramos-Fernandez et al., 2007). As a
result, this method has not found widespread use in quantitative proteomics
(Ong et al., 2004).
Efficient chemical tagging approaches can alleviate some of the negative
constraints of enzymatic tagging procedures. The first of these was devel-
oped by Ruedi Aebersold’s laboratory and termed isotope-coded affinity
tags (ICAT) (Gygi et al., 1999). The ICAT reagent was initially composed
of a thiol reactive group that targets cysteines, a polyether linker region with
either zero or eight deuterium atoms, and a biotin group for affinity
purification with an avidin column. Since deuterium isotopes can behave
differentially in their chromatographic response relative to the light forms,
13C and 15N tagging reagents have been developed (Bottari et al., 2004).

The ICAT reaction is performed on a reduced sample to chemically label it

with either the heavy or light ICAT reagent. Then, the two samples are
mixed and digested with a protease of choice. Labeled peptides are
recovered with the affinity tag of the ICAT group and quantified with
Protein Posttranslational Modifications 753

mass spectrometry. This approach effectively simplifies the peptide mixture

by selecting only peptides containing a cysteine. However, since most
peptides do not contain cysteine residues, ICAT has not found widespread
use for the quantification of PTMs since a majority of PTMs are discarded at
the affinity purification step (Ross et al., 2004).
Another set of chemical tagging strategies utilizes the reactivity of the
peptide N-terminus and the epsilon-amino group of lysine residues. Iso-
tope-coded protein labeling (ICPL), isotope tags for relative and absolute
quantification (iTRAQ), and tandem mass tags (TMT) are all approaches
that were developed to exploit the reactivity of amines to specific chemical
groups (Ross et al., 2004; Schmidt et al., 2005; Thompson et al., 2003). The
most recognized of these approaches, iTRAQ, incorporates isobaric tags
that fragment in the tandem mass spectrometer to generate unique reporter
ions at m/z values of 113–121 (Ong et al., 2004). The peak areas of the low-
mass-fragment ions are integrated to determine the quantification values.
Mass spectrometers such as quadrupoles and TOF instruments which have
the inherent ability to detect low m/z fragment ions are typically used for
iTRAQ experiments. Commercially available kits allow examination of up
to eight states in one single experiment. Since the labeled peptides are
isobaric, they should behave similarly in chromatographic separations and
reveal quantitative differences in their fragmentation spectra. Using the
iTRAQ approach, different samples are isolated under different conditions,
trypsin digested, and labeled with the iTRAQ reagents. Samples are then
combined and analyzed in a single MS/MS experiment. Figure 40.6 shows
an example of tandem mass spectra generated using four iTRAQ reagents in
varying ratios (Ross et al., 2004). As with other MS-based approaches,
sufficient peptide separation is crucial since coeluting peptides of similar
mass can contribute to the observed reporter ion series and interfere with
the quantification. Ion trap mass spectrometers are typically not useful for
iTRAQ analysis because the reporter ions generated from the fragmentation

A 1:1:1:1 mixture B 1:5:2:10 mixture

117.1
114.1
115.1

116.1

117.1

100 6135.0 100 2.3E + 4

90 90
80 80
115.1
% Intensity
% Intensity

70 70
60 60
116.1

50 50
114.1

40 40
30 30
20 20
10 10
0 0
111.0 112.8 114.6 116.4 118.2 120.0 111.0 112.8 114.6 116.4 118.2 120.0
Mass (m/z) Mass (m/z)

Figure 40.6 iTRAQ reporter ion series for two mixtures of varied ratios of a six
protein digest. Reprinted with permission from Ross et al. (2004).
754 Adam R. Farley and Andrew J. Link

of the iTRAQ tags are lost with the lower 1/3 of the MS/MS data. The use
of pulsed q-dissociation (PQD) was introduced to alleviate this problem but
has found limited use in proteomics (Bantscheff et al., 2008; Cunningham
et al., 2006; Griffin et al., 2007).
Using the iTRAQ approach, it is possible to compare the degree
of phosphorylation of the proteome under different conditions. Studies
performed by Forest White and coworkers used antiphosphotyrosine anti-
bodies followed by IMAC enrichment and iTRAQ labeling to investigate
the effect of epidermal growth factor (EGF) stimulation on the phosphory-
lation state of cells over time (Zhang et al., 2005). This study probed four
time points (0, 5, 10, and 30 min) in a single analysis. They were able to
quantify relative changes in the phosphorylation of 78 sites on 58 proteins
(Zhang et al., 2005). This experimental approach highlights the utility of
combining enrichment for PTMs with quantitation to gain additional
insight into the biology of cells. It is difficult to determine what fraction
of a peptide population is phosphorylated under certain conditions. This is
because the unphosphorylated counterpart will behave differently in its MS-
detector response. So while it is feasible to compare unmodified forms and
other unmodified forms or different phosphorylated forms of the same
peptide under different conditions, it is not reliable to compare modified
versus unmodified forms based on the ratio of reporter ion peaks generated.
A limited number of additional tagging methods aimed specifically at
defined PTMs have been developed. These utilize the selective chemical
reactivity of the chemically modified amino acid side chains. For the
quantitation of phosphorylations, b-elimination of phosphoric acid, and
subsequent Michael addition of ethanedithiol derivatives can be employed
to incorporate isotope tags (Tao et al., 2005). Hydrazine chemistry can be
applied to glycosylated peptides to replace the carbohydrate group with an
isotopically labeled tag (Zhang et al., 2003). These approaches are only
effective for a limited number of PTMs. In addition, if the chemical
reactions involved are not highly efficient, the resulting peptide mixture
can increase in complexity. The increase in complexity can lead to problems
in identifying the peptides and their associated PTMs as well as inaccurate
quantitative information.
Isotope labeling and tagging strategies can be tedious, cumbersome, and
expensive. In addition to the chromatographic peak integration method
discussed previously, other label-free quantitation methods exist that can be
applied to quantifying PTMs. These methods typically involve some form of
spectral counting and normalization for protein length (Washburn et al.,
2001). It is debatable whether these methods are truly quantitative, and as a
result, these methods are more commonly referred to as differential proteo-
mic techniques. The spectral counting methods are based on the principle
that, as the abundance of a protein increases in a given sample, more MS/MS
spectra are isolated for peptides derived from that protein. Relative
Protein Posttranslational Modifications 755

quantitation is inferred by comparing the number of spectra collected for a

particular protein between experiments. The utility of this approach is ques-
tionable because it does not directly measure any physical property of the
peptide and assumes a linear response for every protein (Bantscheff et al.,
2007). As previously mentioned, the physiochemical properties of peptides
vary widely in a sample digest and their resulting chromatographic and MS-
detector behavior can vary greatly. These methods provide a larger dynamic
range than labeling and tagging procedures and are most applicable when
investigating large or global protein changes among samples. However, the
uncertain linear response among peptides and the potentially poor accuracy of
spectral counting approaches limits their widespread use (Old et al., 2005).
In an attempt to minimize the variability in label-free quantitation of
proteins in a sample, it is possible to compare specific and multiple peptides
from the same protein across multiple samples. This has lead to the devel-
opment of selected reaction monitoring (SRM) and multiple reaction
monitoring (MRM) for PTMs (Unwin et al., 2005, 2009; Williamson
et al., 2006; Wolf-Yadlin et al., 2007). In an SRM experiment, a unique
peptide precursor and its fragment ion m/z value or transition is selectively
monitored across multiple experiments, typically using a triple quadrupole
mass spectrometer (Lange et al., 2008). For proteins to be quantified with a
high degree of statistical confidence, 3–5 peptides and their transitions are
typically monitored during an MRM experiment. For these experiments,
triple quadrupole mass spectrometers are utilized due to the high degree of
selectivity afforded by their first and third quadrupoles to isolate the selected
transitions and their high duty cycles. Multiple transitions are monitored
over time to produce a set of chromatographic profiles with a retention time
and signal intensity for each specific transition. The integrated areas of the
transitions are used to quantify the proteins. MRM enables the detection of
low abundance proteins in complex mixtures and yields a linear response
over a dynamic range up to five orders in magnitude (Lange et al., 2008).
MRM can be used to selectively monitor proteins containing PTMs
(Williamson et al., 2006). However, MRM lacks the ability to identify
proteins in a mixture, and transitions are typically selected based on previous
experimental data or literature searches. Transitions must be optimized for
efficient fragmentation in the second quadrupole of the triple quadrupole, and
the investigator can be limited by sample availability, although computational
tools can compensate for this. In addition, triple quadrupoles use CID to
fragment peptides and are subject to neutral loss events for labile PTMs. An
extension of MRM includes the use of isotopically labeled synthetic peptides
in a manner analogous to AQUA to achieve absolute quantitation of peptides
(Wolf-Yadlin et al., 2007). Taken together, this emerging technology in mass
spectrometry has the potential to provide quantitative information for many
proteins and PTMs across varied biological samples.
756 Adam R. Farley and Andrew J. Link

8. Future Directions
Traditional proteomic studies employ a bottom-up approach whereby
whole proteins are digested with a protease and analyzed using MS/MS
strategies to determine the peptide sequences. From these results, the
identities of the proteins can be inferred. However, complete coverage of
a protein is rarely accomplished. As a result, if peptides harboring the
modification are not detected, the PTMs cannot be identified. Top-down
investigations aim to determine the entire primary structure of a single
protein by performing tandem mass spectrometry on the intact protein to
produce fragmentation information. The combination of precursor mass
measurement and fragmentation sequencing data allows the identity of the
protein to be determined. Top-down methods can be useful in determining
specific PTMs and identifying splice variants and degradation products not
typically identified in bottom-up approaches (Waanders et al., 2007). Top-
down mass spectrometry also can discern gene products with a high degree
of sequence identity (Parks et al., 2007; Siuti et al., 2006). Top-down
approaches are readily applicable to the identification of PTMs. The initial
MS event on the whole protein determines if it has been modified from its
native form based upon any mass shift that may be present. The change in
the observed mass can be accounted for by only a limited number of
modifications. Additional fragmentation data generated by MS/MS can
help to further characterize the modification based on the observed mass
shift(s) for the peptides.
Top-down proteomics requires the use of instruments with high reso-
lution and high mass accuracy. This has relegated top-down studies to
FT-ICR and Orbitrap mass spectrometers with typical resolving powers
> 500,000 and 100,000, respectively (Macek et al., 2006). The high resolv-
ing power allows for the accurate determination of charge states, which is
essential for measuring the intact protein’s mass. The sub- to low-ppm mass
accuracy enables the discrimination between PTMs with the same nominal
mass, such as acetylation and trimethylation. Recently, studies by Andrea
Armirotti and coworkers and Scott McLuckey and his group have demon-
strated the ability to perform top-down proteomics on a q-TOF mass
spectrometer (Armirotti et al., 2009; Liu et al., 2009). This extending the
utility of this approach beyond the generally more expensive FT-ICR and
Orbitrap mass spectrometers and opens this methodology to a broader
section of investigators. Top-down approaches possess the ability to handle
a dynamic range of protein sizes. Parks et al. (2007) used such an approach to
identify 22 proteins ranging from 14 to 35 kDa in S. cerevisiae. One group
extended this range to characterize the PTMs of a 115-kDa cardiac myosin-
binding protein (Ge et al., 2009).
Protein Posttranslational Modifications 757

In addition to being useful for determining the identities of proteins

within a biological sample, top-down proteomics is also applicable to
studies focusing on quantifying proteins and PTMs. Top-down studies
are particularly well suited for quantitating PTMs, since the ionization
efficiencies of intact proteins are much less affected by the presence of
PTMs as compared to peptides (Ge et al., 2009). 15N labeling strategies
were used by Neil Kelleher and coworkers for intact protein quantitation
and identification generating 50 abundance ratios (Du et al., 2006).
Matthias Mann extended the applicability of his SILAC approach for
proteins up to 220 kDa and quantified a 28-kDa signaling protein
(Waanders et al., 2007). Mann also determined that the quantitation of a
one to one mixture of the 28 kDa species had a standard deviation of 6%
and was primarily limited by the signal-to-noise ratio in the protein
measurements as compared to peptide measurements using SILAC
(Waanders et al., 2007).
Widespread use of top-down proteomics has not yet been achieved. There
are technological challenges in measuring and identifying large whole pro-
teins. The implementation of top-down techniques has been slowed by the
laborious separations of whole proteins, the requirements for high sensitivity,
efficient fragmentation, and the lack of computational tools for data interpre-
tation (Collier et al., 2008; Kelleher et al., 1999; Meng et al., 2002; Siuti and
Kelleher, 2007).
The preferred method for fragmenting proteins in bottom-up proteo-
mics utilizes CID of peptides to generate fragmentation information. CID
of intact proteins is dependent on protein structure and size (Wu et al., 2004;
Zabrouskov et al., 2005). CID produces fragments preferentially at the
termini of proteins. One approach to overcome the low internal sequence
coverage of CID is to employ a middle-down approach. This technique
uses limited proteolysis to generate larger peptides in the 3–20 kDa range.
However, the recovery of peptides digested in this manner is unpredictable
and as a result has gained little momentum in proteomics (Ge et al., 2009).
The use of instruments capable of ECD (FT-ICRs) and, more recently,
ETD (Orbitraps), is alleviating some of the problems faced when using CID
to fragment whole proteins in top-down proteomics (Bunger et al., 2008;
Coon et al., 2005; Ge et al., 2002; McAlister et al., 2007, 2008).
As previously mentioned, ECD and ETD are size independent and have
the ability to produce fragmentation patterns of whole proteins sufficient to
determine protein sequence. For top-down approaches to gain more wide
spread use, proteome coverage and instrument sensitivity both need to be
improved (Ferguson et al., 2009).
Determining PTMs on proteins can be a daunting task. However, as out-
lined earlier, there are a variety of tools available that utilize mass spectrometry
to assist in identifying and quantifying PTMs. Sample preparation, enrichment
strategies, quantitative approaches, instrument configurations, and data analysis
758 Adam R. Farley and Andrew J. Link

tools are all crucial in determining the successfulness of the approach. This
chapter has presented many of the techniques and technologies that exist today
in the field of proteomics that can be tailored toward the goal of determining
PTMs. The chemical modifications of proteins are extremely important
because they potentially change many properties of a protein. Therefore, it is
important to take advantage of the approaches offered in mass spectrometry to
determine sites of PTMs and elucidate their biological function.

ACKNOWLEDGMENTS
We acknowledge Elizabeth M. Link and Kevin Schey for helpful comments and suggestions
in the preparation of this manuscript. A. R. F. was supported by NIH training grant T32
CA009385 and GM64779. A. J. L. was supported by NIH grants GM64779 and AR055231.

REFERENCES
Ahmad, I., Hoessli, D. C., Walker-Nasir, E., Choudhary, M. I., Rafik, S. M., and
Shakoori, A. R. (2006). J. Cell. Biochem. 99, 706–718.
Anderson, N. L., and Anderson, N. G. (1998). Electrophoresis 19, 1853–1861.
Armirotti, A., Benatti, U., and Damonte, G. (2009). Rapid Commun. Mass Spectrom. 23,
661–666.
Bantscheff, M., Schirle, M., Sweetman, G., Rick, J., and Kuster, B. (2007). Anal. Bioanal.
Chem. 389, 1017–1031.
Bantscheff, M., Boesche, M., Eberhard, D., Matthieson, T., Sweetman, G., and Kuster, B.
(2008). Mol. Cell. Proteomics 7, 1702–1713.
Barnes, P. J., Shapiro, S. D., and Pauwels, R. A. (2003). Eur. Respir. J. 22, 672–688.
Bayer, P., Arndt, A., Metzger, S., Mahajan, R., Melchior, F., Jaenicke, R., and Becker, J.
(1998). J. Mol. Biol. 280, 275–286.
Beausoleil, S. A., Jedrychowski, M., Schwartz, D., Elias, J. E., Villen, J., Li, J., Cohn, M. A.,
Cantley, L. C., and Gygi, S. P. (2004). Proc. Natl. Acad. Sci. USA 101, 12130–12135.
Beckman, J. S., Beckman, T. W., Chen, J., Marshall, P. A., and Freeman, B. A. (1990). Proc.
Natl. Acad. Sci. USA 87, 1620–1624.
Bennett, K. L., Stensballe, A., Podtelejnikov, A. V., Moniatte, M., and Jensen, O. N. (2002).
J. Mass Spectrom. 37, 179–190.
Berger, S. L. (2007). Nature 447, 407–412.
Bernstein, E., and Allis, C. D. (2005). Genes Dev. 19, 1635–1655.
Blagoev, B., Ong, S. E., Kratchmarova, I., and Mann, M. (2004). Nat. Biotechnol. 22,
1139–1145.
Blume-Jensen, P., and Hunter, T. (2001). Nature 411, 355–365.
Bodenmiller, B., Mueller, L. N., et al. (2007). Nat. Methods 4, 231–237.
Bottari, P., Aebersold, R., et al. (2004). Bioconjug. Chem. 15, 380–388.
Bunger, M. K., Cargile, B. J., et al. (2008). Anal. Chem. 80, 1459–1467.
Cantin, G. T., Shock, T. R., Park, S. K., Madhani, H. D., and Yates, J. R. 3rd (2007). Anal.
Chem. 79, 4666–4673.
Carr, S. A., Huddleston, M. J., and Annan, R. S. (1996). Anal. Biochem. 239, 180–192.
Chalkley, R. J., Baker, P. R., Medzihradszky, K. F., Lynn, A. J., and Burlingame, A. L.
(2008). Mol. Cell. Proteomics 7, 2386–2398.
Chang, I. F. (2006). Proteomics 6, 6158–6166.
Protein Posttranslational Modifications 759

Chen, Y. R., Chen, C. L., Chen, W., Zweier, J. L., Augusto, O., Radi, R., and
Mason, R. P. (2004). J. Biol. Chem. 279, 18054–18062.
Cirulli, C., Chiappetta, G., Marino, G., Mauri, P., and Amoresano, A. (2008). Anal. Bioanal.
Chem. 392, 147–159.
Clauser, K. R., Baker, P., and Burlingame, A. L. (1999). Anal. Chem. 71, 2871–2882.
Cohen, P. (2001). Eur. J. Biochem. 268, 5001–5010.
Collier, T. S., Hawkridge, A. M., Georgianna, D. R., Payne, G. A., and Muddiman, D. C.
(2008). Anal. Chem. 80, 4994–5001.
Coon, J. J., Syka, J. E. P., Schwartz, J. C., Shabanowitz, J., and Hunt, D. F. (2004). Int. J.
Mass Spectrom. 236, 33–42.
Coon, J. J., Ueberheide, B., et al. (2005). Proc. Natl. Acad. Sci. USA 102, 9463–9468.
Covey, T. R., Huang, E. C., and Henion, J. D. (1991). Anal. Chem. 63, 1193–1200.
Creasy, D. M., and Cottrell, J. S. (2004). Proteomics 4, 1534–1536.
Cunningham, C. Jr., Glish, G. L., and Burinsky, D. J. (2006). J. Am. Soc. Mass Spectrom. 17,
81–84.
Denison, C., Kirkpatrick, D. S., and Gygi, S. P. (2005). Curr. Opin. Chem. Biol. 9, 69–75.
Dongre, A. R., Jones, J. L., Somogyi, A., and Wysocki, V. H. (1996). J. Am. Chem. Soc. 118,
8365–8374.
Du, Y., Parks, B. A., Sohn, S., Kwast, K. E., and Kelleher, N. L. (2006). Anal. Chem. 78,
686–694.
Durand, G., and Seta, N. (2000). Clin. Chem. 46, 795–805.
Fang, S., and Weissman, A. M. (2004). Cell. Mol. Life Sci. 61, 1546–1561.
Ferguson, J. T., Wenger, C. D., Metcalf, W. W., and Kelleher, N. L. (2009). J. Am. Soc.
Mass Spectrom. 20, 1743–1750.
Ficarro, S. B., McCleland, M. L., Stukenberg, P. T., Burke, D. J., Ross, M. M.,
Shabanowitz, J., Hunt, D. F., and White, F. M. (2002). Nat. Biotechnol. 20, 301–305.
Fuchs, S. M., Laribee, R. N., and Strahl, B. D. (2009). Biochim. Biophys. Acta 1789, 26–36.
Garcia, B. A., Shabanowitz, J., and Hunt, D. F. (2005). Methods 35, 256–264.
Garcia, B. A., Shabanowitz, J., and Hunt, D. F. (2007). Curr. Opin. Chem. Biol. 11, 66–73.
Gatzka, M., and Walsh, C. M. (2007). Autoimmunity 40, 442–452.
Ge, Y., Lawhorn, B. G., et al. (2002). J. Am. Chem. Soc. 124, 672–678.
Ge, Y., Rajkumar, L., Guzman, R. C., Nandi, S., Patton, W. F., and Agnew, B. J. (2004).
Proteomics 4, 3464–3467.
Ge, Y., Rybakova, I. N., Xu, Q., and Moss, R. L. (2009). Proc. Natl. Acad. Sci. USA 106,
12658–12663.
Gerber, S. A., Rush, J., Stemman, O., Kirschner, M. W., and Gygi, S. P. (2003). Proc. Natl.
Acad. Sci. USA 100, 6940–6945.
Gingras, A. C., Raught, B., Gygi, S. P., Niedzwiecka, A., Miron, M., Burley, S. K.,
Polakiewicz, R. D., Wyslouch-Cieszynska, A., Aebersold, R., and Sonenberg, N.
(2001). Genes Dev. 15, 2852–2864.
Glozak, M. A., Sengupta, N., Zhang, X., and Seto, E. (2005). Gene 363, 15–23.
Godde, J. S., and Ura, K. (2008). J. Biochem. 143, 287–293.
Good, D. M., Wirtala, M., McAlister, G. C., and Coon, J. J. (2007). Mol. Cell. Proteomics 6,
1942–1951.
Goodlett, D. R., Keller, A., Watts, J. D., Newitt, R., Yi, E. C., Purvine, S., Eng, J. K., von
Haller, P., Aebersold, R., and Kolker, E. (2001). Rapid Commun. Mass Spectrom. 15,
1214–1221.
Goshe, M. B., Conrads, T. P., Panisko, E. A., Angell, N. H., Veenstra, T. D., and
Smith, R. D. (2001). Anal. Chem. 73, 2578–2586.
Grewal, S. I., and Rice, J. C. (2004). Curr. Opin. Cell Biol. 16, 230–238.
Griffin, T. J., Xie, H., Bandhakavi, S., Popko, J., Mohan, A., Carlis, J. V., and Higgins, L.
(2007). J. Proteome Res. 6, 4200–4209.
760 Adam R. Farley and Andrew J. Link

Gruhler, A., Olsen, J. V., Mohammed, S., Mortensen, P., Faergeman, N. J., Mann, M., and
Jensen, O. N. (2005). Mol. Cell. Proteomics 4, 310–327.
Gu, W., and Roeder, R. G. (1997). Cell 90, 595–606.
Gygi, S. P., Rist, B., Gerber, S. A., Turecek, F., Gelb, M. H., and Aebersold, R. (1999).
Nat. Biotechnol. 17, 994–999.
Haas, W., Faherty, B. K., Gerber, S. A., Elias, J. E., Beausoleil, S. A., Bakalarski, C. E.,
Li, X., Villen, J., and Gygi, S. P. (2006). Mol. Cell. Proteomics 5, 1326–1337.
Hansen, B. T., Jones, J. A., Mason, D. E., and Liebler, D. C. (2001). Anal. Chem. 73,
1676–1683.
Hansen, B. T., Davey, S. W., Ham, A. J., and Liebler, D. C. (2005). J. Proteome Res. 4,
358–368.
Hayduk, E. J., Choe, L. H., and Lee, K. H. (2004). Electrophoresis 25, 2545–2556.
Holt, G. D., and Hart, G. W. (1986). J. Biol. Chem. 261, 8049–8057.
Huang, Y., Triscari, J. M., Tseng, G. C., Pasa-Tolic, L., Lipton, M. S., Smith, R. D., and
Wysocki, V. H. (2005). Anal. Chem. 77, 5800–5813.
Hubbard, M. J., and Cohen, P. (1993). Trends Biochem. Sci. 18, 172–177.
Ikeda, F., and Dikic, I. (2008). EMBO Rep. 9, 536–542.
Issad, T., and Kuo, M. (2008). Trends Endocrinol. Metab. 19, 380–389.
Ito, S., Hayama, K., and Hirabayashi, J. (2009). Methods Mol. Biol. 534, 195–203.
Jenuwein, T., and Allis, C. D. (2001). Science 293, 1074–1080.
Kelleher, N. L., Lin, H. Y., Valaskovic, G. A., Aaserud, D. J., Fridriksson, E. K., and
McLafferty, F. W. (1999). J. Am. Chem. Soc. 121, 806–812.
Kirkpatrick, D. S., Denison, C., and Gygi, S. P. (2005). Nat. Cell Biol. 7, 750–757.
Lange, V., Picotti, P., Domon, B., and Aebersold, R. (2008). Mol. Syst. Biol. 4, 222.
Larsen, M. R., Sorensen, G. L., Fey, S. J., Larsen, P. M., and Roepstorff, P. (2001).
Proteomics 1, 223–238.
Larsen, M. R., Thingholm, T. E., Jensen, O. N., Roepstorff, P., and Jorgensen, T. J. (2005).
Mol. Cell. Proteomics 4, 873–886.
Lee, C. H., McComb, M. E., Bromirski, M., Jilkine, A., Ens, W., Standing, K. G., and
Perreault, H. (2001). Rapid Commun. Mass Spectrom. 15, 191–202.
Lim, K. B., and Kassel, D. B. (2006). Anal. Biochem. 354, 213–219.
Lochnit, G., and Geyer, R. (2004). Biomed. Chromatogr. 18, 841–848.
Lu, B., Motoyama, A., Ruse, C., Venable, J., and Yates, J. R. 3rd (2008). Anal. Chem. 80,
2018–2025.
Ma, Y., Lu, Y., Zeng, H., Ron, D., Mo, W., and Neubert, T. A. (2001). Rapid Commun.
Mass Spectrom. 15, 1693–1700.
MacCoss, M. J., McDonald, W. H., Saraf, A., Sadygov, R., Clark, J. M., Tasto, J. J.,
Gould, K. L., Wolters, D., Washburn, M., Weiss, A., Clark, J. I., and Yates, J. R. 3rd
(2002). Proc. Natl. Acad. Sci. USA 99, 7900–7905.
Macek, B., Waanders, L. F., Olsen, J. V., and Mann, M. (2006). Mol. Cell. Proteomics 5,
949–958.
Makarov, A. (2000). Anal. Chem. 72, 1156–1162.
Makarov, A., Denisov, E., Kholomeev, A., Balschun, W., Lange, O., Strupat, K., and
Horning, S. (2006). Anal. Chem. 78, 2113–2120.
Mann, M., and Kelleher, N. L. (2008). Proc. Natl. Acad. Sci. USA 105, 18132–18138.
Manza, L. L., Codreanu, S. G., Stamer, S. L., Smith, D. L., Wells, K. S., Roberts, R. L., and
Liebler, D. C. (2004). Chem. Res. Toxicol. 17, 1706–1715.
McAlister, G. C., Berggren, W. T., et al. (2008). J. Proteome. Res. 7, 3127–3136.
McAlister, G. C., Phanstiel, D., et al. (2007). Anal. Chem. 79, 3525–3534.
McLachlin, D. T., and Chait, B. T. (2001). Curr. Opin. Chem. Biol. 5, 591–602.
McLachlin, D. T., and Chait, B. T. (2003). Anal. Chem. 75, 6826–6836.
Mechref, Y., Madera, M., and Novotny, M. V. (2008). Methods Mol. Biol. 424, 373–396.
Protein Posttranslational Modifications 761

Medzihradszky, K. F. (2005). Methods Enzymol. 405, 116–138.

Medzihradszky, K. F., Campbell, J. M., Baldwin, M. A., Falick, A. M., Juhasz, P.,
Vestal, M. L., and Burlingame, A. L. (2000). Anal. Chem. 72, 552–558.
Melchior, F., Schergaut, M., and Pichler, A. (2003). Trends Biochem. Sci. 28, 612–618.
Meng, F., Cargile, B. J., Patrie, S. M., Johnson, J. R., McLoughlin, S. M., and
Kelleher, N. L. (2002). Anal. Chem. 74, 2923–2929.
Mikesh, L. M., Ueberheide, B., Chi, A., Coon, J. J., Syka, J. E., Shabanowitz, J., and
Hunt, D. F. (2006). Biochim. Biophys. Acta 1764, 1811–1822.
Miyagi, M., and Rao, K. C. (2007). Mass Spectrom. Rev. 26, 121–136.
Oda, Y., Huang, K., Cross, F. R., Cowburn, D., and Chait, B. T. (1999). Proc. Natl. Acad.
Sci. USA 96, 6591–6596.
Oda, Y., Nagasu, T., and Chait, B. T. (2001). Nat. Biotechnol. 19, 379–382.
Old, W. M., Meyer-Arendt, K., Aveline-Wolf, L., Pierce, K. G., Mendoza, A.,
Sevinsky, J. R., Resing, K. A., and Ahn, N. G. (2005). Mol. Cell. Proteomics 4, 1487–1502.
Olsen, J. V., and Mann, M. (2004). Proc. Natl. Acad. Sci. USA 101, 13417–13422.
Olsen, J. V., Blagoev, B., Gnad, F., Macek, B., Kumar, C., Mortensen, P., and Mann, M.
(2006). Cell 127, 635–648.
Olsson, A., Manzl, C., Strasser, A., and Villunger, A. (2007). Cell Death Differ. 14, 1561–1575.
Ong, S. E., and Mann, M. (2005). Nat. Chem. Biol. 1, 252–262.
Ong, S. E., Blagoev, B., Kratchmarova, I., Kristensen, D. B., Steen, H., Pandey, A., and
Mann, M. (2002). Mol. Cell. Proteomics 1, 376–386.
Ong, S. E., Mittler, G., and Mann, M. (2004). Nat. Methods 1, 119–126.
Pan, C., Gnad, F., Olsen, J. V., and Mann, M. (2008). Proteomics 8, 4534–4546.
Paoletti, A. C., and Washburn, M. P. (2006). Biotechnol. Genet. Eng. Rev. 22, 1–19.
Parks, B. A., Jiang, L., Thomas, P. M., Wenger, C. D., Roth, M. J., Boyne, M. T. 2nd,
Burke, P. V., Kwast, K. E., and Kelleher, N. L. (2007). Anal. Chem. 79, 7984–7991.
Patton, W. F. (2002). J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 771, 3–31.
Pedrioli, P. G., Raught, B., Zhang, X. D., Rogers, R., Aitchison, J., Matunis, M., and
Aebersold, R. (2006). Nat. Methods 3, 533–539.
Peng, J., Schwartz, D., Elias, J. E., Thoreen, C. C., Cheng, D., Marsischky, G., Roelofs, J.,
Finley, D., and Gygi, S. P. (2003). Nat. Biotechnol. 21, 921–926.
Pflieger, D., Junger, M. A., Muller, M., Rinner, O., Lee, H., Gehrig, P. M., Gstaiger, M.,
and Aebersold, R. (2008). Mol. Cell. Proteomics 7, 326–346.
Pickart, C. M., and Fushman, D. (2004). Curr. Opin. Chem. Biol. 8, 610–616.
Pinkse, M. W., Uitto, P. M., Hilhorst, M. J., Ooms, B., and Heck, A. J. (2004). Anal. Chem.
76, 3935–3943.
Pitteri, S. J., Chrisman, P. A., Hogan, J. M., and McLuckey, S. A. (2005). Anal. Chem. 77,
1831–1839.
Pless, D. D., and Lennarz, W. J. (1977). Proc. Natl. Acad. Sci. USA 74, 134–138.
Polevoda, B., and Sherman, F. (2000). J. Biol. Chem. 275, 36479–36482.
Ramos-Fernandez, A., Lopez-Ferrer, D., and Vazquez, J. (2007). Mol. Cell. Proteomics 6,
1274–1286.
Rao, K. C., Palamalai, V., Dunlevy, J. R., and Miyagi, M. (2005). Mol. Cell. Proteomics 4,
1550–1557.
Reid, G., Gallais, R., and Metivier, R. (2009). Int. J. Biochem. Cell Biol. 41, 155–163.
Rodriguez, M. S., Dargemont, C., and Hay, R. T. (2001). J. Biol. Chem. 276, 12654–12659.
Ross, P. L., Huang, Y. N., Marchese, J. N., Williamson, B., Parker, K., Hattan, S.,
Khainovski, N., Pillai, S., Dey, S., Daniels, S., Purkayastha, S., Juhasz, P., et al. (2004).
Mol. Cell. Proteomics 3, 1154–1169.
Roth, S. Y., Denu, J. M., and Allis, C. D. (2001). Annu. Rev. Biochem. 70, 81–120.
Rush, J., Moritz, A., Lee, K. A., Guo, A., Goss, V. L., Spek, E. J., Zhang, H., Zha, X. M.,
Polakiewicz, R. D., and Comb, M. J. (2005). Nat. Biotechnol. 23, 94–101.
Sadoul, K., Boyault, C., Pabion, M., and Khochbin, S. (2008). Biochimie 90, 306–312.
762 Adam R. Farley and Andrew J. Link

Schmidt, A., Kellermann, J., and Lottspeich, F. (2005). Proteomics 5, 4–15.

Schroeder, M. J., Shabanowitz, J., Schwartz, J. C., Hunt, D. F., and Coon, J. J. (2004). Anal.
Chem. 76, 3590–3598.
Schulenberg, B., Aggeler, R., Beechem, J. M., Capaldi, R. A., and Patton, W. F. (2003a).
J. Biol. Chem. 278, 27251–27255.
Schulenberg, B., Beechem, J. M., and Patton, W. F. (2003b). J. Chromatogr. B Analyt.
Technol. Biomed. Life Sci. 793, 127–139.
Schulenberg, B., Goodman, T. N., Aggeler, R., Capaldi, R. A., and Patton, W. F. (2004).
Electrophoresis 25, 2526–2532.
Scigelova, M., and Makarov, A. (2006). Proteomics 6(Suppl. 2), 16–21.
Sirard, J. C., Vignal, C., Dessein, R., and Chamaillard, M. (2007). PLoS Pathog. 3, e152.
Siuti, N., and Kelleher, N. L. (2007). Nat. Methods 4, 817–821.
Siuti, N., Roth, M. J., Mizzen, C. A., Kelleher, N. L., and Pesavento, J. J. (2006). J. Proteome
Res. 5, 233–239.
Sliter, D. A., Kubota, K., Kirkpatrick, D. S., Alzayady, K. J., Gygi, S. P., and
Wojcikiewicz, R. J. (2008). J. Biol. Chem. 283, 35319–35328.
Sobel, R. E., Cook, R. G., and Allis, C. D. (1994). J. Biol. Chem. 269, 18576–18582.
Sokolovsky, M., Riordan, J. F., and Vallee, B. L. (1967). Biochem. Biophys. Res. Commun.
27, 20–25.
Soufi, B., Gnad, F., Jensen, P. R., Petranovic, D., Mann, M., Mijakovic, I., and Macek, B.
(2008). Proteomics 8, 3486–3493.
Spiro, R. G. (2002). Glycobiology 12, 43R–56R.
Steen, H., Jebanathirajah, J. A., et al. (2006). Mol. Cell Proteomics 5, 172–181.
Steinberg, T. H., Pretty On Top, K., Berggren, K. N., Kemper, C., Jones, L., Diwu, Z.,
Haugland, R. P., and Patton, W. F. (2001). Proteomics 1, 841–855.
Steinberg, T. H., Agnew, B. J., Gee, K. R., Leung, W. Y., Goodman, T., Schulenberg, B.,
Hendrickson, J., Beechem, J. M., Haugland, R. P., and Patton, W. F. (2003). Proteomics
3, 1128–1144.
Stemmann, O., Zou, H., Gerber, S. A., Gygi, S. P., and Kirschner, M. W. (2001). Cell 107,
715–726.
Strahl, B. D., and Allis, C. D. (2000). Nature 403, 41–45.
Swaney, D. L., McAlister, G. C., and Coon, J. J. (2008). Nat. Methods 5, 959–964.
Swaney, D. L., McAlister, G. C., Wirtala, M., Schwartz, J. C., Syka, J. E., and Coon, J. J.
(2007). Anal. Chem. 79, 477–485.
Syka, J. E., Coon, J. J., Schroeder, M. J., Shabanowitz, J., and Hunt, D. F. (2004). Proc. Natl.
Acad. Sci. USA 101, 9528–9533.
Sykora, C., Hoffmann, R., and Hoffmann, P. (2007). Protein Pept. Lett. 14, 489–496.
Tagwerker, C., Flick, K., Cui, M., Guerrero, C., Dou, Y., Auer, B., Baldi, P., Huang, L.,
and Kaiser, P. (2006). Mol. Cell. Proteomics 5, 737–748.
Tang, W. H., Halpern, B. R., Shilov, I. V., Seymour, S. L., Keating, S. P., Loboda, A.,
Patel, A. A., Schaeffer, D. A., and Nuwaysir, L. M. (2005). Anal. Chem. 77, 3931–3946.
Taniguchi, C. M., Emanuelli, B., and Kahn, C. R. (2006). Nat. Rev. Mol. Cell. Biol. 7,
85–96.
Tao, W. A., Wollscheid, B., O’Brien, R., Eng, J. K., Li, X. J., Bodenmiller, B., Watts, J. D.,
Hood, L., and Aebersold, R. (2005). Nat. Methods 2, 591–598.
Thingholm, T. E., Jorgensen, T. J., Jensen, O. N., and Larsen, M. R. (2006). Nat. Protoc.
1, 1929–1935.
Thompson, A., Schafer, J., Kuhn, K., Kienle, S., Schwarz, J., Schmidt, G., Neumann, T.,
Johnstone, R., Mohammed, A. K., and Hamon, C. (2003). Anal. Chem. 75, 1895–1904.
Tian, Y., Zhou, Y., Elliott, S., Aebersold, R., and Zhang, H. (2007). Nat. Protoc. 2,
334–339.
Trelle, M. B., and Jensen, O. N. (2007). Expert Rev. Proteomics 4, 491–503.
Protein Posttranslational Modifications 763

Tsaprailis, G., Nair, H., Somogyi, A., Wysocki, V. H., Zhong, W., Futrell, J. H.,
Summerfield, S. G., and Gaskell, S. J. (1999). J. Am. Chem. Soc. 121, 5142–5154.
Udeshi, N. D., Shabanowitz, J., Hunt, D. F., and Rose, K. L. (2007). FEBS J. 274,
6269–6276.
Ulintz, P. J., Yocum, A. K., Bodenmiller, B., Aebersold, R., Andrews, P. C., and
Nesvizhskii, A. I. (2009). J. Proteome Res. 8, 887–899.
Unwin, R. D., Griffiths, J. R., Leverentz, M. K., Grallert, A., Hagan, I. M., and
Whetton, A. D. (2005). Mol. Cell. Proteomics 4, 1134–1144.
Unwin, R. D., Griffiths, J. R., and Whetton, A. D. (2009). Nat. Protoc. 4, 870–877.
Vestal, M. L., and Campbell, J. M. (2005). Methods Enzymol. 402, 79–108.
Waanders, L. F., Hanke, S., and Mann, M. (2007). J. Am. Soc. Mass Spectrom. 18,
2058–2064.
Washburn, M. P., Wolters, D., and Yates, J. R. 3rd. (2001). Nat. Biotechnol. 19, 242–247.
Wells, L., Vosseller, K., et al. (2002). Mol. Cell Proteomics 1, 791–804.
White, F. M. (2008). Curr. Opin. Biotechnol. 19, 404–409.
Williamson, B. L., Marchese, J., and Morrice, N. A. (2006). Mol. Cell Proteomics 5, 337–346.
Witze, E. S., Old, W. M., Resing, K. A., and Ahn, N. G. (2007). Nat. Methods 4, 798–806.
Wolf-Yadlin, A., Hautaniemi, S., Lauffenburger, D. A., and White, F. M. (2007). Proc. Natl.
Acad. Sci. USA 104, 5860–5865.
Wu, S. L., Jardine, I., et al. (2004). Rapid Commun. Mass Spectrom. 18, 2201–2207.
Wu, C. C., MacCoss, M. J., Howell, K. E., and Yates, J. R. 3rd (2003). Nat. Biotechnol. 21,
532–538.
Yang, Z., and Hancock, W. S. (2005). J. Chromatogr. A 1070, 57–64.
Yang, X. J., and Seto, E. (2008). Mol. Cell. 31, 449–461.
Yeo, W. S., Lee, S. J., Lee, J. R., and Kim, K. P. (2008). BMB Rep. 41, 194–203.
Zabrouskov, V., Senko, M. W., et al. (2005). J. Am. Soc Mass Spectrom. 16, 2027–2038.
Zappacosta, F., Collingwood, T. S., Huddleston, M. J., and Annan, R. S. (2006). Mol. Cell.
Proteomics 5, 2019–2030.
Zhan, X., and Desiderio, D. M. (2004). Biochem. Biophys. Res. Commun. 325, 1180–1186.
Zhang, R., and Regnier, F. E. (2002). J. Proteome Res. 1, 139–147.
Zhang, X., Herring, C. J., Romano, P. R., Szczepanowska, J., Brzeska, H.,
Hinnebusch, A. G., and Qin, J. (1998). Anal. Chem. 70, 2050–2059.
Zhang, K., Williams, K. E., Huang, L., Yau, P., Siino, J. S., Bradbury, E. M., Jones, P. R.,
Minch, M. J., and Burlingame, A. L. (2002). Mol. Cell. Proteomics 1, 500–508.
Zhang, H., Li, X. J., Martin, D. B., and Aebersold, R. (2003). Nat. Biotechnol. 21, 660–666.
Zhang, K., Yau, P. M., Chandrasekhar, B., New, R., Kondrat, R., Imai, B. S., and
Bradbury, M. E. (2004). Proteomics 4, 1–10.
Zhang, Y., Wolf-Yadlin, A., Ross, P. L., Pappin, D. J., Rush, J., Lauffenburger, D. A., and
White, F. M. (2005). Mol. Cell. Proteomics 4, 1240–1250.
Zhen, Y., Xu, N., Richardson, B., Becklin, R., Savage, J. R., Blake, K., and Peltier, J. M.
(2004). J. Am. Soc. Mass Spectrom. 15, 803–822.
 C H A P T E R F O R T Y- O N E

Parallel Methods for Expression

and Purification
Scott A. Lesley

Contents
1. Introduction 767
2. Strategies Based on End-Use 768
3. Parallel Cloning Strategies for Creating Expression Constructs 771
4. Small-Scale Expression Screening to Identify Suitable Constructs 774
5. Analytical Testing of Proteins for Selection 778
6. Large-Scale Parallel Expression 780
7. Conclusion 783
Acknowledgments 783
References 784

Abstract
Protein properties are highly diverse, making parallel expression and purifica-
tion a particular challenge. Parallel methods are typically used when a number
derivatives of a target protein are desired or when multiple homologs are
needed. A typical scenario involves target evaluation, cloning and mutagenesis
of the target, expression screening, large-scale expression and purification, and
analytical and biophysical testing of the resulting protein. This chapter
describes some of the strategies and methods employed for parallel protein
expression and purification.

1. Introduction
Parallel protein purification is mainly a challenge of logistics. There are
many purification approaches, which can be employed to isolate proteins.
Some of these are inherently more amenable to parallel processing than
others. (Graslund et al., 2008a; Kim et al., 2004; Lesley et al., 2002b) Most
parallel purification strategies employ the use of purification tags to facilitate
simple affinity purification with standardization of methods and protein

Genomics Institute of the Novartis Research Foundation, San Diego, California, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63041-X All rights reserved.

767
768 Scott A. Lesley

behavior. Although a single approach using a protease-cleavable his-tag

sequence is described in this protocol, other tags and even nontagged
approaches can be developed into parallel processes suitable for many
applications. Rather than try to be all inclusive, this article attempts to
describe a process currently in use in the author’s laboratory as a forum to
illustrate the key aspects of parallel purification. This process has been
applied to thousands of proteins to date and delivers protein suitable for
many applications.

2. Strategies Based on End-Use

Successful protein expression and purification typically requires the
evaluation of multiple expression constructs. Many iterative truncations,
and sometimes orthologous proteins, are required to identify a suitable
construct for recombinant expression. Making and characterizing these
variants is a significant barrier to success which requires the ability to
perform parallel expression screening and protein characterization. The
definition of success in this regard depends largely on the end-use applica-
tion. For use as a simple antigen, expression of sufficient amounts for
immunization, possibly even in aggregated form, is sufficient to claim
success. More often, demanding applications like biochemical studies or
protein structure require not only the production of a purified protein but
one which is properly folded, containing the appropriate cofactors or
posttranslational modifications and largely homogeneous in nature. Prior
to defining any purification strategy or evaluation process, the end-use must
be fully understood and protocols designed to achieve the required para-
meters. One of the most demanding end-use applications is crystallography.
Generally speaking, protein which is suitable for crystallization is also
sufficient for biochemical, immunological, and other functional studies.
Suitable protein should be properly folded (not just soluble), nonaggregated,
homogeneous, lacking unstructured regions and expressed at levels suffi-
cient for multiple crystallization trials. In most cases, achieving these criteria
is the result of screening multiple derivatives of targets and related targets
(Fig. 41.1). This requires parallel processing of purified proteins and the
parallel characterization of them.
A bioinformatic evaluation of the protein of interest should be per-
formed before any experimentation. With the vast amount of genome
sequence information available, a comparison of proteins with even little
to no functional information can provide useful insight used to guide
expression design. Software changes so rapidly that there is no point to
providing here a detailed method for analysis. Instead, some guidance will
be provided as a general approach. The first task, therefore, is to identify
 Truncation series of NP_663012.1 hypothetical protein from Chlorobium tepidum
Soluble Monodisperse
1 174 + +
5 174 + −
9 174 + +
13 174 + −
17 174 + −
21 174 −
25 174 −
29 174 −
33 174 −
37 174 −
41 174 −
45 174 −
49 174 −

1 170 + +
1 166 −
1 162 + +
1 158 + +
1 154 + +
1 150 −
1 146 −
1 142 −
1 138 −
1 134 −
1 130 − 2HR2
1 126 −

6 174 −
20 174 + −
6 162 + +
20 162 + +

Figure 41.1 Identifying suitable protein expression constructs through parallel protein expression, purification, and characterization. Nested
N- and C-terminal truncations were evaluated in parallel for suitability for crystallization trials. Proteins showing both soluble expression with
sufficient protein yield and monodisperse behavior by ANSEC are indicated in gray. One such construct was entered into crystallization trials
and yielded a high-resolution structure.
770 Scott A. Lesley

related proteins and align them to identify conserved regions. These regions
typically constitute the core structural domains which need to be retained
within expression constructs and can be used to identify domain boundaries.
The basic local alignment search tool (BLAST) (Altschul et al., 1990) is the most
common tool for simple retrieval of related sequences. The NCBI (http://blast.
ncbi.nlm.nih.gov/Blast.cgi?CMD¼Web&PAGE_TYPE¼BlastHome) pro-
vides many versions of this essential tool along with access to the majority of
the genome sequences available to query. A hidden Markov model (HMM)
can be developed for a more sophisticated search but a simple PSI-BLAST
query is usually sufficient to find the most highly related sequences which
are the most useful.
Most proteins have already been aligned and placed into families.
Pfam (http://pfam.sanger.ac.uk/) (Finn et al., 2008) is a database of these
alignments and provides much useful information on comparative studies of
related proteins including domain boundaries and sequence conservation.
UniProt (http://www.uniprot.org/) (Consortium, 2008) also provides a
useful source of summary data including secondary structure predictions
for most proteins. Ligands can also be a useful tool for stabilizing proteins
and activity testing. Ligand predictions may be obvious from annotations,
but several databases such as KEGG (http://www.genome.jp/kegg/)
(Kanehisa et al., 2008) and BioCyc (http://www.biocyc.org) (Karp et al.,
2005) can provide additional ligand suggestions. Finally, a simple search of
the literature can often identify related proteins which have been expressed
previously to give guidance for expression. While seemingly an obvious
task, the rush to get an expression construct into whatever system is readily
available often leads to cursory or no review of past publications. There are
many other useful and easily accessible bioinformatic tools available. The
basic advice is to know what is known about your protein and those which
are closely related and to use that information for your experimental design.
Having gathered the available information on the target of interest,
expression constructs can now be designed. The most successful approach
is to evaluate not only the full-length open reading frame (ORF), but also to
create and test multiple truncations of this ORF in parallel. Defining the
N- and C-terminal boundaries is critical. We have observed many cases of
one or a few amino acid differences at the ends completely altering protein
behavior. Even with information on sequence conservation, multiple
N- and C-termini should be tried.
1. Query existing genome sequences for related proteins using BLAST and
perform sequence alignments. Tools such as Pfam can provide a conve-
nient forum for these queries and provides prevetted alignments and
domain boundaries. Incorporate information from literature searches.
2. Identify locations of N- and C-terminal sequence conservation as initial
truncation boundaries. In some cases, internal domain boundaries may
Parallel Methods for Expression and Purification 771

be preferable. For example, kinase researchers typically study the cata-

lytic domain itself.
3. Starting from these initial boundaries select 5–10 additional N- and
C-termini. The final expression construct collection should be a matrix
of these termini. We have found that increments of approximately four
amino acids provides a useful range of termini without adding exces-
sively to the number of constructs to be screened. A 10  9 matrix
(90 constructs) is a convenient number to screen allowing for full-length
and empty vector controls.
4. Review the boundary selections with the end-use in mind making sure
to retain essential regions.

3. Parallel Cloning Strategies for Creating

Expression Constructs
There are many different approaches to cloning and mutagenesis.
Individual laboratories have their own preferences based on past experience.
Creating large numbers of unique expression clones can be daunting,
however. Some cloning approaches are inherently more amenable to paral-
lel processing. A common method utilizes the lambda Int/Xis/IHF recom-
bination at att sites to move ORFs into various recipient vectors (Hatley
et al., 2000). The commercial embodiment of this protocol, GatewayÒ , is
popular for creating expression vectors with varying tags but typically results
in extra amino acids which are encoded by the recombination sites.
T-vectors (Harrison et al., 1994), topoisomerase-linked vectors (Shuman,
1994), and cre-lox recombinatorial vectors (Liu et al., 1998) also are
amenable to parallel cloning projects, but are not particularly flexible and
typically require substantial effort to generate a custom vector. Two meth-
ods which have been used in our laboratory to generate thousands of
expression clones are ligation-independent-cloning (LIC) (Aslanidis and
de Jong, 1990) and polymerase incomplete primer extension (PIPE)
(Klock et al., 2008). In both cases these protocols provide a substantial
amount of flexibility for cloning ORFs into expression vectors, can be set
up to utilize common primers, are rapid and require a minimum of reagent
costs. The PIPE method will be described here since it also provides a
convenient protocol for creating truncations.
The PIPE method is based on the observation that normal PCR ampli-
fications result in mixtures of products which are not fully extended thereby
leaving the 50 ends of such products variably unpaired. This is similar to LIC
which utilizes a proofreading polymerase to remove nucleotides from the
30 -terminus. These unpaired 50 ends on the PCR products derive from
the synthetic amplification primers. Therefore, oligo design defines the
772 Scott A. Lesley

Vector PCR – can be prepared in advance, in bulk

5′ annealing site Negative selection 3′ annealing site

Insert PCR – 5' ends on insert primers include sequences complementary

to vector annealing sites

Mix PCR products – unpurified PCR products will anneal directionally

across the primer-encoded complementary sequences

Transform cells – nicks and gaps are repaired in vivo

Screen colonies for insert-containing clones by dPCR
1 2 3 4 5 6 7 8 9 10 11 12
A X X X X X X X X

B X X X X X

C X X X X X

D X X X X X X

E X X X X

F X X X X X X

G X X X X X X
H X X X X X X

Figure 41.2 Parallel cloning process. The PIPE cloning method allows for simple
parallel cloning of many expression constructs, including deletions and point mutations
for screening. Appropriately designed primers are used for vector amplification and
insert amplification. Resulting amplifiers are simply mixed and transformed. Isolated
colonies are screened by a dPCR amplification and detection of PCR product by
fluorescent dye. Typically, two colonies are screened for each attempted construct
which is sufficient to identify insert-containing clones (shaded X) for most targets.
Insert-containing clones are then sequenced for confirmation.

resulting 50 -overhangs which can be incorporated into a cloning strategy.

By simply preparing PCR products of inserts and vector with complemen-
tary 50 regions of approximately 15 bases, these products will anneal and can
be transformed to produce a stable expression plasmid (Fig. 41.2).
1. Design vector and insert primers with appropriate 15 base complemen-
tary regions (see Klock et al., 2008) for design criteria. The first 15 bases
Parallel Methods for Expression and Purification 773

on the 50 ends of the primers are designed to be directionally comple-

mentary such that the resultant PCR fragment(s) can anneal as desired
and become viable plasmids upon transformation.
2. Set up PCR reactions containing 5 ml of both 10 mM forward and reverse
primers, 5 ml of template (20–100 pg per PCR amplification) and 5 ml of
10 Cloned Pfu DNA Polymerase (Stratagene) Reaction Buffer, 2.5
units of PfuTurboÒ DNA polymerase, 1 ml of 10 mM dNTPs and water
to a volume of 50 ml.
3. Perform PCR by incubating the reaction at 95  C for 2 min, then
25 cycles of 95  C for 30 s, 55  C for 45 s and 68  C for 14 min, and
finally a 4  C hold.
4. Confirm successful amplifications by gel electrophoresis. PCR products
may be faint and smeared, but this is acceptable.
5. Thaw 2 ml aliquot of competent cells on ice for 10–15 min. Chill a
96-well microtiter plate on ice for 10–15 min. Transfer 2 ml of each of
the vector and insert PCR reactions into wells of a prechilled microtiter
plate and mix by pipetting. Dispense 20 ml of competent cells into each
well. Pipette up and down once to ensure DNA has mixed with the cells
and incubate for 15 min on ice.
6. Heat shock the cells by floating the microtiter plate in a 42  C water bath
for 45 s. Immediately return the microtiter plate to ice. Dispense 100 ml
of LB Broth (no antibiotic) into each well. Recover the transformed cells
by incubating at 37  C while shaking at 250 rpm for 1 h. Dispense 100 or
40 ml of the recovered cells into the respective wells of the selective LB
agar trays with glass beads. By hand, gently shake the trays enough to
evenly distribute the glass beads across the entire well. Invert the tray
to drop the glass beads off of the LB agar and onto the lid and then
remove the glass beads. Incubate the inverted trays 12–16 h at 37  C.
Isolated colonies can be screened by classic procedures such as restriction
mapping. Alternatively, large numbers of colonies can be screened by
diagnostic PCR (dPCR) or SYBR-PCR.
1. Dispense 200 ml of LB Broth with appropriate antibiotic into the wells of
flat-bottom 96-well plate. Using aseptic technique, pick and transfer 1–4
isolated colonies per transformation into unique wells of the microtiter
plate. Incubate the plate at 37  C while shaking at 250 rpm for 3–12 h.
2. Transfer 3 ml samples from each culture into a 96-well PCR plate.
Put the remainder of the cultures back into the shaking incubator to
continue growth for future glycerol stock archival.
3. Add 47 ml of master mix containing PCR reagents and a gene-specific
and a vector-specific primer to each well containing cells in the PCR
plate. Primers should be designed such that amplification only occurs
when the vector contains the desired insert. Amplify the DNA fragments
for 30 cycles using cycling conditions appropriate for the primers.
774 Scott A. Lesley

4. Dispense 50 ml of 10 mM EDTA in a flat-bottom 96-well plate. Transfer

5 ml of each PCR product into the wells and dispense 150 ml of SYBR
Buffer [40 ml of 10,000 SYBR Green I Dye (Sigma) is diluted in 20 ml
of 50 mM Tris–HCl (pH 8.0) to make a 20 solution].
5. Using an unamplified sample for a negative control, measure
fluorescence of each sample using a microtiter plate fluorescence reader.
Excitation: 485 nm. Emission: 525 nm.
6. SYBR results are determined on a relative scale with the positive wells
having at least fourfold higher fluorescence than the negatives or the
control. dPCR-positive reactions can then be confirmed by gel analysis
or direct sequencing procedures.

4. Small-Scale Expression Screening to Identify

Suitable Constructs
Small-scale evaluation of clones is an integral part of the clone selec-
tion process. Obtaining recombinant protein in good yield, which is in the
soluble fraction of a lysate, is often one of the first major hurdles encoun-
tered. However, all too often this single criterion dominates the decision
process leading to a false sense of success and much more work in the later
stages. Screening approaches such as reporter fusions (Lesley et al., 2002a;
Waldo et al., 1999) and colony blots (Martinez Molina et al., 2008) provide a
means to screen thousands of protein derivatives. Fusion to partners which
enhance solubility (Kapust and Waugh, 1999) is another strategy which has
been employed to define appropriate construct boundaries for solubility.
In many cases, these soluble proteins are aggregates and heterogeneous
mixtures of partially folded species (Fig. 41.3A and B). This becomes
problematic during purification, often resulting in poor yields and instabil-
ity, and also presents problems in obtaining high specific activity or crystal-
lizing protein. Partial proteolysis (Fig. 41.3D) can be used to identify stable
boundaries. The resulting stable fragments are identified via mass spectrom-
etry and can be used to define subsequent expression constructs. Addition of
ligands to the protein can have a stabilizing effect as well (Fig. 41.3C).
Ligands can be identified through screening methods such as Thermofluor
(Pantoliano et al., 2001) or predicted through annotation and experimenta-
tion. To the degree possible, additional protein analytics should be per-
formed at an early stage and the results factored into decisions for selecting
expression constructs for larger scale studies. With these caveats in mind,
expression screening should be performed at a scale to allow for such tests to
be implemented.
Small-scale expression screening is performed using deep-well microti-
ter plates as a convenient way to grow enough cells to evaluate protein
Parallel Methods for Expression and Purification 775

expression. Maximizing cell growth requires optimizing expression condi-

tions. In particular, aeration of cultures can prove limiting. Short-throw
plate shakers and gas-permeable plate sealers improve final yields. Care must
be taken at each step to minimize volume loss, as even small volume can
have a large impact on final yields.
1. Prepare an overnight culture of the desired clones in a 96 deep-well
block containing 850 ml of TB containing 2% glycerol and desired
antibiotic.
2. In a deep-well block, add 17 ml of overnight culture to 1.25 ml of
TB containing 2% glycerol and the desired antibiotic. Incubate 3 h at
37  C shaking at 900 rpm.

aa 1–195

aa 1–191

aa 1–187

aa 1–183

aa 1–195

aa 1–191

aa 1–187

0 2 4 6 8 10 12 min

Figure 41.3 (Continued)

776 Scott A. Lesley

C
100,000

90,000

80,000

70,000
Fluorescence units

60,000

50,000

40,000

30,000

20,000

10,000

0
0 20 40 60 80 100 120
Temperature (C)
D
1 MRGMMLGMLAETHIHSGAGRSEGFVDLPVA 30
31 REAVTSYPVIAGSSLKGALRDAARERGMDE 60
61 SIFGDQDRAGDVLVSDARLLLLPVRSLTGS 90
Uncleaved

91 YRWVTCPHILERLSRDMRLCGISDGFEGAS 120
Cleaved

121 VERGKACCTDDLNQIFLEEREFQRSNGIDG 150

151 ALIDALKKMVPHKQTASRLERQLVIISDDD 180
181 FGWFASYGLPVIARNKLDDNKKSKNLWYEE 210
211 ALAPDTLMYAMVFERKDGALGKVQSMFETK 240
241 PYLQLGGNETVGMGWFAVKILEQGEGR 267

31790.801
1 MRGMMLGMLAETHIHSGAGRSEGFVDLPVA 30
31 REAVTSYPVIAGSSLKGALRDAARERGMDE 60
18035.201 61 SIFGDQDRAGDVLVSDARLLLLPVRSLTGS 90
91 YRWVTCPHILERLSRDMRLCGISDGFEGAS 120
121 VERGKACCTDDLNQIFLEEREFQRSNGIDG 150
9017.601 151 ALIDALKKMVPHKQTASRLERQLVIISDDD 180
181 FGWFASYGLPVIARNKLDDNKKSKNLWYEE 210
211 ALAPDTLMYAMVFERKDGALGKVQSMFETK 240
241 PYLQLGGNETVGMGWFAVKILEQGEGR 267

Figure 41.3 Analytical profiling of proteins derived from parallel expression and purifica-
tion. (A) Capillary electrophoresis results of a truncation series of YP_958225.1. (B) ANSEC
profile of a truncation series of YP_958225.1 showing variation in the amount of soluble
monomer (white box) and soluble aggregate (gray box) between samples. (C) Thermofluor
ligand binding assay showing increased thermostability YP_164977.1 (threonine aldolase)
in the presence of pyridoxal phosphate (filled symbols) versus the apo enzyme. (D) Partial
trypsin proteolysis of YP_843077.1 shows stable fragments visible by SDS–PAGE (cleaved
lane). Mass spectrometry can be used to define stable fragments of the original protein. Two
masses (18032.201 and 9017.601 Da) correspond to peptide fragments (underlined) which
can be targeted for subsequent subcloning and further expression studies.
Parallel Methods for Expression and Purification 777

3. Add appropriate inducer for protein expression (i.e., 1 mM IPTG) and

continue incubation at 37  C with shaking for 5 h. Note the final
OD600 nm of the culture. Another popular method for expression
utilized autoinduction media (Studier, 2005) which does not require
addition of expression inducer.
4. Pellet cells at 3000 rpm for 15 min. Discard supernatant and freeze cells
overnight at  20  C. Freezing is important to facilitate cell lysis.
5. Thaw cells by incubating at room temperature briefly then refreeze the
pellet for 2 h at  20  C. After the second freeze, thaw cells in Extraction
Buffer (12 ml [300 kU] Ready Lyse [Epicentre] in 50 ml buffer [50 mM
Tris pH 7.5, 50 mM Sucrose, 1 mM EDTA]) using a proportion of
50 ml for every observed OD600 nm of the original expression culture.
Resuspend the pellet well using a multichannel pipette. Incubate 30 min
at room temperature.
6. After lysis the viscosity of the solution is often quite high and can prevent
efficient transfer of the lysate and protein capture on the resin. To reduce
viscosity, a Benzonase is used to degrade nucleic acids which contribute
the bulk of the viscosity to the lysate. Prepare Benzonase Solution by
diluting 1.5 ml [375 units] Benzonase [Sigma] in 50 ml of buffer [10 mM
Tris pH 7.5, 50 mM NaCl, 1 mM EDTA, 10 mM MgCl2]. Add
Benzonase Solution at 1.25 times the amount of Extraction Buffer
used in step 5. Seal the plate and mix by inversion approximately five
times or until the viscosity begins to reduce. Incubate an additional
15 min at room temperature. Pellet cells at 4000 rpm for 30 min.
7. Prepare a 96 deep-well plate containing 150 ml of a 50% slurry of affinity
resin (Nickel-NTA for his-tag purification) in Equilibration Buffer
(50 mM HEPES pH 8.0, 50 mM NaCl, 10 mM imidazole, 1 mM
TCEP). Add the supernatant from the centrifuged lysate to the resin
plate being careful to avoid the residual cell debris pellet. Seal the plate
and incubate with mixing for 60 min at 4  C.
8. Wet a 96 deep-well filter plate with 300 ml of Equilibration Buffer,
pulling sufficient vacuum to leave a thin layer of liquid above the frit.
Add the bound resin from the previous step and allow the supernatant
to flow-through the filter plate. Add 700 ml of Wash Buffer (50 mM
HEPES pH 8.0, 300 mM NaCl, 40 mM imidazole, 10% glycerol, 1 mM
TCEP) to each well and allow to drain. Centrifuge the filter plate at
200g for 1 min to remove the remaining Wash Buffer. Place a collec-
tion plate beneath the filter plate. Add 150 ml of Elution Buffer
(50 mM HEPES pH 8.0, 300 mM imidazole, 10% glycerol, 1 mM
TCEP) and incubate for 5 min. Centrifuge the filter plate at 200g for
5 min to collect the purified protein. It is important in this step to avoid
over-drying the resin which can lead to channeling (cracking) of the
resin bed and poor wash and elution of the resin.
778 Scott A. Lesley

5. Analytical Testing of Proteins for Selection

There are many methods for characterizing the quality and homoge-
neity of a protein prep. Data from these methods are needed to make
informed decisions about the suitability of a particular expression construct
or purification method. The protein consumption requirements for many of
these tests are substantial. However, in some cases miniaturized or abbre-
viated methods can be used at the level of small-scale screening to provide
useful protein characterization. Table 41.1 lists several analytical methods
which can be applied to small amounts of protein.
The production of soluble protein is the first step toward successful
purification. In many cases alteration of growth conditions such as reduced

Table 41.1 Analytical methods suitable for small-scale protein screens

Method Protein characteristics Comments

SDS–PAGE/CE MW and purity Capillary electrophoresis
methods are rapid, require
minimal protein volumes,
and provide electronic
records for ease of storage
and comparison
ANSEC Protein aggregation, Trade-offs of resolution versus
monodispersity, time and sample
oligomeric state
Protease stability Identification of Useful for defining domain
stable fragments, boundaries and formulation
proper folding
Thermal stability Proper folding, ThermofluorTM and
optimizing StargazerTM methods utilize
formulations, small amounts of protein
ligand binding relative to differential
scanning calorimetry
Protein Protein Various methods with different
concentration concentration sensitivities
UV/Vis Protein NanodropTM instrumentation
spectroscopy concentration, allows wavelength scans
cofactor/ligand using microliter volumes
predictions
Tryptophan Protein folding Useful for construct
fluorescence comparison
Parallel Methods for Expression and Purification 779

expression temperature or alteration of the expression construct such as

fusion to a highly soluble partner like maltose binding protein will greatly
improve the soluble yield of the target protein. However, these are often
soluble aggregates of the protein which have reduced activity and stability
relative to their properly folded counterparts. It is best to identify and avoid
these issues as early as possible. Analytical size exclusion chromatography
(ANSEC) is a convenient and predictive method which can be employed
in a parallel testing mode. By scaling-down column size and increasing
flow rates, a screening method with a throughput of 13–18 min per sample
can be achieved. By combining this approach with an HPLC autosampler, a
full plate of 96 purified proteins produced from the microscale protocol
can be screened in a day. A method is presented here for chromatography
in a 13.3 min run using an Agilent HP1100 with a refrigerated, 384-well
formatted autosampler. A slightly higher resolution method uses an 18 min
run, which may also be more gentle on very high molecular weight samples.
1. Equilibrate appropriate size exclusion column (Shodex 8  300 mm
Protein KW-803 column with 6  50 mm Protein KW-G guard
column) in 20 mM Tris pH 7.5, 200 mM NaCl, 0.25 mM TCEP,
3 mM NaN3. The flow rate should be 1.0–1.5 ml/min. This will take
about 30 min.
2. After the column is appropriately equilibrated and A280 baseline has
stabilized, 5–15 ml (depending on concentration) injections of sample
proteins are performed. Appropriate protein MW standards and
buffer blanks should be run prior to the sample sequence and after
approximately 50 samples.
3. A comparison of typical results is shown in Fig. 41.3A and B. Although
the rapid flow rate and relatively short column do not allow high-
resolution size determination, it is quite easy to see the relative propor-
tion of high molecular weight aggregates and the monomeric state of
the protein. Generally speaking, proteins which show a uniform and
monomeric profile are highly preferable to those containing aggregate,
even if the initial purification yield is smaller.
Selecting the correct construct from among many requires understand-
ing the end-use requirements. For example, protein crystallization requires
protein to be monodisperse, homogeneous, and lacking unstructured
regions. While many constructs might provide sufficient soluble protein
to enter into crystal trials, evaluation of SDS–PAGE, ANSEC, ligand
binding, and protease stability can effectively narrow the selection to the
preferred candidates. The more information that can be gathered at the
small-scale screening stage, the better the selection decisions that will
be made.
780 Scott A. Lesley

6. Large-Scale Parallel Expression

Large-scale parallel expression and purification of proteins often
requires some specialized equipment. Scale should be appropriate for the
end-use. One liter bacterial expression is typically enough for most labora-
tory needs. Simply utilizing multiple 1 l shake flasks is sufficient in most
cases, but is difficult to employ when pursuing 10s or 100s of targets.
An airlift fermentor (http://www.gnfsystems.com) provides a convenient
means to express up to 96 different proteins in parallel at a scale comparable
to typical 1 l shake flasks (Fig. 41.4) (Lesley, 2001). Bacterial cultures can be
grown to OD600 values of 25–30 resulting in a few grams of cell paste for
processing.
Lysis and purification of multiple cultures presents the biggest logistical
challenge to parallel processing at larger scale. Custom automation can
facilitate this effort but much can be accomplished by simplification of
the lysis and purification approach. A protocol is given here for purifying
his-tagged proteins containing a TEV cleavage site for tag removal.
We have found this approach to be reliable and easily scaled to many
proteins in parallel, even without the use of automation. Our laboratory
processes over 3000 proteins per year by this approach.
1. Cells containing expressed protein directly from fermentation are
incubated for 30 min at room temperature with 0.25 mg hen egg
white lysozyme per ml of culture, pelleted by centrifugation and the

Figure 41.4 (Continued)

Parallel Methods for Expression and Purification 781

Transfer pump T1

Buffer pump B1

Waste pump W1 Buffer pump B2

Buffer
Air regulator

Sample
tube Cleaning
station
Waste Solenoid
value
Toe fitting
Heat Microfluidizer
exchanger

Figure 41.4 Custom instrumentation for parallel expression and purification.

(A) GNFermentor (GNF Systems) provides high cell density parallel fermentation of
96 cultures simultaneously. (B) Schematic of an automated cell resuspension and lysis
apparatus. (C) Automated suspension and lysis apparatus provides parallel processing of
samples expressed in the GNFermentor.

resulting pellets frozen before use. For purification, pellets ( 3 g) are

thawed in 80 ml of Lysis Buffer (50 mM HEPES pH 8.0, 50 mM NaCl,
10 mM imidazole, 1 mM Tris(2-carboxyethyl)phosphine hydrochloride
[TCEP]). Each of the cell pellets is lysed as follows.
2. Homogenize the pellet and Lysis Buffer for approximately 80 s at
35,000 rpm (maximum setting) using a homogenizer (Omni Interna-
tional). Each cycle consists of 6 s with the probe near the bottom of the
homogenate followed by 3 s with the probe near the surface. These both
782 Scott A. Lesley

resuspend the pellet and effectively lyses the cells. If insufficient lysis is
observed, passage through a microfluidizer (Microfluidics) or French
press can be used.
3. Centrifuge the lysates for 30 min at 32,000g. The clarified lysates are
ready for affinity purification.
In parallel purification, conventional purification approaches can be a
significant bottleneck. Instrumentation and protocols for multisample pro-
cessing using HPLC instrumentation has been described (McMullan et al.,
2005). For most applications, however, an optimized affinity purification
will provide sufficient purity along with the requisite throughput. We have
found that a two-step nickel purification combined with a proteolytic
removal of the purification tag provides excellent purification and can be
performed in parallel for hundreds of samples at a diversity of scale. This
approach provides material that has proved suitable for crystallization trials
and enzymatic screening studies. Our vector contains a his-tag followed by a
TEV protease cleavage site. This protocol can be adapted for other tags and
cleavage enzymes.
1. Decant each clarified lysates onto two gravity-fed 1.5-ml nickel-chelat-
ing resin columns pre-equilibrated with Lysis Buffer. Collect each flow-
through into a waste tray.
2. Add 7.5 ml of Wash Buffer (50 mM HEPES pH 8.0, 300 mM NaCl,
40 mM imidazole, 10% glycerol, 1 mM TCEP). Add 0.5 ml of Elution
Buffer (20 mM HEPES pH 8.0, 300 mM imidazole, 10% glycerol, 1 mM
TCEP). This volume of elution buffer should be no more than one third
of the bed resin. The purpose is to push the elution front of the buffer to
the bottom of the resin to minimize the elution volume for the next step.
3. Place each column over a PD10 desalting column which has been pre-
equilibrated with Digestion Buffer (20 mM HEPES pH 8.0, 200 mM
NaCl, 40 mM imidazole, 1 mM TCEP). Add 2.5 ml of Elution Buffer to
the Ni-chelate resin and collect the elution directly onto the PD-10
column.
4. Elute the protein from the PD-10 column by further addition of 3.5 ml
of Digestion Buffer, collecting the elution in a 15 ml disposable tube.
Remove a small sample for protein assay and SDS–PAGE analysis.
5. The next step involves removal of the his-tag through proteolytic
cleavage. The protease itself is his-tagged. Therefore, subsequent passage
of the protein over a nickel resin will remove the protease, as well as
uncleaved proteins and those which stick nonspecifically to the nickel
resin while the protein of interest passes in the column flow-through.
For each 15 mg of protein obtained from the PD-10 elution, add 1 mg of
his-tagged TEV protease (Tropea et al., 2009). Bring the total volume of
the digestion to 9 ml using Digestion Buffer. Mix by inversion at room
temperature for 2 h or overnight at 4  C.
Parallel Methods for Expression and Purification 783

6. Pour each of the digested protein mixtures onto a gravity-fed 1.5-ml

nickel-chelating resin column pre-equilibrated with Digestion Buffer.
Collect each flow-through into a 50 ml conical tube. After the digestion
mixture has passed through the column, add an additional 1.5 ml of
Digestion Buffer to the resin to remove residual cleaved protein and
collect into the same tube as above. This 10.5 ml elution volume
contains the purified protein with the his-tag sequence removed.

7. Conclusion
Parallel protein expression and purification is necessary to adequately
evaluate multiple targets or multiple derivatives of single targets. The
methods described here utilize bacterial expression systems. These systems
are by far the most convenient and robust platforms for expression, but often
suffer problems when applied to mammalian proteins. In many cases,
focusing on domains of interest and systematically optimizing expression
constructs to identify the proper domain boundaries can result in excellent
expression levels of active proteins (Graslund et al., 2008b; Klock et al.,
2008). Identifying suitable domain boundaries is one of the best applications
of parallel expression and purification. For example, kinase catalytic
domains have been defined and expressed in bacteria with great success
and have contributed significantly to drug design (Marsden and Knapp,
2008). Despite the difficulties and tedium of mammalian and baculovirus
expression systems, these platforms can also be implemented in a parallel
processing way (Peng et al., 1993). Platforms such as BacMam (Kost and
Condreay, 2002) offer additional flexibility in exploring multiple expression
systems. Regardless of the expression platform, the key to successful protein
expression is to understand the ultimate requirements of the protein in
terms of its physical properties and activities. Too often, simple solubility
and total expression levels are used as the sole selection criteria when
evaluating expression options. Even very small changes to the domain
boundaries of the expression construct can cause drastic changes to protein
behavior without visibly changing apparent soluble yield (Klock et al.,
2008). Combining parallel expression and purification with parallel analyti-
cal techniques and data analysis provides the best opportunity to identify a
successful expression construct and purification combination.

ACKNOWLEDGMENTS
The protocols described here were developed through the contributions of many people.
The author would particularly like to acknowledge Heath Klock, Mark Knuth, Carol Farr,
Anna Grzechnik, Julie Feuerhelm, Daniel McMullan, and Dennis Carlton for their
contributions to this chapter.
784 Scott A. Lesley

REFERENCES
Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990). Basic local
alignment search tool. J. Mol. Biol. 215, 403–410.
Aslanidis, C., and de Jong, P. J. (1990). Ligation-independent cloning of PCR products
(LIC-PCR). Nucleic Acids Res. 18, 6069–6074.
Consortium, U. (2008). The universal protein resource (UniProt). Nucleic Acids Res. 36,
D190–D195.
Finn, R. D., Tate, J., Mistry, J., Coggill, P. C., Sammut, S. J., Hotz, H. R., Ceric, G.,
Forslund, K., Eddy, S. R., Sonnhammer, E. L., and Bateman, A. (2008). The Pfam
protein families database. Nucleic Acids Res. 36, D281–D288.
Graslund, S., Nordlund, P., Weigelt, J., Hallberg, B. M., Bray, J., Gileadi, O., Knapp, S.,
Oppermann, U., Arrowsmith, C., Hui, R., Ming, J., dhe-Paganon, S., et al. (2008a).
Protein production and purification. Nat. Methods 5, 135–146.
Graslund, S., Sagemark, J., Berglund, H., Dahlgren, L. G., Flores, A., Hammarstrom, M.,
Johansson, I., Kotenyova, T., Nilsson, M., Nordlund, P., and Weigelt, J. (2008b). The
use of systematic N- and C-terminal deletions to promote production and structural
studies of recombinant proteins. Protein Expr. Purif. 58, 210–221.
Harrison, J., Molloy, P. L., and Clark, S. J. (1994). Direct cloning of polymerase chain
reaction products in an XcmI T-vector. Anal. Biochem. 216, 235–236.
Hatley, J. L., Temple, G. F., and Brasch, M. A. (2000). DNA cloning using in vivo site-
specific recombination. Genome Res. 10, 1788–1795.
Kanehisa, M., Araki, M., Goto, S., Hattori, M., Hirakawa, M., Itoh, M., Katayama, T.,
Kawashima, S., Okuda, S., Tokimatsu, T., and Yamanishi, Y. (2008). KEGG for linking
genomes to life and the environment. Nucleic Acids Res. 36, D480–D484.
Kapust, R. B., and Waugh, D. S. (1999). Escherichia coli maltose-binding protein is uncom-
monly effective at promoting the solubility of polypeptides to which it is fused. Protein
Sci. 8, 1668–1674.
Karp, P. D., Ouzounis, C. A., Moore-Kochlacs, C., Goldovsky, L., Kaipa, P., Ahren, D.,
Tsoka, S., Darzentas, N., Kunin, V., and Lopez-Bigas, N. (2005). Expansion of the
BioCyc collection of pathway/genome databases to 160 genomes. Nucleic Acids Res. 33,
6083–6089.
Kim, Y., Dementieva, I., Zhou, M., Wu, R., Lezondra, L., Quartey, P., Joachimiak, G.,
Korolev, O., Li, H., and Joachimiak, A. (2004). Automation of protein purification for
structural genomics. J. Struct. Funct. Genomics 5, 111–118.
Klock, H. E., Koesema, E. J., Knuth, M. W., and Lesley, S. A. (2008). Combining the
polymerase incomplete primer extension method for cloning and mutagenesis with
microscreening to accelerate structural genomics efforts. Proteins 71, 982–994.
Kost, T. A., and Condreay, J. P. (2002). Recombinant baculoviruses as mammalian cell
gene-delivery vectors. Trends Biotechnol. 20, 173–180.
Lesley, S. A. (2001). High-throughput proteomics: Protein expression and purification in
the postgenomic world. Protein Expr. Purif. 22, 159–164.
Lesley, S. A., Graziano, J., Cho, C. Y., Knuth, M. W., and Klock, H. E. (2002a). Gene
expression response to misfolded protein as a screen for soluble recombinant protein.
Protein Eng. 15, 153–160.
Lesley, S. A., Kuhn, P., Godzik, A., Deacon, A. M., Mathews, I., Kreusch, A., Spraggon, G.,
Klock, H. E., McMullan, D., Shin, T., Vincent, J., Robb, A., et al. (2002b). Structural
genomics of the Thermotoga maritima proteome implemented in a high-throughput
structure determination pipeline. Proc. Natl. Acad. Sci. USA 99, 11664–11669.
Liu, Q., Li, M. Z., Leibham, D., Cortez, D., and Elledge, S. J. (1998). The univector
plasmid-fusion system, a method for rapid construction of recombinant DNA without
restriction enzymes. Curr. Biol. 8, 1300–1309.
Parallel Methods for Expression and Purification 785

Marsden, B. D., and Knapp, S. (2008). Doing more than just the structure-structural
genomics in kinase drug discovery. Curr. Opin. Chem. Biol. 12, 40–45.
Martinez Molina, D., Cornvik, T., Eshaghi, S., Haeggstrom, J. Z., Nordlund, P., and
Sabet, M. I. (2008). Engineering membrane protein overproduction in Escherichia coli.
Protein Sci. 17, 673–680.
McMullan, D., Canaves, J. M., Quijano, K., Abdubek, P., Nigoghossian, E., Haugen, J.,
Klock, H. E., Vincent, J., Hale, J., Paulsen, J., and Lesley, S. A. (2005). High-throughput
protein production for X-ray crystallography and use of size exclusion chromatography
to validate or refute computational biological unit predictions. J. Struct. Funct. Genomics 6,
135–141.
Pantoliano, M. W., Petrella, E. C., Kwasnoski, J. D., Lobanove, V. S., Myslik, J., Graf, E.,
Carver, T., Asel, E., Springer, B. A., Lane, P., and Salemme, F. R. (2001). High-density
miniaturized thermal shift assays as a general strategy for drug discovery. J. Biomol. Screen.
6, 429–440.
Peng, S., Sommerfelt, M., Logan, J., Huang, Z., Jilling, T., Kirk, K., Hunter, E., and
Sorscher, E. (1993). One-step affinity isolation of recombinant protein using the bacu-
lovirus/insect cell expression system. Protein Expr. Purif. 4, 95–100.
Shuman, S. (1994). Novel approach to molecular cloning and polynucleotide synthesis using
vaccinia DNA topoisomerase. J. Biol. Chem. 269, 32678–32684.
Studier, F. W. (2005). Protein production by auto-induction in high density shaking
cultures. Protein Expr. Purif. 41, 207–234.
Tropea, J. E., Cherry, S., and Waugh, D. S. (2009). Expression and purification of soluble
His(6)-tagged TEV protease. Methods Mol. Biol. 498, 297–307.
Waldo, G. S., Standish, B. M., Berendzen, J., and Terwilliger, T. C. (1999). Rapid protein-
folding assay using green fluorescent protein. Nat. Biotechnol. 17, 691–695.
 C H A P T E R F O R T Y- T W O

Techniques to Isolate O2-Sensitive

Proteins: [4Fe–4S]-FNR as an Example
Aixin Yan* and Patricia J. Kiley†

Contents
1. Introduction 788
2. Anaerobic Isolation of 4Fe-FNR 790
2.1. Principles and procedures for purifying O2-labile 4Fe-FNR 790
2.2. Protocols 794
2.3. Isolation of protein for special applications 797
3. Characterization of [4Fe–4S]2þ Cluster Containing FNR 799
3.1. Fe and sulfide analysis 800
3.2. ICP-MS analysis of the 57Fe content in the 57Fe
labeled 4Fe-FNR 801
3.3. UV–visible spectrophotometric analysis of the [4Fe–4S]2þ
cluster 802
3.4. Mössbauer spectroscopic analysis 802
3.5. Low-temperature EPR 802
4. Summary 803
References 803

Abstract
Many key enzymes in biological redox reactions require metal centers or
cofactors for optimum activity and function. While the metal centers provide
unique properties for protein structure and function, some also render protein
activity sensitive to environmental O2 and cause experimental challenges to
isolation and biochemical analysis. Iron–sulfur (Fe–S) clusters represent an
important class of such metal centers and Fe–S proteins are widely distributed
in nature. Here, we utilize FNR, a regulatory Fe–S protein from Escherichia coli,
as an example to describe the techniques essential to purifying O2-labile
proteins and summarize various approaches for their biochemical analysis.
These methods can be readily adapted to purify other O2-labile proteins and
advance our understanding of this interesting class of proteins.

* School of Biological Sciences, The University of Hong Kong, Hong Kong, Hong Kong SAR
{
Department of Biomolecular Chemistry, University of Wisconsin, Madison, Wisconsin, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63042-1 All rights reserved.

787
788 Aixin Yan and Patricia J. Kiley

1. Introduction
Many key proteins in biology contain O2-labile metal centers or
cofactors. Such proteins are found widely distributed in various physiologi-
cal processes in nature, such as oxidoreductases in the life-sustaining process
of respiration or photosynthesis, nitrogenases for N2 fixation, dehydratases
in metabolic pathways, and DNA helicases in DNA replication and repair,
etc. (Beinert et al., 1997; Brzoska et al., 2006; Kiley and Beinert, 2003;
Meyer, 2008). While the presence of such metal centers or cofactors
provides extraordinary diversity in protein structure and function, it may
also render the activity of a protein sensitive to oxidants, such as atmo-
spheric oxygen (O2) or its reactive oxygen species. These properties present
technical challenges in isolating proteins in their active state and in
performing subsequent in vitro analyses.
Creating and maintaining an anaerobic environment throughout the
entire process of protein purification is the key to obtain active O2-labile
proteins. Successful methods for isolating O2-sensitive proteins in their
active state were first reported in the 1960s and employed sealed benchtop
glove boxes in which O2 was removed by evacuating and flushing with an
inert gas (Chase and Rabinowitz, 1968; Kajiyama et al., 1969; Vandecasteele
and Burris, 1970). This type of device, while significantly advancing the
study of O2-labile metalloproteins, was limited in its ability to maintain an
O2-free environment in procedures requiring sequential transfers of mate-
rials into the chamber. These early glove boxes were also limited in the type
of equipment that could be enclosed in a chamber, such as those required
for centrifugation, chromatography, spectroscopy, etc. As a result, the yield
of protein isolated was often low and the purified proteins often exhibited
heterogeneity, resulting in some ambiguity in the subsequent biophysical
and biochemical characterizations (Gotto and Yoch, 1982; Saari et al.,
1984). The invention and construction of the anaerobic laboratory chamber
in the late 1970s solved many of these problems by providing greater
flexibility in adapting many standard laboratory techniques for anaerobic
use (Gunsalus et al., 1980; Poston et al., 1971). Introduction of this device
has greatly advanced the isolation and characterization of O2-labile proteins
in the last several decades and has led to an exponential increase in our
knowledge of this class of proteins.
In this chapter, we describe the methods used in our laboratory to isolate
the O2-sensitive iron–sulfur (Fe–S) protein, FNR, emphasizing current
practices for creating and maintaining an O2-free environment for the
purification and study of O2-labile proteins. Fe–S proteins are an ancient
class of proteins that exist across all biological kingdoms (Beinert et al., 1997;
Brzoska et al., 2006; Johnson et al., 2005; Kiley and Beinert, 2003;
Purification of O2-Labile Proteins 789

Meyer, 2008) and many contain O2-sensitive Fe–S clusters. Fe–S proteins
contain tetrahedral iron usually coordinated with sulfur, provided either by
inorganic sulfide or the thiolates of cysteine residues in the polypeptide
backbone. The basic types of these metal centers include the mononuclear
[Fe(Cys)4] complex and the bi-, tri-, and tetranucleated [2Fe–2S], [3Fe–4S],
and [4Fe–4S] clusters (Meyer, 2008), although more complex cluster types
can be found in nitrogenases and hydrogenases. While the cluster types are
relatively simple in structure, the protein environment surrounding these Fe–S
clusters can be remarkably different, resulting in great variation in cluster
stability or the redox potential of Fe–S centers (Dey et al., 2007). In addition
to their well-known roles in catalyzing electron transfer reactions, the known
functions of Fe–S proteins have expanded to also include regulatory factors that
mediate transcriptional and translational control of gene expression (Cabiscol
et al., 2000; Crack et al., 2008; Kiley and Beinert, 2003). Examples include the
well-characterized c-aconitase in eukaryotes, which binds specific RNA
sequences (iron regulatory elements) when it is in its apoform (the cytosolic
apoform of c-aconitase is also called iron regulatory protein, IRP1) and
controls protein translation in response to iron status (Kennedy et al., 1992),
and several bacterial transcription factors, such as FNR, IscR, and SoxR. FNR
and perhaps IscR appear to exploit the stability of their Fe–S clusters to sense
environmental signals and regulate the expression of a large number of genes in
E. coli ( Johnson et al., 2005; Kiley and Beinert, 2003).
In E. coli, FNR is a global transcription factor that mediates the transition
from an aerobic to anaerobic lifestyle by altering the expression profile of
hundreds of genes under anaerobic growth conditions (Constantinidou
et al., 2006; Grainger et al., 2007; Kang et al., 2005; Salmon et al., 2003).
The ability of FNR to directly and rapidly sense O2 levels is attributed to its
[4Fe–4S]2þ cluster that is ligated to FNR through four essential cysteine
residues: Cys20, Cys23, Cys29, and Cys120 (Khoroshilova et al., 1995).
This form of the protein is referred to as 4Fe-FNR. Under anaerobic
conditions, the [4Fe–4S]2þ cluster promotes FNR dimerization and
subsequent site-specific DNA binding and transcription regulation
(Lazazzera et al., 1993, 1996). Upon a switch to aerobic conditions, O2
rapidly reacts with the [4Fe–4S]2þ cluster and converts it to a [2Fe–2S]2þ
cluster. The resulting protein is referred to as 2Fe-FNR. The 2Fe-FNR is
unable to dimerize, and thus is inactive in site-specific DNA binding and
transcription regulation (Khoroshilova et al., 1997; Sutton et al., 2004).
Since the conversion from a [4Fe–4S]2þ cluster to a [2Fe–2S]2þ cluster by
O2 is sufficient to inactivate FNR in vivo and in vitro, isolation of the natively
active FNR must be carried out exclusively under anaerobic conditions.
Here, we describe the techniques and protocols necessary to isolate
native FNR protein in high yield and cluster occupancy that is suitable
for kinetic and various spectroscopic studies. We also summarize various
790 Aixin Yan and Patricia J. Kiley

approaches to characterize the purified 4Fe-FNR in the absence of O2. We

expect that these techniques and approaches will be broadly applicable to
other O2-labile proteins and promote further characterization of this class of
proteins.

2. Anaerobic Isolation of 4Fe-FNR

2.1. Principles and procedures for purifying O2-labile 4Fe-FNR
2.1.1. General preparation for achieving and maintaining
an oxygen-free environment
Prior to protein purification, common procedures must be adapted to create
and maintain an O2-free environment for all the chemicals, solutions, and
apparatus used during protein purification (Sutton and Kiley, 2003). This
can be generally achieved by construction of a sparging station and use of an
anaerobic glove-box vinyl chamber (Coy Laboratory Products, Model A).
A sparging station is usually composed of an argon or nitrogen gas tank, an
apparatus to provide copper-scrubbed (Sargent-Welch furnace) gas, and a
custom-built manifold (Fig. 42.1). The sparging system is used to remove
trace amount of O2 from chemicals by dispersing inert gas into butyl
rubber-sealed glass vials which contain O2-sensitive solid reagents [such as
dithionite (DTH)] or to remove dissolved O2 from solutions to be used for
protein isolation or characterization. For this purpose, the manifold is
equipped with multiple lines that are capped by either gas dispersion tubes

Gas dispersion head Needle

Figure 42.1 Diagram of a sparging station connected to an argon tank. A, argon tank;
B, copper-furnace apparatus; C, manifold attached with gas dispersion tubes or needles.
Purification of O2-Labile Proteins 791

(Upchurch Scientific) or needles (Fig. 42.1). Standard needles are also used
to vent gas during the sparging. The glove-box chamber provides sufficient
space for storage and operation during the process of protein purification.
The vinyl chamber is filled with a gas mixture composed of 80% N2, 10%
CO2, and 10% H2 such that any O2 that enters the chamber will be reduced
by H2 using a palladium catalyst and the generated H2O will be absorbed by
a desiccant. To purify proteins, FPLC or HPLC purification systems are
advantageous because the seals are resistant to O2 diffusion and the newer
designs are sufficiently compact to operate within a chamber (assuming
minimal heat production). In addition, their largely automated design
makes it easy to manipulate within the chamber. However, the design of
our older model FPLC requires that most parts including the pump,
column, and detection system are placed outside of the chamber and
connected to the buffer and fraction collector inside the chamber through
PEEK tubing (Upchurch Scientific) (Fig. 42.2), which is impermeable to
O2, so that the entire contents of the system remain anaerobic. The path of
solutions through the FPLC is from liquid containers located inside of the
chamber, to the pumps and/or column, located outside of the chamber, and
then back into the chamber to the fraction collector (Fig. 42.2). Most
solutions such as buffers are first sparged with argon gas for a minimum of
20 min before introduction into the anaerobic chamber via the airlock and
then equilibrated for at least 12 h in the chamber before use. All glassware
and plastic items are equilibrated in the anaerobic chamber for a minimum
of 24 h before being used to ensure any O2 inadvertently introduced into
the chamber is reduced.

2.1.2. Expression of FNR and assembly of [4Fe–4S] clusters

E. coli strain PK872 is used to obtain overexpression of 4Fe-FNR. This
strain is a derivative of PK22 (BL21(DE3) Dfnr, Dcrp) (Lazazzera et al., 1993).
It contains plasmid pPK823, carrying the fnr coding sequence cloned into

E B
C A

Figure 42.2 Diagram of an anaerobic chamber with an attached FPLC purification

system. A, air lock; B, anaerobic chamber; C, fraction collector; D, FPLC pumps and
controller (Pharmacia LKB Controller LCC-501 plus); E, column.
792 Aixin Yan and Patricia J. Kiley

the NdeI–BamHI sites of pET11a (Novagen) so that fnr transcription is

under the control of the T7 promoter, and translation is controlled from
the gene 10 ribosome-binding site. Cell growth and the synthesis/matura-
tion of 4Fe-FNR follows the procedures described by Sutton and Kiley
(2003). Typically, strain PK872 is grown aerobically with shaking at
200 rpm at 37  C in four 2-l flasks, each of which contains 1 l of M9-
minimal medium supplemented with 0.2% glucose, 0.2% casamino acids,
1 mM MgCl2, 100 mM CaCl2, 2 mg/ml thiamine, 20 mM ferric ammonium
citrate, and 50 mg/ml ampicillin. When the cell culture reaches an OD600 of
0.3 (Spectronic 20Dþ spectrophotometer), the expression of FNR is
induced with 0.4 mM isopropyl-b-D-thiogalactoside (IPTG) for 1 h. To
promote insertion of the [4Fe–4S]2þ cluster by the endogenous Fe–S cluster
machinery (Raulfs et al., 2008) into the apo-FNR that largely accumulates
during the induction period, these cultures are combined into a 9-l glass
carboy and sparged with argon overnight at 4  C. While we fortuitously
discovered that this process is more efficient than expression under steady
state anaerobic conditions, our recent discovery that the expression of both
Fe–S cluster biogenesis pathways is reduced under anaerobic conditions as
compared to aerobic conditions (Mettert et al., 2008) may provide some
insight for this observation. The glass carboy for sparging the cultures is
fitted with a two-hole stopper containing two pieces of glass tubing: the first
piece connects a line from an argon gas tank to a gas dispersion tube, which
reaches nearly to the bottom of the carboy, and the second short (3–4 in.)
piece permits gas release from the sparging vessel. After the overnight
sparging step, cells are quickly poured into a container of suitable size (4-l
beaker) to fit within the airlock and transferred immediately into the
anaerobic chamber, where the cells are dispensed into 500-ml centrifuge
bottles that have been equilibrated in the anaerobic chamber for at least
24 h. The centrifuge bottles are sealed tightly with lids containing O rings
(Beckman) before being removed from the anaerobic chamber. Cells are
pelleted by centrifugation (Beckman Avanti J-25 centrifuge, JLA10.500
rotor) at 4  C for 15 min at 8000 rpm. The centrifuge bottles are then
returned to the chamber, where the supernatant is poured off and the
process is repeated until all of the cells have been collected. At this point,
the anoxic cell pellets can either be stored at  80  C or used immediately
for cell lysis.

2.1.3. Preparation of cell lysate

Cell lysis is achieved by passage of an anaerobic cell suspension containing
reducing agents through a French press at 20,000 psi with the collection vial
continually gassed with argon, essentially as described by Sutton and Kiley
(2003). This step is the most technically challenging for maintaining anoxic
conditions. To minimize O2 exposure to the cells, a chilled French press cell
and the bottles containing the anoxic cell pellets are first brought into the
Purification of O2-Labile Proteins 793

anaerobic chamber through two cycles in the airlock. The French press cell
is immediately rinsed twice with anaerobic buffer. Typically, the cell pellet
is resuspended (to 0.5% of the original volume) in anaerobic buffer A
supplemented with dithionite (DTH) and dithiothreitol (DTT) (see
Table 42.1) and then pipetted immediately into the rinsed cold French
press cell. The French press cell containing the cell suspension is sealed and
the sample outlet tubing is fitted with a capped 18-gauge needle before the
entire cell is removed from the anaerobic chamber. Once the cell is
assembled onto the French press, the needle is inserted into a butyl rub-
ber-stoppered glass vial, previously sparged with argon gas, for collection of
the cell lysate. After a single pass through the French press at 20,000 psi, the
anaerobic lysate is collected into the sealed glass vial and is immediately
brought back into the anaerobic chamber, transferred into ultracentrifuge
tubes and sealed with a gas-tight lid before being removed from the
chamber. Cell debris is removed by ultracentrifugation (Beckman Optima
LE-80K, 70.l Ti rotor) at 4  C for 60 min at 45,000 rpm. The samples are
then returned to the chamber, where the supernatant is used as the source of
4Fe-FNR for protein isolation.

2.1.4. Protein purification

[4Fe–4S]-FNR is routinely purified using an FPLC (Pharmacia LCC-501
Plus system) equipped with a 5-ml BioRex-70 cation-exchange column
(BioRad Laboratories). Buffer solutions are located within the anaerobic
chamber and are pumped out to the external FPLC via PEEK tubing
through the chamber port. Prior to fractionating the cell lysate, the
BioRex-70 column is washed with anaerobic buffers B and A
(Table 42.1), containing DTT and DTH for 1 h each and equilibrated in
buffer A. The cell lysate containing FNR is first transferred to a superloop
and sealed before being brought outside of the anaerobic chamber and
connected with the sample loading loop. The 4Fe-FNR is eluted with a
gradient from 0.1 to 0.55 M KCl at a flow rate of 10 ml/h and is pumped

Table 42.1 Composition of buffers for purifying FNR

Buffer Composition
A (low salt buffer) 50 mM phosphate (pH 6.8), 10% glycerol,
0.1 M KCl
B (high salt buffer) 50 mM phosphate (pH 6.8), 10% glycerol,
1 M KCl
C (no salt buffer) 50 mM phosphate (pH 6.8), 10% glycerol
D (elution buffer) 50 mM phosphate (pH 6.8), 10% glycerol,
0.4 M KCl
E (buffer to dissolve DTH) 0.1 M Tris–HCl (pH 7.9)
794 Aixin Yan and Patricia J. Kiley

back into the chamber through PEEK tubing, which is connected to a

fraction collector located within the chamber. FNR is eluted at  0.4 M
KCl as measured with a conductivity meter. Because the [4Fe–4S]2þ cluster
absorbs visible light maximally at a wavelength of about 420 nm, 4Fe-FNR
displays a characteristic green color. The colored fractions are subsequently
pooled and concentrated by passing through a gravity-flow column
equipped with a 1-ml BioRex-70 cation-exchange resin. The column is
stored in anaerobic chamber and is equilibrated with anaerobic buffer A. In
addition, since the presence of reducing agents, DTT and DTH, interferes
with some subsequent biochemical characterizations of 4Fe-FNR, this step
is also used to remove these agents from the protein solution. The 4Fe-
FNR obtained from this method typically displays a dark green color and
contains  1.0 mM protein with a cluster occupancy of more than 80%.

2.2. Protocols
A step-by-step protocol for isolating 4Fe-FNR is summarized as follows1

Day 1
1. Inoculate 100 ml of Luria Broth containing 1 ml of 20% glucose and
50 ml of 200 mg/ml Ampicillin (Ap) with strain PK872 in the early
afternoon and grow overnight at 37  C.
2. Prepare and autoclave four 2-l flasks containing 1 l of 1 M9 (Miller,
1972) minimal media.
Day 2
1. To each 1-l flask of M9-minimal medium, add:
 10 ml 20% glucose
 10 ml 20% casamino acids
 2 ml vitamin B1 (thiamine) (2 mg/ml stock)
 1 ml MgSO4 (1 M stock)
 100 ml CaCl2 (1 M stock)
 1 ml ferric ammonium citrate (10 mg/ml stock, filter sterilized)
 250 ml Ap (200 mg/ml stock, filter sterilized)
2. For each of the four flasks, inoculate 10 ml of the overnight culture into
1 l of the complete M9 media. Grow at 37  C with shaking at  200 rpm
to an OD600 of 0.3. Induce FNR synthesis by adding 1 ml of 0.4 M
IPTG to each flask (final concentration is 0.4 mM ) and shaking for 1
additional hour.

1
This protocol was created and amended by several members in the Kiley group: Lazzaera, B., Khoroshilova,
K., Voet, K., Sutton, V., Moore, L., Mettert, E., Yan, A.
Purification of O2-Labile Proteins 795

3. Pool cells into a 9-l flask and sparge cells with argon gas at 4  C for 14–16 h.
4. Prepare solutions and glassware that will be used next day and bring
them into the anaerobic chamber: (the chamber is normally equipped
with standard pipetting devices and tips, a microcentrifuge, a vortex
machine and eppendorf tubes.)
(a) Solutions:
 Buffer A: prepare 600 ml, divide into a 500- and 100-ml screw-
capped glass bottles (recipe as shown in Table 42.1)
 Buffer B: prepare 600 ml, divide into a 500- and 100-ml screw-
capped glass bottles (recipe as shown in Table 42.1)
 Buffer C: 100 ml in a screw-capped glass bottle (recipe as shown in
Table 42.1)
 Buffer E: 100 ml in a screw-capped glass bottle (recipe as shown in
Table 42.1)
 Water: 100 ml bottle

Sparge solutions with argon gas for 20 min by adjusting the argon flow to
produce steady gas-flow through the custom-built sparging system as
described in 2.1 (Fig. 42.1) with the gas dispersion tubes (Upchurch Scientific)
placed at the bottom of the buffer solutions.
(b) Glassware and plastic items to be brought into anaerobic chamber:
 Two ultracentrifuge tubes and caps
 Fraction collector tubes
 Beaker for waste
 Superloop with piston and cap
 Glass vials (5–6 small, 2 medium) and butyl rubber caps
 Glass Pasteur pipettes
 Four to six centrifuge bottles and caps with O rings
 500 ml graduated cylinder
 2 and 4-l plastic beakers
 25 ml glass pipettes
 Pasteur pipette bulb
 Needle and tubing for French Press
Bring these items into anaerobic chamber through air lock, leave the
buffers uncapped and equilibrate for 12 h.
Day 3
1. Pour sparged cells from the 9-l flask to a plastic 4-l beaker and immedi-
ately cover it with Saran wrap and tape to seal. Bring the beaker into
the anaerobic chamber and aliquot 450 ml into each of the 500-ml
centrifuge bottles and seal the bottles with the lids.
2. Bring the bottles out of the anaerobic chamber and measure the weight
to balance. Unbalanced bottles are brought back to the anaerobic
chamber to adjust the weight. Spin at 8000 rpm for 15 min at 4  C.
796 Aixin Yan and Patricia J. Kiley

3. After centrifugation, bring the bottles back to the chamber and remove
the supernatant. Repeat the centrifugation step until all cells are cen-
trifuged. The cell pellet should have an army green color.
4. During the centrifugation steps, weigh 154 mg dithiothreitol (DTT) and
150 mg dithionite (DTH) into individual 1.5 ml eppendorf tubes and
seal tightly. Bring them into the chamber together with the bottles after
centrifugation.
5. In the chamber, add 1.5 ml buffer E to DTH and 1.0 ml H2O to DTT.
Add 300 ml of DTH (final concentration 1.7 mM) and 100 ml of DTT
(final concentration 1.0 mM) to the 100 ml bottle of buffer A. Resus-
pend the cells in 20 ml of buffer A containing 1.7 mM DTH and 1.0 mM
DTT. The cell suspension can either be used immediately for cell lysis or
transferred into a medium glass vial, sealed with a butyl rubber stopper,
frozen on dry ice and then removed from the chamber and stored in a
80  C freezer until ready for purification.
Day 4
1. To the 500 ml of buffers A and B, add 1.5 ml of DTH and 0.5 ml of
DTT (both prepared freshly as described above).
2. Place pump leads into the buffers such that buffer A is connected to
pump A and buffer B is connected to pump B.
3. Connect BioRex-70 column to FPLC system, making sure that there
are no air bubbles in the column.
4. Run FPLC system program to wash the column with 100% buffer B
(contained within the chamber) for 1 h at a flow rate of 0.17 ml/min
followed by 100% buffer A for another 1 h.
5. Bring the frozen cells into the chamber and thaw the cell suspension
under anaerobic conditions (do not thaw cells outside the chamber as O2
will be drawn into vial).
6. Bring French press cell into the chamber. Rinse the sample cell twice
with anaerobic buffer A containing DTT and DTH. Fill cell with the
resuspended cells, seal, and attach sample outlet with the tubing fitted
with a capped 18-gauge needle, and remove from chamber along with a
butyl rubber-capped glass vial for collection of cell lysate.
7. Assemble cell onto French Press located preferably in a cold room. To
exclude O2 from the lysed extract, insert the needle from the cell into
the collection vial and insert two additional needles into the vial, using
one to gently flow argon gas and the second to vent the gas. Lyse cells at
20,000 psi at 4  C.
8. Remove needles and bring the cell lysate into anaerobic chamber and
divide the lysate into the two ultracentrifuge tubes. Bring them out of
the chamber and weigh the tubes to balance before ultracentrifugation.
Unbalanced tubes are brought back into the chamber to adjust the weight.
9. Centrifuge the lysate with 70.1 Ti rotor for 60 min at 45,000 rpm, 4  C.
Purification of O2-Labile Proteins 797

10. After centrifuging, bring tubes into the chamber, pour supernatant into
superloop and seal. Fill the other end of the superloop with buffer A
and seal. Remove the superloop from the chamber and connect super-
loop to the FPLC system.
11. Run FPLC program to load the sample and initiate the appropriate
sequence of buffer solutions to allow FNR to elute from the BioRex-
70 column.
12. Program the fraction collector located in the anaerobic chamber to
collect the column eluate.
Day 5
Concentration and storage
1. Fractions containing 4Fe-FNR are easily recognized by their green
color. Pool these fractions into a medium sized flask and dilute 1:4
with buffer C.
2. Construct a 1-ml BioRex-70 gravity column and wash with 2 column
volumes of buffer B followed by 6 column volumes of buffer C.
3. Load protein solution onto column, wash with 2 column volumes of
buffer A followed by elution with 0.8 M KCl buffer (made from buffers
B and C). Collect only colored drops.
4. Aliquot the concentrated 4Fe-FNR into 0.5 ml eppendorf tubes, which
have had their caps removed, and place tubes into a small glass vial. The
glass vial is topped with butyl rubber stoppers and crimp sealed with
aluminum caps (Bellco). The protein samples stored in the glass vial are
immediately frozen in a dry ice-ethanol slurry, which is introduced into
the anaerobic chamber though the air lock. Samples can be safely stored
in a 80  C freezer for many months using this method. Before thaw-
ing, it is critical that the sealed glass vials be placed in dry ice-ethanol
slurry during transfer to the anaerobic chamber to prevent O2 from
being drawn into the vial while thawing.

2.3. Isolation of protein for special applications

2.3.1. Removal of trace RNAses from purified 4Fe-FNR for in vitro
transcription assays
After passage through the two BioRex-70 cation-exchange columns, the
purified FNR protein is more than 95% pure as estimated by SDS–PAGE.
However, these preparations contain sufficient RNases to degrade RNA
products in an in vitro transcription assay. RNase activity was subsequently
removed by separation via a size exclusion column (HR-12 Superose, GE
Healthcare) attached to a Beckman HPLC system (Mettert and Kiley,
2007). The entire HPLC system is housed within an anaerobic chamber
and is also fitted with PEEK tubing (Upchurch Scientific). Since FNR
isolated this way is to be used in in vitro transcription assays, all buffer
798 Aixin Yan and Patricia J. Kiley

solutions and fraction collection tubes were treated with diethylpyrocarbo-

nate (DEPC) and/or autoclaved. Purification by size exclusion chromatog-
raphy also yields  100% dimeric 4Fe-FNR as judged by sulfide analysis
(Moore and Kiley, 2001).

2.3.2. Purification of 57Fe labeled 4Fe-FNR for Mössbauer

spectroscopic analysis
Mössbauer spectroscopy is a powerful method that can both qualitatively and
quantitatively characterize Fe–S clusters (Beinert et al., 1997). It differenti-
ates Fe–S cluster types based on the hyperfine splitting pattern of Fe nuclear
energy levels, as well as energy required for the Fe nuclei to transit from the
ground to the excited states in the absorption or emission spectrum of g-rays.
57Fe is by far the most common element studied with Mössbauer spectros-

copy and because of the vital roles of Fe–S containing proteins in biological
systems, Mössbauer spectroscopy has been widely used to characterize this
class of proteins. In order to prepare 57Fe enriched 4Fe-FNR, we remove the
naturally abundant 56Fe from the growth media by passing the 1 M9 media
through a Chelex 100 (200–400 mesh, BioRad) column (40 g of resin per 1 l
of medium) and then supplement the growth media with 57Fe in the form of
ferrous ethylenediammonium sulfate. The protocol for preparing Fe-free
growth media and supplementation with 57Fe is as follows:
1. Prepare Chelex 100 column by placing 200 g of Chelex resin and
 200 ml H2O into a beaker. After incubating at room temperature
for 30 min, pour resin into a 250-ml gravity column. Wash the column
with 5 column volumes of H2O before use. To regenerate column, wash
with 2 column volumes of 1 N HCl followed by 5 column volumes of
H2O; then wash with 2 column volumes of 1 N NaOH and another
5 column volumes of H2O till the pH of the column eluate is  7.0.
2. Deferrate the growth media: Pass 5 l of 1 M9 media through the newly
prepared or regenerated Chelex 100 column. Collect the flow-through
and autoclave. After use, the column is stored in 0.5 M ammonium
sulfate.
3. Cell growth and isolation of FNR containing 57Fe labeled [4Fe–4S]2þ
cluster:
 Inoculate 100 ml of Luria Broth with strain PK872. Add 1 ml of 20%
glucose and 50 ml of 200 mg/ml Ampicillin to the medium and grow
overnight at 37  C.
 Centrifuge the overnight cell culture to remove the LB media. Add
100 ml of deferrated 1 M9 media to resuspend the cell pellet,
centrifuge at 6000 rpm at 4  C and decant the supernatant.
 Resuspend the pellet in 100 ml of deferrated 1 M9 media, and
inoculate 10 ml of cell suspension to 1 l of deferrated M9-minimal
medium containing the same supplements as described earlier, except
Purification of O2-Labile Proteins 799

that the iron is replaced with 15 ml 57Fe ferrous ethylenediammonium

sulfate (1 M stock). Cultures are grown as described for the isolation of
non-57Fe labeled FNR including sparging cells with argon to allow
the assembly of 57Fe labeled [4Fe–4S]2þ cluster into FNR.
 Since we have observed that once 57Fe is incorporated into 4Fe-FNR,
the exchange between 57Fe in the [4Fe–4S] cluster and 56Fe in
solution is unlikely to occur, buffers (A, B, and C) for subsequent
protein purification are not deferrated and the purification procedures
are the same as those for non-57Fe labeled 4Fe-FNR (section 2.2).

3. Characterization of [4Fe–4S]2þ Cluster

Containing FNR
Several assays are performed to characterize the purified 4Fe-FNR,
including determination of protein concentration, Fe and sulfide content,
cluster occupancy and the type, and oxidation state of the cluster. The purity
of each protein preparation is estimated by SDS–PAGE and is typically >95%.
The protein concentration is measured by the Coomassie Plus protein assay
reagent (Pierce) and corrected by dividing by a factor of 1.33, to standardize to
the protein concentration determined from amino acid analysis (The Protein
Facility Center of the Iowa State University). The Fe and sulfide content of the
protein is measured by the methods developed by Beinert and coworkers
(Beinert, 1983; Kennedy et al., 1984). The [4Fe–4S]2þ cluster content of the
purified protein is calculated from the moles of sulfide determined per mole of
FNR protein. It can also be determined by the absorbance of [4Fe–4S]2þ
cluster at 405 nm from the extinction coefficient of 16,125 M 1 cm 1 (Sutton
et al., 2004). The [4Fe–4S]2þ cluster occupancy is defined by the percentage of
FNR subunits containing a [4Fe–4S]2þ cluster. In our measurements of >100
FNR preparations, we have typically observed a slightly higher Fe concentra-
tion than the sulfide concentration, which may be due to the presence of a
small amount of contaminating iron present in the buffer solutions but not
bound in clusters. Therefore, we use the sulfide concentration to determine
the concentration of [4Fe–4S]2þ cluster in the protein preparation which is
unlikely to be introduced artificially.
While the [4Fe–4S]2þ cluster exhibits a characteristic visible absorption
spectrum (Khoroshilova et al., 1995), the cluster type and the oxidation state
of the Fe–S cluster can only be determined by Mössbauer spectroscopy
using 57Fe labeled FNR protein (Khoroshilova et al., 1997). Recently,
Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometry
has been used to identify the apo-, [2Fe–2S]-, and [4Fe–4S]-forms of
adenosine 50 -phosphosulfate (APS) reductase isolated from Mycobacterium
800 Aixin Yan and Patricia J. Kiley

tuberculosis (Carroll et al., 2005), although it is not yet clear how generally
applicable this method will be.

3.1. Fe and sulfide analysis2

Fe is determined by a rapid procedure described by Kennedy et al. (1984)
without ashing. The principle of the method is to reduce all the Fe in the
protein sample to the Fe2þ state and then add the Fe2þ-specific chelator
ferene (disodium salt of 5,50 -[3-(2-pyridyl)-1,2,4-triazine-5,6-diyl]bis-
2-furansulfonic acid, Sigma) to form the Fe2þ-ferene complex which has
a high extinction coefficient factor at 593 nm. Three solutions, a, b, and c,
are prepared prior to the measurement (Table 42.2) (Kennedy et al., 1984).
Protein dilution buffer (buffer D in Table 42.1) and semi-micro cuvettes
(1-cm path length) are introduced into the anaerobic chamber at least 1 day
in advance of the assay. Typically, 5 ml of the protein sample and 95 ml of
buffer D are added to the cuvette in the anaerobic chamber; the cuvette is
covered with parafilm and then removed from the anaerobic chamber.
100 ml of reagent a is then added to the cuvette, mixed by pipetting and
then followed by addition of 100 ml of reagent b, mixing and incubating the
parafilmed cuvette at 30  C for 15 min. Finally, 5 ml of reagent c is added,
mixed, and the absorption measured at 593 nm. The molecular absorption
of 1 ng atom of Fe is 0.119 with this assay.
Sulfide is determined by the semi-micro analysis method developed by
Beinert (1983) using the reagent N,N-dimethyl-p-phenylenediamine
(DMPD) which forms methylene blue with sulfide. In this method, several
preventative measures were adopted to avoid the inadvertent loss of S2 in
the form of H2S gas, a side reaction that often occurs in an environment of

Table 42.2 Composition of buffers for Fe analysis

Reagent Composition
a 1.35 g of sodium dodecyl sulfate dissolved in 30 ml of H2O and
mixed with 0.45 ml of saturated Fe-free sodium acetate
ba 270 mg of ascorbic acid and 9 mg of sodium metabisulfate
(Na2S2O5) dissolved in 5.6 ml of H2O and mixed with 0.4 ml of
saturated acetate
c 18 mg of Ferene in 1 ml of H2O
a
Reagent b is not stable and has to be remade about every 2 weeks or kept frozen.

2
The Fe and S analysis of our protein preparations were performed by our long-time collaborator and friend
Dr. Helmut Beinert who passed away on December 21, 2007.
Purification of O2-Labile Proteins 801

strong acid and/or with the presence of an oxidizing agent (such as FeCl3).
The preventative measures include: (1) perform the entire procedure in a
small glass tube with a 10-mm o.d. and a total volume of 1.7 ml that has a
conical bottom and is fitted with a stir bar (10  3  3 mm) and a glass
stopper; (2) use gentle mixing in steps where the solution is acidic or when an
oxidizing agent is added. This approach minimizes the agitation of the liquid
in the tube and the formation of new air–water interfaces which facilitates
the escape of H2S if it is formed; (3) add a few drops of phenolphthalein in the
solution to follow the efficiency of mixing by gentle mixing; (4) strongly acid
DMPD solution is added to the zinc hydroxide suspension using a micropi-
pet to underlay this heavy solution under the lighter one. The typical
procedure is as follows: prior to the measurement, the glass tubes and H2O
are brought into the anaerobic chamber to equilibrate for 24 h. Then 10 ml of
protein sample and 90 ml of H2O are added to each of the glass tubes, the cap
is added and the samples are removed from the anaerobic chamber. With the
tubes secured above a stir plate, 300 ml of freshly prepared 1% zinc acetate
[Zn(C2H3O2)2] is added and gently stirred. Then 15 ml of 12% NaOH is
added immediately and stirred vigorously until the schlieren of NaOH
disappears or the phenolphthalein color is homogeneous. After incubation
at room temperature for 5 min, 75 ml of 0.1% DMPD solution (in 5 N HCl)
is layered under the protein solution. The solution is gently stirred until only
the top 2-mm layer of liquid has undissolved zinc hydroxide or pink color.
Then 10 ml of 23 mM FeCl3 (in 1.2 N HCl) is added directly to the solution
and quickly stirred to achieve a homogeneous (colorless) solution. After
incubation for 30 min at 20–25  C, the tube is centrifuged for 15–20 min
until protein is packed. The supernatant is transferred to a cuvette and the
absorption values at 670, 710, and 750 nm are measured. The ratio of the
absorbance should be approximately A670:A750:A710 ¼ 3:2:1. The sulfide
content is determined from the extinction coefficient of 34,500 M 1 cm 1
at 670 nm (Beinert, 1983).

3.2. ICP-MS analysis of the 57Fe content in the 57Fe

labeled 4Fe-FNR
Inductively coupled plasma mass spectrometry (ICP-MS) is a type of mass
spectrometry that is highly sensitive for analyzing a range of metals and
capable of distinguishing isotopic speciation for ions of choice. It is an ideal
approach for analyzing the amount of 57Fe enrichment in various forms of
natural and laboratory samples. Our protein preparations are currently
analyzed with a Thermoscientific ELEMENT2 high resolution ICP-MS
housed at the Wisconsin State Laboratory of Hygiene. Typically, 50 ml of a
1-mM protein sample is transferred to a Fe-free tube, sealed and removed
from the anaerobic chamber and submitted for analysis. Protein isolation
802 Aixin Yan and Patricia J. Kiley

following the protocols mentioned above usually yields 80–90% of cluster

occupancy and  80% of the clusters are incorporated with 57Fe.

3.3. UV–visible spectrophotometric analysis

of the [4Fe–4S]2þ cluster
UV–visible spectra are routinely obtained employing a Lambda 2 UV–
visible spectrophotometer (Perkin-Elmer) (Sutton et al., 2004). Screw-
capped quartz cuvettes (1-cm path length) (Starna) are brought into the
anaerobic chamber, left uncapped, and equilibrated for 24 h. These cuvettes
are manufactured to seal with O-ring adapted screw caps to prevent intro-
duction of O2 and thus can be used in a standard spectrophotometer. In the
anaerobic chamber, a protein sample of the appropriate dilution is placed in
a cuvette, sealed with a screw cap and removed from the anaerobic cham-
ber. The sample is scanned from 250 to 700 nm to obtain a full absorption
spectrum of the protein and the absorbance at 405 nm is used to calculate
the concentration of [4Fe–4S]2þ cluster in the protein sample using the
extinction coefficient of 16,125 M 1 cm 1 (Sutton et al., 2004).

3.4. Mössbauer spectroscopic analysis

In order to further identify the type and oxidation state of the Fe–S cluster,
Mössbauer spectroscopic analysis is performed at the Department of Chem-
istry, Carnegie Mellon University (Khoroshilova et al., 1995, 1997; Popescu
et al., 1998). The Mössbauer cups, lids, and the tools to tighten the
Mössbauer cups are placed into anaerobic chamber 24 h before use.
Approximately 400 ml of protein solution is placed into a custom-built
Mössbauer cup, the lid is sealed and the entire cup is immediately frozen
in liquid nitrogen. Liquid nitrogen was carefully brought into the chamber
using the manual cycle on the air lock. After the frozen samples are removed
from the chamber, they are shipped for Mössbauer spectroscopic analysis.

3.5. Low-temperature EPR

Electron paramagnetic resonance (EPR) spectroscopy is a technique for
studying chemical species that have one or more unpaired electrons, such as
organic and inorganic free radicals or inorganic complexes possessing a
transition metal ion. It is useful for study of specific oxidations states of
Fe–S clusters, which contain unpaired electrons (Cammack and Cooper,
1993). As expected, the [4Fe–4S]2þ cluster of FNR is EPR silent, thus EPR
is not suitable to characterize the type and oxidation state of the Fe–S cluster
in the natively purified 4Fe-FNR protein. However, it is a powerful and
widely used tool to study other Fe–S clusters that contain paramagnetic
irons such as aconitase (Kennedy et al., 1984). Furthermore, the
Purification of O2-Labile Proteins 803

combination of Mössbauer and EPR allows the detection of Fe–S cluster

intermediates that occurs during cluster conversion or destruction of some
proteins. EPR samples are prepared similar to those for Mössbauer, except
that a special syringe is required to load the EPR tubes because of their long
neck. A 0.5-ml glass syringe equipped with a long needle tubing is used to
pipette  200 ml of  100 mM protein solution into the tube and then
quickly frozen in liquid nitrogen.

4. Summary
In summary, by adapting common procedures to those that can be
carried out in the absence of O2, we routinely isolate 4Fe-FNR from E. coli
with high purity and cluster occupancy. It is noted that special precautions
must be taken during the entire process of the protein purification, includ-
ing cell growth, protein isolation, and protein storage, in order to achieve a
satisfactory yield of this class of O2-labile proteins. The procedures we
routinely use are based upon well-established approaches for eliminating
O2 exposure and maintaining a reduced environment and thus should be
applicable for isolation and/or characterization of other O2-labile proteins.
However, modification of this approach may be required in order to
optimize expression of a different O2-labile protein, particularly, if it has a
different cofactor or if the cluster cannot be inserted posttranslationally as it
is with FNR (Sutton and Kiley, 2003). Other alternatives to studies of O2-
sensitive metal center proteins include purifying the apoform of the protein
under aerobic conditions and reconstituting the O2-labile centers through
appropriate in vitro systems (Crack et al., 2008) or use of O2-stable protein
variants (Bates et al., 2000; Kiley and Reznikoff, 1991). However, it is an
open question whether reconstituted Fe-S protein exhibits all of the same
properties of natively isolated protein (Saunders et al., 2008). Nonetheless,
there is no doubt that purification and characterization of more O2-labile
proteins and enzymes will help to further broaden our knowledge of this
interesting class of proteins.

REFERENCES
Bates, D. M., Popescu, C. V., Khoroshilova, N., Vogt, K., Beinert, H., Munck, E., and
Kiley, P. J. (2000). Substitution of leucine 28 with histidine in the Escherichia coli
transcription factor FNR results in increased stability of the [4Fe–4S]2þ cluster to oxygen.
J. Biol. Chem. 275, 6234–6240.
Beinert, H. (1983). Semi-micro methods for analysis of labile sulfide and of labile sulfide plus
sulfane sulfur in unusually stable iron–sulfur proteins. Anal. Biochem. 131, 373–378.
804 Aixin Yan and Patricia J. Kiley

Beinert, H., Holm, R. H., and Münck, E. (1997). Iron–sulfur clusters: Nature’s modular,
multipurpose structures. Science 277, 653–659.
Brzoska, K., Meczynska, S., and Kruszewski, M. (2006). Iron–sulfur cluster proteins:
Electron transfer and beyond. Acta Biochim. Pol. 53, 685–691.
Cabiscol, E., Tamarit, J., and Ros, J. (2000). Oxidative stress in bacteria and protein damage
by reactive oxygen species. Int. Microbiol. 3, 3–8.
Cammack, R., and Cooper, C. E. (1993). Electron paramagnetic resonance spectroscopy of
iron complexes and iron-containing proteins. Methods Enzymol. 227, 353–384.
Carroll, K. S., Gao, H., Chen, H., Leary, J. A., and Bertozzi, C. R. (2005). Investigation of
the iron–sulfur cluster in Mycobacterium tuberculosis APS reductase: Implications for sub-
strate binding and catalysis. Biochemistry 44, 14647–14657.
Chase, T. Jr., and Rabinowitz, J. C. (1968). Role of pyruvate and S-adenosylmethioine in
activating the pyruvate formate-lyase of Escherichia coli. J. Bacteriol. 96, 1065–1078.
Constantinidou, C., Hobman, J. L., Griffiths, L., Patel, M. D., Penn, C. W., Cole, J. A., and
Overton, T. W. (2006). A reassessment of the FNR regulon and transcriptomic analysis
of the effects of nitrate, nitrite, NarXL, and NarQP as Escherichia coli K12 adapts from
aerobic to anaerobic growth. J. Biol. Chem. 281, 4802–4815.
Crack, J. C., Le Brun, N. E., Thomson, A. J., Green, J., and Jervis, A. J. (2008). Reactions of
nitric oxide and oxygen with the regulator of fumarate and nitrate reduction, a global
transcriptional regulator, during anaerobic growth of Escherichia coli. Methods Enzymol.
437, 191–209.
Dey, A., Jenney, F. E. , Adams, M. W., Babini, E., Takahashi, Y., Fukuyama, K.,
Hodgson, K. O., Hedman, B., and Solomon, E. I. (2007). Solvent tuning of electro-
chemical potentials in the active sites of HiPIP versus ferredoxin. Science 318, 1464–1468.
Gotto, J. W., and Yoch, D. C. (1982). Purification and Mn2þ activation of Rhodospirillum
rubrum nitrogenase activating enzyme. J. Bacteriol. 152, 714–721.
Grainger, D. C., Aiba, H., Hurd, D., Browning, D. F., and Busby, S. J. (2007). Transcrip-
tion factor distribution in Escherichia coli: Studies with FNR protein. Nucleic Acids Res. 35,
269–278.
Gunsalus, R. P., Tandon, S. M., and Wolfe, R. S. (1980). A procedure for anaerobic column
chromatography employing an anaerobic Freter-type chamber. Anal. Biochem. 101,
327–331.
Johnson, D. C., Dean, D. R., Smith, A. D., and Johnson, M. K. (2005). Structure, function,
and formation of biological iron–sulfur clusters. Annu. Rev. Biochem. 74, 247–281.
Kajiyama, S., Matsuki, T., and Nosoh, Y. (1969). Separation of the nitrogenase system of
azotobacter into three components and purification of one of the components. Biochem.
Biophys. Res. Commun. 37, 711–717.
Kang, Y., Weber, K. D., Qiu, Y., Kiley, P. J., and Blattner, F. R. (2005). Genome-wide
expression analysis indicates that FNR of Escherichia coli K-12 regulates a large number of
genes of unknown function. J. Bacteriol. 187, 1135–1160.
Kennedy, M. C., Kent, T. A., Emptage, M., Merkle, H., Beinert, H., and Münck, E. (1984).
Evidence for the formation of a linear [3Fe-4S] cluster in partially unfolded aconitase.
J. Biol. Chem. 259, 14463–14471.
Kennedy, M. C., Mende-Mueller, L., Blondin, G. A., and Beinert, H. (1992). Purification
and characterization of cytosolic aconitase from beef liver and its relationship to the
iron-responsive element binding protein. Proc. Natl. Acad. Sci. USA 89, 11730–11734.
Khoroshilova, N., Beinert, H., and Kiley, P. J. (1995). Association of a polynuclear iron–
sulfur center with a mutant FNR protein enhances DNA binding. Proc. Natl. Acad. Sci.
USA 92, 2499–2503.
Khoroshilova, N., Popescu, C., Münck, E., Beinert, H., and Kiley, P. J. (1997). Iron–sulfur
cluster disassembly in the FNR protein of Escherichia coli by O2: [4Fe–4S] to [2Fe–2S]
conversion with loss of biological activity. Proc. Natl. Acad. Sci. USA 94, 6087–6092.
Purification of O2-Labile Proteins 805

Kiley, P. J., and Beinert, H. (2003). The role of Fe–S proteins in sensing and regulation in
bacteria. Curr. Opin. Microbiol. 6, 181–185.
Kiley, P. J., and Reznikoff, W. S. (1991). Fnr mutants that activate gene expression in the
presence of oxygen. J. Bacteriol. 173, 16–22.
Lazazzera, B. A., Bates, D. M., and Kiley, P. J. (1993). The activity of the Escherichia coli
transcription factor FNR is regulated by a change in oligomeric state. Genes Dev. 7,
1993–2005.
Lazazzera, B. A., Beinert, H., Khoroshilova, N., Kennedy, M. C., and Kiley, P. J. (1996).
DNA binding and dimerization of the Fe–S-containing FNR protein from Escherichia coli
are regulated by oxygen. J. Biol. Chem. 271, 2762–2768.
Mettert, E. L., and Kiley, P. J. (2007). Contributions of [4Fe–4S]-FNR and integration host
factor to FNR transcriptional regulation. J. Bacteriol. 189, 3036–3043.
Mettert, E. L., Outten, F. W., Wanta, B., and Kiley, P. J. (2008). The impact of O2 on the
Fe–S cluster biogenesis requirements of Escherichia coli FNR. J. Mol. Biol. 384, 798–811.
Meyer, J. (2008). Iron–sulfur protein folds, iron–sulfur chemistry, and evolution. J. Biol.
Inorg. Chem. 13, 157–170.
Miller, J. H. (ed.) (1972). In ‘‘Experiments in Molecular Genetics’’, Cold Spring Harbor
Laboratory Press, Plainview, NY.
Moore, L. J., and Kiley, P. J. (2001). Characterization of the dimerization domain in the
FNR transcription factor. J. Biol. Chem. 276, 45744–45750.
Popescu, C. V., Bates, D. M., Beinert, H., Münck, E., and Kiley, P. J. (1998). Mössbauer
spectroscopy as a tool for the study of activation/inactivation of the transcription regula-
tor FNR in whole cells of Escherichia coli. Proc. Natl. Acad. Sci. USA 95, 13431–13435.
Poston, J. M.,, Stadtman, T. C., and Stadtman, E. R. (eds.) (1971). In An Anaerobic
Laboratory, Academic Press, New York.
Raulfs, E. C., O’Carroll, I. P., Dos Santos, P. C., Unciuleac, M. C., and Dean, D. R. (2008).
In vivo iron–sulfur cluster formation. Proc. Natl. Acad. Sci. USA 105, 8591–8596.
Saari, L. L., Triplett, E. W., and Ludden, P. W. (1984). Purification and properties of the
activating enzyme for iron protein of nitrogenase from the photosynthetic bacterium
Rhodospirillum rubrum. J. Biol. Chem. 259, 15502–15508.
Salmon, K., Hung, S. P., Mekjian, K., Baldi, P., Hatfield, G. W., and Gunsalus, R. P.
(2003). Global gene expression profiling in Escherichia coli K12. The effects of oxygen
availability and FNR. J. Biol. Chem. 278, 29837–29855.
Saunders, A. H., Griffiths, A. E., Lee, K. H., Cicchillo, R. M., Tu, L., Stromberg, J. A.,
Krebs, C., and Booker, S. J. (2008). Characterization of quinolinate synthases from
Escherichia coli, Mycobacterium tuberculosis, and Pyrococcus horikoshii indicates that [4Fe–4S]
clusters are common cofactors throughout this class of enzymes. Biochemistry 47,
10999–11012.
Sutton, V. R., and Kiley, P. J. (2003). Techniques for studying the oxygen-sensitive
transcription factor FNR from Escherichia coli. Methods Enzymol. 370, 300–312.
Sutton, V. R., Mettert, E. L., Beinert, H., and Kiley, P. J. (2004). Kinetic analysis of the
oxidative conversion of the [4Fe–4S]2þ cluster of FNR to a [2Fe–2S]2þ cluster.
J. Bacteriol. 186, 8018–8025.
Vandecasteele, J. P., and Burris, R. H. (1970). Purification and properties of the constituents
of the nitrogenase complex from Clostridium pasteurianum. J. Bacteriol. 101, 794–801.
 C H A P T E R F O R T Y- T H R E E

Rethinking Your Purification

Procedure
Murray P. Deutscher

Contents
1. Introduction 809

1. Introduction
Every protein purification that you undertake should provide you not
only with purified material but also with considerable information about the
protein. Thus, during the course of purification you most likely will learn
about the stability of the protein under a variety of conditions, as well as about
its size, charge, and, perhaps, its affinity properties. You will have learned
whether the protein can be concentrated, diluted, dialyzed, or exposed to a
variety of agents. In addition, you may have prepared an antibody against the
protein, subjected it to a limited sequence analysis or mass spec analysis or
determined whether it has any covalent modifications. Finally, the sequence
information obtained from the purified protein may allow you to identify
the gene encoding it, opening up the possibility of overexpression and the
generation of mutants for physiological and mechanistic studies.
All of this information can be of great help in deciding whether you have
developed an optimal purification scheme. Obviously, in some cases you
may not care. However, if this protein is one you plan on studying in some
detail, and you can foresee many purifications ahead, a rapid and efficient
(meaning high purity and high yield) purification scheme can save you an
enormous amount of work in the long run. By taking advantage of what
you have learned about the protein, it generally should be possible to
streamline and optimize the procedure a great deal. There is a natural
tendency, especially after having spent many mouths learning to purify a

Department of Biochemistry and Molecular Biology, University of Miami School of Medicine, Miami,
Florida, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63043-3 All rights reserved.

809
810 Murray P. Deutscher

protein, to go with what you know works. Nevertheless, spending some

time rethinking your purification procedure will be a worthwhile exercise.
Clearly, in many situations, overexpressing and tagging the protein of
interest will be the simplest route to obtaining additional material. How-
ever, in some cases, depending on the organism, regulation of the gene,
physiological problems due to overexpression, or negative effects of the tag,
this may not be possible or easy to accomplish. While there may be ways to
circumvent these particular issues, carefully examining the details of your
purification protocol will usually provide benefits.
Some things you might want to consider, with relevant chapters to help
you in this process, are as follows: [chapters to be added later, when chapter
numbers known]
1. Am I using the best source of material?
(a) Is the source readily available in large quantities?
(b) Is the protein associated with a specific subcellular structure that
might be a better starting material?
(c) Would it be possible to develop a better source by cloning the gene
for this protein?
(d) Can the protein be overexpressed from the cloned gene?
2. Are all the steps in the purification scheme necessary and useful?
(a) Do I extract most of the available activity?
(b) Can I pool fractions more broadly at a particular step knowing
that newly included impurities can be removed at a subsequent
fractionation step?
(c) Do any steps lead to unnecessarily large losses or relatively poor
purification, and is there a special reason for including that step?
(d) If there are large losses at any step, do I know why, and can it be
prevented?
(e) Are all the steps cost effective? Would a cheaper purification
medium or procedure suffice?
(f) Can any time-consuming procedures be eliminated?
3. Are the various purification steps carried out in the most optimal
sequence?
(a) Are procedures best for larger amounts of material in the beginning
of the scheme, and those better for smaller amounts later in the
procedure?
(b) Can I avoid a concentration or dialysis step by changing the order of
steps? Is it necessary to concentrate a sample prior to a column to
which the protein binds?
(c) Are the solution conditions of the last step in the procedure com-
patible with storage of the protein, or is a solution change necessary?
Rethinking Your Purification Procedure 811

4. Should I introduce any new steps into the purification procedure?

(a) Might a new step that takes advantage of the protein’s binding
properties (i.e., affinity chromatography) be effective?
(b) Can the antibody I have prepared be of use for purification?
(c) Am I taking advantage of all the protein’s structural properties in
deciding on purification steps?
5. Is the scale of the purification appropriate to the planned uses of the
material?
(a) Can enough material for my needs be obtained in fewer steps by using
two-dimensional gel electrophoresis or immunoprecipitation?
6. Am I using the best assay for my protein considering speed, cost, and
accuracy? Is a high degree of accuracy necessary during the purification?
7. If I have not already done so, can I learn anything from the literature by
examining purification schemes for related proteins?
The answers to these questions will give you a good idea whether
modification of your purification procedure or storage conditions might
be warranted.
 C H A P T E R F O R T Y- F O U R

Important but Little Known

(or Forgotten) Artifacts
in Protein Biochemistry
Richard R. Burgess

Contents
1. Introduction 814
2. SDS Gel Electrophoresis Sample Preparation 814
2.1. Proteases at room temperature 815
2.2. Asp–Pro bond cleavage at 100
C 815
2.3. Keratin contamination in sample or SDS sample buffer 816
3. Buffers 816
3.1. Reducing agent can become an oxidizing agent 816
3.2. Contaminant in sulfonylethyl buffers 817
4. Chromatography 817
4.1. EDTA binding to anion-exchange columns 817
4.2. EDTA in sample can strip nickel from a Ni-chelate
affinity column 818
5. Protein Absorption During Filtration 818
6. Chemical Leaching from Plasticware 819
7. Cyanate in Urea 819
References 820

Abstract
Many subtle and sometimes obscure artifacts exist that can have major effects
on the outcomes of otherwise carefully performed experiments. This chapter
focuses on a few of these artifacts involved in processes such as SDS gel
sample preparation, buffers, Ni-chelate affinity and ion-exchange chromatogra-
phy, urea preparation, use of plasticware, and filtration of protein samples.

McArdle Laboratory for Cancer Research, University of Wisconsin–Madison, Madison, Wisconsin, USA

Methods in Enzymology, Volume 463 # 2009 Elsevier Inc.

ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)63044-5 All rights reserved.

813
814 Richard R. Burgess

1. Introduction
Many of us who are experienced in protein biochemistry and protein
purification are called on regularly to troubleshoot the problems of our
colleagues. I often find that the answer to their problem is a little known bit
of knowledge that they have never heard of, but which I have encountered
many times. Some of us teach courses in protein purification and add little
tidbits of information to our lectures that we feel will help our students
avoid mysterious problems with their experiments. Many of us review
multiple manuscripts on protein purification and occasionally find that
certain artifacts reappear again and again, even when the effect has been
seen and reported many years earlier. One of the problems is that while
these sorts of artifact are many times reported, they are seldom the subject of
the paper, but merely a sentence or paragraph buried deep in the paper or
figure legend where they are rarely seen by a wider audience.
Other times, the importance has not been sufficiently emphasized in the
original paper, and the novice scientist may have heard of the problem, but
has chosen to ignore it or has assumed it really is not a problem most of the
time. Finally, young scientists often assume that there is nothing of impor-
tance in literature older than a few years old and simply are unaware of issues
that were well known 10–20 years ago.
Here, I present a few such potential artifacts and how to avoid them.
I hope that some of this obscure and often forgotten information might help
readers. These are just a few examples that come to mind, or have made a
particularly strong impression on me, but I am sure that there are many,
many more that are equally if not more important in various subareas of
protein biochemistry.
If you are aware of other such artifacts or have tips, or cautions that you
want to share, I would love to hear from you about them (burgess@oncology.
wisc.edu).

2. SDS Gel Electrophoresis Sample Preparation

It is quite common to think you have a nearly pure protein and then
find multiple bands upon analysis on SDS polyacrylamide gels. In my
experience, there are three significant and common ways in which you
can actually make your protein look less pure than it really is. While many
scientists are quite aware of these problems, it still amazes me how many are
not. Some of this material was presented in a short article in the Novagen
Newsletter inNovations (Grabski and Burgess, 2001).
Important Artifacts in Protein Chemistry 815

2.1. Proteases at room temperature

In the early days of SDS gel electrophoresis, it was common to put
proteins into a sample buffer consisting of a buffer, glycerol to increase
sample density, SDS to denature most proteins, and 2-mercaptoethanol to
reduce disulfide bonds. It was soon recognized that many proteases have
disulfide bonds and are quite stable and can only be inactivated by heating
the sample to better denature and reduce the protease (Pringle, 1975). So, it
became standard to heat samples to 95–100
C for 5 min or more. It was
learned a bit more slowly that the heating must be done immediately, since
the SDS easily unfolds most proteins, including usually the protein of
interest, but not some proteases. As a result, if a sample is added to sample
buffer, but the heating occurs sometime later, then the protease will have time to digest
the protein of interest. To see if this is a problem in your case, a simple
experiment is to add a partially purified protein sample to two portions of
sample buffer, mix, and immediately heat one. Leave the other at room
temperature for 2–4 h and then heat it. Run both samples on an SDS gel and
see if there is significant degradation of your protein in the sample not
heated immediately. I have heard that as little as 1 pg of protease in a protein
sample can cause major degradation of a protein if the sample is added to
sample buffer, but not heated immediately.

2.2. Asp–Pro bond cleavage at 100
C

Another related problem is the inadvertent cleavage of a protein at aspartic
acid–proline peptide bonds (DP bonds) due to heating at too high tempera-
ture for too long. The DP peptide bond is the most easily cleaved peptide
bond by heat or acidic conditions. I work on a protein that, if heated in sample
buffer at 100
C for 5 min, will undergo about 5% degradation to several smaller
bands. This is particularly evident if the gel is subjected to Western blot
analysis, which is easily able to detect a few nanogram of degradation frag-
ments. As a result, I always caution students to do the following experiment.
Prepare six identical tubes of your protein in sample buffer. Immediately heat
one set of three tubes at 95–100
C for 5, 20, and 60 min. Heat the other set
of three at 75
C for 5, 20, and 60 min. Put heated samples on ice and run the
samples on a SDS gel. Often, but not always, significant cleavage will occur at
the higher temperature, especially at the longer time. Of course, one can also
analyze the sequence to see if it contains the DP dipeptide. As a result, our lab
routinely heats samples for 5 min at 75
C. We avoid DP peptide bond
cleavage and have been able to completely inactivate proteases. The DP bond
is known to be particularly labile; 8–20 times more labile that other
DX sequences and 100-fold more labile than dipeptide bonds lacking D
(Volkin et al., 1995). However, many proteins have been found to be stable
in sample buffer for hours at 100
C (Deutsch, 1976).
816 Richard R. Burgess

2.3. Keratin contamination in sample or SDS sample buffer

Many times someone has shown me a stained SDS gel pattern and lamented
that they just cannot seem to remove the impurities from their otherwise
pure protein. Often this impurity is a cluster of bands around 55–65 kDa in
apparent molecular weight (Ochs, 1983). Keratin is omnipresent in skin,
dander, etc. When run on an SDS gel under nonreducing conditions (no 2-
mercaptoethanol in the sample buffer), keratin runs near the top of the gel,
but under reducing conditions that are most commonly used, keratin runs as
a heterogeneous cluster of bands around 55–65 kDa on the reducing SDS gels. If
you suspect that you have a keratin contamination, you can confirm this
suspicion by simply running a nonreducing gel and the characteristic keratin
bands should not be seen. The most common problem is that the sample
buffer itself becomes contaminated by contact with skin or a flake of
dandruff. Usually, this is quite a minor contaminant and can only be seen
in silver stained gels. Occasionally, it can be seen on Coomassie Brilliant
Blue stained gels. If you suspect this is a problem, simply apply sample buffer
alone, with no other added protein, to a gel lane. If you see keratin, it means
it is in your sample buffer and the buffer should be remade. We usually
prepare batches of 4 SDS sample buffer, aliquot them into 1-ml portions,
and store them frozen at  80
C. Then, we thaw one portion, dilute it
appropriately, and use it within a day or two. Keratin is also occasionally
seen on Western blots where the antigen used to prepare a polyclonal
antibody is contaminated with keratin and reacts with keratin in the proteins
transferred from SDS gel to nitrocellulose membranes.

3. Buffers
3.1. Reducing agent can become an oxidizing agent
One very commonly adds a reducing agent to buffers to help prevent air
oxidation of enzymes and other proteins. It is typical to add 0.1–1 mM 1,4-
dithiothreitol (DTT) (HSCH2CH2OHCH2OHCH2SH) (Cleland, 1964).
Usually such buffers are stored in the cold and used within a few days, but
often buffers are used that are months old. The problem is that slowly over
time, the reducing agent becomes oxidized by oxygen in the air and becomes an
oxidizing agent; effectively, it is now much worse that having no thiol agent
in the buffer. How do you know whether a buffer is still good and has
effective reducing capability? A simple test is to treat a small portion of the
buffer with Ellman’s reagent (5, 50 -dithiobis-(2-nitrobenzoic acid) or
DTNB) (Ellman, 1959). This chemical reacts with free thiol groups, such
as the sulfhydryl of the DTT and releases 2-nitro-5-thiobenzoic acid, which
is yellow colored and absorbs light with a molar extinction coefficient at
Important Artifacts in Protein Chemistry 817

412 nm of 14,150 M 1 cm 1 (Riddles et al., 1983). Read the A412 nm of the
solution and a control buffer carefully prepared with fresh DTT. If all the
DTT is active then there should be two sulfhydryls per mole of DTT. Buffers
should be made fresh if more than 10–20% of the DTT is oxidized. While this
effect has been known for decades it is still surprising that this simple test is
not more commonly used. The problem is even more pronounced with
2-mercaptoethanol which is more prone than DTT to air oxidation and is
usually used at 5–50 times higher concentrations than DTT.

3.2. Contaminant in sulfonylethyl buffers

An unexpected and potentially critical technical observation was made in
the Ron Raines lab at the University of Wisconsin–Madison and carefully
documented in a paper by Smith et al. (2003). In studying ribonuclease A
activity in a MES buffer, they noticed that the activity was progressively
lower as the salt in the buffer was decreased. What they eventually found
was that MES and, in fact, many of the sulfonylethyl group-containing
buffers (MES, BES, CHES, and PIPES) contain a contaminant, oligo
(vinylsulfonic acid), that is a side product in the chemical synthesis of the
chemical buffer. This polyanionic contaminant mimicks RNA and at low
salt (0.05 M NaCl) binds to and inhibits the enzyme with a Ki of 11 pM.
This has implications to anyone studying nucleic acid-binding proteins.

4. Chromatography
4.1. EDTA binding to anion-exchange columns
Recently, Robert Chumanov in my lab observed a sharp peak of
215 nm-absorbing material eluting during a salt gradient elution of a
MonoQ anion exchange resin (GE Healthcare). It turned out that it was
EDTA, bound quite tightly to MonoQ and eluting at a NaCl concentration of about
250 mM (R. Chumanov and R. Burgess, unpublished). We routinely
monitor column outputs at 215 nm when our protein/peptide lacks absorp-
tion at 280 nm or if we are loading low amounts of protein and want more
sensitivity (most proteins absorb at least 15 times better at 215 nm than
280 nm). It is not surprising that the di- or trivalent anion EDTA binds to
MonoQ column resin. We were surprised that so much bound that it
completely depleted EDTA from our low salt equilibration buffer originally
containing 0.1 mM EDTA. When the salt increased, the EDTA eluted with
a peak EDTA concentration of up to 40 mM. This could have unexpected
effects on many experiments. The fractions where the EDTA elutes will, for
example, bind Mg2þ needed for many in vitro enzyme assays (such as
transcription and replication assays) and appear to contain a mysterious
818 Richard R. Burgess

enzyme inhibitor. It is also possible that a column nearly saturated with

EDTA would have a lower protein-binding capacity, especially for proteins
that only bind weakly to the column and elute at NaCl concentrations
lower than 250 mM. This type of binding had been observed and published
over 10 years ago (Sharpe and London, 1997), but we had not seen the
paper and had not realized how dramatic this effect could be. Further
reflection also suggests that other multivalent anions such as sulfate and
phosphate will also bind to and elute from anion exchange resins with
potentially artifactual consequences.

4.2. EDTA in sample can strip nickel from a Ni-chelate

affinity column
One of the three methods for eluting hexa-His-tagged proteins from Ni2þ-
chelating columns is by passing EDTA through the column to effectively
chelate and strip out the Ni2þ and release bound tagged proteins (see
Chapter 27). However, many people are used to breaking Escherichia coli
and other biological materials in lysis buffers containing 0.1–5 mM EDTA.
In some cases, they centrifuge out insoluble material and load the crude
extract directly on an immobilized Ni2þ column. The EDTA in the sample,
especially at 1 mM or higher, could easily decrease or remove completely
the chelated Ni2þ and thus dramatically decrease the capacity of the column
to bind His-tagged protein. When the Ni2þ has been stripped from the
column, the column will become white instead of the light blue color. Such
a column can easily be recharged with Ni2þ.

5. Protein Absorption During Filtration

When we refold proteins by dilution out of denaturants, we often get
some aggregation that causes light scattering and turbidity of the sample (see
Chapter 17). Before we apply the diluted sample onto an ion-exchange
column, we filter the sample through a 0.22-mm cellulose acetate filter to
remove any particulate material. One time we ran out of cellulose filters and
a student used another filter that was in the lab. Unfortunately, it was a
cellulose nitrate (or nitrocellulose) filter. All the soluble protein was lost, due to
protein binding to the filter. This was not too surprising, since proteins gener-
ally bind very well to nitrocellulose filters. In fact, this is the principle of
protein transfer from SDS gels to nitrocellulose membranes for Western blot
analysis. Now we routinely use special PVDF filter units (Stericup-GV
0.22 mm, 500 ml, Millipore #SCGVU05RE) with good recovery of pro-
tein. Most filter membranes, no matter how ‘‘low protein binding’’ they are
designed to be, will bind mg amounts of protein per cm2, but some
Important Artifacts in Protein Chemistry 819

materials, such as cellulose nitrate and polystyrene are able to bind much
more material and should be avoided for filtration or storage of protein.

6. Chemical Leaching from Plasticware

It is worth being aware of a recent report (McDonald et al., 2008) that
certain chemicals can leach from common disposable laboratory plasticware into
standard aqueous buffers, and in some cases can have strong effects on the results of
biological and biochemical experiments. Some chemicals, like oleamide, are used
as lubricating agents in the molding process and others, like certain cationic
biocides, are used to help prevent bacterial colonization of the plastic
surface. Washing the plasticware in water or better in methanol or
DMSO can remove much of this material.

7. Cyanate in Urea
Urea is widely used as a protein-denaturing agent, but it is not widely
appreciated that urea solutions contain substantial amounts of ammonium
cyanate, which is in chemical equilibrium with urea (H2N)2C¼O in equilib-
rium with NH4þ þ NCO). Isocyanic acid H–N¼C¼O can react with
amino groups on proteins (e-amino group of lysine and the amino terminus,
and to a lesser extent with arginine and cysteine) to form a carbamylated
protein (Stark, 1965). This carbamylation can alter charge and in some cases
interfere with enzyme function, block certain protease cleavage reactions, and
add 43 Da per carbamylation event to the mass as measured by mass spec-
trometry. How can one diminish or prevent this from happening? One can
treat a urea solution with a mixed bed resin such as Bio-Rad AG 501-X8 to
remove these contaminant ions. The progress of deionization is easily moni-
tored by measuring conductivity of the solution. Unfortunately, this is a
chemical equilibrium and relatively soon (within a few days) the ammonium
cyanate builds back up again to levels in the 0.5–3 mM range in an 8 M urea
solution and can reach values of 20 mM (Lin et al., 2004). Certain chemical
scavengers such as ethylenediamine, glycylglycine, or glycinamide in the
5–25 mM range can also reduce cyanate to less than 0.1 mM in 8 M urea,
Tris pH 8 (Lin et al., 2004). Perhaps the best general practice if you want to
use urea is to replace some of the NaCl in the buffer with some ammonium
salt (such as 25–50 mM ammonium chloride) to push the equilibrium back by
the common ion effect toward less cyanate. Reaction of isocyanic acid with
amino groups in proteins is slowed by lower temperature and by acidic
conditions and can be minimized by restricting the exposure of the protein
to the urea solution to the shortest time possible.
820 Richard R. Burgess

REFERENCES
Cleland, W. W. (1964). Dithiothreitol, a new protective reagent for SH groups. Biochemistry
3, 480–482.
Deutsch, D. G. (1976). Effect of prolonged 100
C heat treatment in SDS upon peptide
bond cleavage. Anal. Biochem. 71, 300–303.
Ellman, G. L. (1959). Tissue sulfhydryl groups. Arch. Biochem. Biophys. 82, 70–77.
Grabski, A., and Burgess, R. R. (2001). Preparation of protein samples for SDS-polyacryl-
amide gel electrophoresis: Procedures and tips. inNovations 13, 10–12.
Lin, M.-F., Williams, C., Murray, M. V., Conn, G., and Ropp, P. A. (2004). Ion chro-
matographic quantification of cyanate in urea solutions: Estimation of the efficiency of
cyanate scavengers for use in recombinant protein manufacturing. J. Chromatogr. B 803,
353–362.
McDonald, G., Hudson, A., Dunn, S., You, H., Baker, G., Whittal, R., Martin, J., Jha, A.,
Edmondson, D., and Holt, A. (2008). Bioactive contaminants leach from disposable
laboratory plasticware. Science 322, 917.
Ochs, D. (1983). Protein contaminants of sodium dodecyl sulfate-polyacrylamide gels. Anal.
Biochem. 135, 470–474.
Pringle, J. R. (1975). Methods for avoiding proteolytic artefacts in studies of enzymes and
other proteins from yeasts. Methods Cell Biol. 12, 149–184.
Riddles, P. W., Blakeley, R. L., and Zerner, B. (1983). Reassessment of Ellman’s reagent.
Meth. Enzymol. 91, 49–60.
Sharpe, J. C., and London, E. (1997). Inadvertent concentration of EDTA by ion exchange
chromatography: Avoiding artifacts that can interfere with protein purification. Anal.
Biochem. 250, 124–125.
Smith, B. D., Soellner, M. B., and Raines, R. T. (2003). Potent inhibition of ribonuclease A
by oligo(vinylsulfonic acid). J. Biol. Chem. 278, 20934–20938.
Stark, G. R. (1965). Reactions of cyanate with functional groups of proteins: Reaction with
amino and carboxyl groups. Biochemistry 4, 1030–1036.
Volkin, D. B., Mach, H., and Middaugh, C. R. (1995). Degradative covalent reactions
important to protein stability. (B. A. Shirley, ed.)Protein stability and folding: Theory and
practiceMethods Mol. Biol. 40, 35–63.
Author Index

A Aslanidis, C. V., 153, 155

Atha, D. H., 118
Abdolzade-Bavil, A., 317 Attri, A. K., 699–700
Abe, M., 649 Aumiller, J. J., 213
Ackers, G. K., 706–707, 721 Azzouz, N., 196
Acton, T. B., 456, 460
Afeyan, N. B., 355 B
Agnew, B., 557, 559
Aguiar, R. W., 142 Bachmair, A., 245
Ahmad, I., 737 Backliwal, G., 227, 231–232, 234
Akermoun, M., 632 Bailey, M. J. A., 73–93
Akutsu, H., 164 Balaji, P. V., 25
Al-Ibraheem, J., 293, 298 Baldi, L., 138, 224–225
Al-Rubeai, M., 139 Baldwin, M. A., 712–713
Alegria-Schaffer, A., 573–598 Bamber, L., 635
Alexander, C., 424 Baneyx, F., 133, 150, 156, 249, 262, 276
Ali-Khan, 103 Bantan-Polak, T., 89
Allen, R. C., 498, 500–501, 506–509, 511–512 Bantscheff, M., 751, 753, 755
Allis, C. D., 726, 741 Barbieri, C. M., 557
Alterman, M., 85 Bardwell, J. C., 134
Ames, B., 702, 705 Barnes, P. J., 740
Ames, G. F., 162 Batard, P., 229
Anderson, N. G., 517, 750 Bates, D. M., 803
Anderson, N. L., 517, 750 Bayer, M. H., 250
Anderson, T. R., 247–248 Bayer, P., 739
Andersson, C. L., 134 Baynes, B. M., 273
Andersson, D. I., 133 Beausoleil, S. A., 732, 735
Andersson, L., 444 Becker, D. M., 176–177, 179
Andre, N., 632 Beckman, T. W., 741
Andrews, A. T., 498–501, 503, 506–509, Beidler, D. E., 447
511–512, 710 Beinert, H., 788–789, 798–801
Andrews, B. A., 286, 290 Bell, P. J. L., 91, 554
Anfinsen, C. B., 260 Bellehumeur, T. G., 575
Angal, S., 293 Ben-Naim, A., 407
Antal, J., 571 Bennett, K. L., 735
Antharavally, B. S., 586, 614 Benov, L., 293, 298
Anthony, J. R., 478 Bergendahl, V., 478, 481, 490, 575
Anthony, L. C., 269 Berger, I., 196
Antoine, M. D., 582 Berger, S. L., 726, 741
Anusubel, E. M., 170 Berggren, K. N., 553
Arakawa, T., 273, 704 Berkelman, T. R., 555
Armirotti, A., 757 Berlet, H. H., 86
Armstrong, N., 271–272 Bernardi, G., 389
Arnau, J., 243, 252, 457 Bernstein, E., 741
Arnold, T., 608, 611–613 Bertschinger, M., 229, 231
Arnold, U., 531 Bessette, P. H., 134, 279
Asenjo, J. A., 286, 290 Bhalget, M. K., 553
Asermely, K. E., 89 Bhatia, P. K., 139, 144
Ashani, Y., 614 Bier, M., 498

821
822 Author Index

Biggs, R., 681 Campbell, D. H., 419

Billecke, C., 307, 310 Campbell, J. M., 745
Bjellqvist, B., 515, 519 Candiano, G., 547
Blackshear, P. J., 498, 501, 503, 506–508, 511 Cannon, J. G., 586
Blagoev, B., 728 Cantin, G. T., 733
Blakeley, R. L., 83 Cantor, C. R., 704
Blanchard, J. S., 43–56 Carpenter, G., 579
Blau, H. M., 142 Carr, S. A., 737
Block, H., 439–468 Carroll, I. P., 800
Blommel, P. G., 160, 651–652, 656 Carson, M., 246
Blumberg, P. M., 428 Carstens, C. P., 156
Blume-Jensen, P., 731 Carta, G., 364
Bock, P., 699–700 Cass, B., 456
Boden, V., 447 Castellanos-Serra, L., 567
Boer, E., 171 Catley, B. J., 289
Bogyo, M., 543 Catravas, G. N., 614
Bolanos-Garcia, V. M., 246, 449, 455–456 Cèbe, R., 450
Bonander, N., 633 Cecı́lia, A., 428
Booker, C. J., 103 Celik, E., 136
Bornhorst, J. A., 454 Chaga, G., 442
Bossio, R. E., 691–721 Chait, B. T., 731–735
Bottomley, S. P., 249, 262, 277 Chalkley, R. J., 730
Boussif, O., 231 Chan, J. K., 86
Boyer, H. W., 154 Chance, M. R., 447, 463
Boyer, M. E., 649 Chandran, S. P. 433
Bradford, M. M., 12, 79, 85, 624 Chang, J. Y., 680
Bradley, M. K., 196, 214 Chappel, T., 169–187
Braun, P., 460 Chase, T., Jr., 788
Breier, J., 390 Chasin, L. A., 226
Breitbach, K., 213 Chatterjee, D. K., 134, 141, 243
Brena, B., 432 Chazenbalk, G. D., 199
Brendel, M. F., 177, 179 Chelikani, P., 635
Brierley, R., 185 Chen, C. L., 741
Brinker, N., 439–468 Chen, W., 137, 741
Brisack, K., 387–402 Chen, Y. R., 741
Brondyk, W. H., 131–144 Cheng, Y., 251
Brooks, C. A., 352 Chenuet, S., 229, 232
Brown, N., 118 Cherezov, V., 634
Browne, S. M., 139 Cheryan, M., 108
Brzoska, K., 788 Chiari, M., 567
Buchanan, S. K., 632 Chirgwin, J. M., 244
Burgess, R. R., 21–27, 29–34, 259–279, 287, Cho, M. S., 225–226
293, 297, 331–342, 475–493, 565–571, 577, Chong, S., 253
813–819 Chopra, I., 162
Burkitt, W. I., 75, 85 Chotia, C., 408
Burris, R. H., 788 Chou, 162
Bury, A. F., 498, 501 Choudhary, J. S., 244
Busso, D., 250, 450 Choudhary, S., 575
Büssow, K., 460 Chow, M. K., 277
Butler, M., 139 Chrambach, A., 498, 500, 506–508, 511–512,
Butt, T. R., 253 710
Byrne, B., 463 Christensen, K., 229
C Chu, B., 716
Chu, R., 287, 289
Cabiscol, E., 789 Chumanov, R., 265, 267
Cabrita, L. D., 244, 249, 262 Cirulli, C., 733
Calandra, B., 633 Clark, P. L., 261–262
Cammack, R., 802 Clark, P. M., 559
Author Index 823

Clark, W. M., 55 De Mey, M., 293

Clauser, K. R., 744 de Moreno, M. R., 75, 85–86
Cleland, J. L., 273 Dean, D., 224, 226
Cleland, W. W., 62, 816 Deb, J. K. 133
Cline, D. J., 294 Deber, C. M., 244
Cline, G. B., 311 D. G.ip, W. J., 613
Clonis, Y. D., 428 D. J.ng, J., 455
Coghlan, D. R., 91, 554 Delory, G. E., 55
Cohen, P., 726 Demeler, B., 720
Cole, J. L., 687, 698, 700, 702–703 Dempsey, 44–45, 48, 50
Collart, F. R., 149–166 Denison, C., 739
Collier, T. S., 757 Denisov, G. A., 108
Collins-Racie, L. A., 244 Derewenda, Z. S., 440, 451
Condreay, J. P., 783 Derman, A. I., 154
Consortium, U., 770 Desai, S., 584
Constantinidou, C., 789 Desiderio, D. M., 741
Cook, P. F., 62 Deupi, X., 634
Coon, J. J., 728 Deutsch, D. G., 815
Cooper, C. E., 802 Deutscher, M. P., 37–42, 121–127, 343–345,
Cornish-Bowden, A., 62 809–811
Costantino, H. R., 114 di Salvo, M. L., 401
Cottingham, K., 716 Dieckman, L., 150, 152, 156, 286
Cottrell, J. S., 726 Dieckman, L. J., 150, 152, 156
Covey, T. R., 745–746 Dignam, J. D., 286
Cowieson, N. P., 271–272 diGuan, C., 249
Crack, J. C., 789, 803 Dikic, I., 738
Cramer, P., 477, 479, 490 Dixon, R. A., 162
Cramer, S. M., 352 Döbeli, H., 253
Cravatt, B. F., 543 Dong, M. S., 142
Creasy, D. M., 726 Dong, X., 389
Cregg, J. M., 135, 169–187 Dongre, A. R., 747
Creighton, T., 275 Doolittle, R. F., 24
Crowe, J., 455 Dorrier, T. E., 83
Cuatrecasas, P., 419 Douris, V., 194
Cull, M., 286 Drew, D. E., 651
Cummings, L. J., 387–402 Drews, J., 461
Cunningham, F., 244 Drott, D., 289
Curtis, C. A. M., 632, 635 Duellman, S. J., 340, 478, 491–492
Cussler, E. L., 106 Dunbar, B. S., 497–498, 508, 512, 542
Czerney, P. T., 555 Durand, G., 737
Durocher, Y., 138, 227
D Durschlag, J., 703
Dyson, M. R., 150, 244, 249–250
Daban, J.-R., 91, 551
Dalbge, H., 449 E
Daly, R., 135–136, 141–142
Dang, J. M., 229 Eason, R., 687
Danson, M., 65–66, 68 Edelman, M., 311
Das, T. K., 273 Edwards, A. M., 477–478, 490
Davies, O. R., 449, 455–456 Einhauer, A., 247
Davis, B. J., 506, 508 Eisenthal, R., 65–66, 68
Davis, G. D., 250 Elias, C. B., 211
Davis, T. R., 211 Eliyahu, H., 229
Dawson, R. M. C., 294 Ellman, G. L., 816
De Grip, W. J., 634 Endo, Y., 649–650, 663, 669
de Jong, P. J., 153, 155 Englard, S., 332, 335, 341
De Marco, V., 155 Engler, C. R., 286
de Marco, A., 249–250 Ernst, W. J., 207
824 Author Index

Eschenfeldt, W. H., 153, 156 Garzia, L., 460

Eshaghi, S., 301, 462 Gasteiger, E., 27
Esposito, D., 134, 141, 153, 243 Gatzka, M., 731
Etzel, M., 118 Gauchet, C., 433
Evans, D. R. H., 97–118 Ge, Y., 728–729, 757
Gegenheimer, P., 286
F Geiser, M., 450
Geisse, S., 223–234
Fabis, R., 439–468 Gellissen, G., 170
Falke, J. J., 454 Gellman, S. H., 273
Fang, S., 738 Gemmill, T. R. 136
Farley, A. R., 725–758 Georgiou, G., 278
Fasman, G. R., 692 Gether, U., 632
Fath-Goodin, A., 210, 213–214 Getz, E. B., 294
Faull, K. F., 712, 714 Ghosh, R., 113
Faulmann, E. L., 425 Gibbons, R. A., 703
Fausnaugh, J. L., 407 Giglione, C., 142
Felsenfeld, G., 324 Gilbert, H. F., 269
Ferguson, J. T., 758 Gill, P., 448
Ferguson, W. J., 43, 390 Gill, S. C., 23, 83
Fernandez, C., 615 Gillies, A. R., 253
Fernandez-Patron, C., 549 Gingras, A. C., 731
Ficarro, S. B., 733 Girard, P., 138
Finn, R. D., 770 Glatter, O., 694
Fish, W. W., 698, 708 Gleeson, M. A. G., 136, 184
Fisher, W. D., 311 Glimcher, M. J., 392
Flugge, U. I., 117, 531 Glossman, H., 498–499
Foley, K. M., 475–493 Glozak, M. A., 741
Folta-Stogniew, E., 708 Godde, J. S., 726
Ford, T., 314 Goerke, A. R., 649
Foster, D., 286 Golks, A., 559
Fountoulakis, M., 75, 85 Gomori, G., 52, 54
Fox, B. G., 647–670 Good, D. M., 749
Franco, A. T., 526 Good, N. E., 43, 47
Frank, A. M., 149–166 Goodlett, D. R., 729
Franken, K. L. M. C., 453, 460 Gorbunoff, M. J., 388–389–390, 392
Freifelder, D., 702 Goren, M. A., 647–670
Freitag, R., 390, 457 Görg, A., 510, 519
Frey, P. A., 65 Goshe, M. B., 733
Friedenauer, S., 86 Goto, M., 136
Friedman, D. B., 515–538 Gotto, J. W., 788
Friesen, P. D., 194 Goustin, A. S., 133
Fritze, C. E., 247–248 Gräslund, S., 243
Fromme, P., 615 Grabski, A. C., 285–300, 814
Fuchs, S. M., 726 Grace, T. D. C., 211
Furth, A. J., 513, 613 Graessmann, A., 226
Fushman, D., 738 Graham, F. L., 224
Fussenegger, M., 234 Graham, J., 314
Fux, C., 223–234 Grainger, D. C., 789
G Graslund, S., 261, 456, 767
Grewal, S. I., 741
Gagnon, P., 389, 392, 402 Grey, M., 177, 179
Galbraith, D. J., 234 Gribskov, M., 267
Gallagher, S. R., 585 Griffith, D. A., 633
Garcı́a González, L. A., 448 Grisshammer, R., 463, 631–642
Garcia, B. A., 728, 742 Groves, W. E., 82
Garfin, D. E., 497–513, 542, 551 Gu, T. R., 611
Garrett, P. E., 292 Gu, W., 742
Author Index 825

Guarante, L., 176–177, 179 Hearon, J., 440

Guarino, L. A., 194–195, 215 Heerikhuisen, M., 171
Guerini, D., 559 Hegeman, A. D., 65
Guerlava, P., 293 Helenius, A., 608
Guerrier, L., 402 Hengen, P., 154
Guidotti, G., 619–629 Heppel, L. A., 162
Guilbault, G. G., 424 Hermanson, G. T., 425, 427, 430, 432
Gunsalus, R. P., 788 Hernan, R., 248
Guruprasad, K., 24 Higgin, D., 176
Guss, B., 425 Higgin, J. M., 176
Gustafsson, C., 140 Hilbrig, F., 457
Gustavsson, P.-E., 419 Hill, D. R., 210
Gusman, L. M., 154 Hill-Perkins, M. S., 199
Gygi, S. P., 729, 750–752 Hinkkanen, A., 431
Hinze, W. L., 608
H Hirabayashi, J., 419, 427
Hiraide, T., 390
Haas, W., 744 Hirel, P.-H., 449
Habig, W. H., 247 Hitchcock, A., 262
Hage, D. S., 419, 422, 424–425, 428 Hitchman, R. B., 137
Hager, D. A., 566, 568–570 Hjelmeland, L. M., 623
Hahn, R., 349–370 Hjertén, S., 409–410
Haldenwang, W. G., 569 Ho, C. W., 293
Hale, J. E., 447 Hober, S., 461
Hall, J. F., 649 Hochuli, E., 440, 442, 449, 451, 458
Halvorson, H., 699–700 Hofmeister, F., 117, 407
Hames, B. D., 498–501, 503, 506–509, 511 Hollister, J. R., 213
Hammer, K. D., 79, 89 Holmes, W., 54
Hampsey, M., 477 Holt, 737
Han, G. Y., 294 Hopkins, T. R., 286, 290–291, 293
Hancock, 738 Horvath, C., 407
Hansen, B. T., 730 House, J. E., 58
Hanson, M. A., 634–636 Hoving, S., 515–538
Harding, I. S., 394 Howlett, G. J., 699
Harding, S. E., 716 Hsieh, H.-C., 101
Hardwicke, P. I., 211 Hsieh, H.-Y., 401
Hardy, E., 567 Hu, J., 253
Harlow, E., 479 Hu, S. M., 253
Harrington, M. G., 512, 566, 570 Hubbard, M. J., 726
Harris, E. L. V., 293 Huber, L. A., 447
Harris, T. K., 57–71 Hubert, P., 447
Harrison, J. L., 771 Hulme, E. C., 632, 635, 637
Harrison, R. G., 25, 116 Hulo, N., 245
Harrison, R. L., 137, 194, 212, 215 Hunt, I., 460
Harrison, S. T. L., 286, 290–291 Hunter, T., 731
Hart, C., 558
Hart, G. W., 559, 737 I
Hartinger, J., 516
Hartley, D. L., 265 Ibel, K., 522
Hartley, J. L., 155 Ichitsuka, T., 388
Hassaine, G., 632 Idicula-Thomas, S., 25
Hatakeyama, H., 526 Ikeda, F., 738
Hatley, J. L., 771 Ikonomou, L., 137
Haugland, R. P., 552, 556, 558 Imagawa, W., 578
Haun, R. S., 153, 155 Ingham, K. C., 118, 341
Hawtin, R. E., 208 Ishihama, Y., 615
Hayduk, E. J., 729 Issad, T., 726
Hearn, M. T. W., 135–136, 141–142 Ito, S., 728
826 Author Index

J Kelleher, N. L., 744, 757

Kenig, M., 253
Jaakola, V. P., 634–635 Kennedy, M. C. 789, 799–800, 802
Jana, S., 133 Kennedy, S. W., 89
Janson, J. -C., 354, 406, 411 Kersteen, E. F., 269–270
Jarvis, D. L., 137, 191–218 Keshwani, M. M., 57–71
Jazwinski, S. M., 286 Khoroshilova, N., 789, 799, 802
Jendrisak, J. J., 338, 340 Kiga, D., 649
Jenkins, N., 137, 139, 142 Kigawa, T., 649
Jennings, T., 114 Kiley, A. Y., 787–803
Jensen, M. R., 449 Kim, C. Y., 246
Jensen, O. N., 742 Kim, D.-M., 457
Jenuwein, T., 726, 741 Kim, J., 446, 457
Jessani, N., 543 Kim, M. J., 446, 457
Jesse, J., 179 Kim, T.-W., 457
Jia, W., 101 Kim,Y., 767
Jiang, W., 165 King, E. J., 55
Jiang, Y., 477 King, L. A., 199, 215
Jiang, Z., 229
Kinoshita, E., 557
Jin, S., 445
Kitts, P. A., 200
Johnson, J. F., 711
Kivimäe, S., 226
Johnson, M. K., 788–789
Klaassen, C. H., 635
Johnson, M. L., 699, 720
Klammt, C., 462, 649, 669
Jones, D. D., 408
Kleinig, A. R., 291
Jones, L. J., 91
Klock, H. E., 150, 771–772, 783
Jordan, M., 229
Knight, R. D., 463
Jormakka, M., 463
Knox, J. R., 694
Jovin, T. M., 498, 506
Knuth, M. W., 27, 263, 265–266, 270, 338, 340
Jungbauer, A., 247, 249, 262, 276, 349–370, 451
Kobilka, B. K., 632, 634–636
Koehn, J., 460
K
Koike, T., 557
Kaar, W., 249, 262, 276 Komoriya, A., 408
Kaba, S. A., 208–209 Koneracka, M., 423
Kadoya, T., 388 Kool, M., 216
Kadwell, S. H., 211 Kopaciewicz, W., 352
Kajiyama, S., 788 Kornberg, A., 3–6
Kaltenbrunner, O., 365 Korpela, T., 431
Kandori, K., 389 Korth, K. L., 142
Kane, J. F., 140 Kost, T. A., 136, 196, 783
Kanehisa, M., 770 Kovalska, V., 555
Kang, C., 555 Kowal, R., 431
Kang, Y., 789 Krane, S. M., 392
Kanno, T., 649–650 Kratky, O., 694
Kao, F.-T., 226 Krause, F., 566, 570
Kaplan, W., 247 Kresin, L. M., 567
Kapust, R. B., 155, 249, 252, 774 Krigbaum, W. R., 408
Karp, N. A., 525 Kroe, R. R., 700
Karp, P. D., 770 Kubicek, J., 439–468
Karplus, M., 261 Kuck, U., 171
Karuso, P., 91, 554 Kunaparaju, R., 226
Kashino, Y., 289 Kunitani, M. G., 567
Kaslow, D. C., 458 Kunji, E. R., 165
Kassel, D. B., 732 Kuo, M., 726
Kataeva, I., 244, 249–250 Kupke, D. W., 83, 693
Kato, A., 250–251 Kurokawa, M., 313
Kaufmann, S. H., 595 Kusari, A., 169–187
Kawasaki, A., 402 Kuster, B., 244
Kawasaki, T., 389 Kusumoto, K., 229
Author Index 827

Kutter, K., 566 Lin, M.-F., 819

Kuznedelov, K., 490 Lin, S.-H., 619–629
Kwaks, T. H. J., 139 Lin-Cereghino, G. P., 174, 177
Kwok, T. T., 579 Link, A. J., 725
Kwon, M. S., 137 Linke, D., 603–616
Linn, S., 9–19
L Listwan, P., 244
Liu, J. W., 244, 251, 711–712
Labahn, J., 439–468 Liu, L., 244, 251
Labrou, N. E., 419, 428 Liu, Q., 771
Ladish, R. G., 116 Lodge, A., 573–598
Laemmli, U. K., 498, 501, 506–507, 521–522 Londer, Y. Y., 150, 153
Laengle-Rouault, F., 224 London, E., 818
Laidler, K. J., 58 Loo, J. A., 447
Lakowicz, J. R., 66 Lopez, J. M., 86
Lane, D., 479 Lorenzen, A., 89
Langone, J. J., 424 Lorow-Murray, D., 179
Lanio, T., 460 Los, G. V., 251, 428
Largaespada, D. A., 479 Lowe, C. R., 428
Larsen, M. R., 733, 746 Lowry, O. H., 86–87
Larsson, B., 409 Lu, A., 194, 207
Larsson, P.-O., 419 Lu, Y., 728
Lata, S., 445–446 Lubs, H. A., 55
Laue, T. M., 677–689, 691–721 Luckow, V. A., 202
L. V.llie, E. R., 134, 250 Luellau, E., 401
Lazazzera, B. A., 789, 791 Lungwitz, U., 229
Le Hir, H., 227 Lunney, J., 681
Leach, S. J., 82 Lv, G. S., 446
Leckband, D. E., 432 Lynch, N. A., 477
Lee, C., 510, 512 Lynn, D. E., 211
Lee, C. D., 246, 251, 253
Lee, C. H., 735 M
Lee, H., 735
Lee, J., 226 Ma, D., 455
Lee, J. R., 735 Ma, Q. Y., 394
Lee, K. A., 735 Ma, Y., 735
Lee, S. B., 246, 251, 253 Mac Coss, M. J., 744
Lee, S. J., 735 Macek, B., 756–758
Lee, Y. C., 209, 622 MacKenzie, D., 635
Lei, Z., 570 Mackintosh, J. A., 554
Leibl, H., 402 Madden, E. A., 311, 314
Leirmo, S., 49 Madden, K., 169–187
Lennarz, W. J., 737 Madin, K., 649–651, 669
Leong, K. W., 229 Madoery, R., 427
Lerman, L. S., 419 Madzak, C., 171
Lesley, S. A., 286, 767–783, Maertens, B., 439–468
Lettner, H., 366 Magnani, F., 634
Levings, C. S., 142 Majors, B. S., 234
Levinthal, C., 260 Makarov, A., 728
Levy, D., 667 Makino, S.-I., 647–670
Lewinson, O., 462 Makrides, S. C., 155, 226, 265
Laible, P. D., 165 Malakhov, M. P., 253
Lichty, J. J., 243 Maldonado, E., 490
Lieve, L., 289 Malhotra, A., 239–254
Lihoradova, O. A., 207 Mallia, K., 289
Lilley, K. S., 516, 524–526, 528 Mallik, R., 422, 432
Lillie, R. D., 50 Manavalan, P., 408
Lim, K. B., 732 Manchenko, G. P., 543, 566
828 Author Index

Mann, M., 735, 744, 750–752, 757 Mohanty, A. K., 449, 463
Manza, L. L., 739 Molday, R. S., 635
Marblestone, J. G., 251 Monsma, S. A., 137
Margulies, M. M., 500 Mooney, R. A., 612
Marin, M., 140 Moore, L. J., 798
Markley, J. L., 649–650 Morita, E. H., 669
Markwell, J., 75, 93 Mortensen, K. K., 265
Marlor, C. W., 344 Mueller, E., 354–355
Marshak, D. R., 289, 293 Mueller, U., 244
Marshall, K., 703 Mujacic, M., 133, 249, 262, 276
Marshall, T., 507 Mukhopadhyay, A., 139, 144
Marston, F. A. O., 265 Muller, N., 232
Martin, R., 702, 705 Murby, M., 265
Martinez Molina, D., 774 Murhammer, D. W., 215
Maslennikov, I., 668 Muszynska, G., 444
Mateo, C., 456
Mathrubutham, M., 578, 590 N
Matiasson, B., 457
Matsuda, T., 649 Nagel, K., 79, 89
Matsumoto, K., 227, 649 Nagy, P. L., 477
Matthey, B., 155 Nakamura, E., 320
Mattson, D. L., 575 Nall, B., 261
Mc Donald, G., 819 Nallamsetty, S., 249, 252
M. C.slin, D. R., 703 Necina, R., 358
M. C.e, J. T., 395–396, 405–413 Nelson, P. N., 476
M. H.nry, C. S., 286 Neophytou, I., 164
M. I.vaine, T. C., 51 Nestle, M., 290
M. L.chlin, D. T., 731–735 Neu, H. C., 162
M. L.chlin, J. R., 195 Neugebauer, J. M., 289, 605, 608–609
M. M.ster, M. C., 67 Neuhoff, V., 546
M. M.ekin, T. L., 703 Neville, D. M., Jr., 498–499, 621
M. M.llan, D., 782 Nguyen, H., 287
Mechref, Y., 738 Nguyen, L., 263, 265
Medzihradszky, K. F., 737, 745 Nguyen, N. Y., 512
Meiss, G., 290 Nieba, L., 445, 448
Meissner, P., 226–227, 229 Nielsen, T. B., 499
Melander, W., 407 Nilsson, J., 457
Melchior, F., 739 Noble, J. E., 73–93
Merril, C. R., 508–509, 542, 547 Nomura, S. M., 652
Mészárosová, K., 447 Norris, J. L., 615
Mettert, E. L., 792, 797 Northcote, D. H., 86
Meyer, J., 788–789 Nosjean, O., 652
Michael, M. Z., 154 Nossal, N. G., 162
Michaelis, L., 53 Nozaki, Y., 408, 696
Michelsen, U., 305–327 Nozawa, A., 647–670
Middelberg, A. P. J., 262, 272, 286, 290–291 Nuss, J. E., 401
Mikesh, L. M., 728, 747, 749
Miles, B. J., 388 O
Miller, I., 543 O’Farrell, P. H., 515–516, 519
Miller, J. V., 326 O’Farrell, P. Z., 523
Miller, L. K., 193–195, 207 O’Reilly, D. R., 196, 198, 215–216
Minchiotti, M., 427 O’Shannessy, D. J., 432
Minden, J., 556 Oates, M. R., 432
Minton, A. P., 699–700 Ochs, D., 510, 816
Miroux, B., 154, 164 Oda, Y., 729, 733
Mistretta, T. A., 195 Oganesyan, N., 276
Miyagi, M., 751–752 Ogawa, T., 390
Moffatt, B. A., 154, 156 Ohana, R. F., 428
Author Index 829

Ohana, R. F., 251 Plievaa, F. M., 422

Ohtaki, T., 636 Plumel, M., 52
Old, W. M., 755 Podgornik, A., 367
Olin, B., 447 Podgornik, H., 367
Olsen, J. V., 729 Poggeler, S., 171
Olson, B. J., 75, 93 Poland, J., 547
Olsson, A., 726 Polayes, D. A., 457
Olsson, I., 516, 523–524, 534 Polevoda, 742
Ong, S. E., 729, 750–753 Ponnuswamy, P. K., 408
Ooi, B. G., 195 Popescu, C. V., 802
Oprian, D. D., 635 Porath, J., 409, 430, 440, 444, 446–447
Ornstein, L., 506, 508 Posewitz, M. C., 444
Orr, G. A., 328 Possee, R. D., 199–200, 204, 215
Osborn, M., 499 Poston, J. M., 788
Ostrove, S., 419 Potts, J. T., 248
Otte, A. P., 139 Prakash, V., 704
Ozols, J., 75 Pramauro, E., 608
Presta, L. G., 147
P Prickett, K. S., 247
Pabst, T. M., 364–365 Pringle, J. R., 815
Pace, C. N., 23, 83, 266 Prinz, B., 456
Padan, E., 445 Prinz, W. A., 156
Påhlman, S., 407 Prive, G. G., 614
Palumbo, J. L., 156 Probasco, M. D., 478, 490
Pan, C., 731 Prokipcak, R. D., 570
Panda, A. K., 262, 267, 272, 276–277 Przic, D. S., 114
Pandolfe, W. D., 291 Puig, O., 244
Pantoliano, M. W., 774 Puri, N. K., 267
Panvas, T., 251
Parks, B. A., 756–757 Q
Parsons, R. G., 431 Qasba, P. K., 559
Partridge, S. M., 352 Qing, G., 165
Passarelli, A. L., 194 Quronfleh, M. W., 249, 262, 271–272, 276
Patonay, G., 582
Patton, W. F., 543, 728 R
Paulick, M. G., 543
Payne, C. K., 229 Rabilloud, T., 527–529, 547, 553
Pédelacq, J. D., 243 Rabinowitz, J. C., 788
Pedersen, J., 449–450, 457 Raines, R. T., 248
Pedrioli, P. G., 740 Ralston, G., 687, 698, 700, 702–703
Peng, J., 739 Ramakrishnan, B., 559
Peng, S., 783 Ramanan, R. N., 291
Pennock, G., 193–194 Ramsby, M. L., 317
Perler, F. B., 253 Rankl, N. B., 199
Perrin, D. D., 43–45, 48–50 Rao, K. C., 751–752
Pertoft, H., 313 Rao, L., 492
Peter-Katalinic, J., 139 Rapoport, B., 199
Peterson, G. L., 79, 86–87, 624 Raulfs, E. C., 792
Pflieger, D., 746 Read, S. M., 86
Pham, P. L., 224–225, 227–,232 Reddy, R. C. K., 273
Phillips, G. N., Jr., 669 Reeves, P. J., 635
Pickart, C. M., 738 Regnier, F. E., 407, 751
Piehler, J., 445 Reichel, A., 446
Pijlman, G. P., 204–205 Reid, G., 726
Pikal, M. J., 114 Renkin, E. M., 420
Pinkse, M. W., 733 Reynolds, J. A., 499, 703
Pitteri, S. J., 728 Reznikof, W. F., 803
Pless, D. D., 737 Rhodes, D. G., 677–689, 691–721
830 Author Index

Rice, J. C., 741 Schimmel, P. R., 704

Richards, F. M., 248 Schindler, J., 622
Richardson, C. D., 200, 215 Schirch, V., 390, 394
Rickwood, D., 313 Schlaeger, E.-J., 229
Riddles, P. W., 817 Schley, C., 101
Rigaud, J. L., 613, 667 Schmidt, F. R., 164
Righetti, P. G., 519 Schmidt, T. G., 248
Rinas, U., 262, 267, 269, 272 Schmitt, J., 454
Roberts, W. K., 290 Schnitzler, G. R., 323
Robinson, A. S., 402 Schroder, E., 394
Rodbard, D., 500, 710 Schroeder, M. J., 735
Roeder, R. G., 455, 742 Schuck, P., 720
Romanos, M. A., 170 Schulenberg, B., 729
Romero, J. K., 97–118 Schumann, C., 401
Rong, W., 110 Schwarz, D., 649
Roodveldt, C., 244 Scigelova, M., 762
Rose, G. D., 408 Scopes, R. K., 82, 287, 293–294, 332, 341, 394
Ross, P. L., 314, 729, 751, 753–754 Scorer, C. A., 186
Rossi, F. M., 142 Scott, M., 137
Rossomando, E. F., 65 Seddon, A. M., 668
Roth, C. B., 634 Seelert, H., 566, 570
Roth, M. J., 726 Segel, I. H., 62, 65, 68
Roulland-Dussoix, D., 154 Sehgal, P. B., 320
Roux, B., 652 Seifter, S., 332, 335, 341
Rozema, D., 273 Seo, N. S., 213
Rumbley, J. N., 463 Serpa, G., 447
Rush, J., 728 Serrano-Vega, M. J., 634
Rydén, L., 354, 406, 411 Seta, N., 737
Seto, E., 741
S Shadforth, I., 315
Shah, M. B. 320
Saari, L. L., 788 Shapiro, A. L., 499
Sachdev, D., 244 Sharfstein, S. T., 229
Sadoul, K., 741 Sharpe, J. C., 818
Sadowski, P. G., 315 Shaw, A. Z., 164
Sahdev, S., 133, 140, 144, 249 Shaw, G., 224
Saiyed, Z., 423 Sheen, H., 103
Sakena, S., 113 Shen, J. W., 389
Saleh, L., 253 Shen, S., 186
Sali, A., 718 Shepard, S. R., 394
Salmon, K., 789 Sherman, 742
Sambrook, J., 170, 182 Sherrier, D. J., 314
Sanchez, J. C., 535 Shi, X.-L., 403
Sapan, C. V., 85 Shi, W., 447, 463
Sarkar, C. A., 634 Shi, X., 194, 212
Sarramegna, V., 632 Shi, Y., 455
Sasse, J., 585 Shiloach, J., 458
Sato, M., 578 Shimamoto, N., 273
Saunders, A. H., 803 Shimeld, S. M., 463
Sawasaki, T., 649–650, 652, 662–663, 667, 669 Shinkawa, T., 147
Schäfer, F., 439–468, 458, 460, 466 Shirgaonkar, I. Z., 291
Schaffner, W., 637, 642 Silla, T., 226
Schagger, H., 516 Simon, R. H., 324
Scheich, C., 460 Simpson, D., 417–435
Schein, C. H., 272, 278 Sinacola, J. R., 402
Scheraga, H. A., 408 Singh, S. M., 262, 267, 272, 276
Scheraga, H. A., 82 Sirard, J. C., 731
Author Index 831

Sittampalam, G. S., 83 Suehara, Y., 526

Sivaraman, T., 117 Sugyiama, K., 137
Sjoblom, J., 611 Sulkowski, E., 440
Skartsila, K., 394 Summers, M. D., 193, 196, 211, 215
Skerra, A., 248 Sun, L., 579
Slack, J. M., 208 Sun, X., 229, 447
Slentz, B. E., 447 Sunga, J., 169–187
Sletyr, U. B., 718 Surzycki, S., 292
Slotboom, D. J., 668 Sutton, V. R., 789–790, 792, 799, 802–803
Smejkal, G. B., 543 Svensson, H., 518
Smith, B. D., 817 Svensson, J., 450
Smith, G. E., 193–194, 196–197, 211, 214–215 Swaney, D. L., 747, 749–750
Smith, P. K., 79, 88, 624 Swart, R., 101
Smyth, D. R., 252 Swartz, J. R., 649, 669
Snyder, M. A., 387–402 Swiderek, K. M., 289
Sobel, R. E., 742 Swietnicki, W., 262, 276
Sobrado, P., 652 Syka, J. E., 729, 749
Sokolovsky, M., 741 Sykora, C., 732
Song, J., 267 T
Sorensen, H. P., 265
Sorensen, S. P. L., 50, 53, 55 Tait, A. S., 234
Soufi, B., 731 Takayama, Y., 164
Sowell, J., 582 Tanaka, K., 637
Spanos, N., 394 Tanford, C., 407–408
Speers, A. E., 667 Tang, W. H., 730
Spencer, D. M., 286 Taniguchi, C. M., 726, 731
Spencer, J. F. T., 286 Tao, W. A., 754
Spiess, C., 133 Tate, C. G., 632
Spinola, S. M., 586 Tempst, P., 444
Spirin, A. S., 649, 669 Tennikova, T. B., 355
Spiro, R. G., 737 Terp, K., 150, 156, 243
Spriestersbach, A., 439–468 Thao, S., 652
Sridhar, P., 199 Thatcher, D. R., 262
Stafford, W. F., 720 Thill, E. R., 185
Stalder, E. S., 475–493 Thingholm, T. E., 733
Stark, G. R., 819 Thomas, C. J., 208
Starkenstein, E., 419 Thompson, A. R., 388
Stasyk, T., 447 Thompson, E. B., 326
Steele, J. H. C., 703 Thompson, N., 265, 267
Steen, H., 733, 736 Thompson, N. E., 475–493
Steen, J., 461 Thomsen, D. R., 137
Steinberg, T. H., 541–559, 729, 757 Tian, Y., 738
Stellwagen, E., 373–385, 427 Tiffany, H. L., 500
Stemmann, O., 751 Tigges, M., 234
Stettler, M., 232 Timasheff, S. N., 390, 692, 694, 704, 716
Stevens, R. C., 286–287 Timmons, T. M., 498
Stock, A. M., 557 Tinoco, I., 697
Stoll, S. V., 43–56 Tipsmark, C. K., 575
Storrie, B., 311, 314 Tiselius, A., 388, 399
Stoscheck, C. M., 79, 83 Tolstorukov, I., 169–187
Stowers, A. W., 458 Tolun, A. A., 249
Stránská, J., 401 Tombs, M. P., 13
Strausberg, R. L., 135 Tomiya, N., 210
Strausberg, S. L., 135 Tonge, R., 556
Strömberg, P., 461 Topf, M., 694, 718
Strominger, J. L., 428 Towbin, H., 574
Studier, F. W., 154, 156, 160, 263, 265, 777 Townend, R., 694, 716
Stuhfelder, C., 401 Trésaugues, L., 271–272, 274
832 Author Index

Trelle, M. B., 742 Walpole, G. S., 51

Trimble, R. B., 136 Walsh, C. M., 731
Tropea, J. E., 252, 782 Walsh, P. S., 444, 448
Tsaprailis, G., 747 Wang, K. H., 245
Tsien, R. Y., 492 Warne, T., 634–636
Tsumoto, K., 262 Washburn, M. P., 307, 310
Tucker, J., 463, 634–635, 637–641 Waters, N. J., 107
Tyler, R. C., 651, 662, 669 Waugh, D. S., 134, 155, 243, 249, 774
Waziri, S. M., 107
U Weber, G., 307, 310
Weber, K., 499, 566
Udeshi, N. D., 748–750
Weigel, P. H., 144
Uhlén, M., 461
Weiß, H. M., 632, 634–636
Ulbrich, V., 409
Weissman, A. M., 738
Ulbrich-Hofmann, R., 531
Weissmann, C., 637, 642
Unger, T., 249
Wells, K. S., 733, 738
Unlu, M., 516, 524–525
Wen, J., 718
Unwin, R. D., 755
Wertz, D. H., 408
Ura, K., 726
Wessel, D., 117, 531
Urh, M., 417–435
Westermeier, R., 515–538, 547
Urlaub, G., 226
Westoby, M., 97–118
V Wetlaufer, D. B., 272
Wheat, T. E., 101
Vaillancourt, P., 249
White, F., 731–732, 753
Vallejo, L. F., 262, 267, 269, 272
White, J. F., 632, 634–638, 640–641
van Bruggen, E. J. F., 718
Wiebenga, E. H., 718
Van Craenenbroeck, K., 226
Wiechelman, K. J., 88
van Ooyen, A. J., 171
Wiener, M. C., 449, 463
van Veldhoven, P. P., 313
Wiggs, J., 569
Vancan, S., 447
Wilchek, M., 432
Vandecasteele, J. P., 788
Wilkinson, D. L., 25
Varshavsky, A., 24
Willemsen, O., 310
Vattem, K., 573–598
Williamson, B. L., 755
Vaughn, J. L., 211
Willis, M. S., 271–272
Veldkamp, C. T., 262, 276
Wilson, C. M., 509
Ventura, S., 260
Wilson, I. A., 286
Vera, A., 133, 265, 278
Wilt, F. H., 133
Vestal, M. L., 745
Witt, H. T., 615
Vesterberg, O., 518–519
Woestenenk, E. A., 451
Vialard, J. E., 198, 200
Wolf, B. D., 555
Vicennati, P., 229
Wolf-Yadlin, A., 755–756
Vijayendran, R. A., 432
Womack, M. D., 668
Villaverde, A., 260
Woycechowsky, K. J., 270
Vinarov, D. A., 649–651, 662–663
Woychik, N. A., 477
Vincentelli, R., 271–272, 274
Wrobel, R. L., 647–670
Vithayarhil, P. J., 248
Wu, A. M., 87
Volkin, D. B., 815
Wu, C. C., 667–668, 744
von Hagen, J., 305–327
Wu, J. C., 87
von Hippel, P. H., 23, 83
Wurm, F. M., 138
von Jagow, G., 516
Wuthrich, K., 615
W Wuu, J. J., 649
Wyckoff, M., 501, 506
Waddell, W. J., 13
Waggoner, A. S., 556 X
Wagner, S., 633
Waldo, G. S., 244, 774 Xia, W., 226–227
Walker, G. R., 575 Xie, Y., 272
Walker, J. E., 154 Xing, C., 578
Walker, J. M., 543 Xu, C.-G., 451
Author Index 833

Y Zappacosta, F., 737

Zeheb, R., 328
Yamada, S., 557 Zeman, L., 108, 111
Yamamoto, S., 352, 354, 361, 365, 412 Zerbs, S., 149–166
Yan, A., 787–803 Zerner, B., 83
Yang, J. T., 693, 711 Zhan, X., 741
Yang, X. J., 738, 741 Zhang, J., 224–226
Yeliseev, A. A., 633, 635 Zhang, Y. B., 251, 728–729, 741–743,
Yeo, W. S., 740–741 746, 751, 754
Yeung, K. K. C., 103 Zhao, K., 417–435
Yik, J. H. N., 144 Zhaohua, H., 445
Yoch, D. C., 788 Zhou, H., 444
Yokoyama, S., 243, 649 Zhu, B., 153
You, W. W., 79, 89, 91 Zhu, H., 286
Young, J. C., 133 Zhu, X. D., 388
Yphantis, D. A., 698, 701 Zimmerman, J. M., 408
Yue, S. T., 552 Zydney, A. L., 108, 111
Z
Zachariou, M., 419
Zacharius, R. M., 558
Zagariya, A., 324
Subject Index

A AmpholinesÒ , 519
Analytical size exclusion chromatography
Absolute quantitation (AQUA), 751 (ANSEC), 777
A431 cells (ATCC), 590 Antibody-affinity chromatography, 627
Acetate buffer, 51 Arginine, 273
Affinity chromatography Assay selection quantitation, 75–76
amine reactive linkages, 432 Autoinduction media protocol, 160
covalent attachment, 430
ligand selection B
characterization, 424–425
covalent affinity chromatography, 428–429 Bacterial systems, heterologous proteins
de novo development, 423 production
group specific ligands, 425 analysis
lectins, 426–427 autoinduction method, 159–160
specificity, 426 expression, 159
matrix selection solubility, 158–159
examples of, 421 autoinduction media protocol, 160
magnetic affinity beads, 422–423 cloning
pore size, 419 expression hosts and vectors, 153–155
Renkin equation, 420 ligation independent cloning (LIC)
specific ligand, 420 methods, 153
stability, 422 cytoplasmic expression, 156
purification method cytoplasmic targets expression and solubility,
elution, 434–435 157
GPCRs, 634–636 Escherichia coli, 150–151
sample preparation, 433 gram-negative and positive bacterium,
surface activation, 430 164–165
Affinity tags osmotic shock protocol, 162–164
basis, 243 periplasmic targets expression, 160–162
calmodulin-binding peptide (CBP)-tag, 248–249 planning and sequential steps, 151–152
epitope, 247–248 production scale, 166
GST, 246–247 project requirements evaluation, 152
His-tag, 245–246 target analysis, 152–153
S-tag, 248 T4 DNA polymerase-treated DNA fragments,
STREP-II, 248 155–156
Alkaline copper reduction assays. See Lowry assay vector and induction conditions, 165
Amine derivatization assay, 89–91 Bacteria lysis, 289
Amino acid sequence. See Bioinformatics, protein BaculoDirect approach, 205–206
purification Baculovirus–insect cell expression systems
2-Amino-2-methyl-1,3-propanediol (ammediol) baculovirus biology and molecular biology,
buffer, 54–55 193–195
Amino-terminal epitope tags, 739 basic protocols
Ammonium sulfate (AS) precipitation genomic DNA isolation, 215–216
concentration, 335 insect cell maintenance, 215
principles, 332–334 plaque assays, 216–218
problems/solutions, 337 conditions and methods, 211–212
procedure, 334 expression vector, 195–196
solubility curve, 333 foreign gene deliver and protein
test, 336 encoding, 211

835
836 Subject Index

Baculovirus–insect cell expression systems (cont.) denatured state, 27

genome modifications resources, 27
accessory genes, 208 Biological extracts preparation
BaculoDirect approach, 205–206 cell disruption
b-galactosidase substrate, 207 buffer composition, 293–294, 296
flashBAC approach, 204 chemical and enzymes, 287–290
linearized/gapped parental viral genome, mechanical, 290–292
201–202 E. coli cell lysis
linearized parental viral genome, 200–201 gram-scale mechanical disruption, 299–300
nonessential viral genes, 210 microscale protocols, 296–299
recombinant protein production, 209 Bio-Rad whole gel eluter, 570
transfer plasmid/bacmid, 203 N,N-Bis(2-hydroxyethyl)–2-
viral chitinase gene, 208–209 aminoethanesulfonic acid (BES), 390, 401
insect cell hosts, new generation of Blotting and optimization protocols,
eukaryotic protein process, 212 chemiluminescent substrate
foreign glycoproteins, 214 antigen concentration, 596
transgenic lepidopteran, 213 detection method, 598
modification categories, 198 enzyme conjugate concentration, 597–598
transfer plasmid modifications, 198–199 membrane blocking and washing, 596–597
vector technology primary antibody concentration, 597
homologous recombination, 197 Western blotting protocol
transfer plasmid, 196–197 electrophoretic transfer, 594
Baculovirus/insect cells, recombinant protein, primary and secondary antibody, 594
136–137 proteins separation, 593
Barbital buffer, 53–54 stripping protocol, 595–596
BCA assay. See Bicinchoninic acid assay Borax–NaOH buffer, 55
Bead homogenization, 291 Boric acid–borax buffer, 54
Bead milling, 291 Bradford protein assay. See Coomassie blue assay
BES. See N,N-Bis(2-hydroxyethyl)–2- Broad-range buffers, 50
aminoethanesulfonic acid Buffer capacity (b) vs. DpH, 45
Bicinchoninic acid (BCA) assay Buffers, principles and practice
procedure and comments, 89 broad-range buffers, 50
reagents, 88 preparation, 48–49
Bilayer reaction vs. dialysis reaction, 662–663 selection
Binding conditions, ion-exchange pH optimum, 45
chromatography pK values, 45–47
capacity vs. concentration, protein, 359 temperature-sensitive, 48
cation exchangers, 356 stock solutions, recipes
column volume, 358 acetate, 51
isoelectric point, 358, 360 2-amino-2-methyl-1,3-propanediol
ligand density, 355 (ammediol), 54–55
prepacked minicolumns, 360 barbital, 53–54
Biogrammatics expression vector map, 174–175 borax–NaOH, 55
Bioinformatics, protein purification boric acid–borax, 54
amino acid sequence cacodylate, 52–53
charge vs. pH, 22–23 carbonate–bicarbonate, 55–56
cysteine content, 23–24 citrate, 50–51
E. coli overexpression, solubility, 25 citrate-phosphate, 51–52
homology and cofactor affinity, 24–25 glycine–HCl, 50
hydrophobicity and membrane-spanning glycine–NaOH, 55
regions, 24 phosphate, 53
molar extinction coefficient, 23 succinate, 52
polypeptide chain molecular weight, 22 tris(hydroxymethyl)aminomethane (tris), 54
potential modification sites, 25 theory
secondary structure, 24 buffer capacity (b) vs. DpH, 45
stability, 24 Henderson–Hasselbalch equation, 44
titration curve, isoelectric point, 22–23 volatile buffers, 49–50
critical problems, 26 Bulk/batch purifications procedures, 17
Subject Index 837

C protocol, 659–660
reagents, 659
Cacodylate buffer, 52–53 proteoliposomes characterization, 667–668
Calcium-phosphate-mediated transfection, 229 scale-up considerations, 669
Calmodulin-binding peptide (CBP)-tag, 248–249 steps, 650–651
Carbonate–bicarbonate buffer, 55–56 transformation reaction
Carboxymethyl aspartate (CM-Asp), 442 materials and reagents, 658
Catalytic activity, principles protocol, 658–659
cell-free translation, 667–668 wheat germ translation reaction
chemical kinetics bilayer reaction protocol, 664
basic kinetic mechanism, 58–59 bilayer reaction vs. dialysis reaction,
first-order rate constant, 60–61 662–663
rate law, order, and rate constant, k, dialysis reaction protocol, 664–665
59–60
materials and reagents, 663–664
reactant concentration, 61–62
Ceramic fluoroapatite (CFT), 391
reaction velocity, 58, 60
Ceramic hydroxyapatite type I (CHT I),
enzyme kinetics
391, 397
initial velocity, 64
CFT. See Ceramic fluoroapatite
Michaelis constant, 63–64
Chaotropic agents, 273
steady-state approximation, 63
Charged-coupled device (CCD) cameras, 575
turnover number, 62
Chemical additives, 273
Cell disruption
Chemical and enzymatic cell disruption
buffer composition
bacteria lysis, 289
basic criteria, 294
high-throughput (HT) extraction, 287
protease inhibitors, 295
lytic enzyme treatments, 290
volume, 294, 296
reagents and enzymes, 288
chemical and enzymes, 287–290
yeast lysis, 289
bacteria lysis, 289
Chemiluminescence signal
high-throughput (HT) extraction, 287
signal capture, 583–584
lytic enzyme treatments, 290
signal intensity and duration, 584–585
reagents, 288
Western blot optimizing
yeast lysis, 289
antibodies, 586–587
E. coli cell lysis, 296–300
blotting membrane, 585
high-pressure homogenizers mechanism,
detection method, 587–588
290–292
Cell-free translation, integral membrane proteins signal reduction, chemical treatment, 587
density gradient ultracentrifugation, 665–667 target protein, 585–586
expression vectors, 651–652 CHO cell lines. See HEK293 cell lines
flexivector and PCR product digestion CHT I. See Ceramic hydroxyapatite type I
reaction Citrate-phosphate buffer, 50–52
protocol, 657 Cleavage sites, 245
reagents, 656 Click chemistry reagents, 433, 559
gene cloning Cloning, bacterial systems
materials and reagents, 654 expression hosts and vectors, 153–155
protocol, 654–655 ligation independent cloning (LIC) methods,
sacB-CAT cassette, 652–653 153
isotopic labeling, 669 Cloud point and phase separation, 611–612
ligation reaction, 657–658 CMC. See Critical micellization concentration
liposomes preparation CNBr. See Cyanogen bromide
materials and reagents, 661 Collision induced dissociation (CID), 747
protocol, 661–662 Column chromatography, 708
mRNA preparation Conacanavalin A (ConA), lectins, 427
protocol, 661 Continuous assays, 65–67
reagents, 660 Coomassie blue assay
PCR product cleanup observation, 86
materials and reagents, 655–656 procedure, 85–86
protocol, 656 reagents, 85
plasmid DNA purification Coomassie brilliant blue (CBB), 546–547
838 Subject Index

Copurifying proteins, IMAC temperature effects, 611–612

affinity precipitation, 457 Dialysis
bind–wash–elute procedure, 456 affinity binding applications and concentration,
chelating amino acids, 455 107
removal of, 459 buffer concentration, 104–105
size exclusion chromatography (SEC) columns, mechanical design, 106
455 molecular weight cutoff (MWCO), 104
Core histone purification method, 324–325 Difference gel electrophoresis (DIGE), 525–526
Covalent affinity chromatography, 428–429 Differential detergent fractionation (DDF)
Critical micellization concentration (CMC), extraction efficacy, 317
609–610 procedure, 319–320
Crystallization, 118 reagents and equipment, 318–319
Culture media and microbial manipulation Diffusion coefficient, 703
techniques, 171–172 Diffusion, protein elution
Cyanogen bromide (CNBr), 487 Bio-Rad whole gel eluter, 570
Cytoplasmic proteins, expressing method, 140–141 electrophoretic elution, 570–571
Cytoplasmic targets expression and solubility, 157 gel preparation, 567
location, 567
D renaturation, 569
reverse phase HPLC, 570
DDF. See Differential detergent fractionation SDS removal and concentration, 568–569
Denaturing PAGE analysis, cell-free translation, Discontinuous assays, 67–69
665–666 Disposable WaveTM bioreactor, 232
Density velocity centrifugation, 311 Dithionite (DTH), 790, 793
advantage, 308 Dithiothreitol (DTT), 793, 796
mechanical homogenate, 310 Divalent cations, 272–273
procedure, 309–310 DNA preparation, Pichia pastoris expression
proteome complexity reduction electroporation procedure, 178–179
free-flow electrophoresis, 310 protocol, 177–178
multidimensional protein identification solutions prepare, 177
technology (MudPIT), 310 Drip dilution, 267–268
Detection methods, immunodetection Dye-based protein assay, 75, 80
assay requirements, 39–40
chemiluminescence detection, 582–583 E
colorimetric detection, 581
enzyme conjugate, 579 EBNA-1 gene, 224
fluorescence detection E. coli cell lysis
single blot and high sensitivity, 581 gram-scale mechanical disruption
spectral properties, 582 high-pressure homogenization, 300
Detergent-resistant fractionation. See Lipid raft sonication, 299–300
enrichment microscale protocols
Detergents, 273 detergent-based reagent method, 296–298
fluorescent detection, 91 freeze-thaw enzymatic method, 298–299
GPCRs analysis, 636–637 recombinant protein expressing method
lipid–receptor, 633 disulfide bond formation, 134
mass spectrometry and nuclear magnetic fusion partners, 134
resonance (NMR) measurements, 615 inclusion bodies, 133
membrane protein suitable detergent, 613–614 posttranslational modifications, 134
optical spectroscopy, 614 temperature and molecular chaperones, 133
phase diagrams and critical micellization Electron-capture dissociation (ECD), 747–748
concentration, 609–610 Electron microscopy, 718
physicochemical parameters, 612–613 Electron transfer dissociation (ETD), 748
properties, 605–609 Electrophoretic elution, 570–571
protein crystallization, 615 Electrophoretic method, 103–104
reagent lysis, 296–298 principle, 709
removal and exchange, 613 protein purity determination
solubility, 611–612 contaminants, 681
structure, 604–605 front-line, 682
Subject Index 839

gradient gels, 682 small-scale expression screening

isoelectric focusing gel, 683 analytical methods, 776
protein stains, 684 benzonase, 775
SDS gels, 710 capillary electrophoresis result, 774
Electroporation procedure and parameters, Thermofluor ligand, 772, 774
178–179 Expression vector technology
Electrospray ionization (ESI), 744–745 homologous recombination, 197
ELISA-elution assay, 480–482 transfer plasmid, 196–197
Elution volume (Ev), 707
Endomembrane enrichment F
enzyme and substrate markers, 315
material and equipment, 315–316 Field-flow fractionation (FFF), 711–712
procedure, 316 First-order rate constant, 60–61
workflow, 314 FLAG-tag, 247–248
Enrichment techniques flashBAC approach, 204
glycosylation Flash dilution, 267
lectin affinity chromatography, 738 Flexivector and PCR product digestion reaction,
O-b-N-acetylglucosamine (O-GlcNAc), 656–657
737 Fluorescent total protein stains
phosphorylation detection sensitivity, 549–550
32 epicocconone, 554
P-labeled ATP, 731
mass spectrometry-based proteomics, flamingo and krypton protein gel stain, 555
731–737 fluorescein derivatives, 555
ubiquitination and sumoylation LUCY protein gel stains, 555
amino-terminal epitope tags, 739 SYPRO stains
characteristic mass shifts, 740 carbazolylvinyl stain, 553
Enzyme activity measurement organometallic ruthenium dyes, 553
catalytic activity, principles protein–SDS–dye complex, 552
chemical kinetics, 58–62 sulfopropylaminostyryl, 552
enzyme kinetics, 62–64 [4Fe-4S]-FNR, O2-sensitive proteins isolation
continuous assays, 65–67 anaerobic glove-box vinyl chamber, 790
discontinuous assays, 67–69 buffer composition, 793
reaction assay mixtures formulation, 69–70 cell lysate preparation, 792–793
Enzyme and substrate markers, 315 dithionite (DTH), 790, 793
Enzyme-catalyzed product, 69 dithiothreitol (DTT), 793, 796
Enzymes purification, 1–5 57Fe labeled 4Fe-FNR purification, 798–799
Epicocconone, 554 Fourier transform ion cyclotron resonance
Epitope tags, 247–248 (FT-ICR), 799
Epstein–Barr virus nuclear antigen 1 (EBNA1), FPLC purification system, 791
341 ICP-MS analysis, 801–802
Ethanol and acetone precipitation, 341 low-temperature EPR, 802–803
Expression and purification, parallel methods Mössbauer spectroscopic analysis, 802
analytical testing, proteins PK872, E. coli strain, 791
ANSEC, 777 protein purification, 793–794
methods, 776 protocols, 793–797
cloning strategies sparging station, 790
gel electrophoresis, 771 trace RNAses removal, 797–798
ligation-independent-cloning (LIC), 769 UV–visible spectrophotometric analysis, 802
parallel cloning process, 770 Foreign gene deliver and protein encoding, 211
polymerase incomplete primer extension Forgotten artifacts, protein biochemistry
(PIPE), 769 buffers
restriction mapping, 771 reducing agent vs. oxidizing agent, 816–817
SYBR results, 772 sulfonylethyl buffers, 817
large-scale parallel expression chemical leaching, plasticware, 819
airlift fermentor, 778 chromatography, EDTA
custom automation, 778–779 anion-exchange columns, 817–818
digestion buffer, 781 Ni-chelate affinity column, 818
protein expression identification, 767 cyanate, urea, 819
840 Subject Index

Forgotten artifacts, protein biochemistry (cont.) transfer plasmid/bacmid, 203

protein absorption, filtration, 818–819 viral chitinase gene, 208–209
SDS gel electrophoresis Genomic DNA isolation, baculovirus, 215–216
Asp–Pro bond cleavage, 815 Glutathione-S-transferase (GST) tag, 246–247
keratin contamination, 816 Glycerol, 273
proteases, 815 Glycine
Fourier transform ion cyclotron resonance (FT- HCl buffer, 50
ICR), 799 NaOH buffer, 55
Fractionation requirements, laboratory setting, Glycoprotein and phosphoprotein detection
40–42 acid fuchsin dye, 558
Free-flow electrophoresis, 310 glycosylation state, 558–559
Freeze-thaw plus enzymatic lysis, 298–299 O-GlcNAc detection, 559
Frictional coefficient ratio (f/f0), 703–704 Phos-tag phosphoprotein stains, 557
Fusion protein. See Protein expression tagging Pro-Q diamond phosphoprotein gel stain, 557
Glycosylation
G lectin affinity chromatography, 738
O-b-N-acetylglucosamine (O-GlcNAc), 737
Gel filtration GPCRs stability, detergent solution, 634
conventional matrix G-protein-coupled receptors (GPCRs). See
absorbance, 383 Recombinant G-protein-coupled receptors
column, 382 Gram-scale cell disruption. See Mechanical cell
desired protein disappearance, 385 disruption
effluent fractions, 381 Green fluorescent protein (GFP), 492
hydrodynamic diameter, 374 GuHCl, 266
low flow rate, 384
matrices H
chromatograhic solvents, 380
conventional, 376 Halotag technology, 251, 428–429
high-performance, 376 HEK293 cell lines
packed column, 379 cultivation, 228
parameters, 377 EBNA-1 gene, 224
powdered and suspended parameters, expression vectors, 226–227
378–379 HKB-11 cell line, 226
sample preparation, 380 overview, 225–226
resolution, 384 SV40 enhancer, 224, 226
scaling upward, 383 transfection methods, 228–234
semilogarithmic plot, 375 Heterologous protein production, E. coli,
size-exclusion matrix, 375, 381 150–151
skewed peaks, 384–385 HIC. See Hydrophobic interaction
zero elution volume, 374 chromatography
Gel-filtration chromatography, 99 High-performance liquid chromatography
method (HPLC), 67
column chromatography, 708 High-pressure homogenization lysis method, 300
elution and void volume, 707 High-pressure homogenizers, 291–292
problems and pitfalls, 708 High-resolution ion-exchange chromatography,
stationary gel volume fraction, 706 268–269
Gene cloning, cell-free translation, 654–655 High-throughput (HT) cell extraction, 287
Genetic strain construct, Pichia pastoris, 172–174 His tagged protein purification, IMAC, 245–246
Genome modifications, baculovirus applications
accessory genes, 208 detection and immobilization, 445–446
BaculoDirect approach, 205–206 protein fractions purification, 446–448
flashBAC approach, 204 carboxymethyl aspartate (CM-Asp), 442
b-galactosidase substrate, 207 chemical compatibility, 452–453
linearized/gapped parental viral genome, copurifying proteins
201–202 affinity precipitation, 457
linearized parental viral genome, 200–201 bind–wash–elute procedure, 456
nonessential viral genes, 210 chelating amino acids, 455
recombinant protein production, 209 removal of, 459
Subject Index 841

size exclusion chromatography (SEC) rEPO, 401

columns, 455 solubility constant, 394
detergent screening, 462 voids, HA solubility, 394
iminodiacetic acid (IDA), 440, 442
industrial-scale protein production, 458–460 I
interaction model, 441
membrane protein purification, 461–463 Iminodiacetic acid (IDA), 440, 442
metal-ion stripping, 467 Immobilized metal affinity chromatography
nitrilotriacetic acid (NTA), 440–441 (IMAC), 732–733
protein expression, 448–449 applications
protein purification protocols detection and immobilization, 445–446
denaturing conditions, 466 protein fractions purification, 446–448
native conditions, 465 carboxymethyl aspartate (CM-Asp), 442
reducing conditions, 454 chemical compatibility, 452–453
sanitization, 466–467 copurifying proteins
sequences, 449 affinity precipitation, 457
n-terminal tags, 449 bind–wash–elute procedure, 456
zinc-finger proteins purification, 463–464 chelating amino acids, 455
HKB-11 cell line, 226 removal of, 459
Homologous recombination vector technology, size exclusion chromatography (SEC)
197 columns, 455
Homology and cofactor affinity, 24–25 detergent screening, 462
Hydrophobic and affinity chromatography, His tagged protein purification
100–102 protein expression, 448–449
Hydrophobic interaction chromatography (HIC) sequences, 449
adsorbents N-termianl tags, 449
choice of, 409 iminodiacetic acid (IDA), 440, 442
feed/load preparation, 410 industrial-scale protein production, 458–460
gradient elutions, 412 interaction model, 441
isocratic elution, 412–413 media
p
p interactions, 409 His-tag, 245–246
preparation, 410 tag use, 243
product elutions, 410–412 membrane protein purification, 461–463
properties of, 408 metal-ion stripping, 467
regeneration and sanitization, 413 nitrilotriacetic acid (NTA), 440–441
chaotropes, 407 protein purification protocols
Hofmeister series, 407 denaturing conditions, 466
lyotropes, 407 native conditions, 465
polarity, 407 reducing conditions, 454
proteins vs. hydrophobic ligand, 406 sanitization, 466–467
Hydroxyapatite (HA) columns, protein zinc-finger proteins purification, 463–464
chromatography Immobilized pH gradient (IPG), 519–520
BES, 390, 401 Immunoaffinity reagents enrichment
calcium ions, 396 antibodies, 326
chemical characteristics material and equipment, 326
protein separation, 392 procedure, 326–327
pyrophosphate (PPi), 392 Immunodetection
effluent pH, 395 blotting and optimization protocols,
hydroxyapatite sources, 399 chemiluminescent substrate
mechanism antigen concentration, 596
CHT I and CFT, elution time, 391 detection method, 598
desorption, 390 enzyme conjugate concentration, 597–598
P-and C-sites, 389 membrane blocking and washing, 596–597
zero net charge, 389 primary antibody concentration, 597
MES, 394 Western blotting protocol, 593–596
metal adsorption, 393–394 chemiluminescence signal
phosphate gradient, 401 signal capture, 583–584
process-scale columns, 399–401 signal intensity and duration, 584–585
842 Subject Index

Immunodetection (cont.) capacity vs. concentration, protein, 359

Western blot optimizing, 585–588 cation exchangers, 356
detection methods column volume, 358
chemiluminescence detection, 582–583 isoelectric point, 358, 360
colorimetric detection, 581 ligand density, 355
enzyme conjugate, 579 prepacked minicolumns, 360
fluorescence detection, 581–582 column operation conditions
problems and explanations equilibration, 363
bands/entire blot glowing, darkroom, 590 gradient volume, 364
brown/yellow bands, 589 oxidizing agents, 366
ghost/hollow bands, 590–593 pH gradient elution, 365
high background, 589 salt loading, 363
no signal, 588 complex protein mixture separation
signal fades quickly, 588–589 cation exchange resins, 366
substrate different, 589 milk, pretreatment, 366
Western blotting elution conditions
advances, 575 diplacement chromatography, 361
charged-coupled device (CCD) cameras, a-lactoglobulin isoform separation, 362
575 resolution, 362
direct and indirect, 575–577 monolithic column, high-resolution separation
far-Western, 577 bovine whey proteins purification, 368
semiquantitative, 577–580 MP production, 367
Inclusion body (IB) proteins, 241 pH vs. concentration gradient, 370
characterization, 270 Superdex 200, 369
disulfide bonds principle
formation methods, 269–270 isocratic elution, 352
reoxidation, 269 pH gradients, 352
folding process, 261 protein retention vs. salt concentration, 353
high-resolution ion-exchange salting-in effect, 350
chromatography, 268–269 stationary phases
native state, 260 binding site vs. molecular mass, 354
overexpression, 264–265 monoliths, 355
procedure steps, 262 pore size, 354
protocol steps, 263–264 Ion trap mass spectrometers, 744
refolding Isoelectric focusing (IEF)
alkaline pH shift, 276–277 high resolution 2D gel, 517
catalysts, 276 pH gradients types
characterization, 270 carrier ampholytes, 518–519
database, 277 different configurations, 521
dilute ways, 267–268 IPG strip technology, 519–520
high-pressure, 276 principle, 517–518
on-column, 275–276 Isoelectric precipitation, 341
problem, 262 Isotope-coded affinity tags (ICAT), 752–753
rational approach, 274–275 Isotope tags for relative and absolute
systematic screens, 271 quantification (iTRAQ), 753–754
variables, 271–274 Isotopic labeling, cell-free translation, 669
solubilizing, 265–267
washing, 265 K
Inductively coupled plasma mass spectrometry
Krafft point, 611–612
(ICP-MS), 801–802
Insect cell hosts, baculovirus L
eukaryotic protein processing, 212
foreign glycoproteins, 214 Laboratory setting
transgenic lepidopteran, 213 chemicals, 38
Insect cell maintenance, 215 detection and assay requirements, 39–40
p
p Interactions, 409 disposables, 38
Ion-exchange chromatography, 99 equipment and apparatus, 39
binding conditions fractionation requirements, 40–42
Subject Index 843

glassware and plasticware, 38 detergent removal and exchange, 627–628

small equipment and accessories, 39 lectin-affinity chromatography, 626–627
b-Lactamase expression, 181 ligand-affinity chromatograph, 627
Lectin affinity chromatography, 626–627, 738 precautions, 625–626
Lectins, ligand selection, 426–427 preparation
Ligand-affinity chromatograph, 627 advantage, 622
Ligation independent cloning (LIC) methods, 153 membrane fractionation methods, 621
Ligation reaction, cell-free translation, 657–658 plasma membranes isolation, 620
Linear map of pEU-HSCB, 652–653 recombinant integral membrane proteins,
Lipid raft enrichment expression and purification, 628–629
material and equipment, 320 solubilization, native membrane proteins
mild detergents, 321 detergents choice, 623
procedure, 322 preparing membrane fractions, 624–625
Lipofection. See Small-scale transfection methods screening, 624
Liposomes preparation, 661–662 Membranes preparation
LOPIT. See Endomembrane enrichment advantage, 622
Loss of activity, protein stability, 125–127 membrane fractionation methods, 621
Lowry assay plasma membranes isolation, 620
comments, 88 2-Mercaptoethanol/dithiothreitol, 499
procedure, 87 MES. See 4-Morpholinepropanesulfonic acid
reagents, 87 Metalloproteome, IMAC, 447
Lyophilization, 114–115 Methylation and acetylation
antibodies, 742
M histone, 741
mAb 8RB13, 490 immonium ions, 742–743
Maltose-binding protein (MBP) tag, 249–250 Michaelis constant, 63–64
Mammalian expression methods, 138–139 Microfluidics microfluidizer, 292
Manganese peroxidase (MP) production, 367 Micro-scale cell disruption. See Chemical and
Marker proteins, 510–511 enzymatic cell disruption
Mass spectrometry Mitochondria enrichment
analysis density gradient formulations, 313
electrospray ionization (ESI), 744–745 drawback, 311
features, 744 material and equipment, 312
matrix assisted laser desorption ionization procedure, 312–313
(MALDI), 746 Molar absorbance extinction
detergents, 615 coefficients (e), 65
methods, 687–688, 713–715 Molar extinction coefficient, 23
Matrices, gel filtration 4-Morpholinepropanesulfonic acid (MES),
chromatograhic solvents, 380 cobuffer, 394–396
conventional, 376 Mössbauer spectroscopic analysis, 798–799, 802
high-performance, 376 mRNA preparation, cell-free translation,
packed column, 379 660–661
parameters, 377 MudPIT. See Multidimensional protein
powdered and suspended parameters, 378–379 identification technology
sample preparation, 380 Multidimensional protein identification
Matrix assisted laser desorption ionization technology (MudPIT), 310
(MALDI), 746 Multiple angle light scattering (MALS), 688–689
Mechanical cell disruption Multiple reaction monitoring (MRM), 755
instrumentation Multiple tags. See Tandem tags
bead milling, 291 Multisubunit features, critical problems, 26
microfluidics microfluidizer, 292 M11V3 cultivation medium
mechanisms, 290–291 large-scale transfection, 232–233
Medium-scale transfection methods, 231–232 medium scale transfection, 231–232
Membrane and secreted proteins, expressing N
method, 141
Membrane fractionation methods, 621 Native membrane proteins solubilization, 623
Membrane proteins purification Nitrilotriacetic acid (NTA), 440–441
antibody-affinity chromatography, 627 Nitrosative protein modifications, 740–741
844 Subject Index

Nonionic polymer precipitation. See 57Fe labeled 4Fe-FNR purification,

Polyethylene glycol precipitation 798–799
Nonultracentrifugation-based organelle FPLC purification system, 791
enrichment technique. See Differential PK872, E. coli strain, 791
detergent fractionation (DDF) protein purification, 793–794
NTS1 fusion protein protocols, 793–797
analysis, 641–642 sparging station, 790
purification trace RNAses removal, 797–798
immobilized metal affinity chromatography, [4Fe–4S]2þ cluster characterization
640 Fourier transform ion cyclotron resonance
neurotensin column, 640–641 (FT-ICR), 799
solubilization, 639–640 ICP-MS analysis, 801–802
Nuclear extract enrichment, 325 low-temperature EPR, 802–803
Nuclear magnetic resonance (NMR) Mössbauer spectroscopic analysis, 802
measurements, 615 UV–visible spectrophotometric analysis,
Nuclei and histone enrichment 802
core histone purification, 324–325 Osmotic shock protocol, 162–164
washed nuclear pellet
materials and equipment, 323 P
mechanism, 322
procedure, 323–324 Parallel methods, expression and purification
Nucleic acid removal, affi-gel blue analytical testing, proteins
Cibacron Blue F3GA dye, 344 ANSEC, 777
column buffer, 345 methods, 776
NusA tag, 250 cloning strategies
gel electrophoresis, 771
O ligation-independent-cloning (LIC), 769
parallel cloning process, 770
O-b-N-acetylglucosamine (O-GlcNAc), 737 polymerase incomplete primer extension
One-dimensional gel electrophoresis (PIPE), 769
marker proteins, 510–511 restriction mapping, 771
2-mercaptoethanol/dithiothreitol, 499 SYBR results, 772
molecular weight determination, 511 large-scale parallel expression
polyacrylamide gels, 500 airlift fermentor, 778
preparative electrophoresis, 511–513 custom automation, 778–779
principle, 501 digestion buffer, 781
protein detection, gel protein expression identification, 767
Coomassie Brilliant Blue R-248, 508–509 small-scale expression screening
copper staining, 510 analytical methods, 776
silver staining, 509–510 benzonase, 775
SDS–PAGE procedure capillary electrophoresis result, 774
casting gels, 503–505 Thermofluor ligand, 772, 774
catalyst, 503 PCR product cleanup, 655–656
electrode buffer, 503 Periplasmic targets expression, 161–162
electrophoresis, 505–506 Phase diagrams, 609–610
2-mercaptoethanol, 507 Phosphate buffer, 53
Ornstein–Davis system, 506 Phosphorylation
sample preparation, 505 mass spectrometry-based proteomics
stock solutions, 502 chemical affinity purification, 733–734
Optical spectroscopy, detergents, 614 immobilized metal affinity chromatography
Ornstein–Davis system, 506 (IMAC), 732–733
O2-sensitive proteins isolation neutral loss artifacts, 735
dithionite (DTH), 790, 793 precursor ion scanning technique, 736–737
dithiothreitol (DTT), 793, 796 strong cation exchange (SCX) resin, 732
32
4Fe-FNR, anaerobic isolation P-labeled ATP, 731
anaerobic glove-box vinyl chamber, 790 Physicochemical parameters, detergents, 612–613
buffer composition, 793 Pichia pastoris expression
cell lysate preparation, 792–793 Arxula adeninivorans, 171
Subject Index 845

culture media and microbial manipulation o-phenylenediamine (OPD), 480

techniques, 171–172 specific antibody production, 480–481
DNA preparation immobilization, 487–488
electroporation procedure, 178–179 mAbs source, 479
protocol, 177–178 properties of, 478–479
solutions preparation, 177 protein purification
expression cassette, multiple copies epitopes, 491–492
general approaches, 185–186 RNA polymerase, 488
selection procedure, drawback, 186–187 SDS–PAGE, 489
gene preparation and vector selection using cross-reacting PR-mAbs, 490–491
biogrammatics expression vector map, Posttranslational modifications (PTMs)
174–175 collision induced dissociation (CID), 747
nucleotide sequences, 176 electron-capture dissociation (ECD), 747–748
genetic strain construct electron transfer dissociation (ETD), 748
mating, creating diploids, 172–173 enrichment techniques
random spore analysis, 173–174 glycosylation, 737–738
Hansenula polymorpha, 170–171 phosphorylation, 731–737
Kluyveromyces lactis, 171 ubiquitination and sumoylation, 738–740
Pichia methanolica, 170–171 identification, 184–185
posttranslational modification, 184–185 mass spectrometry analysis
strains, recombinant protein production electrospray ionization (ESI), 744–745
cell extracts preparation, 182 features, 744
b-lactamase expression, 181 matrix assisted laser desorption ionization
plate activity assay, 180 (MALDI), 746
transformation, electroporation, 176 methylation and acetylation
Yarrowia lipolytica, 171 antibodies, 742
Yeastern blot assay development, 182–184 histone, 741
Pichia pastoris, recombinant protein expressing immonium ions, 742–743
method, 135–136 nitrosative, 740–741
Plaque assays, 216–218 quantification
Plasma membranes isolation, 620 meaurement forms, 750
Plasmid/bacmid transposition, 203 stable isotope dilution theory approach,
Plasmid DNA purification, 659–660 751–756
Plate activity assay, 180 Precipitation
Polyethylene glycol precipitation, 341 critical problems, 26
Polyethyleneimine (PEI) precipitation, 337 proteins concentration, 116–118
vs. AS precipitation, 339–340 Precursor ion scanning technique, 736–737
DNA binding capacity, 340 Primary and secondary drying, lyophilization, 114
procedure, 338–339 PR-mAbs. See Polyol-responsive monoclonal
test, 340 antibodies
usage strategies, 338 PR-mAbs, immunoaffinity chromatography
PEI-mediated transfection, 230. See also antibody purification, 485–487
Large-scale transfection methods; continuous culture, mAbs productions
Medium-scale transfection methods cell compartment harvest, 484
Polymin P. See Polyethyleneimine (PEI) CELLline flask 350 (CL 350), 482–483
precipitation Dulbecco’s modified eagle medium
Polyol-responsive monoclonal antibodies (DMEM), 482
(PR-mAbs) total viable cell count, 485
antibody purification, 485–487 ELISA-elution assay
continuous culture, mAbs productions characterization, 481
cell compartment harvest, 484 false-positive, 482
CELLline flask 350 (CL 350), 482–483 o-phenylenediamine (OPD), 480
Dulbecco’s modified eagle medium specific antibody production, 480–481
(DMEM), 482 immobilization, 487–488
total viable cell count, 485 mAbs source, 479
ELISA-elution assay properties of, 478–479
characterization, 481 protein purification
false-positive, 482 cross-reacting PR-mAbs, 490–491
846 Subject Index

PR-mAbs, immunoaffinity chromatography solubility tags

(cont.) basis, 242–243
epitopes, 491–492 common tags, 241
RNA polymerase, 488 maltose-binding protein (MBP), 249–250
SDS–PAGE, 489 NusA, 250
Protease inhibitors, 125, 295 smaller solubility enhancement tags (SETs),
Proteinases and glycosylation, 184–185 250–251
Protein chromatography, HA columns solubility enhancement peptide (SEP) tags,
BES, 390, 401 250–251
calcium ions, 396 thioredoxin (Trx), 250
chemical characteristics tags removal, 251–253
protein separation, 392 Protein gel staining methods
pyrophosphate (PPi), 392 colorimetric total protein stains
effluent pH, 395 Coomassie brilliant blue (CBB), 546–547
hydroxyapatite sources, 399 silver staining, 547–548
mechanism Zn2þreverse staining, 548–549
CHT I and CFT, elution time, 391 fluorescent total protein stains
desorption, 390 detection sensitivity, 549–550
P-and C-sites, 389 epicocconone, 554
zero net charge, 389 flamingo and krypton protein gel stain, 555
MES, 394 fluorescein derivatives, 555
metal adsorption, 393–394 LUCY protein gel stains, 555
phosphate gradient, 401 SYPRO stains, 552–554
process-scale columns, 399–401 glycoprotein detection
rEPO, 401 acid fuchsin dye, 558
solubility constant, 394 glycosylation state, 558–559
voids, HA solubility, 394 O-GlcNAc detection, 559
Protein colligative phenomena, 697 instrumentation, 543
Protein composition phosphoprotein detection
analysis, 695 Phos-tag phosphoprotein stains, 557
method, 696 Pro-Q Diamond phosphoprotein gel stain,
problems and pitfalls, 696–697 557
Protein crystallization, 615 preelectrophoresis sample labeling, 556
Protein elution, gels zymography, 543
electrophoretic elution, diffusion, 570–571 Protein inactivation, causes, 121–122
gel preparation, 567 Protein posttranslational modifications (PTMs)
location, 567 CID vs. ETD, 749–750
renaturation, 569 collision induced dissociation (CID), 747
reverse phase HPLC, 570 ECD vs. ETD, 749
SDS removal and concentration, 568–569 electron-capture dissociation (ECD), 747–748
Protein expression tagging electron transfer dissociation (ETD), 748
affinity tags identification
basis, 243 enrichment techniques, 731–740
calmodulin-binding peptide (CBP), 248–249 mass spectrometry analysis, 744–746
common tags, 241 methylation and acetylation, 741–743
epitope, 247–248 nitrosative, 740–741
glutathione-S-transferase (GST), 246–247 proteomic techniques
His-tag, 245–246 bottom-up, 757–758
S-tag, 248 mass spectrometry-based, 727–729
STREP-II, 248 top-down, 756–757
designing quantification
affinity and solubility, 241–243 meaurement forms, 750
choice, 243–244 stable isotope dilution theory approach,
N-/C-terminal, 244–245 751–756
tags removal, 245 Protein precipitation techniques
tandem tags, 244 ammonium sulfate
endoproteases, 242 concentration, 335
schematics, 242 principles, 332–334
Subject Index 847

problems/solutions, 337 flow chart, purification procedures, 75–76

procedure, 334 instructions
solubility curve, 333 cuvettes, 91–92
test, 336 interfering substrates, 92–93
ethanol and acetone, 341 microwell plates, 92–93
isoelectric and thermal, 341 Lowry assay
polyethylene glycol, 341 comments, 88
polyethyleneimine (PEI) procedure, 87
vs. AS precipitation, 339–340 reagents, 87
DNA binding capacity, 340 substance compatibility, 75, 77–79
procedure, 338–339 ultraviolet absorbance, 205 nm
test, 340 concentration standards, 83–58
usage strategies, 338 dye-based protein assays, 83
procedures, 341–342 extinction coefficient (e) calculation, 82–83
Protein purifications, strategies and considerations ultraviolet absorption spectroscopy, 280 nm,
assays, 11–13 80–81
bulk/batch procedures, 17 Proteins and solutes removal, concentration
contaminating activities, 15 chromatography
extracts preparation, 16–17 applications, manufacturing-scale, 102–103
properties and sensitivities, 10 gel filtration, 99
refined procedures hydrophobic and affinity, 100–102
high-capacity steps, 18–19 ion-exchange, 99
intermediate-capacity steps, 19 reversed phase, 100
low-capacity steps, 19 crystallization, 118
source dialysis
overexpressed protein, 16 affinity binding applications and
sequence information, 15–16 concentration, 107
storage, protein solutions, 14–15 buffer concentration, 104–105
suspension, storage, and assay buffers, 13–14 mechanical design, 106
use, 10–11 molecular weight cutoff (MWCO), 104
Protein purity determination electrophoresis, 103–104
chromatographic methods lyophilization
gel filtration method, 685–686 process and key system, 114–115
reversed phase HPLC, 686 steps, 114
composition and activity based analyses, precipitation, 116–118
680–681 ultrafiltration
electrophoretic method convective and diffusive solute transport,
contaminants, 681 108–109
front-line, 682 devices, 110–112
gradient gels, 682 membranes, 109–110
isoelectric focusing gel, 683 modeling approach, 108
protein stains, 684 purification applications, 112–113
light scattering methods, 688–689 Protein size determination
mass spectrometry methods, 687–688 categorization, 692
methods for, 679 chemical methods
sedimentation velocity methods, 687 colligative properties, 697
Protein quantitation composition, 695–697
amine derivatization, 89–91 scattering methods
assay selection, 75–76 dynamic, 717
bicinchoninic acid (BCA) electron microscopy, 718
procedure and comments, 89 static, 716–717
reagents, 88 subunits
Coomassie blue associated and dissociated states, 719
observation, 86 stoichiometry, 720–721
procedure, 85–86 techniques comparison, 693–694
reagents, 85 transport methods
detergent-based fluorescent detection, 91 electrophoresis, 708–711
dye-based protein assay, 75, 80 field-flow fractionation (FFF), 711–712
848 Subject Index

Protein size determination (cont.) contaminating activities, 15

gel-filtration chromatography, 706–708 extracts preparation, 16–17
mass spectrometry, 712–716 properties and sensitivities, 10
sedimentation equilibrium, 698–701 refined procedures, 18–19
sedimentation velocity, 701–705 source, 15–16
viscosity, 711 storage, protein solutions, 14–15
Protein stability suspension, storage, and assay buffers, 13–14
concentration and solvent conditions, 122–123 use, 10–11
inactivation causes, 121–122 resources, 27
loss of activity, 125–127 summary table
procedures, 122 footnotes, 32
proteolysis and protease inhibitors, 124–125 mistakes and problems, 32–34
trials and storage conditions, 123–124 SDS–polyacrylamide gel analysis, 32
Proteoliposomes steps, 30–31
characterization, 667–668
purification, 664–665 R
Proteolysis and protease inhibitors, 124–125
Proteomic techniques Reactant concentration, 61–62
bottom-up, 757–758 Reaction assay mixtures formulation, 69–70
mass spectrometry-based, 727–729 Receptor-specific ligand affinity chromatography,
top-down, 756–757 GPCRs, 636
Purification procedures Recombinant erythropoietin (rEPO), 401–402
affinity chromatography Recombinant G-protein-coupled receptors
amine reactive linkages, 432 maltose-binding protein (MBP) fusion
characterization, 424–425 approach, 637
covalent affinity chromatography, 428–429 purification
covalent attachment, 430 affinity purification, 634–636
de novo development, 423 detergent-solubilized GPCRs analysis,
elution, 434–435 636–637
examples of, 421 GPCRs stability, detergent solution, 634
group specific ligands, 425 receptor-specific ligand affinity
lectins, 426–427 chromatography, 636
ligands and specificity, 426 purified NTS1 analysis, 641–642
magnetic affinity beads, 422–423 solubilization, 633–634
pore size, 419 Recombinant integral membrane proteins,
Renkin equation, 420 expression and purification, 628–629
sample preparation, 433 Recombinant protein expressing method
specific ligand, 420 advantages and disadvantages, 143–144
stability, 422 applications, 142, 144
surface activation, 430 baculovirus/insect cells, 136–137
amino acid sequence, bioinformatics Escherichia coli
charge vs. pH, 22–23 disulfide bond formation, 134
cysteine content, 23–24 fusion partners, 134
E. coli overexpression, solubility, 25 inclusion bodies, 133
homology and cofactor affinity, 24–25 posttranslational modifications, 134
hydrophobicity and membrane-spanning temperature and molecular chaperones, 133
regions, 24 mammalian cells, 138–139
molar extinction coefficient, 23 Pichia pastoris, 135–136
polypeptide chain molecular weight, 22 protein characteristics
potential modification sites, 25 cytoplasmic proteins, 140–141
secondary structure, 24 E. coli and codon usage, 140
stability, 24 membrane and secreted proteins, 141
titration curve, isoelectric point, 22–23 toxic proteins, 141–142
critical problems, 26 Recombinant protein production
denatured state, 27 criteria, 224
protein purifications transient gene expression (TGE)
assays, 11–13 HEK293 and CHO cell lines, 224–228
bulk/batch procedures, 17 transfection methods, 228–234
Subject Index 849

Redox agents, 272 Thermofluor ligand, 772, 774

Refolding inclusion body (IBs) transfection methods, 229–230
alkaline pH shift, 276–277 Small ubiquitin-like modifier (SUMO) protein,
catalysts, 276 251
characterization, 270 Sodium dodecyl sulfate-polyacrylamide gel
database, 277 electrophoresis (SDS-PAGE)
dilute ways, 267–268 casting gels, 503–505
high-pressure, 276 catalyst, 503
on-column, 275–276 difference gel electrophoresis (DIGE)
problem, 262 quantitative analytical algorithms, 525–526
rational approach, 274–275 software analytical tools, 524–525
systematic screens, 271 staining and detection, 524–525
variables electrode buffer, 503
arginine, glycerol and sugars, 273 electrophoresis, 505–506, 511–513
chemical additives, 273 gel types, 522–523
detergents and chaotropic agents, 273 marker proteins, 510–511
divalent cations, 272–273 2-mercaptoethanol/dithiothreitol, 499, 507
pH and temperature, 272 molecular weight determination, 511
salt concentration and redox agents, 272 Ornstein–Davis system, 506
target protein-specific additives, 274 polyacrylamide gels, 500
Relative/fold purification, 31 principle, 501
Renkin equation, 420 protein detection, gel
Reverse dilution, 267 Coomassie Brilliant Blue R-248, 508–509
Reversed phase chromatography, 100 copper staining, 510
silver staining, 509–510
S resolution enhancement, 523–524
cathodic drift, 523
Salt concentration, 272 hydroxyethyl disulfide (HED), 524
Scattering methods IPG strip, 524
dynamic, 717 sample preparation, 505
electron microscopy, 718 separation principle, 522
static, 716–717 stock solutions, 502
SDS-PAGE. See Sodium dodecyl sulfate- vertical electrophoresis systems, 523
polyacrylamide gel electrophoresis Solid-phase peptide synthesis, 641
Sedimentation coefficient(s), 701–703 Solubility enhancement peptide (SEP) tags,
Sedimentation equilibrium 250–251
capabilities, 698 Solubility tags
method basis, 242–243
equilibrium centrifugation experiment, common tags, 241
699–700 maltose-binding protein (MBP), 249–250
reduced molecular weight (s), 698–699 NusA, 250
problems and pitfalls, 700–701 smaller solubility enhancement tags (SETs),
Sedimentation velocity 250–251
diffusion coefficient, 703 solubility enhancement peptide (SEP) tags,
frictional coefficient ratio (f/f0 ), 703–704 250–251
vs.gel chromatography, 701 thioredoxin (Trx), 250
methods, 687, 702–704 Soluble multimer, 262
s/D ratio, 703–704 Sonication, 299–300
variations, 705 Stable isotope labeling by amino acids in cell
Selected reaction monitoring (SRM), 755 culture (SILAC), 752
Size exclusion chromatography (SEC), 457. S-tag, 248
See also Gel–filtration chromatography Stationary gel volume fraction, 706
Smaller solubility enhancement tags (SETs), Strains, recombinant protein production
250–251 cell extracts preparation, 182
Small-scale expression screening b-lactamase expression, 181
analytical methods, 776 plate activity assay, 180
benzonase, 775 STREP-II-tag, 248
capillary electrophoresis result, 774 Strong cation exchange (SCX) resin, 732
850 Subject Index

Subcellular organelles Transformation reaction, 658–659

complexity, 307 Transgenic lepidopteran, 213
density and dimensions, 309 Transient gene expression (TGE)
enzyme and substrate markers, 315 HEK293 and CHO cell lines
extraction and prefractionation cultivation, 228
centrifugation, 308–311 EBNA-1 gene, 224
core histone purification, 324–325 expression vectors, 226–227
differential detergent fractionation (DDF), HKB-11 cell line, 226
316–320 overview, 225–225
enrichment, 311–327 SV40 enhancer, 224, 226
isolation transfection methods
complexity, 307 large-scale approach, 232–234
core histone purification, 307 medium scale approach, 231–232
density gradient, centrifugation, 311 small-scale approach, 229–230
density velocity, 308–310 reagents, 229
differential detergent fractionation (DDF), Transport methods
307 electrophoresis, 708–711
endomembrane, enrichment, 314–316 field-flow fractionation (FFF), 711–712
immunoaffinity reagents, 326–327 gel-filtration chromatography, 706–708
lipid raft, 320–322 mass spectrometry, 712–716
mitochondria, 311–313 sedimentation equilibrium, 698–701
nuclear extract, 325 sedimentation velocity, 701–705
nuclei and histone, 322–324 viscosity, 711
schematic workflow, 308 Tris(hydroxymethyl)aminomethane (tris) buffer,
Subproteomes. See Subcellular organelles 54
Subunits Two-dimensional gel electrophoresis, isoelectric
associated and dissociated states, 719 focusing
stoichiometry, 720–721 acidic range IPG gels, 531–534
Succinate buffer, 52 alkaline range IPG gels, 534–535
Sucrose gradient sedimentation method, 705 difference gel electrophoresis (DIGE)
Sucrose gradient separation method. See quantitative analytical algorithms, 525–526
Mitochondria enrichment software analytical tools, 524–525
Sumoylation. See Ubiquitination and sumoylation staining and detection, 524–525
SV40 enhancer, 224, 226 equilibration, IPG gels, 535–536
gel types, 522–523
T high resolution 2D gel, 517
materials
Tags removal equipment, 527
chemical cleavage, 253 solutions and reagents, 527–528
cleavage sites, 252 pH gradients types
endo and exo proteases, 252 carrier ampholytes, 518–519
Tandem affinity purification (TAP), 244 different configurations, 521
Tandem tags, 244 IPG strip technology, 519–520
Target protein-specific additives, 274 principle, 517–518
T4 DNA polymerase-treated DNA fragments, protein sample preparation, 528–530
155–156 resolution enhancement, 523–524
Tetramethylethylenediamine (TEMED), 500 cathodic drift, 523
Thermal precipitation, 341 hydroxyethyl disulfide (HED), 524
Thioredoxin (Trx) tag, 250 IPG strip, 524
Three-dimension structure, critical problems, 26 sample cleanup and precipitation, 530–531
Time-of-flight (TOF) mass analyzer, 745 separation principle, 522
Toxic proteins, expressing method, 141–142 vertical electrophoresis systems, 523
Transfection methods, large-scale
modifications, 234 U
process parameters, 232
protocol, 232–234 Ubiquitination and sumoylation
Transfer plasmid modifications, 198–199 amino-terminal epitope tags, 739
Transformation, electroporation, 176 characteristic mass shifts, 740
Subject Index 851

Ultrafiltration charged-coupled device (CCD) cameras, 575

convective and diffusive solute transport, direct and indirect, 575–577
108–109 far-Western, 577
devices, 110–112 semiquantitative
membranes, 109–110 advantageous, 577
modeling approach, 108 assay methods, 578
purification applications, 112–113 densitometric analysis, results, 579
Ultraviolet absorbance, 205 nm Wheat germ translation reaction
concentration standards, 83–58 bilayer reaction protocol, 664
dye-based protein assays, 83 bilayer reaction vs. dialysis reaction,
extinction coefficient (e) calculation, 82–83 662–663
Ultraviolet absorption spectroscopy, 280 nm, dialysis reaction protocol, 664–665
80–81 materials and reagents, 663–664
Ultraviolet–visible absorption spectroscopy, 66
Urea, 266 X

V 6xHis-tag. See His-tag

Viral chitinase gene, 208–209 Y

Viscosity, 711
Void volume (E0), 707 Yeastern blot assay procedure, 182–184
Yeast expression. See Pichia pastoris expression
W Yeast lysis, 289

Washed nuclear pellet preparation, 322–324 Z

Western blotting
advances, 575 Zymography, 543