Methods in

Molecular Biology 1600

Otto Holst Editor

Microbial
Toxins
Methods and Protocols
Second Edition
Methods in Molecular Biology

Series Editor
John M. Walker
School of Life and Medical Sciences
University of Hertfordshire
Hatfield, Hertfordshire, AL10 9AB, UK

For further volumes:

http://www.springer.com/series/7651
 Microbial Toxins

Methods and Protocols

Second Edition

Edited by

Otto Holst
Research Center Borstel, Leibniz-Center for Medicine and Biosciences,
Borstel, Schleswig-Holstein, Germany
Editor
Otto Holst
Research Center Borstel
Leibniz-Center for Medicine and Biosciences
Borstel, Schleswig-Holstein, Germany

ISSN 1064-3745     ISSN 1940-6029 (electronic)

Methods in Molecular Biology
ISBN 978-1-4939-6956-2    ISBN 978-1-4939-6958-6 (eBook)
DOI 10.1007/978-1-4939-6958-6

Library of Congress Control Number: 2017937053

© Springer Science+Business Media LLC 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to
be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty,
express or implied, with respect to the material contained herein or for any errors or omissions that may have been made.
The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Printed on acid-free paper

This Humana Press imprint is published by Springer Nature

The registered company is Springer Science+Business Media LLC
The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.
Preface

In the year 2000, a first methods collection entitled Bacterial Toxins: Methods and Protocols
was published which contained 20 chapters on protein toxins and endotoxin from bacteria
and cyanobacteria. Then, in 2011, a next such collection was published, entitled Microbial
Toxins: Methods and Protocols, which included both, protocols on (cyano)bacterial and mold
fungus toxins, with some focus on aflatoxins. In both cases, the idea was to support research-
ers of various scientific disciplines with detailed descriptions of state-of-the-art protocols
and, since the books turned out to be quite successful, it is quite obvious that this aim could
be achieved. Based on this success, a second volume entitled Microbial Toxins: Methods and
Protocols is presented now which contains protocols on (cyano)bacterial and mold fungus
toxins, with a rather strong focus on Gram-negative endotoxins (lipopolysaccharides).
The interest of researchers across a broad spectrum of scientific disciplines in the field
of microbial toxins is clearly unbroken. As many other fields do, this field makes use of a
broad variety of biological, chemical, physical, and medical approaches, and researchers
dealing with any microbial toxin should be familiar with various techniques from all these
disciplines. It is our hope that the book Microbial Toxins: Methods and Protocols, Second
Edition can strongly support researchers here.
Microbial Toxins: Methods and Protocols, Second Edition comprises 17 chapters present-
ing state-of-the-­art techniques that are described by authors who have regularly been using
the protocol in their own laboratories. Each chapter begins with a brief introduction to the
method which is followed by a step-by-step description of the particular method. Also, and
importantly, all chapters possess a Notes section in which e.g. difficulties, modifications and
limitations of the techniques are exemplified. Taken together, our volume should prove
useful to many scientists, including those without any previous experience with a particular
technique.

Borstel, Schleswig-Holstein, Germany Otto Holst

v
Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

1 Detection of Cholera Toxin by an Immunochromatographic Test Strip . . . . . . 1

Eiki Yamasaki, Ryuta Sakamoto, Takashi Matsumoto, Biswajit Maiti,
Kayo Okumura, Fumiki Morimatsu, G. Balakrish Nair,
and Hisao Kurazono
2 Electrochemical Aptamer Scaffold Biosensors for Detection
of Botulism and Ricin Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Jessica Daniel, Lisa Fetter, Susan Jett, Teisha J. Rowland,
and Andrew J. Bonham
3 A Cell-Based Fluorescent Assay to Detect the Activity of AB Toxins
that Inhibit Protein Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Patrick Cherubin, Beatriz Quiñones, Salem Elkahoui, Wallace Yokoyama,
and Ken Teter
4 Molecular Methods for Identification of Clostridium tetani
by Targeting Neurotoxin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Basavraj Nagoba, Mahesh Dharne, and Kushal N. Gohil
5 Label-Free Immuno-Sensors for the Fast Detection of Listeria in Food . . . . . . 49
Alexandra Morlay, Agnès Roux, Vincent Templier, Félix Piat,
and Yoann Roupioz
6 Aptamer-Based Trapping: Enrichment of Bacillus cereus Spores
for Real-Time PCR Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
Christin Fischer and Markus Fischer
7 Detection of Yersinia pestis in Complex Matrices by Intact Cell
Immunocapture and Targeted Mass Spectrometry . . . . . . . . . . . . . . . . . . . . . . 69
Jérôme Chenau, François Fenaille, Stéphanie Simon, Sofia Filali,
Hervé Volland, Christophe Junot, Elisabeth Carniel,
and François Becher
8 A Method to Prepare Magnetic Nanosilicate Platelets for Effective
Removal of Microcystis aeruginosa and Microcystin-LR . . . . . . . . . . . . . . . . . . . 85
Shu-Chi Chang, Bo-Li Lu, Jiang-Jen Lin, Yen-Hsien Li,
and Maw-Rong Lee
9 An Immunochromatographic Test Strip to Detect Ochratoxin A
and Zearalenone Simultaneously . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Xiaofei Hu and Gaiping Zhang
10 Endotoxin Removal from Escherichia coli Bacterial Lysate
Using a Biphasic Liquid System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Janusz Boratyński and Bożena Szermer-Olearnik

vii
viii Contents

11 Fourier Transform Infrared Spectroscopy as a Tool in Analysis

of Proteus mirabilis Endotoxins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Paulina Żarnowiec, Grzegorz Czerwonka, and Wiesław Kaca
12 Laser Interferometry Method as a Novel Tool in Endotoxins Research . . . . . . . 125
Michał Arabski and Sławomir Wąsik
13 Endotoxin Entrapment on Glass via C-18 Self-Assembled Monolayers
and Rapid Detection Using Drug-Nanoparticle Bioconjugate Probes . . . . . . . . 133
Prasanta Kalita, Anshuman Dasgupta, and Shalini Gupta
14 A Bioassay for the Determination of Lipopolysaccharides and Lipoproteins . . . . 143
Marcus Peters, Petra Bonowitz, and Albrecht Bufe
15 Capillary Electrophoresis Chips for Fingerprinting Endotoxin
Chemotypes and Subclasses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Béla Kocsis, Lilla Makszin, Anikó Kilár, Zoltán Péterfi,
and Ferenc Kilár
16 Micromethods for Isolation and Structural Characterization
of Lipid A, and Polysaccharide Regions of Bacterial Lipopolysaccharides . . . . . . 167
Alexey Novikov, Aude Breton, and Martine Caroff
17 Mass Spectrometry for Profiling LOS and Lipid A Structures
from Whole-Cell Lysates: Directly from a Few Bacterial Colonies
or from Liquid Broth Cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
Béla Kocsis, Anikó Kilár, Szandra Péter, Ágnes Dörnyei, Viktor Sándor,
and Ferenc Kilár

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Contributors

Michał Arabski • Department of Microbiology, Jan Kochanowski University, Kielce,

Poland
G. Balakrish Nair • World Health Organization, Mahatma Gandhi Marg, Indraprastha
Estate, New Delhi, India
François Becher • CEA, iBiTec-S, Service de Pharmacologie et d’Immunoanalyse, Gif-sur-
Yvette, France
Andrew J. Bonham • Department of Chemistry, Metropolitan State University of Denver,
Denver, CO, USA
Petra Bonowitz • Department of Experimental Pneumology, Ruhr University Bochum,
Bochum, Germany
Janusz Boratyński • Laboratory of Biomedical Chemistry - "Neolek," Ludwik Hirszfeld
Institute of Immunology and Experimental Therapy, Polish Academy of Sciences,
Wroclaw, Poland
Aude Breton • LPS-BioSciences, Institute for Integrative Biology of the Cell (I2BC), CEA,
CNRS, Université Paris-Sud, Université Paris-Saclay, Orsay, France
Albrecht Bufe • Department of Experimental Pneumology, Ruhr University Bochum,
Bochum, Germany
Elisabeth Carniel • Institut Pasteur, Unité de Recherche Yersinia, Paris, France
Martine Caroff • LPS-BioSciences, Institute for Integrative Biology of the Cell (I2BC),
CEA, CNRS, Université Paris-Sud, Université Paris-Saclay, Orsay, France
Shu-Chi Chang • Department of Environmental Engineering, National Chung Hsing
University, Taichung, Taiwan
Jérôme Chenau • CEA, iBiTec-S, Service de Pharmacologie et d’Immunoanalyse, Gif-sur-
Yvette, France
Patrick Cherubin • Burnett School of Biomedical Sciences, College of Medicine, University
of Central Florida, Orlando, FL, USA
Grzegorz Czerwonka • Department of Microbiology, Jan Kochanowski University, Kielce,
Poland
Jessica Daniel • Department of Chemistry, Metropolitan State University of Denver,
Denver, CO, USA
Anshuman Dasgupta • Department of Nanomedicine and Theranostics, Institute for
Experimental Molecular Imaging, RWTH Aachen University Clinic, Aachen, Germany
Mahesh Dharne • NCIM Resource Centre, CSIR-National Chemical Laboratory (NCL),
Pune, Maharashtra, India
Ágnes Dörnyei • Department of Analytical and Environmental Chemistry, University of
Pécs, Pécs, Hungary
Salem Elkahoui • Laboratoire des Substances Bioactives, Le Centre de Biotechnologie à la
Technopole de Borj-Cédria, Hammam-Lif, Tunisia
François Fenaille • CEA, iBiTec-S, Service de Pharmacologie et d’Immunoanalyse,
Gif-sur-Yvette, France

ix
x Contributors

Lisa Fetter • Department of Chemistry, Metropolitan State University of Denver, Denver,

CO, USA
Sofia Filali • Institut Pasteur, Unité de Recherche Yersinia, Paris, France
Markus Fischer • Hamburg School of Food Science, Institute of Food Chemistry, University
of Hamburg, Hamburg, Germany
Christin Fischer • Hamburg School of Food Science, Institute of Food Chemistry,
University of Hamburg, Hamburg, Germany
Kushal N. Gohil • NCIM Resource Centre, CSIR- National Chemical Laboratory
(NCL), Pune, Maharashtra, India
Shalini Gupta • Department of Chemical Engineering, Indian Institute of Technology,
Delhi, India
Xiaofei Hu • Henan Academy of Agriculture Science/Key Laboratory of Animal
Immunology, Ministry of Agriculture/Henan Key Laboratory of Animal Immunology,
Zhengzhou, China
Susan Jett • Department of Chemistry, Metropolitan State University of Denver, Denver,
CO, USA
Christophe Junot • CEA, iBiTec-S, Service de Pharmacologie et d’Immunoanalyse,
Gif-sur-Yvette, France
Wiesław Kaca • Department of Microbiology, Jan Kochanowski University, Kielce, Poland
Prasanta Kalita • Department of Chemical Engineering, Indian Institute of Technology,
Delhi, India
Ferenc Kilár • Institute of Bioanalysis, Faculty of Medicine and Szentágothai Research
Center, University of Pécs, Pécs, Hungary
Anikó Kilár • MTA-PTE, Molecular Interactions in Separation Science Research Group,
Pécs, Hungary
Béla Kocsis • Institute of Medical Microbiology and Immunology Faculty of Medicine,
University of Pécs, Pécs, Hungary
Hisao Kurazono • Division of Food Hygiene, Department of Animal and Food Hygiene,
Obihiro University of Agriculture and Veterinary Medicine, Obihiro, Hokkaido, Japan
Maw-Rong Lee • Department of Chemistry, National Chung Hsing University, Taichung,
Taiwan
Yen-Hsien Li • Department of Chemistry, National Chung Hsing University, Taichung,
Taiwan
Jiang-Jen Lin • Institute of Polymer Science and Engineering, National Taiwan
University, Taipei, Taiwan
Bo-Li Lu • Department of Environmental Engineering, National Chung Hsing
University, Taichung, Taiwan
Biswajit Maiti • Division of Food Hygiene, Department of Animal and Food Hygiene,
Obihiro University of Agriculture and Veterinary Medicine, Obihiro, Hokkaido, Japan
Lilla Makszin • Institute of Bioanalysis, Faculty of Medicine, University of Pécs, Pécs,
Hungary
Takashi Matsumoto • R&D Center, NH Foods Ltd., Ibaraki, Japan
Fumiki Morimatsu • Center for Regional Collaboration in Research and Education,
Obihiro University of Agriculture and Veterinary Medicine, Obihiro, Hokkaido, Japan
Alexandra Morlay • University Grenoble Alpes, SyMMES UMR 5819, CNRS, SyMMES
UMR 5819, CEA, SyMMES UMR 5819, Grenoble, France
Basavraj Nagoba • Maharashtra Institute of Medical Sciences & Research (Medical
College), Latur, Maharashtra, India
 Contributors xi

Alexey Novikov • LPS-BioSciences, Institute for Integrative Biology of the Cell (I2BC),
CEA, CNRS, Université Paris-Sud, Université Paris-Saclay, Orsay, France
Kayo Okumura • Division of Food Hygiene, Department of Animal and Food Hygiene,
Obihiro University of Agriculture and Veterinary Medicine, Obihiro, Hokkaido, Japan
Szandra Péter • Department of Analytical and Environmental Chemistry, University of
Pécs, Pécs, Hungary
Zoltán Péterfi • First Department of Internal Medicine, Infectology, Faculty of Medicine,
University of Pécs, Pécs, Hungary
Marcus Peters • Department of Experimental Pneumology, Ruhr University Bochum,
Bochum, Germany
Félix Piat • Prestodiag, Villejuif, France
Beatriz Quiñones • USDA-ARS, Produce Safety and Microbiology Research Unit,
Western Regional Research Center, Albany, CA, USA
Yoann Roupioz • University Grenoble Alpes, SyMMES UMR 5819, CNRS, SyMMES
UMR 5819, CEA, SyMMES UMR 5819, Grenoble, France
Agnès Roux • University Grenoble Alpes, SyMMES UMR 5819, CNRS, SyMMES UMR
5819, CEA, SyMMES UMR 5819, Grenoble, France
Teisha J. Rowland • Cardiovascular Institute and Adult Medical Genetics Program,
University of Colorado Denver Anschutz Medical Campus, Aurora, CO, USA
Ryuta Sakamoto • R&D Center, NH Foods Ltd., Ibaraki, Japan
Viktor Sándor • Faculty of Medicine, Szentágothai Research Center, Institute of
Bioanalysis, University of Pécs, Pécs, Hungary
Stéphanie Simon • CEA, iBiTec-S, Service de Pharmacologie et d’Immunoanalyse,
Gif-sur-Yvette, France
Bożena Szermer-Olearnik • Laboratory of Biomedical Chemistry - "Neolek," Ludwik
Hirszfeld Institute of Immunology and Experimental Therapy, Polish Academy of
Sciences, Wroclaw, Poland
Vincent Templier • University Grenoble Alpes, SyMMES UMR 5819, CNRS, SyMMES
UMR 5819, CEA, SyMMES UMR 5819, Grenoble, France
Ken Teter • Burnett School of Biomedical Sciences, College of Medicine, University of
Central Florida, Orlando, FL, USA
Hervé Volland • CEA, iBiTec-S, Service de Pharmacologie et d’Immunoanalyse, Gif-sur-
Yvette, France
Sławomir Wąsik • Department of Molecular Physics, Jan Kochanowski University, Kielce,
Poland
Eiki Yamasaki • Division of Food Hygiene, Department of Animal and Food Hygiene,
Obihiro University of Agriculture and Veterinary Medicine, Obihiro, Hokkaido, Japan
Wallace Yokoyama • USDA-ARS, Healthy Processed Foods Research Unit, Western
Regional Research Center, Albany, CA, USA
Paulina Żarnowiec • Department of Microbiology, Jan Kochanowski University, Kielce,
Poland
Gaiping Zhang • Henan Academy of Agricultural Science/Key Laboratory of Animal
Immunology, Ministry of Agriculture/Henan Key Laboratory of Animal Immunology,
Zhengzhou, China
 Chapter 1

Detection of Cholera Toxin by an Immunochromatographic

Test Strip
Eiki Yamasaki, Ryuta Sakamoto, Takashi Matsumoto, Biswajit Maiti,
Kayo Okumura, Fumiki Morimatsu, G. Balakrish Nair, and Hisao Kurazono

Abstract
As cholera toxin (CT) is responsible for most of the symptoms induced by Vibrio cholerae O1 or O139
infection, detection of CT is an important biomarker for diagnosis of the disease. The procedure for patho-
genicity analysis of V. cholerae isolates must be carefully developed for the reason that the amount of CT
produced by V. cholerae varies according to the medium used and culture conditions (i.e. temperature and
aeration status) applied. Here we describe a reproducible rapid method for analysis of CT production by
toxigenic V. cholerae with an immunochromatographic test strip that can detect as low as 10 ng/mL of
purified recombinant CT.

Key words Immunochromatographic test strip, V. cholerae, Cholera toxin, Toxigenicity, Rapid diag-
nostic tests, AKI medium

1 Introduction

Cholera remains a major public health problem, especially in devel-

oping countries, and the seventh pandemic of cholera, which began
in 1961 is still ongoing. In the case of diagnosis of cholera, after or
along with the detection of the causative agent Vibrio cholerae, veri-
fication of cholera toxin (CT) production is of added significance,
because only V. cholerae produces CT which is responsible for chol-
era symptoms such as acute “rice watery” diarrhea. Various meth-
ods, including immunoassays like immunochromatography (IC),
enzyme-linked immunosorbent assay (ELISA), reversed passive
latex agglutination (RPLA) etc., DNA-based assays like polymerase
chain reaction (PCR), quantitative PCR (qPCR), loop mediated
isothermal amplification (LAMP) etc., and bioassays including rab-
bit ileal loop test, rabbit skin test, cultured Chinese hamster ovary
(CHO) cell assay, etc. for toxigenicity investigation of V. cholerae
isolates have been established [1, 2]. Such ­immunoassays and DNA-
based assays contribute to the rapid detection of CT and facilitate

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_1, © Springer Science+Business Media LLC 2017

1
2 Eiki Yamasaki et al.

timely, in some cases, on-site responses. While DNA-based assays

may be more sensitive than immunoassays, the latter have an impor-
tant advantage for the detection of extracellular bacterial toxin and
analysis of the toxin expression level. Recently, some novel method-
ology of immunoassays with extremely high sensitivity has been
reported [3, 4]. However, IC is still one of the most commonly
utilized immunoassay because it is rapid and very easy to conduct.
When IC is used, careful consideration has to be given to the way
samples are prepared to allow an optimal production of the target
protein.
Previously, we have established a toxigenic V. cholerae-specific
immunochromatographic test strip (CT-IC) [5] which could
detect CT in V. cholerae cultures in which at least 10 ng/mL of CT
was expressed. The amount of CT produced in V. cholerae El Tor,
which is the causative bacterium of the ongoing 7th cholera pan-
demic varies according to the medium used and culture conditions
(i.e. temperature and aeration status), and the optimal condition is
significantly different from that for the classical biotype which was
responsible for the earlier cholera pandemics [6–9]. The AKI
medium is one of the most efficient media to induce CT expression
in V. cholerae El Tor. It was reported that if V. cholerae El Tor
strains were cultured in AKI medium under biphasic culture condi-
tions, i.e. 4 h cultivation in a stationary phase followed by 16 h
cultivation in a shaking flask at 37 °C, most of the strains produced
more than 10 ng/mL of CT [8]. Therefore, the combination of
CT-IC with AKI medium is advantageous in analyzing the ability
of V. cholerae isolates to produce CT.

2 Materials

Prepare all media and solutions using ultrapure water (deionized)

at room temperature.

2.1 Culture Media 1. 5% NaHCO3: Weigh 1.5 g NaHCO3 and transfer to the cylin-
der. Add distilled water to a volume of 20 mL. Dissolve all
powder of NaHCO3 completely. Make up to 30 mL with addi-
tional distilled water. Filter the solution by using DISMIC
Mixed Cellulose Ester Syringe Filter Unit (25AS Type) having
a pore size of 0.45 μm (Advantec Co. Ltd.) immediately before
use (see Note 1).
2. AKI medium: Weigh 4.5 g of Bacto™ Peptone, 1.2 g of yeast
extract and 1.5 g of NaCl and transfer to 300 mL glass beaker
containing about 200 mL of distilled water. Dissolve all pow-
der completely. Transfer the solution to 300 mL graduated
cylinder and make up to 282 mL with distilled water. Transfer
the ­solution into the autoclavable container and sterilize the
 Cholera Toxin Detection by Immunochromatographic Strip 3

solution with autoclave unit at 121 °C for 15 min. After cool-

ing the sterilized medium to room temperature, add 18 mL of
5% NaHCO3 aseptically (see Note 2).
3. Luria-Bertani (LB) agar plate: Weigh 4.0 g of Bacto™
Tryptone, 2.0 g of Bacto™ Yeast extract and 4.0 g of NaCl and
transfer to 500 mL beaker containing about 350 mL of dis-
tilled water. Dissolve all powder completely. Transfer the solu-
tion to 500 mL graduated cylinder and make up to 400 mL
with distilled water. Transfer the solution into the autoclavable
container containing 6.0 g of agar (powder). After mixing the
solution, sterilize with autoclave unit at 121 °C for 15 min.
After cooling the sterilized medium to 50–55 °C, pour it into
sterile Petri plates.

2.2 Immuno­ 1. The test strip was prepared with rabbit polyclonal antibodies
chromatographic Test raised against recombinant purified CT as reported previously
Strip [5, 10].

3 Methods

3.1 Bacterial Cell 1. Inoculate a V. cholerae isolate (isolated by the established pro-
Culture Preparation cedures) onto a non-selective agar medium such as LB agar
for Immuno­ plate (see Note 3).
chromatographic 2. Incubate the LB agar plate at 37 °C for 18–24 h until colonies
Analysis can be observed.
3. Pick a well isolated colony and inoculate a culture tube con-
taining 10 mL of AKI broth (see Note 4).
4. Let the tube stand at 37 °C for 4 h.
5. Transfer entire culture into a sterilized 100 mL Erlenmeyer
flask (see Note 5).
6. Incubate in the shaking incubator at 37 °C for about 16 h
(see Note 6).
7. The obtained bacterial cell culture is used in the analysis with
CT-IC (see Note 7).

3.2 Cholera Toxin 1. Place CT-IC on horizontal table (see Note 8).
Detection by Immuno­ 2. Apply 100 μL of the bacterial cell culture obtained in step 7 in
chromatographic Test Subheading 3.1 to the sample application section (Fig. 1a)
Strip (see Note 9).
3. Leave the strip for about 15 min at room temperature.
4. Observe the test result. A positive result shows reddish purple
lines in the test position and control position. A negative result
shows a reddish purple line only in the control position. If no
reddish purple line appears in the control position, the result is
considered to be invalid.
4 Eiki Yamasaki et al.

Fig. 1 Illustration of an immunochromatographic test strip for detection of CT

(CT-IC). CT-IC can be used either on a flat bench (a) or in a test tube (b). a,
Sample application section. b, Reagent containing section. c, Detection section.
d, Absorbent pad. e, Test line for CT detection position. f, Control line appearance
position

3.3 Cholera Toxin 1. Transfer 150 μL of the bacterial cell culture obtained in step 7
Detection by Immuno­ in Subheading 3.1 to fresh 1.5 mL micro tube.
chromatographic Test 2. Put the CT-IC in the micro tube to immerse the sample appli-
Strip (Alternative cation section in the cell culture (Fig. 1b) (see Note 10).
Procedure) 3. Leave the strip for about 15 min at room temperature.
4. Observe the test result. Criteria for result judgment are same as
above (see Subheading 3.2, step 4).

4 Notes

1. The filtration step is done for the purpose of sterilization.

Therefore, the filtered solution must be collected and stored in
a sterilized bottle. If the sterilized solution cannot be used
immediately, store the solution in a sealed bottle at 4 °C.
2. It is better to prepare the medium immediately before use. If
the prepared medium cannot be used immediately, store the
medium in a sealed bottle at 4 °C. The medium must be
brought to room temperature before use.
3. Selective or differential media such as TCBS agar, Vibrio agar
or CHROMagar™ Vibrio can be used for isolation of V. chol-
erae. Before the analysis with CT-IC, it is better to do subcul-
tivation on non-selective agar medium to obtain completely
isolated colony.
 Cholera Toxin Detection by Immunochromatographic Strip 5

4. For the stationary cultivation phase, a small sterilized container

such as 15 mL centrifuge tube (height, 150 mm; diameter,
15 mm) can be used. Degassing is not needed.
5. For the shaking cultivation phase, Erlenmeyer flask with the
volume more than 100 mL (i.e. more than 10 times of volume
of the culture) must be used to enforce adequate aeration.
6. Expression of CT not only in V. cholerae O1 El Tor but also in
other serotypes is known to vary depending on the culture
conditions. AKI medium is known as an effective medium to
induce CT expression. Two culture conditions are known for
effective induction of CT expression in AKI medium: AKI-SW
condition (4 h cultivation in a stationary test tube followed by
>16 h cultivation in a shaking flask at 37 °C) and AKI condi-
tion (>20 h cultivation in a stationary test tube at 37 °C) [6–
9]. Quantitative analysis with 15 independent ct-gene positive
V. cholerae strains revealed that CT expression under AKI-SW
condition was considerably higher than under AKI condition
(Fig. 2). We strongly recommend AKI-SW condition for use in

Fig. 2 Expression of CT under AKI-SW and AKI conditions. Concentration of CT in the culture supernatant were
obtained after cultivation of various V. cholerae isolates under AKI-SW (black bar) or AKI (gray bar) and were
then analyzed by bead-ELISA for CT quantification [11]. Data are mean ± SD of values from three independent
experiments
6 Eiki Yamasaki et al.

the analysis with CT-IC. Among the 15 isolates we analyzed,

one isolate (strain No. 15 in Fig. 2) expressed CT at a concen-
tration of substantially lower than the detection limit of CT-IC
(10 ng/mL) under AKI condition. However, CT expression
could be detected by CT-IC with AKI-SW condition (concen-
tration of CT was 0.28 ± 0.12 ng/mL in AKI condition
whereas 16.1 ± 6.97 ng/mL in AKI-SW condition).
7. We confirmed that centrifugation to remove bacterial cells did
not affect CT-IC results as indicated in Table 1. In case No. 1
(moderate CT expression with low cell density), case No. 2
(high CT expression with high cell density) and case No. 3
(low CT expression with high cell density), whole cell cultures
and cleared supernatants that were obtained after centrifuga-
tion (900 × g, 5 min) gave the same results in CT-IC analysis.
These results indicated that bacterial cells did not inhibit reac-
tions developing on the immunochromatographic test strip. In
addition, in case No. 4, ct gene-negative V. cholerae El Tor
Ogawa strain with high cell density did not give false-positive
results even if the whole cell culture was applied to the immu-
nochromatographic test strip.
8. The test strips are normally provided with light shielding pack-
age and stored at 4 °C. Allow the test strips to come to room
temperature (20–25 °C) before opening the package to pre-
vent moisture absorption. Do not touch with bare fingers on

Table 1 Effect of centrifugation before immunochromatographic analysis on CT-IC results

Case Strain ct Culture CT expression Cell CT-IC

No. No.*1 gene condition [ng/mL]*2 density*3 Centrifugation result*4
1 8 + AKI 127.5 ± 102.1 + +++
− +++
Low
2 AKI-SW 2009.6 ± 77.3 + +++
− +++
High
3 15 + AKI-SW 16.1 ± 7.0 + +
− +
High
4 16 − AKI-SW < 0.1 + −
− −
High
*1: The strain No. are matched with the number in Fig. 2
*2: The concentration of CT in the clear supernatant of the cultures measured with bead-ELISA are indi-
cated. Data are mean ± SD of values from three independent experiments
*3: Relative cell density and pictures of cell cultures in cuvettes are shown
*4: Results of CT-IC analyses are shown. The “+++”, “+” or “-” symbols are placed on the left side of the
strips developing “strong”, “faint” or “no” bands at test lines respectively. T: test line, C: control line
 Cholera Toxin Detection by Immunochromatographic Strip 7

the sample application section and detection section (Fig. 1).

It is better to hold the absorbent pad with tweezers or gloved
fingers when handling the test strips.
9. Be careful not to overload the sample application zone to pre-
vent spillage. Apply the culture in two batches as appropriate.
10. Any type of the container can be used in step 1 in Subheading
3.3. Adjust the volume of cell culture to keep the surface of the
sample below the reagent containing sections of the test strips.

Acknowledgements

This study was supported in part by a Grant-­in-­Aid of Ministry of

Health, Labor and Welfare (H26-Shinkou-Shitei-002).

References
1. Dick MH, Guillerm M, Moussy F et al (2012) 7. Iwanaga M, Yamamoto K (1985) New medium
Review of two decades of cholera diagnostics— for the production of cholera toxin by Vibrio
how far have we really come? PLoS Negl Trop cholerae O1 biotype El Tor. J Clin Microbiol
Dis 6:e1845 22:405–408
2. CDC (1999) Laboratory methods for the diag- 8. Iwanaga M, Yamamoto K, Higa N et al
nosis of Vibrio cholerae. Chapter VII (1986) Culture conditions production for
3. Palchetti I, Mascini M (2008) Electroanalytical stimulating cholera toxin by Vibrio cholerae
biosensors and their potential for food patho- OI El Tor. Microbiol Immunol 30:1075–
gen and toxin detection. Anal Bioanal Chem 1083
391:455–471 9. Sánchez J, Medina G, Buhse T et al (2004)
4. Shlyapnikov YM, Shlyapnikova EA, Simonova Expression of cholera toxin under non-AKI
MA et al (2012) Rapid simultaneous ultrasen- conditions in Vibrio cholerae El Tor induced by
sitive immunodetection of five bacterial toxins. increasing the exposed surface of cultures.
Anal Chem 84:5596–5603 J Bacteriol 186:1355–1361
5. Yamasaki E, Sakamoto R, Matsumoto T et al 10. Yonekita T, Fujimura T, Morishita N et al
(2013) Development of an immunochromato- (2013) Simple, rapid, and reliable detection of
graphic test strip for detection of cholera toxin. Escherichia coli O26 using immunochromatog-
Biomed Res Int 2013:679038 raphy. J Food Prot 76:748–754
6. Iwanaga M, Kuyyakanond T (1987) Large 11. Uesaka Y, Otsuka Y, Kashida M et al (1992)
production of cholera toxin by Vibrio cholerae Detection of cholera toxin by a highly sensitive
O1 in yeast extract peptone water. J Clin linked immunosorbent assay. Microbiol
Microbiol 25:2314–2316 Immunol 36:43–53
 Chapter 2

Electrochemical Aptamer Scaffold Biosensors

for Detection of Botulism and Ricin Proteins
Jessica Daniel, Lisa Fetter, Susan Jett, Teisha J. Rowland,
and Andrew J. Bonham

Abstract
Electrochemical DNA (E-DNA) biosensors enable the detection and quantification of a variety of molecu-
lar targets, including oligonucleotides, small molecules, heavy metals, antibodies, and proteins. Here we
describe the design, electrode preparation and sensor attachment, and voltammetry conditions needed to
generate and perform measurements using E-DNA biosensors against two protein targets, the biological
toxins ricin and botulinum neurotoxin. This method can be applied to generate E-DNA biosensors for the
detection of many other protein targets, with potential advantages over other systems including sensitive
detection limits typically in the nanomolar range, real-time monitoring, and reusable biosensors.

Key words Biosensors, Toxins, Electrochemical, Aptamer, Botulism, Ricin, Voltammetry, E-DNA,
Gold electrodes, Proteins

1 Introduction

Accurate and rapid detection of biomarkers is useful in many appli-

cations, ranging from food safety [1] and environmental sampling
to medical diagnostics [2] and small-molecule drug discovery.
Biosensors, which are devices that incorporate biological interac-
tions as the basis of their sensing mechanisms [3], are uniquely
suited to overcoming challenges associated with detecting a spe-
cific biomolecule in dense, complex, biological liquid matrices [4]
(e.g., whole blood or river water samples). In addition, biosensors
have several other appealing features that allow them to be used
successfully in unique and challenging situations, including high
specificity of detection, high reproducibility, relative ease of manu-
facturing and affordability, rapid throughput, direct readout, and
minimal invasiveness.
One prominent and successful class of biosensors is electro-
chemical DNA (E-DNA) biosensors [5]. E-DNA biosensors
rely on the changing conformational dynamics of a synthetic

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_2, © Springer Science+Business Media LLC 2017

9
10 Jessica Daniel et al.

deoxyoligonucleotide (DNA) scaffold containing an aptamer or

transcription factor-binding motif that recognizes the target
biomolecule [6–9] (Fig. 1). The DNA scaffold is modified with
functional groups to enable attachment to an electrode surface
(typically through a thiol-gold bond) and to an electrochemi-
cally active reporter molecule (e.g., methylene blue) [10, 11].
When the biosensor is subjected to voltammetric analysis, the
scaffold conformation changes depending on whether or not it
is bound to its target biomolecule, and this affects the dynamics
and the position (and thus observed current) of the electro-
chemically active reporter molecule relative to the electrode
surface [12] (Fig. 1). This principle enables E-DNA biosensors
to effectively function in complex matrices [4], including real-
time monitoring in animal blood [2], and be successfully used
against a range of targets, including oligonucleotides [6], small-
molecule drugs [13], heavy metals [14], antibodies [15], DNA-
binding proteins [16], and protein toxins [9].
In theory, E-DNA biosensors can be designed to detect any
molecule for which oligonucleotide-binding interactions are
known or discoverable (such as via systematic evolution of ligands
by exponential enrichment [SELEX]). Recently, our group gener-
ated E-DNA biosensors for the detection of protein toxins respon-
sible for ricin and botulism toxicity [9]. Here we describe the
design and use of novel E-DNA biosensors against these and simi-
lar targets. The biosensors described here can detect nanomolar
concentrations of ricin chain A and botulinum neurotoxin variant
A (Fig. 2), with high specificity and negligible off-target signals,
and function when challenged with complex matrices such as blood
serum albumin or other proteins (Fig. 3).

Fig. 1 Schematic of E-DNA biosensor, illustrating the change in position and dynamics of the reporter molecule
(methylene blue, represented by a blue star) attached to the DNA scaffold in response to binding of the biomol-
ecule target. Shown are biosensors directed toward the biomolecular targets (a) botulinum neurotoxin variant
A (BoNTA) and (b) ricin toxin chain A (RTA). The gold electrode surface (yellow disc) is passivated with a mono-
layer of 6-mercapto-1-hexanol (not shown) to prevent nonspecific binding of biomolecules (Reproduced from
ref. [9] with permission of the Royal Society of Chemistry)
 Protein Toxin E-DNA Biosensors 11

Fig. 2 Representative dose-responsive curves of peak current vs. toxin concentration for botulinum neurotoxin
variant A (BoNTA, a) and ricin toxin chain A (RTA, b). Both E-DNA biosensors display robust equilibrium signal
change in response to target concentration, with apparent dissociation constant (KD) values of 0.4 ± 0.2 nM
for BoNTA and 0.7 ± 0.5 nM for RTA (Reproduced from ref. [9] with permission of the Royal Society of Chemistry)

Fig. 3 The botulinum (BoNTA) and ricin (RTA) biosensors display minimal off-­
target responses when challenged with off-target proteins, including bovine
serum albumin (BSA) and other biomolecular targets, such as the unrelated DNA-­
binding protein complex Myc/Max. Student’s t-test was performed to compare
on-target to off-target response (* for p < 0.05, *** for p < 0.0001) (Reproduced
from ref. [9] with permission of the Royal Society of Chemistry)

2 Materials

Prepare all solutions using ultrapure water (prepared by purifying

deionized water, to attain a sensitivity of 18 MΩ-cm at 25 °C) and
analytical grade reagents. Prepare and store all reagents at room
temperature (unless indicated otherwise). Diligently follow all
waste disposal regulations when disposing of waste materials.
12 Jessica Daniel et al.

2.1 Biosensor 1. Biosensor DNA: Synthetic DNA scaffold with 5′ terminal

Design and Synthesis disulfide (e.g., 5″ thio C6 modifier/trityl-6-thiohexyl amidite)
and internal thymidine-methylene blue to be used as an
electrochemically active reporter molecule (methylene blue
­
succinimidyl ester coupled to amino modifier C6 T amidite/5″-
DMT-T[acrylamido-C6-NH-TFA]) (see Note 1). Resuspend
DNA in ultrapure water at 100 μM. Store aliquoted at −20 °C
wrapped in aluminum foil. The ricin biosensor sequence used
here is 5′- AGAG CGT AGG TTC G C[T(Methylene Blue)]C
GGG AA CGG AGT GGT CCG TTATTA ACC ACT ATTT
GAA CCT ACC -3′, and the botulinum toxin biosensor
sequence is 5′- TTT CA[T(Methylene Blue)] AGG GA AA
ATTTGACACT TT TCAAAC T GTCCTATGAC A GTCCA
TAGG -3′ [9].
2. Quickfold application from the DINAMelt web server, hosted
by the RNA Institute at the State University of New York at
Albany [17], available at http://unafold.rna.albany.edu/?q=
DINAMelt/Quickfold (see Note 2).
3. OPTIONAL: Fealden DNA biosensor algorithm [8], available at
http://www.bonhamlab.com/wp-content/uploads/2016/05/
Fealden-0.2_04232016.zip (see Note 3).
4. PCR tubes: 0.5 mL flat-cap PCR tubes, RNase- and DNase-­
free, polypropylene.

2.2 Electrode 1. Pine Research Instrumentation WaveNano USB Potentiostat

Preparation (see Note 4).
2. Pine Research Instrumentation Compact Voltammetry Cell
Grip Mount.
3. Pine Research Instrumentation WaveNano Shielded Cell
Cable.
4. Pine Research Instrumentation Compact Voltammetry Cable.
5. Pine Research Instrumentation Ceramic Patterned Gold
Electrode.
6. Pine Research Instrumentation AfterMath Scientific Data
Organizer Software.
7. Alkaline cleaning solution: 0.5 M NaOH.
8. Acid cleaning solution: 0.5 M H2SO4.
9. Etch solution: 0.1 M H2SO4, 0.01 M KCl.
10. Evaluation solution: 0.05 M H2SO4.

2.3 Biosensor 1. TCEP solution: 1 M Tris(2-carboxyethyl)phosphine hydro-

Attachment chloride. Store aliquoted at −20 °C.
and Surface 2. Phosphate-buffered saline (PBS): 8 g NaCl, 0.2 g KCl, 1.44 g
Passivation Na2HPO4, 0.24 g KH2PO4, and pH adjusted to 7.4 with HCl,
adjusted to 1 L with ultrapure water.
 Protein Toxin E-DNA Biosensors 13

3. Mercaptohexanol solution: 0.001 M 6-mercapto-1-hexanol in

PBS. Prepare and work with solution in a chemical fume hood.
Store at 4 °C for up to 1 month.
4. PCR tubes: 0.5 mL flat-cap PCR tubes, RNase- and DNase-­
free, polypropylene.
5. Petri dish: 100 mm × 15 mm, polystyrene.

2.4 Botulism 1. PBS: see Subheading 2.3.

and Ricin Protein 2. Ricin solution: Ricin A chain from Ricinus communis (castor-
Preparation bean or castor-oil-plant, from Sigma-Aldrich). Resuspend at
1 mg/mL in PBS. Store aliquoted at 4 °C.
3. Botulism solution: Botulinum neurotoxin variant A1 atoxic
derivative [18]. Resuspend at 1 mg/mL in PBS. Store ali-
quoted at −80 °C.

2.5 Electrochemical 1. PBS: see Subheading 2.3.

Biosensing Experiment 2. Target biomolecule solution: see Subheading 2.4.
3. Prepared electrode: see Subheading 2.3.
4. AnyPeakFinder software program (source code available at
http://www.bonhamlab.com/tools/code/) or AfterMath pro-
gram with built-in peak height analysis functions (see Note 5).

3 Methods

Carry out all procedures at room temperature unless otherwise

specified.

3.1 Sensor Design 1. Identify a DNA-binding motif that recognizes your biomole-
and Synthesis cule target of interest. We have used previously identified tran-
scription factor binding sites [16] or aptamers [2, 9] or
aptamers that we identified in-house [9].
2. Identify regions of the motif that are presumed to be “essen-
tial” for target binding interactions (Fig. 4). For aptamers,
detailed mechanistic binding studies are often available in the
literature; the regions of interest will typically be predicted to
form “loops” in their secondary structure. Confirmation via
Quickfold may be useful.
3. Design a synthetic DNA scaffold that incorporates the motif
region(s) identified to be essential for target binding
­interactions and allows for potential disruption of these bind-
ing interactions. To do this, design the essential regions to be
flanked on either or both of its 5′ and 3′ ends with deoxyoli-
gonucleotides that are partially complementary to the essen-
tial regions, facilitating the formation of secondary structures
14 Jessica Daniel et al.

Fig. 4 Schematic of biosensor design workflow process. An initial aptamer is truncated to essential regions
and then flanked by a random scaffold of novel oligonucleotides. Secondary structure predictions are used to
guide changes to the scaffold sequence to promote the formation of isoenergetic states that either present or
obscure the aptamer essential regions. The addition of a reporter molecule (e.g., methylene blue, “MB”) and
surface attachment modifications (i.e., thiol-gold bond, “HS”) leads to a completed biosensor design

or folding patterns that likely disrupt target binding. Multiple

rounds of confirmation via Quickfold or other secondary
structure prediction services may be useful (see Note 6).
4. Continue designing the scaffold by iteratively adding, remov-
ing, or changing oligonucleotides in the nonessential regions
to ultimately create a scaffold with two potential, equally favor-
able (i.e., isoenergetic) states: one state in which the essential
regions are available for target binding interactions (i.e., in
their native form) and one in which the essential regions are
unavailable due to being base paired with nonessential regions
(Fig. 4) [8]. The Fealden DNA biosensor algorithm is a design-
ing tool that may be used to help automate this process.
5. Once a scaffold has been designed with the two desirable
isoenergetic states, modify the scaffold design to include an
electrode attachment point. A thiol group located at the 5′
terminus of the entire scaffold should serve as the attachment
point by forming a thiol-gold bond between the scaffold and
the gold electrode surface.
6. Further modify the scaffold design to include an electrochemi-
cally active reporter molecule; here a methylene blue is used.
The methylene blue can be easily covalently appended to a
modified thymine. Examine the scaffold’s two isoenergetic
states to identify a thymine that is nonessential in a significantly
different folded environment and has significant distance
 Protein Toxin E-DNA Biosensors 15

change from the 5′ terminus between the two folded states

(Fig. 1). Again, Fealden may be used to help automate this
selection process (see Note 6).
7. Synthesize the designed scaffold using a DNA synthesis com-
pany or in-house phosphoramidite deoxyoligonucleotide
synthesis.
8. Resuspend the DNA in ultrapure water upon receipt at a con-
centration of 100 μM, aliquot it into PCR tubes (typically 4 μL
per tube), and store aliquots at −20 °C.

3.2 Electrode 1. Connect the WaveNano USB Potentiostat to a computer via a

Preparation USB cable.
2. Connect the Compact Voltammetry Cell Grip Mount to the
potentiostat using the WaveNano Shielded Cell Cable and
Compact Voltammetry Cable, being sure that the alligator
clips of the Shielded Cell Cable do not touch each other.
3. Place the Ceramic Patterned Gold Electrode face up in the grip
mount, and add a plastic adaptor spacer (included with elec-
trode) at the bottom of the grip mount to ensure solid contact
between the grip mount and electrode. Ensure that the black
ground electrode of the Shielded Cell Cable is connected to
outlet ground (see Fig. 5).
4. Power on the potentiostat and ensure that the status light is
green.
5. Open and log in to the AfterMath Scientific Data Organizer
Software. Ensure that the WaveNano Potentiostat is recognized

Fig. 5 Image of Pine Research Instrumentation (a) WaveNano instrument with

correct cables and (b) ceramic-patterned electrode with exposed gold electrode
surfaces. The biosensor attaches to the central, circular gold electrode
16 Jessica Daniel et al.

and communicating with AfterMath; the potentiostat’s status

should be listed as “idle” (see AfterMath support site for guid-
ance; http://wiki.voltammetry.net/pine/aftermath).
6. Insert the electrode into a 30 mL beaker, and add 15 mL of
alkaline cleaning solution, ensuring that the exposed gold sur-
faces of the electrode are submerged and the grip mount and
contacts on the electrode remain dry.
7. Create and run a new cyclic voltammetry experiment to per-
form 100 scans from −0.4 V to −1.35 V at a sweep rate of
2 V/s. This will reductively desorb any sulfur-linked molecules
on the electrode surface.
8. Remove the electrode from the alkaline cleaning solution,
rinse with ultrapure water, and repeat step 6 using 15 mL of
acid cleaning solution (instead of alkaline cleaning solution).
9. Create and run a new bulk electrolysis experiment to perform
oxidation using 2 V applied for 5 s followed by reduction
using −0.35 V applied for 10 s. This will oxidize any organic
contaminants and then reduce any gold oxide formed.
10. Create and run a new cyclic voltammetry experiment to per-
form cyclic oxidation and reduction voltammetric scans, per-
forming 20 scans with a scan rate of 4 V/s, followed by a
further 4 scans at 0.1 V/s, from 0.35 V to 1.5 V. This step will
sequentially oxidize and then reduce any remaining contami-
nants on the electrode surface.
11. Remove the electrode from the acid cleaning solution, rinse
with ultrapure water, and repeat step 6 using 15 mL of etch
solution (instead of alkaline cleaning solution).
12. Create a new cyclic voltammetry experiment, and perform
scans over four different potential ranges, each for ten scans at
scan rate of 0.1 V/s: 0.2–0.75 V, 0.2–1.0 V, 0.2–1.25 V, and
0.2–1.5 V. This will etch away the surface layer of the electrode
as gold chloride complexes, resulting in a substantially cleaned
surface.
13. Remove the electrode from the etch solution, rinse with ultra-
pure water, and repeat step 6 using 15 mL of evaluation solu-
tion (instead of alkaline cleaning solution).
14. Create a new cyclic voltammetry experiment, and perform
four scans from −0.35 V to 1.5 V at a scan rate of 0.1 V/s.
This will oxidize a gold oxide layer on the electrode and
then completely reduce it. The area under the reduction
peak can be used to calculate the available surface area of
the electrode [10, 19].
15. Store the cleaned electrode submerged in evaluation solution for
up to 1 h before proceeding with using it in Subheading 3.3.
 Protein Toxin E-DNA Biosensors 17

3.3 Biosensor 1. While performing the electrode preparation protocol

Attachment (Subheading 3.2), thaw aliquots of sensor DNA (prepared in
and Surface Subheading 3.1) and TCEP solution at room temperature.
Passivation Avoid exposing the sensor DNA to light.
2. To a clean PCR tube, add 3 μL sensor DNA and 3 μL TCEP
solution. Allow the sensor DNA/TCEP mixture to react for at
least 15 min, until it has changed from light blue to clear in
color [10] (see Note 7).
3. Mix 44 μL of PBS with the sensor DNA/TCEP mixture.
4. Remove the electrode from the evaluation solution (from step
15 of Subheading 3.3) and rinse it with ultrapure water. Using
a clean, delicate task wiper (e.g., a Kimwipe), dry the electrode
by wicking it dry, touching only the ceramic portions of the
electrode and taking care not to touch the exposed gold
surfaces.
5. Add the entire 50 μL of sensor DNA/TCEP/PBS mixture to
the electrode’s surface, being careful to cover the entire exposed
gold surface. Place the electrode inside a closed petri dish for
60 min, which will minimize evaporation and allow the reaction
to proceed. In arid climates, we have found it is important to
also add 250 μL of PBS to the bottom of the petri dish and rest
the electrode on top of a small, upside-down weigh boat in the
dish to help prevent premature drying (see Note 8).
6. Using a delicate task wiper, dry the electrode as described in
step 4. Immediately proceed with the next step to prevent the
electrode from completely drying.
7. Add 100 μL of mercaptohexanol solution [20] to the elec-
trode, being careful to cover the entire exposed gold surface.
Place the electrode inside a petri dish. Allow the reaction to
proceed for 1–24 h at 4 °C (see Note 9).
8. Equilibrate the prepared biosensor in PBS for at least 20 min
before use (in Subheading 3.5). This can be accomplished by
repeating step 6 of Subheading 3.2 using 15 mL of PBS
(instead of alkaline cleaning solution).

3.4 Botulism 1. Remove aliquoted protein solution (either ricin or botulinum

and Ricin Protein solution, depending on desired target biomolecule) from stor-
Preparation age and place on ice.
2. In microcentrifuge tubes, prepare a total of approximately
ten serial dilutions of the protein solution in PBS, each dilution
containing 100 μL, ranging from 0.01 nM to 1 μM. Keep the
dilutions on ice and use them within 2 h.

3.5 Electrochemical 1. Following PBS equilibration at the end of Subheading 3.3,

Biosensing Experiment remove the electrode from the PBS. Using a delicate task wiper,
dry the electrode as described in step 4 of Subheading 3.3.
18 Jessica Daniel et al.

2. Place the grip mount and electrode in a horizontal position

with the exposed gold surfaces facing up.
3. Cover the electrode’s exposed gold surfaces with 100 μL of
PBS, being careful to cover the working, counter, and refer-
ence elements of the electrode. Allow the electrode to
equilibrate for at least 10 min. Note: Instead of PBS, bovine
blood serum or whole bovine blood may alternatively be
used (see Note 10).
4. Ensure that the Compact Voltammetry Cable and Shielded
Cell Cable are correctly attached to the grip mount and that
the potentiostat’s status in the AfterMath software is shown as
“idle.”
5. Create a new square wave voltammetry experiment in the
AfterMath program, with voltage ranging from −0.5 to 0.1 V,
an amplitude of 50 mV, and a step size of 1 mV. The optimal
square wave frequency should be experimentally derived as it
can dramatically affect the signaling of the biosensor [21];
100 Hz is a typically useful frequency for a wide variety of
biosensors.
6. Run the experiment; a rounded peak in current at approxi-
mately −0.3 V should be present, due to the redox potential of
the methylene blue modification (see Fig. 6). The peak height
is proportional to the effective efficiency of electron transfer
between the surface and the methylene blue modification [22]
(see Note 11).
7. Rinse electrode with ultrapure water. Using a delicate task wiper,
dry the electrode as described in step 4 of Subheading 3.3.
8. Cover the electrode’s exposed gold surface with 100 μL of
0.01 nM ricin or botulinum solution (prepared in Subheading
3.4), being sure to cover the working, counter, and reference

Fig. 6 Square wave voltammograms of BoNTA biosensor equilibrated in PBS with

100 nM BoNTA (BoNTA) or with PBS only (Blank) (Reproduced from ref. [9] with
permission of the Royal Society of Chemistry)
 Protein Toxin E-DNA Biosensors 19

elements of the electrode. Allow the electrode to equilibrate

for at least 10 min.
9. Create a new square wave voltammetry experiment from −0.5
to 0.1 V with an amplitude of 50 mV and a step size of 1 mV
using the experimentally determined optimal frequency.
10. Run the experiment; the methylene blue-derived peak in cur-
rent at approximately −0.3 V should remain present. The mag-
nitude of any change in peak height reflects changes in
biosensor signaling due to the presence of the target
biomolecule.
11. Repeat steps 7–10 using the other prepared serial dilutions
(prepared in Subheading 3.4), using them in order of increas-
ing concentration. For each dilution, measure the peak in cur-
rent using AfterMath’s peak height tool, or export the data as
a comma-separated value (csv) file format, and use the
AnyPeakFinder program to determine the peak heights.
12. Using the peak current observed in step 6 as the baseline cur-
rent, calculate the relative change in current for each dilution
as a percentage increase or decrease in signal. For example, for
each dilution, the baseline peak height could be subtracted
from the dilution’s peak height, and this value could be divided
by the baseline peak height to calculate the percentage change.
13. Use the resulting data to construct a saturation binding curve,
allowing visualization of the apparent dissociation constant for
the target.
14. Following establishment of the target concentration depen-
dent response, this section’s procedure can be repeated using
samples of unknown protein concentration to allow for quan-
tification of the protein concentration in solution, enabling
biosensing applications.

4 Notes

1. DNA synthesis is performed by standard phosphoramidite cou-

pling on a solid support, which is available from many compa-
nies, such as Biosearch Technologies or Integrated DNA
Technologies. Briefly, a 5′ dimethoxytrityl (DMT)-protected
deoxynucleotide phosphoramidite is attached to a controlled
pore glass support through the 3′ hydroxyl. Acid treatment is
then used to remove DMT, followed by coupling to the next
deoxynucleotide phosphoramidite, protective acetylation, and
oxidation and then a repeated cycle of deprotection and cou-
pling. Modified deoxynucleotide phosphoramidites can be easily
included in this synthesis process.
20 Jessica Daniel et al.

2. The online Quickfold module is convenient, but there are sev-

eral other tools available that predict DNA secondary structure
folding, and any of these other tools should, in principle, be
sufficient for the necessary analysis. Examples include
RNAstructure from the University of Rochester Medical
Center (http://rna.urmc.rochester.edu/RNAstructureWeb/
Servers/Predict1/Predict1.html) and Integrated DNA
Technologies’ OligoAnalyzer Hairpin module (https://www.
idtdna.com/calc/analyzer).
3. The Fealden software significantly automates the task of evalu-
ating predicted DNA secondary structures for correct biosen-
sor conformational states, but it is optional and may require
Python programming experience to customize it for new appli-
cations. Fealden requires a UNIX-like environment and has
been confirmed to work on Ubuntu Linux and Mac OSX.
4. Potentiostats and analysis software are available from several
vendors; here we use Pine Research Instrumentation. Other
vendors that could provide suitable instrumentation packages
include CH Instruments, Inc., and Metrohm Autolab Nova.
5. Analysis of square wave voltammetric data requires accurately
measuring the height of observed current peaks (when plot-
ting current vs. voltage). AfterMath software includes a man-
ual tool for this measurement, and as the data can be exported
in csv format, a variety of computational tools can be used to
identify and measure current peaks, including Mathematica
and Matlab. Our lab provides source code for AnyPeakFinder,
a Python program that can automatically read csv formats and
extract peak height values; see http://www.bonhamlab.com/
tools/code/any-peak-finder-interactive/.
6. The core principles of selecting correct regions and optimizing
folded structures have been explored in a number of studies
[7–9, 23–26] and demonstrate that an iterative, trial-and-­error
approach can often yield good results. Generally, ­structures
with predicted free energies within 1 kJ/mol are more likely to
be meaningful. Minimizing the number of predicted states
helps avoid inconsistent results.
7. For sensor DNA solutions, the DNA sensor concentration is
known to affect biosensor performance [12, 27]. The added
TCEP solution must be sufficient to fully reduce the disulfide
modification present in the sensor DNA solution, and conse-
quently in this protocol the amount of TCEP solution added is
in high excess. The observed color change is due to reversible
reduction of the methylene blue modification. Although this
change has no impact on the final performance of the biosen-
sor, it is a convenient marker for the progress of reduction of
the solution.
 Protein Toxin E-DNA Biosensors 21

8. This process allows the sensor DNA to attach to the surface in

an incomplete monolayer, with average spacing between mol-
ecules that minimizes or eliminates interactions between
neighboring sensors, which is important for reproducible per-
formance. Optimizations of this surface packing have been
previously explored [12, 28].
9. The mercaptohexanol solution addition acts to form a stable,
mixed surface monolayer with the attached sensor DNA. While
6-mercapto-1-hexanol is the most common of these “passiv-
ation” chemicals, our lab has additionally found success with
the use of (11-mercaptoundecyl)tetra(ethylene glycol), which
presents a more biocompatible monolayer for studies in com-
plex matrices. The monolayer formed prevents nonspecific
interactions of the biomolecule target with the electrode’s gold
surface and provides a more reproducible current response.
Typically, 100 μL is added to the electrode surface and allowed
to adhere for 1 h. Using larger volumes, such as 200 μL, fol-
lowed by sealing the electrode in a petri dish and storing it
overnight at 4 °C, has also been successful. Our lab has also
attempted to briefly wash the mercaptohexanol-­ passivated
electrodes with saline sodium citrate (SSC) buffer or 1% bovine
serum albumin (BSA) buffer with minimal successes. This was
performed by allowing the mercaptohexanol to adhere to the
electrode overnight, then removing it with a pipette, and add-
ing 50 μL of either the SSC or BSA. The solution was allowed
to sit for 10 min before beginning trials.
10. To serve as a test bed for complex matrices uses of these sen-
sors, we have employed both adult bovine serum and bovine
whole blood (citrate stabilized) in place of PBS. In both matri-
ces, sensors still performed well, although the magnitude of
current changes is often reduced.
11. The precise voltage where the peak in current is found for
methylene blue will vary based on solution conditions (e.g.,
pH and ionic content). Different reporter dyes will have a dif-
ferent characteristic voltage for peak current.

Acknowledgment

This work would not be possible without ideas from Kevin Plaxco,
University of California Santa Barbara, and Ryan White, University
of Maryland Baltimore County. Support for this work was pro-
vided by the Metropolitan State University of Denver’s College of
Letters, Arts, and Sciences Dean’s office, Provost’s office, and the
Applied Learning Center.
22 Jessica Daniel et al.

References
1. Miranda-Castro R, de-los-Santos-Álvarez N, lead via an electrode-bound DNAzyme assem-
Lobo-Castañón MJ (2016) Aptamers as synthetic bly. J Am Chem Soc 129:262–263
receptors for food quality and safety control. 15. Vallée-Bélisle A, Ricci F, Uzawa T et al (2012)
Compr Anal Chem 74:155–191. doi:10.1016/ Bioelectrochemical switches for the quantita-
bs.coac.2016.03.021 tive detection of antibodies directly in whole
2. Ferguson BS, Hoggarth DA, Maliniak D et al blood. J Am Chem Soc 134:15197–15200
(2013) Real-time, aptamer-based tracking of 16. Bonham AJ, Hsieh K, Ferguson BS et al (2012)
circulating therapeutic agents in living animals. Quantification of transcription factor binding
Sci Transl Med 5:213ra165 in cell extracts using an electrochemical,
3. Lee TM-H (2008) Over-the-counter biosensors: structure-­switching biosensor. J Am Chem Soc
past, present, and future. Sensors 8:5535–5559 134:3346–3348
4. Lubin AA, Lai RY, Baker BR et al (2006) 17. Markham NR, Zuker M (2005) DINAMelt
Sequence-specific, electronic detection of oli- web server for nucleic acid melting prediction.
gonucleotides in blood, soil, and foodstuffs Nucleic Acids Res 33:W577–W581
with the reagentless, reusable E-DNA sensor. 18. Vazquez-Cintron EJ, Vakulenko M, Band PA
Anal Chem 78:5671–5677 et al (2014) Atoxic derivative of botulinum
5. Hasanzadeh M, Shadjou N (2016) neurotoxin a as a prototype molecular vehicle
Electrochemical nanobiosensing in whole for targeted delivery to the neuronal cyto-
blood: recent advances. TrAC Trends Anal plasm. PLoS One 9:e85517
Chem 80:167–176 19. Xiao Y, Lai RY, Plaxco KW (2007) Preparation of
6. Lubin AA, Plaxco KW (2010) Folding-based electrode-immobilized, redox-modified oligonu-
electrochemical biosensors: the case for respon- cleotides for electrochemical DNA and aptamer-
sive nucleic acid architectures. Acc Chem Res based sensing. Nat Protoc 2:2875–2880
43:496–505 20. Creager SE, Olsen KG (1995) Self-assembled
7. Vallee-Belisle A, Bonham AJ, Reich NO et al monolayers and enzyme electrodes: progress,
(2011) Transcription factor beacons for the problems and prospects. Anal Chim Acta
quantitative detection of DNA binding activity. 307:277–289
J Am Chem Soc 133:13836–13839 21. White RJ, Plaxco KW (2009) Exploiting
8. Schaffner SR, Norquest K, Baravik E et al binding-­ induced changes in probe flexibility
(2014) Conformational design optimization of for the optimization of electrochemical biosen-
transcription factor beacon DNA biosensors. sors. Anal Chem 82:73–76
Sens Bio-Sensing Res 2:49–54 22. Uzawa T, Cheng RR, White RJ et al (2010) A
9. Fetter L, Richards J, Daniel J et al (2015) mechanistic study of electron transfer from the
Electrochemical aptamer scaffold biosensors distal termini of electrode-bound, single-­stranded
for detection of botulism and ricin toxins. DNAs. J Am Chem Soc 132:16120–16126
Chem Commun (Camb) 51:15137–15140 23. Vallée-Bélisle A, Ricci F, Plaxco KW (2012)
10. Rowe AA, White RJ, Bonham AJ, Plaxco KW Engineering biosensors with extended, nar-
(2011) Fabrication of electrochemical-DNA rowed, or arbitrarily edited dynamic range.
biosensors for the reagentless detection of J Am Chem Soc 134:2876–2879
nucleic acids, proteins and small molecules. 24. Vallée-Bélisle A, Plaxco KW (2010) Structure-­
J Vis Exp 52:e2922 switching biosensors: inspired by Nature. Curr
11. Ricci F, Plaxco KW (2008) E-DNA sensors for Opin Struct Biol 20:518–526
convenient, label-free electrochemical detection 25. White RJ, Plaxco KW (2009) Engineering new
of hybridization. Microchim Acta 163:149–155 aptamer geometries for electrochemical
12. Xiao Y, Uzawa T, White RJ et al (2009) On the aptamer-based sensors. In: Fell NF, Jr,
signaling of electrochemical aptamer-based Swaminathan VS (eds) Proc Soc Photo Opt
sensors: collision- and folding-based mecha- Instrum Eng. SPIE, Department of Chemistry
nisms. Electroanalysis 21:1267–1271 and Biochemistry University of California,
13. Liu J, Wagan S, Dávila-Morris M et al (2014) Santa Barbara, Santa Barbara, CA 93106-9510,
Achieving reproducible performance of elec- p 732105
trochemical folding aptamer-based sensors on 26. Schoukroun-Barnes LR, Wagan S, White RJ
microelectrodes: challenges and prospects. (2014) Enhancing the analytical performance
Anal Chem 86:11417–11424 of electrochemical RNA aptamer-based sensors
14. Xiao Y, Rowe AA, Plaxco KW (2007) for sensitive detection of aminoglycoside anti-
Electrochemical detection of parts-per-billion biotics. Anal Chem 86:1131–1137
 Protein Toxin E-DNA Biosensors 23

27. White RJ, Phares N, Lubin AA et al (2008) 28. Huang K-C, White RJ (2013) Random walk
Optimization of electrochemical aptamer-­ on a leash: a simple single-molecule diffusion
based sensors via optimization of probe pack- model for surface-tethered redox molecules
ing density and surface chemistry. Langmuir with flexible linkers. J Am Chem Soc
24:10513–10518 135:12808–12817
 Chapter 3

A Cell-Based Fluorescent Assay to Detect the Activity

of AB Toxins that Inhibit Protein Synthesis
Patrick Cherubin, Beatriz Quiñones, Salem Elkahoui, Wallace Yokoyama,
and Ken Teter

Abstract
Many AB toxins elicit a cytotoxic effect involving the inhibition of protein synthesis. In this chapter,
we describe a simple cell-based fluorescent assay to detect and quantify the inhibition of protein synthesis.
The assay can also identify and characterize toxin inhibitors.

Key words AB toxin, Ricin, Shiga toxin, Toxin detection, Toxin inhibitors, Toxicity assay, Vero cells

1 Introduction

AB-type protein toxins are produced by numerous bacterial patho-

gens and some plants [1]. These toxins contain a catalytically active
A subunit and a cell-binding B subunit. The A and B moieties can
encompass different regions of a single polypeptide chain or may
represent distinct proteins in various stoichiometries (e.g., AB,
AB2, AB5, A3B7). Many AB toxins inhibit protein synthesis through
inactivation of elongation factor 2 or the 28S rRNA [2–5].
Several methods can detect the toxin-induced inhibition of
protein synthesis or resulting cell death. A common procedure
measures the viability of intoxicated cells by dye exclusion,
MTT/MTS assay, or similar protocols [6–11]. This strategy can
require several days of toxin exposure and often involves additional
processing steps for data collection. Furthermore, as discussed
later, there is a temporal disconnect between the inhibition of pro-
tein synthesis and the loss of cell viability. A direct method to quan-
tify the toxin-induced inhibition of protein synthesis measures the
incorporation of radiolabeled amino acids into newly synthesized
proteins [12, 13]. This requires the handling of radioisotopes,
which is laborious, potentially hazardous, and can only accommo-
date a limited number of samples. Quantitative luciferase-based

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_3, © Springer Science+Business Media LLC 2017

25
26 Patrick Cherubin et al.

assays have been described that are similar to the system reported
here, but these systems require several preparatory and/or process-
ing steps to enact the detection method [14, 15]. A recently
described assay that monitors the production and secretion of ace-
tylcholinesterase likewise requires additional processing steps for
data acquisition [16].
As an alternative to existing technologies, we developed a sim-
ple and quantitative cell-based assay for the detection of toxins that
inhibit protein synthesis. A Vero cell line with constitutive expres-
sion of a destabilized variant (t1/2 = 2 h) of the enhanced green
fluorescent protein (d2EGFP) is challenged with toxin for 18–24 h.
Intoxicated cells degrade d2EGFP and do not replenish the lost
protein due to the toxin-induced block of protein synthesis. The
fluorescent signal from Vero-d2EGFP cells is accordingly lost in
proportion to the applied dose of toxin. This assay provides repro-
ducible data with minimal hands-on effort (see Note 1). The pro-
cedure does not require radioisotopes, commercial kits, or
additional processing steps. A plate reader is required for reading
fluorescent samples, but the only major recurring cost is the use of
black-walled, clear-bottom 96-well tissue culture microplates. As
described below, the noninvasive nature of the fluorescent mea-
surement allows the Vero-d2EGFP cells to be used for additional
purposes. Furthermore, the protocol can be adapted to screen for
toxin inhibitors.

2 Materials

2.1 Cell Culture 1. Parental Vero cells (ATCC #CCL-81) and a clonal population
of Vero-2dEGFP cells with stable, constitutive expression of
the d2EGFP reporter (see Notes 2–5).
2. Complete Dulbecco’s Modified Eagle Medium (DMEM) for
carrying cells: DMEM, high glucose (4.5 g/L d-glucose) with
584 mg/L l-glutamine and 110 mg/L sodium pyruvate
(GIBCO), supplemented with 10% fetal bovine serum (Atlanta
Biologicals, Flowery Branch, GA), 1% antibiotic-antimycotic
solution, and 1 mg/mL Geneticin (G-418).
3. Intoxication medium: F-12 + GlutaMAX-I nutrient mixture
(Ham’s F-12; GIBCO) (see Note 6).
4. HyClone antibiotic-antimycotic 100× solution: 10,000 U/mL
penicillin, 10,000 μg/mL streptomycin, and 25 μg/mL
amphotericin B (GE Healthcare).
5. Trypsin-EDTA 1× solution containing 0.25% trypsin, 0.9 mM
EDTA, and phenol red (GIBCO).
6. HyClone phosphate-buffered saline (PBS) 1× solution:
6.7 mM PO4 without calcium or magnesium (GE Healthcare).
 Cell-Based Detection of Toxin Activity 27

7. Costar® black-walled 96-well polystyrene plates with a clear,

flat bottom (Corning Inc., Kennebunk, ME) (see Note 7).
8. Geneticin/G-418 (GIBCO) (see Note 8).
9. 100 × 20 mm tissue culture dishes (Techno Plastic Products).
10. Fisherbrand 25 mL sterile disposable divided well pipette
basins (Fisher Scientific).
11. Eppendorf Research® plus 8-channel variable 30–300 μL
pipette (p300) (Eppendorf).
12. Costar® 8-channel adapter vacuum aspirator (Corning Inc.).

2.2 Toxins 1. Ricin (Vector Laboratories, Burlingame, CA).

2. Cell-free culture supernatant containing Shiga toxin 1 (Stx1)
and Shiga toxin 2 (Stx2) (see Notes 9 and 10).

3 Methods

We generated a Vero cell line that stably expresses d2EGFP-N1, an

EGFP variant that contains a C-terminal PEST sequence for rapid
degradation by the ubiquitin-proteasome system [17, 18]. Steady-­
state fluorescence in the Vero-d2EGFP cell line is easily detected
by cytofluorometry (Fig. 1) and microscopy or with a plate reader
[17, 19]. When challenged with a toxin that inhibits protein syn-
thesis, toxin-susceptible cells will degrade d2EGFP and will not
produce more of the protein. Productive intoxication accordingly
results in the loss of d2EGFP fluorescence. As shown in Fig. 2,
ricin reduced the Vero-d2EGFP fluorescent signal in a dose-­
dependent manner (see Note 11). The loss of EGFP fluorescence

Fig. 1 Fluorescent output from the Vero-d2EGFP cells. 20,000 parental Vero
(gray) or Vero-d2EGFP (black) cells were subjected to cytofluorometry
28 Patrick Cherubin et al.

Fig. 2 Effect of ricin on protein synthesis and cell viability. Fluorescence (filled circles) and cell viability via MTS
assay (open squares) were measured in the same population of Vero-d2EGFP cells after an 18 h (a) or 42 h (b)
incubation with serial dilutions of ricin. Results were expressed as percentages of the maximal signal obtained
from unintoxicated Vero-d2EGFP cells. The means ± standard errors of the means of at least four independent
experiments with six replicate samples for each condition are shown

was much more dramatic than the loss of cell viability after an 18 h
intoxication: a half-maximal effective ricin concentration (ED50) of
0.03 ng/mL was recorded by the Vero-d2EGFP assay, whereas the
MTS cell viability assay reported an ED50 of 0.7 ng/mL (Fig. 2a).
Both fluorescence and viability were measured in the same cell
population (see Note 12). The loss of viability eventually mirrored
the loss of fluorescence after 42 h of toxin exposure, with both
EGFP and MTS assays documenting an ED50 of 0.02–0.04 ng/mL
(Fig. 2b). These collective results highlight several advantages of
the Vero-d2EGFP system, including (1) relatively rapid detection
of toxin activity, (2) high sensitivity, (3) minimal sample handling
for data acquisition, and (4) a noninvasive/nonterminal measure-
ment that allows the cells to be used for other purposes such as an
MTS assay.
To examine how quickly the Vero-d2EGFP assay can detect
toxin activity, we monitored the time-dependent decay of EGFP
fluorescence from intoxicated cells (Fig. 3, circles) (see Note 13).
Using a single concentration of ricin (1 ng/mL), we found the
EGFP signal begins to decay 4 h after toxin exposure and contin-
ues to decrease until 14 h when a minimal signal of 5–8% is
achieved. With a 2 h half-life for d2EGFP, a signal strength
­corresponding to 6% of the unintoxicated control value could the-
oretically be reached 8 h after exposure to a toxin that inhibits
protein synthesis. The longer time frame required to reach this
point for ricin-treated cells reflects the temporal delay between
toxin binding to the cell surface and A chain delivery to its site of
action in the cytosol [2], as well as the asynchronous nature of
intoxication in a population of cells. These cellular events also
 Cell-Based Detection of Toxin Activity 29

Fig. 3 Time frame for loss of Vero-d2EGFP fluorescence. Measurements of fluo-

rescent intensity were taken at 2 h intervals after incubation of Vero-d2EGFP
cells with 1 ng/mL of ricin (circles) or 20 μg/mL of the protein synthesis inhibitor
cycloheximide (squares). Results were expressed as percentages of the maximal
EGFP signal obtained from untreated Vero-d2EGFP cells. The averages ± stan-
dard deviations of three independent experiments with six replicate samples per
condition are shown

explain why the loss of EGFP fluorescence is slower in toxin-treated

cells than in cells treated with cycloheximide, a membrane-perme-
ant protein synthesis inhibitor (Fig. 3, squares). Nevertheless, our
experiment demonstrated the cellular activity of ricin can be
detected 4 h after toxin exposure and reaches its maximal effect on
protein synthesis 14 h after toxin exposure.
Disruptions to the intoxication process will permit the contin-
ued synthesis of d2EGFP in toxin-treated cells. This principle was
used to identify grape seed extract as a potent inhibitor of both
Stx1 and Stx2 [17]. The antitoxin property of grape seed extract
was then confirmed with an independent toxicity assay that moni-
tored the overall level of protein synthesis in cells exposed to Stx2
[17] (see Note 14). In the presence of grape seed extract, the Vero-­
d2EGFP cells were also protected from ricin, diphtheria toxin, and
exotoxin A [20].
Use of the Vero-d2EGFP assay to screen for toxin inhibitors is
shown in Fig. 4. A previous screen of 26 purified polyphenolic com-
pounds from grape extract failed to identify any individual inhibitor
of Stx1/Stx2 [20], so we next screened fractions from the grape
extract itself for antitoxin effects. Seven polyphenolic fractions,
generated by using a modified normal-phase high-­
­ performance
­liquid chromatography method [21] (Fig. 4a), were mixed with a
cell-free culture supernatant containing Stx1/Stx2 and applied to
30 Patrick Cherubin et al.

Fig. 4 Use of the Vero-d2EGFP assay to screen for toxin inhibitors. (a) A modified method using normal-phase
high-performance liquid chromatography separated grape seed extracts into seven fractions enriched in poly-
phenol monomers (catechin and epicatechin), dimers, trimers, or tetramers as indicated. (b) Vero-d2EGFP cells
were incubated with one of the extract fractions (5% final volume) and a 1:250 dilution of a cell-free culture
supernatant from an E. coli strain that expresses both Stx1 and Stx2. After an 18 h incubation, EGFP fluores-
cence was measured with a plate reader. Results were expressed as percentages of the maximal EGFP signal
obtained from a parallel set of unintoxicated Vero-d2EGFP cells. The means ± standard errors of the means of
at least four independent experiments with six replicate samples per condition are shown. (c) Vero-d2EGFP
cells were incubated with 0.5 ng/mL of ricin and either one of the extract fractions (5% final volume) or 10 μg/mL
of epigallocatechin gallate (EgCg). After an 18 h incubation, EGFP fluorescence was measured with a plate
reader. Results were expressed as percentages of the maximal EGFP signal obtained from a parallel set of
unintoxicated Vero-d2EGFP cells. The averages ± ranges of two to four independent experiments with six
replicate samples per condition are shown. For panels B and C, fraction 1 (F1) represents material eluted from
0 to 5 min; F2 represents material eluted from 5 to 10 min; etc.

Vero-d2EGFP cells for 18 h (see Note 15). The fluorescent signal

from cells exposed to toxin alone was reduced to 25% of the signal
from unintoxicated Vero-d2EGFP cells (Fig. 4b). A similar loss of
fluorescence was recorded for Vero-d2EGFP cells challenged with
 Cell-Based Detection of Toxin Activity 31

toxin in the presence of fractions 1–6. However, intoxicated cells

co-incubated with fraction 7 retained a stronger fluorescent signal,
representing 42% of the unintoxicated control value and a statisti-
cally significant difference from the intoxicated control cells
(p = 0.0217, Student’s t-test). A similar screen indicated fractions
5–7 contain compounds that block ricin activity (Fig. 4c). In this
screen, a known polyphenol inhibitor of ricin—epigallocatechin
gallate (EgCg) [20, 22]—was used as a positive control. Further
separation of the compounds in fractions 5–7, combined with addi-
tional Vero-d2EGFP assays, could potentially identify the specific
polyphenols that confer resistance to Stxs and/or ricin.

3.1 Plating of Cells 1. Maintain the parental Vero and Vero-d2EGFP cells in a
for the Fluorescence DMEM-­ based medium and passage when the cells form
Assay a >90% confluent monolayer on a 10 cm dish (see Note 16).
2. Working in a tissue culture hood, remove the spent medium
from the tissue culture dish and wash the cells with 10 mL of
sterile 1× PBS.
3. Detach cells from the dish by adding 1 mL of trypsin/EDTA
for 10 min at room temperature (see Note 17).
4. Re-suspend detached cells in 9 mL of complete DMEM
medium for a total volume of 10 mL.
5. Determine the cell concentration with a hemocytometer and
alter it to a final concentration of 100,000 cells per mL. For
the parental Vero cells, the final cell suspension should be in a
1.5 mL volume per 96-well plate. For the Vero-d2EGFP cells,
the final cell suspension should be in a 9 mL volume per
96-well plate.
6. Pour each cell dilution (parental Vero and Vero-d2EGFP) into
a separate sterile basin. Using a multichannel pipette (p300),
transfer 100 μL of the parental Vero cell suspension to the first
row (12 wells) of a black-walled 96-well microplate with clear
bottom. Again using a p300 multichannel pipette, transfer
100 μL of the Vero-d2EGFP cell suspension to each of the
remaining wells of the microplate.
7. Grow cells for ~24 h at 37 °C in a 5% CO2 humidified incuba-
tor (see Note 18).

3.2 Treatment 1. Prepare several ten- or twofold serial dilutions of toxin in

of the Cells with Toxin serum-­free Ham’s F-12 medium. The final volume for each
Inhibitors and/or toxin dilution is 0.8 mL for 6 replicate samples or 1.5 mL for
Purified Toxin 12 replicate samples. For a screen of toxin inhibitors, prepare
identical serial dilutions in Ham’s F-12 medium containing
the final concentration of inhibitor.
2. Remove medium from the Vero and Vero-d2EGFP cells with
a multi-well vacuum aspirator.
32 Patrick Cherubin et al.

3. Incubate Vero cells with 100 μL per well of serum- and toxin-­
free Ham’s F-12 medium, which is added with a p300 multi-
channel pipette.
4. Incubate Vero-d2EGFP cells with 100 μL per well of toxin
serial dilutions in the absence or presence of inhibitor.
5. As an unintoxicated control condition, incubate Vero-d2EGFP
cells with 100 μL per well of serum- and toxin-free Ham’s
F-12 medium. When inhibitor screens are performed, incu-
bate additional sets of unintoxicated Vero-d2EGFP cells with
100 μL per well of the inhibitor alone.
6. As a positive control for the loss of fluorescence, treat one set
of unintoxicated Vero-d2EGFP cells with the protein synthesis
inhibitor cycloheximide.
7. Incubate cells with toxin for 18–24 h at 37 °C in a 5% CO2
humidified incubator before fluorescence measurements are
taken.

3.3 Fluorescence 1. Using a multi-well vacuum aspirator, remove 100 μL of spent

Measurements medium from each well (see Notes 19 and 20).
and Data Analysis 2. Using a p300 multichannel pipette, wash the cells with 200 μL
of 1× PBS.
3. Using a p300 multichannel pipette, add another 100 μL of 1×
PBS to the cells. Read measurements from cells bathed in 1×
PBS (see Note 21).
4. Measure the EGFP fluorescence on a Synergy H1 Multi-Mode
Microplate Reader (BioTek, Winooski, VT) with bottom
optics position using a 485 nm excitation and 528 nm emis-
sion filter set. Set the scale value for the automatic sensitivity
adjustment to 20,000.
5. Readings taken from the parental Vero cells represent back-
ground levels of autofluorescence and are accordingly sub-
tracted from the experimental measurements. After background
subtraction, set the fluorescence value obtained from the
unintoxicated (control) Vero-d2EGFP cells arbitrarily at
100%. Express all experimental data as percentages of this
100% control value. Examples of the results produced are
shown in Figs. 2 and 3.
6. When screening toxin inhibitors, apply each inhibitor to Vero-­
d2EGFP cells in the absence of toxin to establish a separate
control (100%) value for that condition. Express readings
from Vero-d2EGFP cells incubated with both toxin and inhib-
itor as percentages of the corresponding control value from
cells treated with the inhibitor alone. This procedure corrects
for variability that could result from inhibitor autofluorescence
and/or inhibitor effects on cell viability in the absence of
toxin. Examples of the results produced are shown in Fig. 4.
 Cell-Based Detection of Toxin Activity 33

4 Notes

1. In our experience, skilled high school and university under-

graduates have been able to generate reproducible data with
this system [20]. Another lab independently developed the
Vero-d2EGFP toxicity assay and likewise reported highly
reproducible data from ricin-treated cells [19].
2. Any cell type could be stably transfected with d2EGFP to gen-
erate a reporter cell line for this assay. We identified Vero cells as
the most appropriate cell line for this work after several cell lines
(Vero, BHK, CHO, Hep-2, HeLa, HEK 293, and COS-­7)
were characterized with dose-response curves to multiple plant
and bacterial toxins. Of the cell lines tested, Vero cells displayed
the highest level of sensitivity to a wide range of toxins.
3. Protocols for the generation of a Vero-d2EGFP cell line can
be found in references [17, 19, 23].
4. The pd2EGFP-N1 plasmid (Clontech, Mountain View, CA)
used to generate the Vero-d2EGFP cell line is no longer com-
mercially available. Clontech offers a destabilized tetrameric
green fluorescent protein (ZsGreen1) as an alternative to
d2EGFP, although the construct is present in a promoterless
plasmid (pZsGreen1-DR).
5. It is possible to run this or related assays with virally trans-
duced cells [14, 24], but such a strategy requires additional
preparatory steps for each experiment.
6. Tissue culture medium contributes a substantial background
signal to the fluorescent measurement. The signal produced
by F-12 medium is weaker than the DMEM signal and is mini-
mized by washing the cells with PBS before measurement. An
18–24 h incubation in serum-free Ham’s F-12 medium does
not affect the viability of Vero or Vero-d2EGFP cells.
7. Black-walled microplates with clear, flat bottoms are required
to prevent well-to-well bleeding of the fluorescent signal.
8. Geneticin was used as the selective pressure for isolation of the
Vero-d2EGFP cell line. Maintaining drug selection (1 mg/
mL) in the passage medium of the Vero-d2EGFP cells ensures
continued, uniform expression of the EGFP reporter.
9. Escherichia coli O157:H7 strain RM1697, which expresses
both Stx1 and Stx2 [25], was grown in Luria-Bertani broth at
37 °C with 200 rpm shaking. After a 15 min centrifugation at
3000 × g, the toxin-containing culture supernatant was filter
sterilized with a 0.45 μm polyvinylidene fluoride syringe filter.
10. Purified Stx1 or Stx2, as well as other AB toxins, can be
obtained from BEI Resources (Manassas, VA) or List Biological
Laboratories (Campbell, CA).
34 Patrick Cherubin et al.

11. The Vero-d2EGFP assay has also been used to generate dose-­
response curves for Stx1 and/or Stx2, Pseudomonas aerugi-
nosa exotoxin A, and diphtheria toxin [17, 20].
12. It takes about 10 min to complete the steps outlined in
Subheading 3.3 for a fluorescent measurement from living
Vero-­d2EGFP cells. Since the cells are viable and have not
been modified by the measurement, they can subsequently be
used for cell viability assays or other purposes. Thus, after
recording the fluorescent intensity from our intoxicated cells,
20 μL of the commercial MTS reagent (Promega, Madison,
WI) was added to the 100 μL of PBS in each well for a 2–3 h
incubation at 37 °C. The MTS reagent is reduced by living
cells to a colored product that is detected with a plate reader
at an absorbance of 490 nm.
13. Separate sets of Vero and Vero-d2EGFP cells were used for
each time point. All Vero or Vero-d2EGFP cells were seeded
at the same time from a common basin of diluted cells (see
Subheading 3.1, step 6). Toxin or cycloheximide was added
simultaneously to five sets of cells for consecutive readings
from 2–10 h. The remaining four sets of cells were challenged
with toxin or cycloheximide at the end of the first day, which
allowed measurements for the later time points to be collected
on the second day. With this procedure, personnel were not
required to remain in the lab for 18+ consecutive hours.
14. When screening for toxin inhibitors, compounds such as pro-
teasome inhibitors that block the turnover of d2EGFP will
produce a false-positive result: in these instances, the d2EGFP
signal will persist despite the toxin-induced inhibition of pro-
tein synthesis. Screens for toxin inhibitors should accordingly
include a secondary assay to validate the hit compounds.
15. Use of the Vero-d2EGFP system to detect toxins in cell-free
bacterial culture supernatants is described in reference [26].
16. For passaging cells, a 1:10 dilution from a confluent 10 cm
dish will take 3–4 days to again reach confluency in a fresh
10 cm dish. A 1:20 dilution from a confluent 10 cm dish will
take about 4–5 days to again reach confluency in a fresh 10 cm
dish. Cells passaged at either dilution will need fresh medium
on the third day after transfer to a new dish.
17. Cells can be detached in less time (≤5 min) by incubating them
with trypsin/EDTA at 37 °C. A longer, room temperature incu-
bation in the tissue culture hood permits time for preparation of
the dishes and plates that will be used in the experiment.
18. Cell density will influence toxin sensitivity; highly confluent
cells are more resistant to intoxication than subconfluent cells.
For this reason, it is important to standardize from experiment
to experiment the number of cells plated and the confluency of
the cells at the time of toxin exposure.
 Cell-Based Detection of Toxin Activity 35

19. If the experiment is terminal at this point, the PBS does not
need to be sterile. In addition, the processing steps for a ter-
minal experiment can be performed at the bench rather than
in a tissue culture hood.
20. Always tilt the plate in the same orientation before aspirating
or adding media to a defined point at the edge of each well.
This will reduce the probability of cell loss during the process-
ing steps.
21. To further reduce the background signal from cell culture
media, final measurements are taken from cells bathed in PBS.

Acknowledgments

This material is based upon work supported by the US Department

of Agriculture, Agricultural Research Service, under the Non-
Assistance Cooperative Agreement (NACA) No. 58-5325-4-024.
Any opinions, findings, conclusions, or recommendations expressed
in this publication are those of the authors and do not necessarily
reflect the view of the US Department of Agriculture. The authors
thank Dr. Lucia Cilenti for technical assistance with the Accuri flow
cytometer (BD Biosciences, San Jose, CA).

References
1. Sandvig K, van Deurs B (2005) Delivery into 8. Gamage SD, McGannon CM, Weiss AA
cells: lessons learned from plant and bacterial (2004) Escherichia coli serogroup O107/
toxins. Gene Ther 12:865–872 O117 lipopolysaccharide binds and neutralizes
2. Spooner RA, Lord JM (2015) Ricin trafficking Shiga toxin 2. J Bacteriol 186:5506–5512
in cells. Toxins (Basel) 7:49–65 9. Sekino T, Kiyokawa N, Taguchi T et al (2004)
3. Murphy JR (2011) Mechanism of diphtheria Characterization of a Shiga-toxin 1-resistant
toxin catalytic domain delivery to the eukary- stock of vero cells. Microbiol Immunol
otic cell cytosol and the cellular factors that 48:377–387
directly participate in the process. Toxins 10. Pauly D, Worbs S, Kirchner S et al (2012)
(Basel) 3:294–308 Real-time cytotoxicity assay for rapid and sen-
4. Lee MS, Koo S, Jeong DG et al (2016) Shiga sitive detection of ricin from complex matrices.
toxins as multi-functional proteins: induction PLoS One 7:e35360
of host cellular stress responses, role in 11. Wahome PG, Bai Y, Neal LM et al (2010)
pathogenesis and therapeutic applications.
­ Identification of small-molecule inhibitors of
Toxins (Basel) 8:E77 ricin and Shiga toxin using a cell-based high-­
5. Michalska M, Wolf P (2015) Pseudomonas exo- throughput screen. Toxicon 56:313–323
toxin A: optimized by evolution for effective 12. Hovde CJ, Calderwood SB, Mekalanos JJ et al
killing. Front Microbiol 6:963 (1988) Evidence that glutamic acid 167 is an
6. Konowalchuk J, Speirs JI, Stavric S (1977) active-site residue of Shiga-like toxin I. Proc
Vero response to a cytotoxin of Escherichia Natl Acad Sci U S A 85:2568–2572
coli. Infect Immun 18:775–779 13. Obrig TG, Louise CB, Lingwood CA et al
7. Paton JC, Paton AW (1998) Pathogenesis and (1993) Endothelial heterogeneity in Shiga
diagnosis of Shiga toxin-producing Escherichia toxin receptors and responses. J Biol Chem
coli infections. Clin Microbiol Rev 11:450–479 268:15484–15488
36 Patrick Cherubin et al.

14. Zhao L, Haslam DB (2005) A quantitative by polyphenolic compounds. PLoS One

and highly sensitive luciferase-based assay for 11:e0166477
bacterial toxins that inhibit protein synthesis. 21. Lazarus SA, Adamson GE, Hammerstone JF
J Med Microbiol 54:1023–1030 et al (1999) High-performance liquid chroma-
15. Gal Y, Alcalay R, Sabo T et al (2015) Rapid tography/mass spectrometry analysis of pro-
assessment of antibody-induced ricin neutral- anthocyanidins in foods and beverages. J Agric
ization by employing a novel functional cell-­ Food Chem 47:3693–3701
based assay. J Immunol Methods 424:136–139 22. Dyer PD, Kotha AK, Gollings AS et al (2016)
16. Cohen O, Mechaly A, Sabo T et al (2014) An in vitro evaluation of epigallocatechin gal-
Characterization and epitope mapping of the late (eGCG) as a biocompatible inhibitor of
polyclonal antibody repertoire elicited by ricin ricin toxin. Biochim Biophys Acta 1860:
holotoxin-based vaccination. Clin Vaccine 1541–1550
Immunol 21:1534–1540 23. Massey S, Quiñones B, Teter K (2011) A cell-­
17. Quiñones B, Massey S, Friedman M et al based fluorescent assay to detect the activity of
(2009) Novel cell-based method to detect Shiga toxin and other toxins that inhibit pro-
Shiga toxin 2 from Escherichia coli O157:H7 tein synthesis. Methods Mol Biol 739:49–59
and inhibitors of toxin activity. Appl Environ 24. Rasooly R, He X (2012) Sensitive bioassay for
Microbiol 75:1410–1416 detection of biologically active ricin in food.
18. Li X, Zhao X, Fang Y et al (1998) Generation J Food Prot 75:951–954
of destabilized green fluorescent protein as a 25. Quiñones B, Swimley MS, Taylor AW et al
transcription reporter. J Biol Chem 273: (2011) Identification of Escherichia coli O157
34970–34975 by using a novel colorimetric detection method
19. Halter M, Almeida JL, Tona A et al (2009) A with DNA microarrays. Foodborne Pathog
mechanistically relevant cytotoxicity assay Dis 8:705–711
based on the detection of cellular GFP. Assay 26. Quiñones B, Swimley MS (2011) Use of a
Drug Dev Technol 7:356–365 vero cell-based fluorescent assay to assess rela-
20. Cherubin P, Garcia MC, Curtis D et al (2016) tive toxicities of Shiga toxin 2 subtypes from
Inhibition of cholera toxin and other AB t­ oxins Escherichia coli. Methods Mol Biol 739:61–71
 Chapter 4

Molecular Methods for Identification of Clostridium tetani

by Targeting Neurotoxin
Basavraj Nagoba, Mahesh Dharne, and Kushal N. Gohil

Abstract
Tetanus is a potentially fatal muscle spasm disease. It is an important public health problem, especially in
rural/tribal areas of developing countries. Tetanus toxin, a neurotoxin (tetanospasmin), is the most impor-
tant virulence factor that plays a key role in the pathogenicity of tetanus. Confirmation of virulence by
confirming the production of tetanospasmin by infecting species forms the most important part in the
diagnosis of tetanus. Various molecular methods have been devised for confirmation of diagnosis by target-
ing different genes. The most common molecular methods are tetanospasmin producing (TetX) gene-­
targeted methods using TetX-specific primers. Here, we describe various molecular methods targeting
TetX gene such as polymerase chain reaction, pulsed-field gel electrophoresis, Southern blotting, loop-­
mediated isothermal amplification assay, etc. to confirm the virulence of Cl. tetani.

Key words Tetanus, Tetanospasmin, TetX gene, PCR

1 Introduction

Clostridium tetani is a motile, spore-forming, anaerobic Gram-­

positive bacillus causing tetanus—a potentially fatal muscle spasm
disease [1]. Tetanus remains an important public health problem
in developing countries, especially in rural/tribal areas [2–5].
Since the introduction of the vaccination program in 1961,
there has been a significant decline in the number of cases in the
Western world [6], and currently it is very rarely reported from the
developed world [7]. Because of increased coverage of immuniza-
tion, the incidence has also declined dramatically from the devel-
oping world. This is more true about the urbanized developing
world where there is an increased coverage of immunization and
where there is realization of the significance of immunization by
population. However, in rural/tribal areas of the developing world,
where the importance of immunization is not yet fully realized by
population, tetanus still exists [8]. Hence, the isolated cases have

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_4, © Springer Science+Business Media LLC 2017

37
38 Basavraj Nagoba et al.

been frequently reported from developing countries, including

India, especially in non-immunized and inadequately immunized
individuals from rural/tribal areas [2–5, 9–12].

1.1 Importance In view of its existence, in isolated cases, isolation, identification,

of Laboratory and to find out its virulence form an important part in the
Diagnosis ­confirmation of infection by Cl. tetani. Although the diagnosis of
tetanus is based on typical clinical features, laboratory methods are
necessary to confirm the clinical diagnosis and to rule out certain
conditions, which clinically mimic generalized tetanus [13]. These
conditions include strychnine poisoning and dystonic reactions to
antidopaminergic drugs. The most important part in the laboratory
diagnosis is the confirmation of virulence to differentiate Cl. tetani
from other morphologically similar nonpathogenic Clostridia,
which include Cl. tetanomorphum and Cl. sphenoides [1].
Toxicity testing in guinea pigs or mice was the only method
available in the past. Recently, a number of molecular methods
have been developed to confirm the toxicity. In view of the strict
rules governing the animal experiments, especially in countries like
India, molecular methods are most suitable for the confirmation of
virulence of the strain by detection of neurotoxin. Also these tests
are very rapid and help in definitive diagnosis as well as in confirma-
tion of virulence and to differentiate Cl. tetani from other morpho-
logically similar nonpathogenic Clostridia mentioned above [14].

1.2 Importance As there is a wide variation in metabolic activities and nutritional

of Molecular Methods requirements, it is difficult to identify bacteria belonging to genus
Clostridium based on phenotypic characters [15]. Moreover, the
confirmation of virulence is still more difficult, and no single test is
available other than toxicity testing in animals, which is not feasible
in most laboratories. Finally, because of ethical rules governing the
use of animals for experimentation, it is very difficult to get permis-
sion to conduct such experiments.
In view of these problems, molecular methods are the best
alternatives. Such molecular identification techniques are the most
important tools for confirmation because they are species specific
or to a particular bacterial phenotype, like genes for secondary
metabolite production, degradation of certain pollutant, and endo-
toxin or exotoxin.
These techniques are highly sensitive and specific and help in
accurate and rapid detection of microbial pathogens resulting in
definitive diagnosis. They include polymerase chain reaction
(PCR), fluorescent amplification-based specific hybridization
(FLASH)-based PCR assay, loop-mediated isothermal ­amplification
(LAMP) assay, pulsed-field gel electrophoresis (PFGE), and
Southern blot analysis targeting neurotoxin gene (TetX gene).
Other molecular methods targeting 16S rRNA and DNA typing
can also be used for the confirmation of Cl. tetani.
 Molecular Methods for Identification of Clostridium tetani by Targeting Neurotoxin 39

The most commonly used molecular methods in the confirma-

tion of diagnosis of tetanus are TetX gene-targeted methods using
TetX-specific primers targeting the Cl. tetani neurotoxin.

1.3 TetX Gene-­ Tetanus is the result of production of powerful exotoxin (neuro-
Targeted Methods toxin) known as tetanus toxin, also known as the tetanospasmin,
which is released during spore germination. This toxin blocks release
of neurotransmitter from the presynaptic membrane of inhibitory
interneurons of the spinal cord and brainstem that regulate muscle
contraction. Thus, the neurotoxin interferes with the normal nerve
functioning, thereby causing continuous muscle contraction [16].
Cl. tetani contains a double-stranded circular DNA molecule
of 2,799,250 bp in length as its genome. In addition to this chro-
mosomal DNA, it shows the presence of an extra chromosomal
plasmid of 74 kbp that encodes for an exotoxin [16]. The produc-
tion of tetanus toxin is regulated by TetR (direct transcriptional
activator of tetanus toxoid). The TetX (a gene of the tetanus toxin)
and TetR are located on the 74 kbp plasmid (also known as mega-
plasmid) [17, 18]. This means that virulence of Cl. tetani is directly
proportional to the presence of megaplasmid [19]. The size of the
TetX gene is 1354 bp. For PCR amplification of this fragment of
1354 bp, the sequences of oligonucleotide primers are used.
Polymerase chain reaction (PCR) is the most commonly used
molecular technique for detection of tetanospasmin.

2 Materials

2.1 PCR Studies 1. TE (Tris/EDTA) buffer: Prepare a stock solution of 1 M Tris

by dissolving 60.57 g of Tris (hydroxymethyl) aminomethane
2.1.1 DNA Extraction
in 500 mL of milli-Q water, and adjust the pH to 7.5 using
HCl (see Notes 1 and 2). Prepare a stock solution of 0.5 M
EDTA by dissolving 18.6 g diaminoethanetetraacetic acid in
100 mL of milli-Q water (see Note 3).
For DNA isolation, take 10 mL of 1 M Tris–HCl (pH 7.5)
and 2 mL of 0.5 M EDTA (pH 8.0) from stock solution and
fill up to the final volume of 1 L.
2. Sodium dodecyl sulfate (SDS): 10% solution in water (see
Note 4).
3. Proteinase K: Required concentration is 20 mg/mL (stored in
small single-use aliquots at −20 °C).
4. Sodium chloride (NaCl): 5 M NaCl in water.
5. Cetyltrimethylammonium bromide (CTAB)/NaCl solution:
Dissolve 4.1 g NaCl in 80 mL water and slowly add 10 g
CTAB while heating and stirring. Heat to approximately
65 °C to dissolve. Adjust volume to 100 mL with water. Store
at room temperature.
40 Basavraj Nagoba et al.

6. Chloroform/isoamyl alcohol 24:1 (v/v, see Notes 5 and 6).

7. Water-saturated phenol/chloroform/isoamyl alcohol 25:24:1
(by vol., see Notes 5 and 6).
8. Isopropanol (see Note 6).
9. 70% ethanol (see Note 6).
Or
10. Bacterial DNA extraction kit (Merck-Genei).
11. TAE buffer: 1× TAE buffer containing 40 mM Tris, 20 mM
acetate, and 2 mM EDTA (see Note 7).
12. 0.8% agarose horizontal gel electrophoresis: Dissolve 0.8 g of
agarose in TAE buffer (see Note 8).
13. Load extracted DNA sample with 2 μL of GelRed™ dye.
Perform the electrophoresis at approximately 100 V till the
end of the gel, and observe the gel in protein simple gel docu-
mentation unit to visualize extracted DNA.
14. Check the quality of extracted DNA with NanoDrop Lite
spectrophotometer (NanoDrop Biotechnologies). To check
the concentration of extracted DNA, take the absorbance at
260 nm (A260). Quality of DNA can be checked by the ratio of
absorbance at 260 and 280 nm. If the ratio is 1.8, then the
DNA is pure, free from contamination of RNA and protein.

2.1.2 PCR Amplification 1. Primer (Eurofins): Primers are single-stranded short oligonu-
and Sequencing cleotides which bind to the template strand and allow DNA
polymerase to add complementary base pairs to its free 3′OH
end. To amplify both strands of the template, two primers are
required, i.e., forward primer and reverse primer which amplify
the DNA in 5′ → 3′ direction. For PCR amplification add
10 nM concentration of each primer to the reaction tube.
Primers used for the amplification of TetX gene are:
5′ CTG GAT TGT TGG GTT GAT AAT G-3′
5′ ATT TGT CCA TCC TTC ATC TGT AGG-3′
2. Deoxynucleoside triphosphate (dNTP) (Merck-Genei):
dNTPs are the nitrogen bases which have to be added during
the amplification reaction. Mixture of dNTPs contains purines,
i.e., adenine (A) and guanine (G), and pyrimidines, i.e., cyto-
sine (C) and thymine (T), in concentration of 200 μM/mL.
3. Taq polymerase (Genei): Taq polymerase is the DNA poly-
merase enzyme which adds dNTPs to the free 3′OH end of
amplifying strand. Concentration required is 1 U/μL.
4. 10× reaction buffer (Genei) comprised of 100 mM Tris–HCl,
15 mM MgCl2, 500 mM KCl, and pH 8.3 (20 °C).
 Molecular Methods for Identification of Clostridium tetani by Targeting Neurotoxin 41

5. Exo-rSAP (New England Biolabs) method: PCR amplicons

normally contain primer dimers and other impurities. To
purify these amplicons, Exo-SAP method is used. PCR ampli-
cons are mixed with Exo-Ant mix (see Note 9) and molecular
grade PCR water.
6. ABI 3500xl genetic analyzer (Thermo Fisher): This is a
sequencing instrument based on Sanger’s sequencing principle.
Sequencing is done according to the manufacturer’s protocol.

2.1.3 Sequence Analysis 1. Sequence analysis software version 5.1 (Applied Biosystems):
Use this software to check the quality of reads.

2.2 Materials 1. Agarose gel: 2% agarose gel (see Note 8).

Required for Pulsed– 2. Buffer I, a lysis buffer, used for the lysis of agar plugs: 6 mM
Field Gel Tris–HCl (pH 7.4), 1 M NaCl, 10 mM Na2EDTA, 0.5% Brij
Electrophoresis (PFGE) 58, 0.2% Na-deoxycholate, and 0.2% Na-lauroylsarcosine,
and Southern Blot supplemented with 5 mg/mL of lysozyme and 1 μg/mL of
Analysis RNase.
3. Buffer II: 10 mM Tris–HCl (pH 7.4), 1 M Na2EDTA, 1% Na-­
lauroyl sarcosine, and 1 mg/mL proteinase K.
4. TE Buffer: 10 mM Tris–HCl, (pH 7.4), 0.1 mM Na2EDTA.
5. Restriction enzymes: 20 U of ApaI or SmaI. ApaI produces
staggered end, while SmaI produces blunt ends.
6. Tris-borate, EDTA: 44.5 mM Tris-borate, 12.5 mM EDTA
(pH 8.0).
7. Nylon membrane: Positively charged nylon membrane probed
with digoxigenin (DIG)-labeled, double-stranded TetX-
specific DNA fragment.
8. UV chamber: For fixation of transferred fragments on nylon
membrane. Fixation is carried out at 312 nm.

2.3 Loop-Mediated 1. Primers: LAMP technique uses 4–6 pairs of primers, primers
Isothermal for conserved region and internal primer specific to gene of
Amplification (LAMP) interest. For the identification of the presence of TetX gene in
Assay the sample are the following pairs of samples:
F3: 5″-GATAAAGATGCATCTTTAGGATT-3″.
B3: 5″-TCTTCTTCATTATCAACCCAAC-3″.
FIP:5″-AGTTGCTTGCAATTAATATATCCCTAGTAGGTA
CCCATAATGGTCA-­3.
BIP:5″-AACATGTGATTGGTACTTTGTACCTTATGTGT
CTATGGTGTGTTG-­3″.
2. Bst polymerase: The principle of LAMP is autocycling strand
displacement DNA synthesis in the presence of Bst DNA
polymerase with high strand displacement activity under
­
42 Basavraj Nagoba et al.

i­sothermal conditions between 60 and 65 °C resulting in 109

copies of target DNA as well as large amount of by-products
(magnesium pyrophosphate) within an hour.
3. Water bath: For amplification. It is set at 60–65 °C.
4. Turbidometer: Measures the turbidity of by-product formed
due to reaction of positive product with magnesium
pyrophosphate.

3 Methods

3.1 PCR Studies Use of PCR for the detection of the TetX gene is a rapid technique
to confirm the presence of Cl. tetani in clinical samples as well as
Cl. tetani grown in culture. The methodology used includes the
following steps.

3.1.1 DNA Extraction Use a specimen obtained from a patient, collected in Robertson
cooked meat broth, for DNA extraction [14]. Alternatively, sus-
pend colonies of Cl. tetani grown on an agar plate in distilled
water, and lyse these by incubation at 95 °C for 10 min, after cen-
trifugation (76 rcf for 60 s). Use the supernatant of the lysate for
DNA extraction [20]. Bacteria grown in broth culture can also be
used for DNA extraction [21]. Extract bacterial DNA applying a
standard procedure [22], i.e., incubate the cells in a CTAB solu-
tion at 65 °C followed by chloroform extraction and isopropanol
precipitation.
1. Inoculate 5 mL of liquid culture with the bacterial strain and
incubate under suitable conditions for 24 h.
2. Centrifuge 1.5 mL of the culture in a microcentrifuge for 2 min
or until a compact pellet forms. Discard the supernatant.
3. Re-suspend the pellet in TE buffer by repeated pipetting, and
add 30 μL of 10% SDS and 3 μL of 20 mg/mL proteinase K
to a final concentration of 100 μg/mL proteinase K in 0.5%
SDS. Mix thoroughly and incubate for 1 h at 37 °C.
4. Add 100 μL of 5 M NaCl and mix thoroughly. This step is
very important as CTAB–nucleic acid precipitate will form, if
salt concentration drops below about 0.5 M at room tempera-
ture. The aim here is to remove cell wall debris, denatured
protein, and polysaccharides complexed to CTAB, while
retaining the nucleic acids in solution.
5. Add 80 μL of CTAB/NaCl solution. Mix thoroughly and
incubate 10 min at 65 °C.
6. Add an approximately equal volume (0.7–0.8 mL) of chloro-
form/isoamyl alcohol, mix thoroughly, and spin 4–5 min in a
microcentrifuge. This extraction removes CTAB–protein/
polysaccharide complexes.
 Molecular Methods for Identification of Clostridium tetani by Targeting Neurotoxin 43

7. Remove aqueous, viscous supernatant to a fresh microcentri-

fuge tube. Add an equal volume of phenol/chloroform/iso-
amyl alcohol, extract thoroughly, and spin in a microcentrifuge
for 5 min.
8. Transfer the supernatant to a fresh tube. Add 0.6 volume of
isopropanol to precipitate the nucleic acids. Shake the tube
back and forth until a stringy white DNA precipitate becomes
clearly visible.
9. Wash the DNA with 70% ethanol to remove residual CTAB
and respin for 5 min at room temperature to re-pellet it.
Carefully remove the supernatant and briefly dry the pellet in
a vacuum concentrator.
10. Re-dissolve the pellet in 100 μL TE buffer. This may take some
time (up to 1 h) since the DNA is of high molecular mass. 15 μL
of this DNA will typically be fully digested by 10 U of EcoRI in
1 h, which is sufficient to be clearly visible on an ­agarose gel.

3.1.2 PCR Amplification 1. Use the Cl. tetani-specific primers targeting a 1354 bp frag-
and Sequencing ment of the TetX gene to amplify the DNA extracted from the
specimen. The PCR reaction containing 10 nM (each) primer
(Eurofins), 200 μM (each) deoxynucleoside triphosphate
(dNTP) (Genei), 1 U of Taq polymerase (Genei) in the appro-
priate reaction buffer, and 50 and 100 ng of DNA extracts as
the templates are carried out in 50 μL of reaction mixture.
2. Use as cycling conditions: 95 °C for 10 min, followed by
25 cycles each consisting of 1 min at 94 °C, 1 min at 52 °C,
and 1.5 min at 72 °C.
3. Document positive PCR amplicons.
4. Purify using Exo-rSAP (New England Biolabs) method. Mix
5 μL of PCR product with 2 μL of Exo-Ant mix (see Note 9).
Heat reaction tube in thermocycler at 37 °C for 30 min and
80 °C for 20 min. Then, cool at 4 °C in thermocycler only.
Check the purity and store samples at −20 °C. Sequence both
strands using 3500xl genetic analyzer (Thermo Fisher) [23].

3.1.3 Detection of PCR The specific PCR amplification product (amplicons) containing
Products the target nucleic acid can be detected by using different tech-
niques such as agarose gel electrophoresis, Southern blot analysis,
dot blot analysis, and other methods. Agarose gel electrophoresis
is the most commonly used method.

3.1.4 Sequence Analysis 1. Assemble and edit the obtained sequences using sequence
and Interpretation analysis software version 5.1 (Thermo Fisher).
2. Submit edited sequences to BLASTN and BLASTX using
default parameters for analysis [24], followed by comparison
with closest homologous sequences retrieved from the
GenBank database.
44 Basavraj Nagoba et al.

3.1.5 Fluorescent This is a FLASH based PCR assay. This PCR assay used for rapid iden-
Amplification-­Based tification of cultured isolates of Cl. tetani appears to be more suitable
Specific Hybridization and rapid [21]. This method is also a TetX gene-targeting method.
(FLASH)-Based PCR Assay The steps used are the same as in PCR assay, except for the last step:
1. Analyze the PCR products by agarose gel electrophoresis.
2. Save the images with the help of the fluorescence detector gene.
3. Analyze the obtained fluorescence data using gene software.
This method gives clear and unambiguous results [21].

3.2 Other Molecular 1. Grow Cl. tetani on solid media, collect, adjust to 2 × 108 CFU/
Methods Used for mL, and then mix 1:1 with 2% agarose to prepare agarose plugs.
Detection of Cl. tetani 2. Lyse the embedded cells in buffer I solution supplemented
by Targeting TetX Gene with 5 mg/mL lysozyme and 1 μg/mL RNase for 5 h at 37 °C.
(See Note 10) 3. Incubate the plugs overnight in 1 mL of buffer II.
3.2.1 PFGE and Southern 4. Wash the plugs thoroughly with TE buffer.
Blot Analysis
5. Digest agarose plugs overnight at 25 °C with 20 U of ApaI or
SmaI.
6. Perform electrophoresis on 1% PFGE-grade agarose using a
CHEF Mapper XA system. Running conditions are 0.5–20 s
for both enzymes, using switch times of 0.5–20 s for 27 h at
6.0 V/cm, 14 °C, in 0.5× TBE (Tris-borate, EDTA).
7. Stain gels either with ethidium bromide and analyze using a
Gel Doc 1000 system (BIO-Rad labs) or prepare Southern
blots, which are subsequently probed with a 1354 bp
digoxigenin (DIG)-labeled, double-stranded probe specific
­
for TetX gene. Prepare this probe by PCR using the DIG
Chem-Link labeling and detection set kit as recommended by
the manufacturer.
8. Fix the DIG-labeled probe specific to TetX gene on a nylon
membrane.
9. Transfer the PFGE-separated DNA fragments to positively
charged nylon membrane, and fix at 312 nm for 2 min.
10. Hybridize the blots against the DIG-labeled TetX probe.
11. Detect the hybridized TetX probe by using a DIG chemilumi-
nescence detection system with chloroplastic drought stress
protein as the substrate [23].

4 Notes

1. Having water at the bottom of the cylinder helps to dissolve

Tris relatively easily, allowing the magnetic stir bar to go to
work immediately. If using a glass beaker, Tris can be dissolved
faster provided the water is warmed to about 37 °C. However,
the downside is that care should be taken to bring the solution
to room temperature before adjusting pH.
Molecular Methods for Identification of Clostridium tetani by Targeting Neurotoxin 45

2. Concentrated HCl (12 N) can be used at first to narrow the

gap from the starting pH to the required pH. However, it
would be better to use a series of HCl (e.g., 6 N and 1 N) with
lower ionic strengths to avoid a sudden drop in pH below the
required pH. Use of safety wears is recommended while using
corrosive reagents.
3. Note that EDTA will not be soluble until pH reaches to 8.0.
Adjust the pH using NaOH. Use vigorous stirring, moderate
heat to dissolve EDTA, if required.
4. SDS precipitates at 4 °C. Therefore, the lysis buffer needs to
be warmed prior to use.
5. Always prepare fresh mixture.
6. It is advisable to use prechilled (≤4 °C) reagents for good
quality DNA isolation.
7. TAE buffer: Prepare 50× stock solution of TAE buffer. To
prepare 50× TAE buffer, dissolve 242 g Tris base, 57.1 mL
glacial acetic acid, and 100 mL of 0.5 M EDTA (pH 8.0) into
600 mL of deionized water. Adjust the final volume to 1 L
with deionized water.
To prepare a 1× working solution from 50× stock buffer,
mix 50× stock buffer with DNAse-free deionized water 1:4
by vol.
8. Add agarose powder in buffer solution and boil it until all the
agarose is melted. Cool it and pour it into casting tray. Allow
it to solidify.
9. Preparation of Exo-Ant mix: Mix 0.2 μL of antarctic phospha-
tase and ten times diluted 0.5 μL of ExoI in 1.3 μL of molecular-­
grade PCR water (HiMedia). Store the reagent at 4 °C.
10. The LAMP assay is another assay in which TetX gene of Cl.
tetani is used as the target gene. LAMP is a rapid, highly spe-
cific, and sensitive technique for the identification of this bac-
terium. It is an amplification-based identification technique,
but unlike PCR, LAMP does not require a thermal cycler and
multiple steps for amplification. It amplifies the template
strand under isothermal condition at 63–67 °C. LAMP assay
uses two pairs of primers, namely, a pair of outer primers and
a pair of inner primers which detect the conserved region of a
gene, and amplification is done using Bst polymerase [25].
The reaction can be identified with the help of a turbidimeter.
Turbidity is due to the formation of the by-product, magne-
sium dihydrogen phosphate. Alternatively, agarose gel electro-
phoresis can also be used or by applying appropriate
chromogenic substances like hydroxynapthol blue. Thus, spe-
cial testing equipment is not required. Jiang et al. used tetanus
toxin gene as target gene for the LAMP assay [25]. The accu-
racy and specificity of LAMP assay has been confirmed by
comparison of LAMP assay with API 20A test [26].
46 Basavraj Nagoba et al.

Acknowledgments

Authors wish to thank Dr. Milind Davane, Asst. Professor in

Microbiology, and Mr. Vinod Jogdand, Asst. Librarian, MIMSR
Medical College, Latur, for their help in the preparation of the
manuscript.

References
1. Nagoba BS, Pichare AP (2016) Medical 13. Longo DL, Fauci AS, Kasper DL, Hauser SL,
microbiology and parasitology, 3rd edn. Jameson JL, Loscalzo J (2012) Harrison’ prin-
Elsevier, New Delhi, pp 283–288 ciples of internal medicine, vol Vol 1, 18th
2. Tullu MS, Deshmukh CT, Kamat JR (2000) edn. The McGraw-Hill Companies, London,
Experience of pediatric tetanus cases from pp 1197–1200
Mumbai. Indian Pediatr 37:765–771 14. Madhu G, Sheikh N, Shah P, Mehetre G,
3. Lodha R, Sareen A, Kumar RM, Arora NK Dharne MS, Nagoba BS (2015) Detection of
(2000) Tetanus in immunized children. Indian Clostridium tetani in human clinical samples
Pediatr 37:223–224 using tetX specific primers targeting the neu-
4. Gibson K, Bonaventure UJ, Kiviri W, Parlow rotoxin. J Infect Public Health 9:105–109
J (2009) Tetanus in developing countries: a 15. Cato EP, Georgew L, Finegolds M (1986)
case series and review. Can J Anaesth Genus Clostridium prazmowski 1880,
56:307–315 23AL. In: PHA S, Mair NS, Sharpe ME, Holt
5. Alagbe-Briggs OT, Tinubu SA (2012) JG (eds) Bergey’s manual of systematic bacte-
Tetanus - a case report with severe autonomic riology, vol vol. 2. Williams & Wilkins,
instability and: a review of the literature. Niger Baltimore, pp 1141–1200
J Med 21:353–356 16. Quantitation of Clostridium tetani genomes.
6. Sanford JP (1995) Tetanus – forgotten but not Genesig Standard kit handbook HB10.04.08.
gone. N Engl J Med 332:812–813 Published Date: 26/04/2016
7. Cook TM, Protheroe RT, Handel JM (2001) 17. Bruggemann H, Bauumer S, Fricke WF,
Tetanus: a review of the literature. Br J Anaesth Wiezer A, Liesegang H, Decker I et al (2003)
87:477–487 The genome sequence of Clostridium tetani,
the causative agent of tetanus disease. Proc
8. Nagoba BS, Jahagirdar VL, Sheikh NK (2016) Natl Acad Sci U S A 100:1316–1321
Tetanus is not the past – It still exists. J Patient
Safety Infect Control. 4: (in press). doi:10.1016/ 18. Marvaud JC, Eisel U, Binz T, Niemann H,
j.jpsic.2015.11.006 Popoff MR (1998) TetR is a positive regulator
of the tetanus toxin gene in Clostridium tetani
9. Chang SC, Wang CL (2010) Neonatal tetanus and is homologous to botR. Infect Immun
after home delivery: report of one case. Pediatr 66:5698–5702
Neonatol 51:182–185
19. Laird WJ, Aaronson W, Silver RP, Habig WH,
10. Njiki Kinkela MN, Nguefack F, Mbassi Awa Hardegnee MC (1980) Plasmid associated
H, Chelo D, Enyama D, Mbollo Kobela M, toxigenicity in Clostridium tetani. J Infect Dis
Koki Ndombo PO (2012) Tetanus in older 142:623
children in a pediatric hospital in Yaounde,
Cameroon. Pan Afr Med J 11:37 20. Nagao K, Mori T, Sawada C, Sasakawa C,
Kanezaki Y (2007) Detection of the tetanus
11. World Health Organization Report. Tetanus toxin gene by polymerase chain reaction: a case
(total) reported cases from India. http://apps. study. Jpn J Infect Dis 60:149–150
who.int/immunization_monitoring/global-
summar y/timeseries/tsincidencettetanus. 21. Mohammad S, Hossein E, Mohammad R,
html. Accessed on 19 May 2016 Mojtaba S, Saeid AJ (2011) Detection of
Clostridium tetani by fluorescent amplification
12. Marulappa VG, Manjunath R, Mahesh Babu based specific hybridization. Afr J Microbiol
N, Maligegowda L (2012) A ten year retro- Res 5:5489–5492
spective study on adult tetanus at the epidemic
disease (ED) hospital, Mysore in Southern 22. Ausubel FM, Bren R, Kingston RE, Moore
India: a review of 512 cases. J Clin Diagn Res DD, Seidman JG, Struhl K (1988) Current
6:1377–1380 protocols in molecular biology. John Wiley &
Sons, Inc., New York, NY
 Molecular Methods for Identification of Clostridium tetani by Targeting Neurotoxin 47

23. Plourde-Owobi L, Seguin D, Baudin MA, 25. Jiang DN, Xiaoyun P, Jiehong W, Meng L,
Moste C, Rokbi B (2005) Molecular charac- Ping L (2013) Rapid, sensitive and specific
terization of Clostridium tetani strains by detection of Clostridium tetani by loop-­
pulsed-­field gel electrophoresis and colony mediated isothermal amplification assay.
PCR. Appl Environ Microbiol 71:5604–5606 J Microbiol Biotechnol 23:1–6
24. Altschul SF, Madden TL, Schaffer AA, Zhang 26. Notomi T, Okayama H, Masubuchi H,
J, Zhang Z, Miller W, Lipman DJ (1997) Yonekawa T, Watanabe K, Amino N, Hase T
Gapped BLAST and PSI- BLAST: a new gen- (2000) Loop mediated isothermal amplifica-
eration of protein database search program. tion of DNA. Nucleic Acids Res 28:E63
Nucleic Acids Res 25:3389–3402
 Chapter 5

Label-Free Immuno-Sensors for the Fast Detection

of Listeria in Food
Alexandra Morlay, Agnès Roux, Vincent Templier, Félix Piat,
and Yoann Roupioz

Abstract
Foodborne diseases are a major concern for both food industry and health organizations due to the eco-
nomic costs and potential threats for human lives. For these reasons, specific regulations impose the
research of pathogenic bacteria in food products. Nevertheless, current methods, references and alterna-
tives, take up to several days and require many handling steps. In order to improve pathogen detection in
food, we developed an immune-sensor, based on Surface Plasmon Resonance imaging (SPRi) and bacterial
growth which allows the detection of a very low number of Listeria monocytogenes in food sample in one
day. Adequate sensitivity is achieved by the deposition of several antibodies in a micro-array format allow-
ing real-time detection. This label-free method thus reduces handling and time to result compared with
current methods.

Key words Listeria monocytogenes, Food safety, Antibody microarrays, Immunosensor, Surface plas-
mon resonance (SPR), Bacterial growth

1 Introduction

Listeria monocytogenes is one of the five most frequent pathogens

involved in death due to foodborne diseases (source CDC, http://
www.cdc.gov/foodborneburden/2011-foodborne-estimates.
html). Moreover, the incidence of listeriosis in the population has
increased in recent years [1] with a mortality rate ranging from 20
to 30% [2]. So far, the standard method for L. monocytogenes detec-
tion in food products is time-consuming both in manipulations
and for getting results. The major drawback of this cultural tech-
nique corresponds to the time necessary for bacterial enrichment,
a step that is even longer for slow growing L. monocytogenes strains.
In fact, according to the ISO 11290, the detection of this bacterial
specie can take up to 6 days [3]. In the last decade, numerous
researches have been investigated and a few alternatives methods
have been developed for the detection of several important

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_5, © Springer Science+Business Media LLC 2017

49
50 Alexandra Morlay et al.

f­ood-­borne pathogens such as L. monocytogenes. For instance,

molecular techniques allowing bacterial DNA detection using
Polymerase-­Chain Reaction (PCR) after specific genetic sequence
amplification have appeared as well as some immunoassays meth-
ods including Enzyme-Linked Immunosorbent Assay (ELISA),
Enzyme-Linked Fluorescent Assay (ELFA) or lateral-flow assays
[4]. Nevertheless, although faster than standards methods, these
bacterial detection systems still require one or two enrichment
steps which may last several days before the bacterial concentration
has reached the limit of detection of the final analysis step. To our
knowledge, only one example of L. monocytogenes detection led
during the enrichment step has been described so far. It consists in
incorporating antibody-functionalized nanoparticles [5]. In this
work, the pathogen detection, using Surface Enhanced Raman
Scattering (SERS), is based on an immunoassay which aggregates
gold and magnetic labeling nanoparticles. In order to going one
step further in the development of easy-to-operate techniques for
the specific detection of L. monocytogenes, we designed a specific
real-time antibody microarray allowing the detection label-free
during the enrichment of food samples. This technology is based
on the use of Surface Plasmon Resonance imaging (SPRi) of the
bacterial growth [6].

2 Materials

2.1 Synthesis 1. 2,5-Dimethoxytetrahydrofuran, mixture of cis and trans iso-

of N-Hydroxy-­ mers (98%).
Succinimidyl-6-(1H- 2. 6-Aminocaproic acid (≥99.0%) (Acros Organics).
Pyrrole-1-­yl)-
3. Acetic acid.
Hexanamide
(Pyrrole-NHS) 4. 1,4-Dioxane (99.8%).
5. N-Hydrosuccinimide (98%).
6. N,N′-Dicyclohexylcarbodiimide (99%).
7. N,N-Dimethylformamide (DMF) (≥99.0%).
8. Dichloromethane (≥99.8%).
9. Ethanol (≥99.8%).
10. Silicagel PF254 containing 15% CaSO4 for chromatography
(Sigma-Aldrich).

2.2 Coupling 1. Use four IgG type polyclonal antibodies, reacting with specific
Antibodies strains of L. monocytogenes: LIM 14 and LIM 16 (Prestodiag),
to Pyrrole–NHS Ab20506 (Abcam) and Ab78729 (Abcam). Re-suspend all
antibodies are re-suspended in storage buffer (PBS-containing
0.1% sodium azide) at 0.5 mg/mL. Follow supplier recom-
mendations regarding storage (4 °C for short term and −20 °C
for long-term storage) (see Notes 1 and 2).
 Label-Free Immuno-Sensors for the Fast Detection of Listeria in Food 51

2. Use one monoclonal antibody, KLH (Keyhole Limpet

Hemocyanin), which not reacts with bacterial models inter-
esting for human health, as negative control (CEA Marcoule)
(see Note 3).
3. 4 mM Pyrrole–NHS in dry DMSO (prepared as described in
Subheading 3.1). Store at −20 °C.
4. Phosphate buffered saline (PBS: 10 mM phosphate buffer,
2.7 mM KCl, 137 mM NaCl, pH 7.4). Dissolve under stirring
1 tablet in 200 mL of ultra-pure water. After preparation, ster-
ilize by autoclaving at 121 °C during 15 min. Store at 4 °C.
5. Vivaspin 500 μL ultracentrifugation spin columns with 30,000
MWCO cutoff membrane in polyethersulfone (Sartorius
Stedim Biotech GmbH).
6. Electropolymerization buffer: 50 mM NaH2PO4, 50 mM
NaCl, 10% (w/v) glycerol. The final pH is adjusted to 6.8 with
NaOH. Store at 4 °C.
7. Spectrophotometer (NanoDrop, ND-1000, ThermoScientific).

2.3 Electrochemical 1. Biochips: glass prisms coated with two metallic layers: 2 nm
Deposition of Pyrrole– chrome and 50 nm gold (Prestodiag).
Antibody Conjugates 2. 1 M Pyrrole (Interchim) in acetonitrile. Store at −20 °C.
on SPRi Biochip
3. 5 μM Pyrrole–antibody conjugates (prepared as described in
Surface Subheading 3.2).
4. The microarrayer used for immobilizing Pyrrole-IgG conju-
gates on the prism is commercially available (OmniGrid Micro
robotic Arrayer).
5. Bovine serum albumin (BSA), 0.1% in PBS.
6. 200 μL small PCR tubes.

2.4 Sample 1. Ready to use Demi-Fraser broth (Biokar diagnostics). Store

Preparation protected from light at 4 °C.
and Detection of 2. Tryptic Soy Broth (TSB). Prepare according to manufacturer
L. monocytogenes instructions.
Strains Using SPRi 3. Buffer peptone water (BPW, Biokar diagnostics). Prepare
according to manufacturer instructions.
4. Tryptone Soy Agar (TSA), ready to use media in 200 mL bot-
tle (bioMérieux). Prepare according to manufacturer
instructions.
5. Compass Listeria Agar (Biokar diagnostics), ready to use. Store
at 4 °C.
6. Two L. monocytogenes strains (A and B, serovars 1/2b and 4b
respectively) were used in this study. They were isolated from
natural contaminated food and provided by the Institut
Scientifique d’Hygiène et d’Analyse (ISHA).
52 Alexandra Morlay et al.

7. Stomacher® Bag containing 280 μm filter membrane.

8. Peristaltic blender (Bag mixer, Interscience).
9. 14 mL sterile culture tubes.
10. Packaged salad (lettuce) bought in a local store.
11. Weighing scale.
12. Densitometer.
13. The SPRi system, called Monopresto™ (Prestodiag) allows the
detection of pathogens in food samples. It is composed of
two distinct parts: (1) the MonoPresto™ reader and the
Bacterics™ software. Detection tests are carried by PlasmIA
™ technology [7], and (2) the Prestokit comprising a six
reaction chambers laid on the antibodies coated prism
(Fig. 1). Each isolated chamber can contain 1000 μL of solu-
tion. In order to prevent any leakage, a silicone gasket is
placed between the prism and reaction chambers, all main-
tained tightened together by a clamping device (see Note 4).
Moreover, ­temperature of the Monopresto™ system is regu-
lated by the software allowing optimal growth of bacteria of
interest (for L. monocytogenes a temperature of 37 °C is used)
(see Note 5).

Fig. 1 (a) Prestokit with biochip between the two plastic parts. The plastic parts on the top present six sepa-
rated chambers. (b) The optical reader, MonoPresto. (c) Presentation of the PlasmIA technology. Viable cells
grow in the wells and the binding of bacteria on specific antibodies is continuously monitored in a real-time
and label-free manner
 Label-Free Immuno-Sensors for the Fast Detection of Listeria in Food 53

3 Methods

3.1 Synthesis 1. Heat under reflux for 4 h a mixture of

of Pyrrole-NHS 2,5-­
dimethoxytetrahydrofuran (490 mmol), 6-aminocaproic
acid (430 mmol), acetic acid (450 mL) and 1,4-dioxane
(600 mL).
2. Stir the previous mixture at room temperature overnight.
3. Remove the volatile compounds under reduced pressure and
eliminate acetic acid by co-evaporation with ethanol (2 × 100 mL).
4. Dissolve the solid residues in 500 mL of dichloromethane and
wash this solution with 2 × 250 mL of ultrapure water to remove
residual impurities and unreacted reagents. This solution is evap-
orated until dryness. The product (N-hydrosuccinimide) is
obtained with a yield of 82%. Mass spectrometry: N-­
Hydrosuccinimide m/z 182.1. It is purified by chromatography
on silica gel (600 × g) column with a gradient of ethanol in
dichloromethane. Start the elution with 500 mL of CH2Cl2,
then 300 mL of a mixture of CH2Cl2/EtOH (98/2), 400 mL of
CH2Cl2/EtOH (95/5) and finally 300 mL of CH2Cl2/EtOH
(90/10) to elute N-hydrosuccinimide. After evaporation of the
solvents, N-hydrosuccinimide looks like a brownish oil.
5. N-Hydrosuccinimide (144 mmol), and N,N′-
Dicyclohexylcarbodiimide (159 mmol) are dissolved in DMF
and stirred at room temperature overnight.
6. Filter the mixture on a sinter glass to eliminate
N,N′-dicyclohexylurea.
7. Remove the volatile compounds as previously described under
reduced pressure. The obtained product (Pyrrole–NHS) has
the form of a white powder. It can be used directly or after
purification by chromatography as in step 4.

3.2 Coupling The transamidification reaction between antibodies and the acti-
Antibodies vated ester of Pyrrole-NHS involves the primary amine functions
to Pyrrole-NHS of antibody molecules. All buffers and antibody solutions are kept
at 4 °C or on ice between each protocol steps.
1. Before starting the coupling reaction, wash the Vivaspin col-
umn membranes once with 200 μL of PBS and centrifuge (all
centrifugations are performed with the following settings:
speed: 15,000 × g; Temperature: 4 °C; Duration: 15 min).
Then, load 200 μL of IgG (0.5 mg/mL) on the columns and
concentrate. In order to remove all traces of sodium azide,
wash IgG then twice with 200 μL of PBS. Finally, collect puri-
fied IgG with 200 μL of PBS. Determine the IgG solution
concentration by measuring the absorbance at 280 nm (molar
54 Alexandra Morlay et al.

extinction coefficient 210,000 L/mol cm), with a

spectrophotometer.
2. In the reacting solution, set the molar ratio of Pyrrole/IgG to
10. Then, adjust volumes of 200 μM Pyrrole–NHS solution
(prepare by dilution of the 4 mM solution in PBS) and PBS to
reach this molar ratio in a final volume of 400 μL. Typically for
200 μL of an IgG solution at 3.3 μM (= 0.5 mg/mL), add
34 μL of Pyrrole–NHS and 166 μL of PBS.
3. Incubate overnight at 4 °C. Protect tubes from light to mini-
mize Pyrrole photo-oxidation.
4. Remove the unreacted Pyrrole-NHS by centrifugation. First of
all, wash the column with 200 μL of PBS. Then, load 400 μL
of the reacting solution and wash it twice with 200 μL of
PBS. Eventually, collect the Pyrrole–antibody conjugates in
80 μL of electropolymerization buffer. Measure the protein
concentration with a spectrophotometer and adjust the vol-
ume with the electropolymerization buffer to reach a 5 μM
final concentration (see Note 6). For short-term utilization,
keep the solution at 4 °C, otherwise divide the solution in sev-
eral aliquots and store at −20 °C.

3.3 Electrochemical 1. Remove any dust from the gold surface by argon flushing. The
Deposition of Pyrrole– surface is then rinsed by running ultrapure water (18.2 MΩ
Antibody Conjugates cm), absolute ethanol and once again ultrapure water before
on SPRi Biochip being dried under argon flow.
Surface 2. Prepare a 100 mM Pyrrole solution by adding 10 μL of the
1 M stock solution to 90 μL of the electropolymerization
buffer.
3. Prepare the spotting solutions by mixing 3 μL of the 100 mM
Pyrrole solution with 3 μL of the antibody–Pyrrole conju-
gates (5 μM) and 9 μL of electropolymerization buffer to
reach a final volume of 15 μL. The final concentrations in
solution are 20 mM Pyrrole and 1 μM antibody-Pyrrole con-
jugates (see Note 7).
4. Homogenize the spotting solutions by several pipetting before
filling a 96-well plate with 11 μL of each sample to be electro-­
chemically grafted on the surface. Make sure that the solution
reaches the bottom of the well and avoid air bubbles formation
(see Note 8).
5. The electropolymerization of the spotting solutions is achieved
using the microarrayer. Briefly, plunge a stainless steel needle
in a specific well where it takes up about 5 μL of solution.
Then, direct it to a desired location on the prism and move it
down to the surface. This step allows the deposit of a drop on
the prism gold surface and the formation of an interface
between two electrodes; the needle (counter electrode) and
 Label-Free Immuno-Sensors for the Fast Detection of Listeria in Food 55

the gold surface (working electrode). At that precise time, trig-

ger an electrical pulse of a difference of potentiel ranging from
0.4 to 2.4 V, and maintainit for 100 ms (see Note 9), between
the two electrodes. This step leads to Pyrrole polymerization
and trappes the antibodies onto the gold surface in a thin and
solid polyPyrrole film. Once all the solutions of an antibody
specie are arrayed, rinse the needle and dry three times using
ultra-pure water. Repeat the same procedure for all antibody
species (see Note 10). Spot diameter is 350 μm. A pitch of
450 nm between spots avoids their coalescence (see Note 11).
Electrodeposition takes place in a closed chamber with a
hygrometry level above 60%.
6. Passivate the functionalized prism then with 0.1% BSA solu-
tion for 20 min to prevent unspecific binding. After three
washings with PBS, keep it finally in PBS at 4 °C before use.

3.4 Sample All the steps described here are made using the two L. monocyto-
Preparation genes strains in parallel.
and Detection of
1. Streak bacteria from stock suspension conserved at −80 °C on
L. monocytogenes TSA plates. Incubate 24 h at 37 °C.
Strains Using SPR
Imaging
2. Take one isolated colony from the TSA plate using a sterile
inoculation loop and re-suspend it in 6 mL of TSB placed in a
14 mL culture tube. Incubate overnight at 37 °C.
3. Dilute 1 mL of the culture in 9 mL of buffered peptone water.
Measure the turbidity of the suspension using the densitome-
ter and adjust it to 1.0 ± 0.2 Mac Farland (approximately
3 × 108 CFU/mL) by adding buffered peptone water. From
this bacterial suspension, proceed to six tenfold serial dilutions
in TSB (100 μL of culture in 900 μL TSB). Spread two TSA
plates with 300 μL of the last dilution (10−6), incubate at 37 °C
and perform colony counting the next day (the number of
CFU is usually around 30 CFU). In the same time, incubate a
14 mL tube with 3 mL of the TSB used for the serial dilution.
The media sterility is guaranteed by the absence of bacterial
growth in this tube.
4. Keep 150 μL of the last dilution and incubate it at 37 °C for
24 ± 1 h. This sample will serve as a positive control containing
only the Listeria strain used for the seeding.
5. Two sample types containing salad are prepared: one negative
control without bacteria and one artificially contaminated with
bacteria of interest. For both bag types, weigh 25 g of lettuce,
put it onto a Stomacher® Bag and add 250 mL of Demi-Fraser
broth. This content is then blended for 1 min at maximal
speed. Add 300 μL of the last bacterial dilution (10−6 corre-
sponding to 20–40 CFU) into dedicated bags. Incubate all
samples 24 ± 1 h at 37 °C.
56 Alexandra Morlay et al.

6. Before the end of the food samples incubation, put up together

the two parts of the Prestokit (including the prism) and add
900 μL of TSB in each chambers. Place it inside the
Monopresto reader. Follow the software instruction leading
to the device calibration to launch the regulation at 37 °C. Let
it stand there for at least 30 min before adding the food sam-
ples (see Notes 11 and 12).
7. After 24 ± 1 h of the incubation, manually homogenize, then
take out about 5 mL of each bag (in the filtered side) and place
it in a 14 mL culture tube. With a 10 μL inoculation loop,
streak each sample onto Compass Listeria agar plates to ensure
the presence (or absence) of the L. monocytogenes species. Put
the plates at 37 °C for 24 h before interpretations. Also, pro-
ceed to seven tenfold serial dilutions of samples in TSB and
spread 2 times 100 μL of the two last dilutions with a sterile
colony spreader onto TSA plates. Incubate 24 h at 37 °C and
count the colony in order to estimate the initial number of
bacteria in each chamber.
8. Once the TSB media presents in the Prestokit has warmed up
sufficiently, add 100 μL of each sample in each chamber.
Typically, the same SPRi experiment monitors the presence of
L. monocytogenes in a negative control (salad without bacteria)
and in a test experiment (salad cultivated in the presence of
20–40 CFU of one of the L. monocytogenes specie). In the same
time, launch analysis with a positive control chamber by add-
ing 100 μL of the pure culture of the same species of L. mono-
cytogenes used for the salad contamination bag. Each sample
should be tested at least in triplicates in two independent
experiments.
9. When all the chambers of the Prestokit are filled with the dif-
ferent samples, start the SPRi experiment (see Notes 13–15).
For L. monocytogenes, the run lasts 4 h at 37 °C.

4 Notes

1. Four antibodies have been tested for L. monocytogenes detec-

tion by SPRi after Pyrrole-NHS coupling. All of them enable
the visualization of specific recognition. However, LIM 14 and
LIM 16 showed a better sensibility than the two other anti-
bodies as indicated by the time-to-results presented in Table 1.
2. Regarding the antibody choice, it is important to choose puri-
fied antibodies with a concentration of at least 0.5 mg/
mL. Lower concentrations can result in important losses dur-
ing centrifugation and concentration steps throughout the
antibody conjugation to Pyrrole.
 Label-Free Immuno-Sensors for the Fast Detection of Listeria in Food 57

Table 1
Detection of L. monocytogenes in salad samples

Bacterial
concentration Time-to-results (min)
Initial at the launch
inoculum of the SPRi Ab Ab
Bacterial (CFU/25 g analysis (CFU/ LIM 14 LIM 16 20506 78729 KLH
Medium strains of salad) mL) antibody antibody antibody antibody antibody
Demi-­ A 27 ± 6 9 E6 25 20 30 20 NR
Fraser
A 27 ± 6 9 E6 35 30 45 30 NR
&
salad A 23 ± 2 1 E7 45 40 115 50 NR
B 23 ± 2 7 E6 45 45 55 55 NR
B 33 ± 3 1.2 E7 65 55 170 65 NR
B 33 ± 3 1.2 E7 65 55 30 25 NR
None NA NA NR NR NR NR NR
None NA NA NR NR NR NR NR
TSB A NA 7 E6 65 55 110 135 NR
A NA 5 E6 120 145 145 130 NR
B NA 1.7 E7 100 90 150 155 NR
B NA 2.3 E7 95 95 195 155 NR
The two Listeria strains in food samples are detected in less than 1 h. The time-to-result is obtained as soon as signals
corresponding to one ROI exceeded the signal of the negative antibody (KLH) by one unit. NA means non applicable
and NR means no recognition

3. KLH antibody recognizes the keyhole limpet hemocyanin pro-

tein from Megathura crenulata. Therefore, it should not inter-
act with Listeria species. However, it can be substituted by any
antibodies that do not interact with Listeria species.
4. The Prestokit, including the prism, is intended for single-use.
This procedure avoids any risk of cross-contaminations
between successive analyses and is easier and quicker.
5. The SPRi device is devoid of any fluidic channels for sample
injection. It eliminates the needs of complex sterilization pro-
cedure and risk of leakage.
6. At the end of the coupling step, the antibody concentration
can be below 5 μM due to loss on the membrane. This is not a
major problem since the solution will be later diluted to a final
concentration of 1 μM in the spotting solution.
7. At the end of the step 4 in Subheading 3.2, the final antibody–
Pyrrole conjugates concentration of 1 μM ensures a good bal-
ance between signal quality and quantity of antibodies used.
58 Alexandra Morlay et al.

However, other concentrations have been tested with success

(data not shown). Usually these solutions are prepared in small
PCR tubes ensuring a sufficient level of liquid for pipetting.
8. Air bubbles in the spotting needle perturb the electric current
flow, preventing a correct electropolymerization. If the electropo-
lymerization process failes, it is necessary to wash and dry the nee-
dle three times before sucking up again the antibody solution.
9. These electropolymerization parameters ensure the formation
of a thick poly-Pyrrole film (around 4 nm). Other electrochem-
ical conditions regarding pulse intensity and duration can be
tested. Nonetheless, reader should keep in mind that a thicker
poly-Pyrrole film can be detrimental for SPR sensitivity. Indeed,
the intensity of the evanescent wave, necessary for the SPR phe-
nomenon, decreases exponentially with thicker layers.
10. Spots are arrayed in triplicate to ensure a good reproducibility
of the interaction. In order to confirm the specificity of the
recognition, at least one spot line should be composed of an
irrelevant antibody (here KLH).
11. During the electrodeposition, the hygrometry level is a crucial
parameter to take into account. In fact, if it is not set up cor-
rectly, the antibody spots could dry too quickly or be spread on
the gold surface which would cause mixes of antibodies.
12. In order to avoid refractive index changes caused by the
medium temperature rise during the detection, the TSB is
heated at 37 °C for 30 min before the experiment beginning.
13. The interactions between bacteria and regions of interest
(ROI) corresponding to each antibody spot are viewable in
real time through the MonoPresto™ interface. These ROI cor-
responding to an established spotting array are automatically
defined by the software at the beginning of the experiment.
14. As the interactions between antibody spots are reproducible, a
mean signal can be calculated.
15. Once the incubation in the SPRi is ended, we subtract the
KLH signal from other antibodies. This post-treatment ­analysis
eliminates the non-specific signal impacting all antibodies
which can be induced by media changes (acidification) or
interactions with the sample endogenous bacteria.

Acknowledgements

The work is part of a PhD thesis (A. M.) funded by the National
Association for Research and Technology (ANRT n° 624/2013). This
project was also partly supported by the Labex ARCANE program
(ANR-11-LABX-0003-01).
 Label-Free Immuno-Sensors for the Fast Detection of Listeria in Food 59

References
1. Magalhães R, Mena C, Ferreira V et al. (2014) 5. Weidemaier K, Carruthers E, Curry A et al
Bacteria: Listeria monocytogenes. Encyclopedia (2015) Real-time pathogen monitoring during
of food safety, Oxford, UK, Elsevier enrichment: a novel nanotechnology-based
2. Todd ECD, Notermans S (2011) Surveillance approach to food safety testing. Int J Food
of listeriosis and its causative pathogen, Listeria Microbiol 198:19–27
monocytogenes. Food Control 22:1484–1490 6. Bouguelia S, Roupioz Y, Slimani S et al (2013)
3. Pusztahelyi T, Szabó J, Dombrádi Z et al (2016) On-chip microbial culture for the specific detec-
Foodborne Listeria monocytogenes: a real chal- tion of very low levels of bacteria. Lab Chip
lenge in quality control. Scientifica (Cairo) 13:4024–4032
2016:5768526 7. Morlay A, Piat F, Mercey T et al (2016)
4. Välimaa T-A, Tilsala-Timisjärvi A, Virtanen E Immunological detection of Cronobacter and
et al (2015) Rapid detection and identification Salmonella in powdered infant formula by plas-
methods for Listeria monocytogenes in the food monic label-free assay. Letter Appl Microbiol
chain—a review. Food Control 55:103–114 62:459–465
 Chapter 6

Aptamer-Based Trapping: Enrichment of Bacillus cereus

Spores for Real-Time PCR Detection
Christin Fischer and Markus Fischer

Abstract
Aptamer-based trapping techniques are in general suitable to replace common antibody-based enrichment
approaches. A time-consuming isolation or clean-up is often necessary during sample preparation, e.g. for
the detection of spores. For the development of bioanalytical routine approaches, aptamers with a high
affinity to B. cereus spores were applied for the establishment and validation of an aptamer-based trapping
technique in milk with fat contents between 0.3 and 3.5%. Thereby, enrichment factors of up to sixfold
were achieved. The combination of an aptamer-based enrichment by magnetic separation and the subse-
quent specific real-time PCR detection represents a reliable and rapid detection system.

Key words Aptamer, Bacillus cereus, Spore trapping, Real-time PCR, Milk, Food poisoning

1 Introduction

Bacillus cereus is a common food poisoning microorganism, which

is closely related to Bacillus anthracis, and Bacillus thuringiensis
[1–6]. The spore-forming B. cereus is a Gram-positive food patho-
gen which can be found in milk and dairy products. The pathoge-
nicity is based on two different types of toxins, the diarrhea-inducing
and the emetic toxin [7–10].
These toxins are harmful, especially for young, old, pregnant
and immunocompromised persons (YOPIs). Thus, it is very
important to prevent food from B. cereus contaminations under
aspects of consumer protection and quality assurance. Concerning
the pathogenicity (toxin producers) various legal regulations are
available. According to article 3 section 1 in conjunction with
annex 1 chapter 2 process hygiene criteria (chapter 2.2 milk and
dairy products) of Commission Regulation (EC) No. 2073/2005
on microbiological criteria for foodstuffs, microorganisms and
the corresponding toxins are strictly regulated so that rapid and
reliable analytical assays are mandatory [11].

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_6, © Springer Science+Business Media LLC 2017

61
62 Christin Fischer and Markus Fischer

The detection of B. cereus spores in milk could be fulfilled by

the utilization of in vitro generated aptamers (i.e. receptors) in
combination with various readout techniques (i.e. PCR or others
like LFD, or sensor chips) [12]. To establish a rapid and cost-­
efficient trapping and detection method for routine analysis with-
out the necessity to perform a time-consuming microbiological
enrichment, an aptamer-based trapping procedure was developed.
In order to validate the trapping technique, a specific optimized
real-time PCR method for detection of B. cereus was applied [13].
The trapping was developed in milk simulating buffer and tested in
sterilized milk (containing 0.3, 1.5, and 3.5% fat) for verification of
suitability under real conditions.

2 Materials

All solutions have to be prepared under the use of ultrapure water

and analytical grade reagents. Prepare and store all reagents at room
temperature (unless it is indicated otherwise). Diligently follow all
waste disposal regulations when disposing waste materials. Generally,
we work under sterile conditions thus there is no need to add
sodium azide to the solutions (unless it is indicated). Pathogenic
microorganisms were autoclaved and disposed afterwards.

2.1 Preparation 1. SiMAG-Carboxyl beads: 10 mg/mL (chemicell GmbH,

of Aptamer-­Linked Berlin, Germany).
Magnetic Beads 2. MES-buffer: 0.1 M 2-(N-morpholino)ethanesulfonic acid,
pH 5.0.
3. Aminated aptamers: 100 μM (Integrated DNA Technologies,
Inc., Leuven, Belgium) with an affinity towards the selected
target molecule (Table 1).
4. PBS buffer: 137 mM NaCl, 2.7 mM KCl, 12 mM Na2HPO4,
pH 7.4.
5. Blocking and Storage-buffer: PBS buffer containing 0.1% BSA,
0.05% sodium azide (see Note 1).

2.2 Aptamer-Based 1. Milk simulating buffer: 55 mM NaCl, 20 mM MgCl2, 67 mM

Trapping CaCl2, 80 mM KCl, 40 mM Tris–HCl, pH 6.6 (see Note 2).
2. Milk samples: three commercial available milk samples with
different fat contents (0.3, 1.5, and 3.5%).
3. B. cereus spores: strains MHI M1 and two wild types [107 CFU/
mL received from Lehrstuhl für Hygiene und Technologie der
Milch (Oberschleißheim, Germany)].

2.3 Spore Lysis 1. Glass beads: acid-washed, diameter = 0.5 mm.

2. TissueLyser: use for mechanical treatment to lyse the spores
(Qiagen GmbH, Hilden, Germany).
Table 1
Sequences with the regarding calculated dissociation constants for aptamers towards B. cereus spores. Primer regions are italicized [14]

KD
Aptamer Aptamer sequence (5′–3′) [nM]
BacApt3 CATCCGTCACACCTGCTCGGTGCAGACCCATAGGGGGGGCGTGCGGATGTAGGAGTAGGGTGTTGGCTCCCGTATC 35.5
BacApt4 CATCCGTCACACCTGCTCCCAGCGTGCGTCGACCCGGACCCCTGTCAGCCCCCTCGCGGGTGTTGGCTCCCGTATC 44.6
BacApt5 CATCCGTCACACCTGCTCCAGGTGGGGGGGCGTATTACTGAGGCAGAGTAGTTGGCCGGGTGTTGGCTCCCGTATC 19.1
Aptamer-Based Trapping: Enrichment of Bacillus cereus Spores…
63
64 Christin Fischer and Markus Fischer

2.4 Real-Time PCR 1. Real time PCR: iCycler iQ5 (BioRad Laboratories, Inc.;
Detection Hercules, CA).
2. DreamTaq buffer: DreamTaq buffer (10×), unknown compo-
sition (Fisher Scientific-Germany GmbH, Schwerte, Germany).
3. dNTPs: equimolar Mix of dGTP, dCTP, dATP, dTTP, each
10 mM (Bioline GmbH, Luckenwalde, Germany).
4. DreamTaq polymerase: 5.0 U/μL (Fisher Scientific-Germany
GmbH, Schwerte, Germany).
5. Primer: mp3L1R1for and mp3L1R1rev (Invitrogen Life
Technologies GmbH, Darmstadt, Germany) [13].
6. SYBR Green: SYBR Green I nucleic acid gel stain (10,000×,
Invitrogen GmbH, Karlsruhe, Germany).

3 Methods

3.1 Preparation 1. Wash 200 μL SiMAG-Carboxyl beads twice with 1 mL MES-­

of Aptamer-­Linked buffer and discard the supernatant.
Magnetic Beads 2. Suspend the resulting bead-pellet in 250 μL MES-buffer con-
taining 10 mg EDC [1-ethyl-3-(3-dimethylaminopropyl)car-
bodiimide hydrochloride] and incubate for 10 min under
agitation to active the carboxylated bead surface.
3. Add 50 μL of aminated aptamer (100 μm) with an affinity towards
the selected target molecule (here: B. cereus spores) and incubate
the solution for 2 h at room temperature under slow agitation.
4. Wash SiMAG-Carboxyl beads three times with 1 mL PBS-­
buffer, dissolve the beads in 1 mL Blocking and Storage-buffer,
and store at 4 °C (see Note 2).

3.2 Aptamer-Based 1. Refold the aptamer-linked magnetic beads in milk simulating

Trapping buffer for 5 min at 95 °C with subsequent cooling to 4 °C
to avoid heteroduplexes and to ensure correct aptamer fold-
ing (see Note 3).
2. Dilute 10 μL of aptamer-linked magnetic bead mixture with
990 μL (1) milk simulating buffer, (2) milk with 0.3% fat,
(3) milk with 1.5% fat, and (4) milk with 3.5% fat containing
the target molecules (here: B. cereus spores up to 107 CFU/
mL). Incubate the resulting solution for 30 min under agi-
tation (see Note 4).
3. Centrifuge the sample for 2 min at 10,000 × g and remove the
supernatant. Wash the resulting bead pellet three times with
1 mL ddH2O each and finally re-suspend it in 100 μL ddH2O.
 Aptamer-Based Trapping: Enrichment of Bacillus cereus Spores… 65

4. Elute the aptamer-bound spores 5 min at 96 °C. Transfer the

reaction tube in a magnetic separator and collect the hot
supernatant, which contains the enriched spores, in a new
reaction tube. Ensure that the magnetic beads remained in
the first reaction tube and get not collected with the super-
natant, which contains the enriched spores. Storing of the
samples at −20 °C until further use is advisible (see Note 5).
A suitable detection method for quantification is subsequent
necessary. Due to the enrichment by reduction of sample vol-
ume (from 1000 μL to 100 μL) and the clean-up by removal of
matrix, a recently published B. cereus specific real-time PCR
assay could be used. The general process is applicable for a
broad range of other spores or targets, solely the used aptamers
and the subsequent method of quantification should be adjusted
(in the present study real-time-PCR or in other studies e.g.
LC-MS/MS).

3.3 Spore Lysis The primers used for spore quantification via real time PCR target
the hbID gene of B. cereus [13]. To obtain the bacterial DNA as
template for PCR, the spore shell was lysed mechanically as
following:
1. Add 250 mg glass beads to the enriched spore solutions (see
Subheading 3.2) to generate a spore lysate.
2. Lyse the solution by mechanical treatment (5 min at 30 Hz,
twice) using a TissueLyser (Qiagen GmbH, Hilden, Germany).
The supernatant with fragmented spores was subsequently
used for real-time PCR (see Note 6).

3.4 Real-Time PCR 1. Real-time PCR assay (template: lysed spore solution) was per-
Detection formed using a iCycler iQ5 (see Note 7).
2. Set up a reaction mixture consisting of following components:
1× DreamTaq buffer, 0.8 mM dNTPs, 0.25 units DreamTaq
polymerase, 0.25 μM of each primer, 0.3125× SYBR Green I,
3 μL enriched and lysed spore solution and fill up to 20 μL
with ddH2O [13].
3. Perform the real-time PCR with the following temperature
program: initial denaturation at 95 °C for 10 min, 35 cycles
(95 °C for 15 s, 60 °C for 10 s, and 72 °C for 10 s each), final
elongation for 10 min at 72 °C [13].
4. For quantitation and comparison, an external calibration with
B. cereus spores (wild type, 103 to 107 CFU/mL) should be
carried along (Fig. 1, see Note 8).
66 Christin Fischer and Markus Fischer

Fig. 1 (a) Example of amplification curves during real time PCR with different B.
cereus MHI M1 spore concentrations (from left to right): 107 CFU/mL, 106 CFU/
mL, 105 CFU/mL, 104 CFU/mL, 103 CFU/mL and 0 CFU/mL. (b) Resulting calibra-
tion line during real time PCR with R2 = 0.99. Spore concentration (from left to
right): 0 CFU/mL, 103 CFU/mL, 104 CFU/mL, 105 CFU/mL, 106 CFU/mL, 107 CFU/
mL. Reprinted with permission from (Fischer, C.; Hünniger, T.; Jarck, J.-H.;
Frohnmeyer, E.; Kallinich, C.; Haase, I.; Hahn, U.; Fischer, M. Food Sensing:
Aptamer-Based Trapping of Bacillus cereus Spores with Specific Detection via
Real Time PCR in Milk. J Agric Food Chem 63:8050–8057). Copyright (2015)
American Chemical Society

4 Notes

1. BSA was added to the Blocking and Storage buffer to prevent

unspecific interactions by activated but not covalent linked
carboxyl-­groups on the bead surface [15].
2. Milk-simulating-buffer was used during aptamer selection, to
mimic the ionic milieu of milk. To simplify the aptamer selec-
tion, a milk simulating buffer was chosen for selection process.
The buffer excludes some ingredients like e.g. milk sugar or
fat, but corresponds to milk in terms of ionic composition and
pH value. Trapping was also performed in milk-simulating
buffer as a positive control, as it was expected that aptamers
exhibit their highest affinity towards the target in the corre-
sponding selection buffer.
3. Before each trapping experiment, the required volume of mag-
netic beads was transferred to a new reaction tube, the super-
natant was removed in a magnetic field and the beads were
taken up in the initial volume of milk simulating buffer.
 Aptamer-Based Trapping: Enrichment of Bacillus cereus Spores… 67

Table 2
Obtained enrichment factors for aptamer-based trapping in milk simulating buffer and milk with
different fat contents. The maximal trapping factor was determined mathematically to 10. Reprinted
(adapted) with permission from (Fischer, C.; Hünniger, T.; Jarck, J.-H.; Frohnmeyer, E.; Kallinich, C.;
Haase, I.; Hahn, U.; Fischer, M. Food Sensing: Aptamer-Based Trapping of Bacillus cereus Spores with
Specific Detection via Real Time PCR in Milk. Journal of agricultural and food chemistry. 2015. 63,
8050–8057). Copyright (2015) American Chemical Society

Matrix strain Milk-simulating buffer Milk 0.3% fat Milk 1.5% fat Milk 3.5% fat
B. cereus wild type 5.2 3.5 4.3 4.0
B. cereus wild type 5.7 3.1 6.7 4.8
B. cereus MHI M1 7.1 3.9 3.2 7.5
Mean value 6.0 ± 1.0 3.5 ± 0.4 4.7 ± 1.8 5.4 ± 1.9
Mean value 2 4.91 ± 1.07

4. The spore surface due to its dimension probably offers several

different binding sites [16]. To cover a maximum range of
potential binding sites, and enable a binding of a multitude of
the heterogenous spores, a mixture of aptamer-linked mag-
netic beads (BacApt3, BacApt4 and BacApt5) was used for the
presented aptamer-based trapping.
5. The resulting enrichment factor (up to 10×, Table 2) is limited
due to the volume reduction (1000 μL to 100 μL) which is
performed during trapping. You may choose a higher initial
volume or a lower end volume to increase your enrichment
factor and to improve the trapping efficiency. Please consider,
that this might mean that a higher aptamer-linked bead con-
centration is needed for trapping.
6. The lysed spore solution should be used quickly to prevent
enzymatic DNA degradation or similar reactions [17].
7. The lysed spores were diluted prior to measurement 1:10 with
ddH2O.
8. Milk simulating buffer/milk samples spiked at a concentration
equal to pre-enrichment were used as control samples. The
assay was additionally performed using B. subtilis and B.
thuringiensis spores as template to demonstrate the specificity
of the protocol. It was shown, that the protocol with the sub-
sequent real-time detection was specific for B. cereus spores.

Acknowledgement

This research project was supported by the German Ministry of

Economics and Technology (via AiF) and the FEI (Forschungskreis der
Ernaehrungsindustrie e. V., Bonn, Germany); Project AiF 331 ZN.
68 Christin Fischer and Markus Fischer

The author’s thank Dr. Tim Hünniger, Jan-Hinnerk Jarck, and

Esther Frohnmeyer for practical support and discussion.

References
1. Helgason E, Økstad OA, Caugant DA, psychrotrophic Bacillus cereus isolates. Int
Johansen HA, Fouet A, Mock M, Hegna I, J Food Microbiol 27:175–183
Kolstø A-B (2000) Bacillus anthracis, Bacillus 11. Regulation C (2005) No. 2073/2005 of 15
cereus, and Bacillus thuringiensis—one species November 2005 on microbiological criteria for
on the basis of genetic evidence. Appl Environ foodstuffs. Official J Eur Union L 338:1–26
Microbiol 66:2627–2630 12. Tombelli S, Minunni M, Mascini M (2007)
2. Drobniewski F (1994) The safety of Bacillus Aptamers-based assays for diagnostics, envi-
species as insect vector control agents. J Appl ronmental and food analysis. Biomol Eng
Bacteriol 76:101–109 24:191–200
3. Bruno JG, Ulvick SJ, Uzzell GL, Tabb JS, 13. Wehrle E, Didier A, Moravek M, Dietrich R,
Valdes ER, Batt CA (2001) Novel immuno-­ Märtlbauer E (2010) Detection of Bacillus
FRET assay method for Bacillus spores and cereus with enteropathogenic potential by mul-
Escherichia coli O157:H7. Biochem Biophys tiplex real-time PCR based on SYBR green
Res Commun 287:875–880 I. Mol Cell Probes 24:24–130
4. Bennet RW, Harmon SM (1990) Bacillus 14. Fischer C, Hünniger T, Jarck J-H, Frohnmeyer
cereus food poisoning. Laboratory diagnosis of E, Kallinich C, Haase I, Hahn U, Fischer M
infectious disesases: principles and practice. (2015) Food sensing: aptamer-based trapping
Bacterial, mycotic and parasitic diseases. of Bacillus cereus spores with specific detection
Springer-Verlag, New York via real time PCR in milk. J Agric Food Chem
5. Lambert B, Peferoen M (1992) Insecticidal 63:8050–8057
promise of Bacillus thuringiensis. Bioscience 15. Hünniger T, Wessels H, Fischer C, Paschke-­
42:112–122 Kratzin A, Fischer M (2014) Just in time-­
6. de Maagd RA, Bravo A, Crickmore N (2001) selection: a rapid semiautomated SELEX of
How Bacillus thuringiensis has evolved specific DNA aptamers using magnetic separation and
toxins to colonize the insect world. Trends BEAMing. Anal Chem 86:10940–10947
Genet 17:193–199 16. Hünniger T, Fischer C, Wessels H, Hoffmann
7. Setlow P (1994) Mechanisms which contribute A, Paschke-Kratzin A, Haase I, Fischer M
to the long-term survival of spores of Bacillus (2015) Food sensing: selection and character-
species. J Appl Bacteriol 76:49S–60S ization of DNA aptamers to Alicyclobacillus
8. Beutling D, Böttcher C (1998) Bacillus cereus: spores for trapping and detection from orange
ein Risikofaktor in Lebensmitteln. Arch juice. J Agric Food Chem 63:2189–2197
Lebensmittelhyg 49:90–96 17. Hünniger T, Felbinger C, Wessels H, Mast S,
9. Mahler H, Pasi A, Kramer JM, Schulte P, Hoffmann A, Schefer A, Martlbauer E,
Scoging AC, Bar W, Krahenbuhl S (1997) Paschke-Kratzin A, Fischer M (2015) Food
Fulminant liver failure in association with the targeting: a real-time PCR assay targeting
emetic toxin of Bacillus cereus. N Engl J Med 16S rDNA for direct quantification of
336:1142–1148 Alicyclobacillus spp. Spores after aptamer-
10. Dufrenne J, Bijwaard M, Te Giffel M, Beumer based enrichment. J Agric Food Chem
R, Notermans S (1995) Characteristics of some 63:4291–4296
 Chapter 7

Detection of Yersinia pestis in Complex Matrices by Intact

Cell Immunocapture and Targeted Mass Spectrometry
Jérôme Chenau, François Fenaille, Stéphanie Simon, Sofia Filali,
Hervé Volland, Christophe Junot, Elisabeth Carniel, and François Becher

Abstract
We describe an immunoaffinity-liquid chromatography-tandem mass spectrometry (immuno-LC-MS/MS)
protocol for the direct (i.e., without prior culture), sensitive and specific detection of Yersinia pestis in
complex matrices. Immunoaffinity enables isolation and concentration of intact bacterial cells from food
and environmental samples. After protein extraction and digestion, suitable proteotypic peptides corre-
sponding to three Y. pestis-specific protein markers (murine toxine, plasminogen activator and pesticin) are
monitored by targeted LC-MS/MS using the selected reaction monitoring (SRM) mode. This immuno-
­LC–MS/MS assay has a limit of detection of 2 × 104 CFU/mL in milk or tap water, and 4.5 × 105 CFU
in 10 mg of soil.

Key words Bacteria, Mass spectrometry, Immunoaffinity, Targeted proteomic, Detection, Matrices

1 Introduction

Yersinia pestis is the causative agent of bubonic and pneumonic

plague, an acute and often fatal disease. Due to its prevalence and
high lethality, Y. pestis is classified as a category A biothreat agent
by the Centers for Disease Control and Prevention (CDC), the
highest rank of potential bioterrorism agents (http://www.bt.cdc.
gov/agent/agentlist-category.asp). Methods for rapid and sensi-
tive detection of the presence of the bacteria in potentially con-
taminated matrices are essential. The high genetic similarity of the
populations composing the Yersinia pseudotuberculosis complex
e.g., 97% nucleotide identity reported between Y. pestis and Y.
pseudotuberculosis [1], has to be taken into consideration for a spe-
cific detection of Y. pestis. The two plasmids pFra (or pMT1) and
pPla (or pPst or pPCP1), which code for virulence factors, are
unique to Y. pestis and thereby represents interesting targets for
specific detection.

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_7, © Springer Science+Business Media LLC 2017

69
70 Jérôme Chenau et al.

Nucleic acid-based assays and immunometric tests have been

developed for Y. pestis detection by targeting respectively genes
and proteins coded by pFra and pPla plasmids [2, 3]. Mass spec-
trometry (MS) offers a complementary approach for biodefense
applications and detection of pathogenic microorganisms [4].
Beside the global bacterial protein fingerprints acquired by matrix-
assisted laser desorption ionization time-of-flight MS (MALDI-
TOF MS), targeted proteomics based on the selected reaction
monitoring mode (SRM) [5] was proposed for direct bacteria
detection. Selective SRM monitoring of bacteria markers was com-
bined with affinity concentration of target cells for enhanced sensi-
tivity and direct detection of bacteria in complex matrices without
any time consuming bacterial culture [6, 7].
We describe here an immunoaffinity-LC–SRM protocol for
the direct detection of Y. pestis in complex environmental and food
samples. It combines the specificity and sensitivity of two comple-
mentary methods, immunoaffinity capture and targeted mass
spectrometry detection of specific Y. pestis markers coded by pPla
and pFra plasmids. Critical parameters for efficient immunocapture
of intact Y. pestis cell and multiplex SRM detection of proteotypic
peptides are emphasized.

2 Materials

Prepare all solutions using ultrapure water (H2O mQ, resistivity

of 18 mΩ cm at 25 °C) and analytical grade reagents. Ultrapure
water used here was from a Milli-Qplus 185 purifier (Millipore,
Bedford, MA).

2.1 Samples 1. Bacterial strains used in this study originated from the collection
of Unité de Recherche Yersinia at the Institut Pasteur (Paris,
France) (see Note 1).
2. Environmental and food samples used in this work: tap water,
commercial milk powder (prepared at 5% v/w in H2O mQ),
soil (sampled in a field around the laboratory).

2.2 Immuno- 1. Magnetic beads: Dynabeads M-280 tosyl-activated (Life

purification Technologies, Saint-Aubin, France), stored at 4 °C.
2. Magnetic particle concentrator for microcentrifuge tubes,
Dynal MPC-S (Life Technologies, Saint-Aubin, France).
3. Coupling buffer: 0.1 M sodium-phosphate buffer, pH 7.4.
Add 2.62 g of NaH2PO4 monohydrate (monobasic) and
14.42 g of Na2HPO4 dihydrate (dibasic) to H2O mQ to make
a volume of 1 L. The pH of the final solution will be 7.4. This
buffer can be stored for up to 1 month at 4 °C.
 Targeted Mass Spectrometry Detection of Yersinia pestis 71

4. Monoclonal antibody Pla35, stock solution at 1.2 mg/mL

(see Note 2). Dilute the antibody to 422 μg/mL in coupling
buffer.
5. Blocking buffer: phosphate-buffered saline (PBS), pH 7.4 with
0.5% (w/v) bovine serum albumin (BSA). Add 0.5 g of BSA
(see Note 3) to 100 mL of PBS (DPBS without Ca2+ and Mg2+,
Lonza, Verviers, Belgium). This buffer can be stored for up to
1 month at 4 °C.
6. Washing and storage buffer: PBS, pH 7.4 with 0.1% (w/v)
BSA. Dilute Blocking buffer fivefold in PBS. For example, add
10 mL of blocking buffer to 40 mL of PBS, pH 7.4. This buffer
can be stored for up to 1 month at 4 °C.
7. Thermomixer compact for 1.5 mL microtubes (Eppendorf,
Montesson, France).
8. PBS with 0.1% Tween-20: for 100 mL, add 0.5 mL of 20%
Tween-20 (v/v) (see Note 4) to 99.5 mL of PBS. This buffer
can be stored for up to 1 month at 4 °C.
9. PBS with 0.1% Tween-20 and 1 mg/mL BSA. Add 100 mg of
BSA to 100 mL of PBS with 0.1% Tween-20. This buffer can
be stored for up to 1 month at 4 °C.
10. Rotating wheel.
11. 1.5 mL microtubes LoBind (Eppendorf, Montesson, France).

2.3 Y. pestis 1. Prepare a 80% trifluoroacetic acid (TFA) solution. For 10 mL,
Inactivation and add 8 mL of TFA (≥99% of purity, Sigma-Aldrich, St. Louis,
Protein Extraction MO) to 2 mL of H2O mQ.
2. Vortex mixer.
3. Laboratory centrifuge for microtubes enabling centrifugation
up to 14,000 × g (ThermoFisher, Villebon sur Yvette, France).
4. Filter tubes of 0.22 μm pore size (Ultrafree MC filters,
Millipore, Billerica, MA) (see Note 5).

2.4 Trypsin Digestion 1. Centrifugal vacuum concentrator (ThermoFisher, Villebon sur

Yvette, France).
2. pH indicator strips.
3. Ammonium bicarbonate solution, 1 M, pH 8.0. Dissolve 8 g
of ammonium bicarbonate in 100 mL of H2O mQ. This buffer
can be stored for up to 1 month at 4 °C.
4. Ammonium bicarbonate solution, 50 mM, pH 8.0 (AB50).
Dilute 20-fold 1 M ammonium bicarbonate solution. For
example, add 0.5 mL of 1 M ammonium bicarbonate to 9.5 mL
of H2O mQ.
5. Dithiothreitol (DTT) solution, 45 mM. Add 6.9 mg of DTT
to 1 mL of AB50. Make fresh before use.
72 Jérôme Chenau et al.

6. Iodoacetamide (IAA) solution, 100 mM. Add 18.5 mg of IAA

to 1 mL of AB50. Make fresh before use (see Note 6).
7. Trypsin solution (1 μg/μL). Reconstitute 20 μg of lyophilized
trypsin (Sequencing Grade Modified Trypsin, Promega Cat#
V5111, Fitchburg, WI) in 20 μL of re-suspension buffer
(50 mM acetic acid) (see Note 7). Keep on ice before use.
8. Thermomixer for 1.5 mL microtubes (Eppendorf, Montesson,
France).

2.5 Internal 1. HPLC-grade acetonitrile (ACN) from SDS (Peypin, France).

Standard Peptides 2. Analytical-grade formic acid (FA).
3. Internal standards (IS): seven stable isotope-labeled peptides
(synthesized by Bachem, Burgdorf, Switzerland) corresponding
to the four targeted proteins, purity ≥95% (determined by
HPLC), supplied as trifluoroacetate salt (see Table 1). Re-­suspend
each peptide in 30% ACN, 0.1% formic acid to have a stock
solution at 4 mg/mL (see Note 8).
4. IS stock mixture (4×): Mix the seven peptides in 20% ACN,
0.4% formic acid at the following final concentrations: peptide
1 of Pla (Pla 1) at 400 ng/mL, peptide 2 of Pla (Pla 2) at
2000 ng/mL, peptide 1 of murine toxin (Ymt 1) at 40 ng/mL,
and the four other peptides (Ymt 2, Psn 1, Ompa 1, Ompa 2)
at 200 ng/mL (see Note 9).

2.6 LC and MS 1. HPLC polypropylen micro-vials (Supelco, France).

Conditions 2. LC system: HP 1100 HPLC system (Agilent Technology, Les
Ulis, France) (see Note 10).
3. LC column: Zorbax SB-C18 column (150 × 2.1 mm i.d., 5 μm
particle size, 300 Å porosity) from Agilent Technology.
4. Use only analytical-grade reagents to constitute mobile phases.
HPLC-grade acetonitrile was from SDS, and analytical-grade
formic acid.
5. Mobile phase A: H2O mQ containing 0.1% (v/v) formic acid.
6. Mobile phase B: Acetonitrile containing 0.1% (v/v) formic
acid.
7. Mass spectrometer: triple quadrupole TSQ Quantum Ultra mass
spectrometer (Thermo Scientific, Les Ulis, France) (see Note 11).
8. XCalibur software (version 1.4 SR1, Thermo Scientific).

3 Methods

The overall immuno-LC-MS/MS procedure for sensitive detec-

tion of Y. pestis in complex environmental and food matrices is
schematized in Fig. 1.
Table 1
Mass spectrometry settings for the multiplex SRM detection

Peptide Segment
Peptide molecular Precursor Product ion Collision Tube Retention time range
Protein name Peptide sequence mass (Da) ion m/z m/z energy (eV) lens (V) time (min) (min)
Plasminogen Pla1 NSGDSVSIGGDAAGISNK 1647.77 825.0 (z = 2) 889.6 (y101+) 27 80 3.2 2.0–3.8
activator (Pla) 1089.9 (y121+) 27 80 3.2 2.0–3.8
IS Pla1 NSGDS[13C5;15N] 1653.77 828.0 (z = 2) 889.6 (y101+) 27 80 3.2 2.0–3.8
VSIGGDAAGISNK 1089.9 (y121+) 27 80 3.2 2.0–3.8
Pla2 INDFELNALFK 1322.69 662.5 (z = 2) 705.5 (y61+) 20 70 13.8 12.0–16.0
834.5 (y71+) 20 70 13.8 12.0–16.0
IS Pla2 INDFELNA[13C6]LFK 1328.69 665.4 (z = 2) 711.6 (y61+) 20 70 13.8 12.0–16.0
840.7 (y71+) 20 70 13.8 12.0–16.0
Pesticin (Psn) Psn1 ENEDILNNNR 1229.56 615.6 (z = 2) 630.3 (y51+) 20 80 2.8 2.0–3.8
517.4 (y41+) 20 80 2.8 2.0–3.8
IS Psn1 ENEDILNNN[13C6;15N4]R 1239.56 620.6 (z = 2) 640.3 (y51+) 20 80 2.8 2.0–3.8
527.4 (y41+) 20 80 2.8 2.0–3.8
Murine Toxin Ymt1 ILIAPFFFTDK 1310.73 656.5 (z = 2) 901.5 (y71+) 18 80 17.8 16.0–19.0
(Ymt) 1085.6 (y91+) 18 80 17.8 16.0–19.0
972.5 (y81+) 18 80 17.8 16.0–19.0
IS Ymt1 ILIAPFFFTD[13C6;15N2]K 1318.73 660.5 (z = 2) 909.5 (y71+) 18 80 17.8 16.0–19.0
1093.6 (y91+) 18 80 17.8 16.0–19.0
980.5 (y81+) 18 80 17.8 16.0–19.0
Ymt2 EIMQQSYLR 1166.58 584.1 (z = 2) 666.4 (y51+) 20 70 4.4 3.8–6.0
538.3 (y41+) 20 70 4.4 3.8–6.0
925.5 (y71+) 20 70 4.4 3.8–6.0
IS Ymt2 EIMQQSYL[13C6;15N4]R 1176.58 589.1 (z = 2) 676.4 (y51+) 20 70 4.4 3.8–6.0
548.3 (y41+) 20 70 4.4 3.8–6.0
935.5 (y71+) 20 70 4.4 3.8–6.0
Targeted Mass Spectrometry Detection of Yersinia pestis

(continued)
73
Table 1
74

(continued)

Peptide Segment
Peptide molecular Precursor Product ion Collision Tube Retention time range
Protein name Peptide sequence mass (Da) ion m/z m/z energy (eV) lens (V) time (min) (min)
Outer Ompa1 LSYPVAQDLDVYTR 1638.83 820.4 (z = 2) 638.8 (y112+) 25 80 3.0 2.0–3.8
membrane 1080.6 (y91+) 25 80 3.0 2.0–3.8
protein A 720.1 (y122+) 15 60 3.0 2.0–3.8
(OmpA) IS Ompa1 LSYPVAQDLDVYT 1648.83 825.4 (z = 2) 643.8 (y112+) 25 80 3.0 2.0–3.8
Jérôme Chenau et al.

[13C6;15N4]R 725.1 (y91+) 25 80 3.0 2.0–3.8

1090.6 (y122+) 25 80 3.0 2.0–3.8
Ompa2 GSFDGGLDR 922.41 462.2 (z = 2) 517.3 (y51+) 18 60 10.0 6.0–12.0
632.3 (y61+) 18 60 10.0 6.0–12.0
IS Ompa2 GSFDGGLD[13C6;15N4]R 932.41 467.2 (z = 2) 527.3 (y51+) 18 60 10.0 2.0–3.8
642.3 (y61+) 18 60 10.0 6.0–12.0
Modified with permission from [7]. Copyright 2016 American Chemical Society
 Targeted Mass Spectrometry Detection of Yersinia pestis 75

Fig. 1 General workflow for Y. pestis detection. Modified with permission from [7]. Copyright 2016 American
Chemical Society

The first step corresponds to the capture of Y. pestis intact cells

using anti-Pla antibodies immobilized on magnetic beads. Proteins
are then extracted from the immunocaptured bacteria with the
80% TFA protocol [8], which allows sample sterilization by total
inactivation of bacteria. After acid neutralization, protein extracts
are digested with trypsin, and the resulting proteotypic peptides
corresponding to three Y. pestis specific targets (Pla, Ymt and Psn)
are simultaneously monitored by SRM mass spectrometry for high
robustness. These makers, coded by pPla and pFra plasmids, are
unique to Y. pestis, and were selected by bottom-up proteomics
(see Note 12).

3.1 Preparation Preparation of 500 μL Pla35 antibody-coated magnetic beads

of Pla35 Antibody- (see Note 13):
Coated Magnetic
1. Re-suspend M-280 tosyl-activated magnetic beads by repeated
Beads gently pipetting until a homogeneous suspension is obtained
and transfer 350 μL of beads suspension to a new tube.
2. Place the tube on the magnet, allow the beads to pellet com-
pletely (liquid will turn clear) and remove supernatant.
3. Remove the tube from the magnet and wash the beads by adding
1 mL of Coupling Buffer. Mix by gently pipetting.
4. Place the tube on the magnet, allow the beads to pellet com-
pletely and remove supernatant.
76 Jérôme Chenau et al.

5. Remove the tube from the magnet and add 520 μL of mono-
clonal antibody Pla35 at 422 μg/mL in Coupling Buffer
(220 μg). Mix by gently pipetting and incubate at 37 °C over-
night (12–18 h) using a thermomixer with shaking at
1200 rpm.
6. Place the tube on the magnet, allow the beads to pellet com-
pletely and remove supernatant.
7. Remove the tube from the magnet and add 1 mL of Blocking
Buffer. Mix by gently pipetting and incubate at 37 °C for 1 h
using a thermomixer with shaking at 1200 rpm (see Note 14).
8. Place the tube on the magnet, allow the beads to pellet com-
pletely and remove supernatant.
9. Remove the tube from the magnet and add 1 mL of Wash
Buffer. Mix by gently pipetting.
10. Place the tube on the magnet, allow the beads to pellet com-
pletely and remove supernatant.
11. Repeat steps 9–11.
12. Re-suspend the beads in 500 μL of Storage Buffer. These anti-
body-coated beads can be stored at 4 °C for up to 1 week.

3.2 Y. pestis Intact Y. pestis immunocapture is illustrated here in tap water sample.
Cell Immunocapture The protocol is adaptable to other complex samples: milk and soil
from Complex (see Note 15).
Samples 1. Safety consideration: Y. pestis is a highly virulent species. All
experiments performed with viable bacteria have to be perform
in Biosafety Level 3 laboratory.
2. Withdraw 10 mL of artificially contaminated tap water (or suspect
tap water sample) into a 50 mL tube.
3. Add 10 mL of PBS containing 0.1% Tween-20 and 1 mg/mL
BSA (ratio 1:1, v/v) (see Note 16).
4. Add 50 μL of IgG-coupled bead solution (prepared at step in
Subheading 3.1) (see Note 17) and incubate for 90 min at
room temperature with gentle rotation on a slowly rotating
wheel (see Note 16).
5. Place the tube on the magnet, allow the beads to pellet com-
pletely and remove supernatant.
6. Remove the tube from the magnet and add 1 mL of PBS buffer
to wash the beads-IgG-bacteria complex and remove weak non-
specific binding. Mix by gently pipetting and transfer the solu-
tion into a 1.5 mL LoBind tube (Eppendorf) (see Note 18).
7. Place the tube on the magnet, allow the beads to pellet com-
pletely and remove supernatant.
8. Repeat steps 4–6 once.
 Targeted Mass Spectrometry Detection of Yersinia pestis 77

3.3 Y. pestis Y. pestis inactivation and protein extraction is performed directly

Inactivation and on immunocaptured bacteria by using the TFA protocol for highly
Protein Extraction pathogenic microorganisms described by Lasch et al. [8].
1. After the last wash of beads-IgG-bacteria complex, place the
tube on the magnet, allow the beads to pellet completely and
remove supernatant.
2. Add 150 μL of 80% TFA solution and mix by strong vortexing
for 10 min.
3. Place the tube on the magnet, allow the beads to pellet com-
pletely and transfer supernatant (containing extracted proteins)
into a new 1.5 mL LoBind tube.
4. Centrifuge at 14,000 × g and 4 °C for 15 min to pellet cellular
debris.
5. Transfer the supernatant to Ultrafree MC filter tubes of
0.22 μm pore size and spun at 10,000 × g for 5 min.
6. Transfer the filtrate containing the total protein extract into a
new 1.5 mL LoBind tube. Protein extract is stored at −20 °C
until use.
7. The total inactivation and elimination of bacteria was checked
by verification of the absence of colonies after culture on LBH
agar as previously described [7, 8] (see Note 19).
8. Extracted samples were further processed outside the Biosafety
Level 3 laboratory.

3.4 Protein Digestion 1. Dry protein extract by vacuum centrifugation to eliminate TFA.
by Trypsin 2. After complete TFA elimination, re-suspend extracted proteins
in 25 μL of 50 mM ammonium bicarbonate. Check pH using
pH indicator strips. The pH of the solution should be ~8.0
(see Note 20).
3. Add 5 μL of reducing agent (45 mM DTT) and incubate at
60 °C for 30 min using a thermomixer with shaking at 900 rpm.
4. Allow sample to cool at room temperature and spin briefly to
collect condensation.
5. Add 5 μL of alkylating agent (100 mM IAA) and incubate at
room temperature for 45 min in the dark (see Note 21).
6. Add 1.5 μL of 1 μg/μL trypsin solution and incubate at 37 °C
overnight (12–18 h) using a thermomixer with shaking at
900 rpm.
7. Allow sample to cool at room temperature and spin briefly to
collect condensation.
8. Transfer 30 μL of this trypsin digest in a HPLC vial containing
10 μL of IS stock mixture 4× (1× final). This sample mixture is
used for LC-MS/MS analysis.
78 Jérôme Chenau et al.

3.5 Multiplex LC-MS conditions used in the original study are described below.
SRM Assay
1. Thirty microliters of IS spiked sample were injected for
LC-MS/MS analysis.
2. Data were acquired using the XCalibur software.
3. LC conditions:
(a) HP 1100 HPLC system
(b) Column: Zorbax SB-C18 maintained at 60 °C.
(c) Flow rate: 200 μL/min.
(d) HPLC Gradient (total run time: 30 min): linear gradient
from 15 to 40% B in 20 min, then a column wash at 95%
B for 4 min and equilibration at 15% B for 6 min.
(e) Column effluent was directly introduced into the electrospray
source of the MS system.
4. MS conditions:
(a) Triple quadrupole TSQ Quantum Ultra MS operated in
the positive ion mode.
(b) Electrospray voltage at 3.9 kV and capillary voltage at 35 V.
(c) Sheat, ion sweep and auxiliary gas flow rates (nitrogen)
optimized at 40, 20 and 15 (arbitrary units) respectively.
(d) Drying gas temperature: 350 °C.
(e) Proteotypic peptide detection in the SRM mode.
5. SRM method:
(a) Common parameters: 0.100 m/z scan width, 0.1 s scan
time, 0.4 Q1, 0.7 Q3, 1.5 mTorr Q2 pressure (using argon
as the collision gas).
(b) SRM transitions: see Table 1 for the SRM transitions
designed and monitored for each proteotypic peptide and
the optimized parameters (tube lens and collision energy
values). (see Note 22)
(c) SRM acquisition time was divided into five segments
(see Table 1)
(d) A typical LC-SRM chromatogram of the monitored
peptides is given in Fig. 2, including proteotypic and IS
peptides.
(e) Data interpretation was performed using the XCalibur
software (see Note 23).
Any modification to the protocol (LC-MS instrument, cap-
ture antibody, sample preparation conditions) could have an
impact on the performance. Specification of the immuno-LC-
MS/MS (SRM) assay in our conditions is described in Subheading 4
(see Note 24).
 Targeted Mass Spectrometry Detection of Yersinia pestis 79

Fig. 2 Typical LC–SRM chromatograms obtained for (a) Y. pestis CO92 and (b) Y. pseudotuberculosis IP32953.
Depicted SRM transitions are as follows: Pla (peptide 1 NSGDSVSIGGDAAGISNK, 825 → 1089; peptide 2 INDFELNALFK,
662.5 → 834.5), Pesticin or Psn (ENEDLNNNR, 615.6 → 630.3), Plasminogen activator or Ymt (peptide 1 ILIAPFFFTDK,
656.5 → 972.5; peptide 2 EIMQQSYLR, 584.1 → 666.4), and OmpA (peptide 1 LSYPVAQDLDVYTR, 820.4 → 638.8;
peptide 2 GSFDGGLDR, 462.2 → 517.3). In each case, the signal of the corresponding isotopically labeled peptide is
indicated by IS (internal standard). Data were obtained using 108 bacteria. Reprinted with permission from [7].
Copyright 2016 American Chemical Society

4 Notes

1. Y. pestis is a highly virulent species. Strains were handled

according to safety considerations applicable at the Institut
Pasteur (Biosafety Level 3 laboratory).
2. The monoclonal antibody Pla35 used in our method was pro-
duced in mouse as previously described [3]. This antibody is
directed against plasminogen activator (Pla) protein. This pro-
tein is a suitable target for intact cell immunocapture, as it is
specific for Y. pestis (coded by the Y. pestis plasmid pPla) and is
located at the bacterial surface. Other monoclonal antibody
could be used but the impact on capture yield and specificity
should be evaluated.
3. For all buffers containing BSA, use purified BSA Fraction V
from Sigma-Aldrich, ≥99% purity.
4. Tween-20 is a highly viscous solution. Prepare a solution of
20% Tween-20 (v/v) in H2O mQ. Hint: to prepare 1 mL of
solution, cut off the extremity of pipet tips before attempting
to pipet the 200 μL. Pipet Tween-20 very slowly and disperse
several times into 800 μL of H2O mQ.
5. Ultrafree filter tubes contain a Durapore PVDF (polyvinyli-
dene fluoride) membrane which exhibits a low protein binding
80 Jérôme Chenau et al.

capacity. The filter material was demonstrated to be resistant to

highly concentrated TFA [8].
6. Iodoacetamide is light sensitive. Always prepare a fresh solution
and protect it from light.
7. Trypsin solution (in 50 mM acetic acid) could be aliquoted
and stored for up to 1 month at −80 °C.
8. Stock solutions at 4 mg/mL were prepared by dissolving 4 mg
of powder peptide in 1 mL of 30% acetonitrile, 0.1% formic
acid. Stock solutions were aliquoted and stored at −20 °C.
9. A simple method to obtain IS stock mixture (4×) consists in
performing an intermediate dilution of each peptide at 40 μg/mL
(100-fold dilution). For 1 mL, add 10 μL of stock peptide at
4 mg/mL to 990 μL of 20% acetonitrile, 0.4% formic acid.
Then use theses diluted solutions to obtain IS stock mixture
(4×) as following: add 100 μL of Pla 1, 500 μL of Pla 2, 10 μL
Ymt 1, 50 μL Ymt 2, 50 μL Psn 1, 50 μL OmpA 1, 50 μL
OmpA 2, qsp 10 mL of 20% acetonitrile, 0.4% formic acid
(9.24 mL). IS stock mixture (4×) could be aliquoted and
stored at −20 °C until use.
10. Another LC system could be used. The critical parameters are:
flow rate at 200 μL/min, mobile phase gradient, and column
compartment at 60 °C.
11. Another triple quadrupole mass spectrometer can be used, but
assay sensitivity will be dependent on the equipment. Tuning
parameters defined for each monitoring transition should be
adapted.
12. In a preliminary work, a bottom-up proteomics approach was
performed and highlighted three relevant protein markers
encoded by the Y. pestis-specific plasmids pFra (murine toxin)
and pPla (plasminogen activator and pesticin). Suitable proteo-
typic peptides were thoroughly selected to monitor the three
protein markers by targeted MS using the SRM mode. In addi-
tion, two peptides from outer membrane protein A (OmpA)
were selected to evaluate the nonspecific capture of the closely
related Y. pseudotuberculosis and Y. similis species. The reader
could refer to Chenau et al. [7] where all details on identifica-
tion and characterization of the selected markers are given.
13. The use of Dynabeads M-280 tosyl-activated (Life Technologies)
is critical to obtain optimal capture conditions. Other magnetic
beads with different diameters (300 nm and 2.8 μm) and
chemistries for antibody coating (covalent binding with tosyl-
activated beads or by protein G affinity) as well as different
homemade monoclonal antibodies [3] directed against the
specific membrane protein Pla were tested. Optimal recovery
was achieved using Dynabeads M-280 tosyl-activated (2.8 μm
diameter) coated with Pla35 antibody [7].
 Targeted Mass Spectrometry Detection of Yersinia pestis 81

14. Blocking of free tosyl-groups is a critical step to avoid non-­specific

capture during the immunocapture of intact Y. pestis cells.
15. Application to other environmental samples was demonstrated
[7]. For milk samples, the same protocol could be used by
replacing the 10 mL of tap water by 10 mL of milk containing
bacteria. For soil analysis, 10 mg of soil sample was directly
mixed with 10 mL of PBS with 0.1% Tween-20 and 1 mg/mL
BSA.
16. Dilution of bacterial samples in PBS buffer containing 0.1%
Tween-20 and 1 mg/mL BSA and incubation for 90 min at
room temperature with gentle shaking on a wheel were
required for a high and specific capture yield, weak nonspecific
binding on the magnetic bead surface or wall tubes, and full
integrity and low autoaggregation of Y. pestis cells. Under
these conditions recovery was around 77% at 5 × 107 CFU/mL,
with CV of 6.8%.
17. Assay conditions are given for a sample volume of 10 mL. Assay
performances were determined in these conditions. If needed,
the volume of IgG-coupled bead solution could be adapted to
lower sample volume. For example, use 35 μL of IgG-coupled
beads solution if detection is performed in 1 mL only of
sample.
18. The use of 1.5 mL microtube LoBind (Eppendorf) allows to
minimize protein loss on tube wall during the protein extrac-
tion step.
19. Y. pestis is a highly virulent species. In our initial study [7], the
total inactivation and elimination of bacteria were systemati-
cally checked after neutralization of TFA with 9% Tween-80,
0.9% lecithin, 3% histidine in TSB and verification of the
absence of viable bacteria in each extract after culture on
LBH agar for 3 days, as described previously [8]. In Lasch
et al. [8], it was demonstrated that Bacillus spores (highly
resistant structure) were reliably sterilized in each of the 67
separate inactivation tests performed with the TFA inactiva-
tion protocol. They also demonstrated, for vegetative cells of
the genera Bacillus, Burkholderia, Escherichia, and Yersinia
(430 separate inactivation tests), a total bacteria inactivation
after the first centrifugation, without the filtration step. We
have confirmed the full efficiency of the TFA protocol for
Bacillus spores inactivation on 55 strains of the B. cereus
group (B. anthracis, B. cereus and B. thuringiensis) [9] and
bacterial inactivation on 17 strains of the Y. pseudotuberculosis
complex [7].
20. The main precaution is to ensure that the pH is around 8.0 for
maximum proteolysis efficiency. If pH is too low, add 1 μL of
1 M ammonium bicarbonate.
82 Jérôme Chenau et al.

21. Aluminum foil can be used to cover the sample.

22. Two proteotypic tryptic peptides were selected (except for Psn)
to unambiguously monitor each protein marker. Peptide
selection was based on the recommendations for targeted
proteomic experiments [5], regarding sequence uniqueness
and amino acid composition.
At least two SRM transitions were monitored for each peptide to
improve detection specificity. Product ions were preferentially
selected with an m/z ratio higher than that of the parent ion.
A signal was considered positive when peak intensity was three
times above the noise level for both selected transitions.
Additionally, two criteria should be met for a given peptide:
(1) the variation of the retention time should be within 0.1 min
in replicate injections with strict co-elution with the IS peptide
and similar peak shape; (2) the relative ratio (i.e., for a given
transition by dividing its peak area by the peak area of another
transition) of the analyte compared with the relative ratio of
the IS peptides should not be significantly different (±30%).
23. At least two peptides were detected at the limit of detection.
The monitoring of the most intense Pla (peptide Pla2) and
Ymt (peptide Ymt1) peptides was sufficient to allow a robust,
sensitive, and specific detection of Y. pestis.
24. Detection specificity for Y. pestis was demonstrated with 17
different strains. Seven Y. pestis strains from the three most
common biovars were used as control for positive detection,
and ten other strains from the Y. pseudotuberculosis complex
were analyzed as negative samples. The first level of specificity
is achieved by the specific antibody capture: immunocapture
yield were determined at 77% for Y. pestis (mean for seven
strains, range: 55–100%), at 1.8% for closely related Y. pseudo-
tuberculosis (mean for eight strains, range: 0.02–7.4%) and
0.07% for closely related Y. similis (mean for two strains,
range: 0.04–0.09%). In this experiment, OmpA proteotypic
peptides were monitored to evaluate the efficiency of the
Pla-dependent immunocapture of Y. pestis and its specificity
toward the Y. pseudotuberculosis and Y. similis species. The
second level of specificity is provided by the SRM monitoring
of unique Y. pestis makers. Positive signals corresponding to
the three markers were observed for Y. pestis. No signal was
detected for any of these markers in Y. pseudotuberculosis or
Y. similis strains (Fig. 2).
Assay sensitivity was evaluated in tap water, milk and soil
samples. Limits of detection were as follow: tap water:
2 × 104 CFU/mL, milk: 2 × 104 CFU/mL, soil: 4.5 × 105 CFU
in 10 mg, with good linearity (r2 > 0.98) over a 4-log concen-
tration range in tap water and milk (2 × 104 to 2 × 108 CFU/mL).
 Targeted Mass Spectrometry Detection of Yersinia pestis 83

The coefficient of variation (CV) was determined at 6.8%

(n = 9) with 5 × 107 CFU/mL of CO92 strain in the PBS-BSA
dilution buffer. Detection of Y. pestis was observed both for
strains grown at 28 °C (optimal growth temperature) and
37 °C (temperature during infection).

References
1. Chain PS, Carniel E, Larimer FW, Lamerdin J, for the selectivity and accuracy of measure-
Stoutland PO, Regala WM, Georgescu AM, ments. Proteomics Clin Appl 6:609–614
Vergez LM, Land ML, Motin VL, Brubaker 6. Chenau J, Fenaille F, Ezan E, Morel N,
RR, Fowler J, Hinnebusch J, Marceau M, Lamourette P, Goossens PL, Becher F (2011)
Medigue C, Simonet M, Chenal-Francisque Sensitive detection of Bacillus anthracis spores
V, Souza B, Dacheux D, Elliott JM, Derbise by immunocapture and liquid chromatography-­
A, Hauser LJ, Garcia E (2004) Insights into tandem mass spectrometry. Anal Chem 83:
the evolution of Yersinia pestis hrough whole-­ 8675–8682
genome comparison with Yersinia pseudotu- 7. Chenau J, Fenaille F, Simon S, Filali S, Volland
berculosis. Proc Natl Acad Sci U S A H, Junot C, Carniel E, Becher F (2014)
101:13826–13831 Detection of Yersinia pestis in environmental
2. Kenny JH, Zhou Y, Schriefer ME, Bearden and food samples by intact cell immunocapture
SWJ (2008) Detection of viable Yersinia pestis and liquid chromatography-tandem mass
by fluorescence in situ hybridization using pep- spectrometry. Anal Chem 86:6144–6152
tide nucleic acid probes. J Microbiol Methods 8. Lasch P, Nattermann H, Erhard M, Stämmler
75:293–301 M, Grunow R, Bannert N, Appel B, Naumann
3. Simon S, Demeure C, Lamourette P, Filali S, D (2008) MALDI-TOF mass spectrometry
Plaisance M, Creminon C, Volland H, Carniel compatible inactivation method for highly
E (2013) Fast and simple detection of Yersinia pathogenic microbial cells and spores. Anal
pestis applicable to field investigation of plague Chem 80:2026–2034
foci. PLoS One 8:e54947 9. Chenau J, Fenaille F, Caro V, Haustant M,
4. Duriez E, Armengaud J, Fenaille F, Ezan E Diancourt L, Klee SR, Junot C, Ezan E,
(2016) Mass spectrometry for the detection of Goossens PL, Becher F (2014) Identification
bioterrorism agents: from environmental to clini- and validation of specific markers of Bacillus
cal applications. J Mass Spectrom 51:183–199 anthracis spores by proteomics and genomics
5. Domon B (2012) Considerations on selected approaches. Mol Cell Proteomics 13:
reaction monitoring experiments: implications 716–732
 Chapter 8

A Method to Prepare Magnetic Nanosilicate Platelets

for Effective Removal of Microcystis aeruginosa
and Microcystin-LR
Shu-Chi Chang, Bo-Li Lu, Jiang-Jen Lin, Yen-Hsien Li,
and Maw-Rong Lee

Abstract
Algal toxin is a unique type of toxin generated with harmful algal blooms in water bodies. This phenom-
enon is worsened by eutrophication caused by excessive discharge of nutrients into surface water bodies.
Since algal toxins are hard to remove after they enter the water treatment processes, an efficient method is
required to inhibit the growth of algal cells, to settle the cells at the bottom of the water body and to
removes the toxin from the water. We report an efficient way to prepare a novel nanohybrid material, i.e.,
magnetic nanosilicate platelet (MNSP), and its effects on the removal of microcystin toxins as well as the
cells of Microcystis aeruginosa. MNSP was fabricated by a special treatment of a clay mineral, montmoril-
lonite, and then its surface was decorated with magnetite nanoparticles by in situ synthesis. The nanohy-
brid can efficiently inhibit the growth of M. aeruginosa—a typical species that can generate one of the most
notorious algal toxins, i.e., microcystins. Algal cells can be settled with minimal 500 ppm MNSP, and the
turbidity can be reduced by more than 67%. The removal of microcystin-LR (MC-LR) was as high as
99.39% at an concentration of 100 ppm, while the pristine nanosilicate platelet could only remove 36.84%
at the same dosage.

Key words Magnetic nanosilicate platelet, Harmful algal bloom, Microcystis aeruginosa,
Microcystin-LR, Eutrophication

1 Introduction

On the earth, about 97.5% of the water is sea water. Only about
0.024% of the earth’s water is available for human usage [1]. Due
to global climate change and development of the modern human
society, the ever-increasing water demand makes water a more and
more scarce natural resource. According to United Nations, by
2050, more than one third of the world’s projected population will
live in river basins with severe water stress [2]. In addition, major
water sources are facing enormous challenges due to algal blooms,
such as Lake Taihu in China [3] or Lake Erie in the USA [4].

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_8, © Springer Science+Business Media LLC 2017

85
86 Shu-Chi Chang et al.

Algal growth in water bodies plays a positive role in nutrient cycling

and fishery operation [5]. However, under special conditions, such
as high nutrient inputs, low predation, and sufficient sunlight, their
rapid proliferation can lead to severe consequences, like water dis-
coloration, foul odor generation, and health impact on animals
(including humans). Harmful algal blooms (HABs) occur when
colonies of algae grow excessively while producing toxins or causing
harmful effects on other organisms and the environment [6]. In the
past, HABs have caused animal and human deaths and drinking
water shortage problems [3, 7–9]. Among all these reported epi-
sodes, microcystins from Microcystin aeruginosa attracted much
attention and have been most frequently studied [10].
Microcystins are generated by M. aeruginosa and can have
more than 70 different types with similar molecular structures
(Fig. 1) [11]. Among them, MC-LR is probably the most toxic
one. In the past few decades, several methods have been proposed
to remove microcystins in drinking water treatment processes,
such as activated carbon, reverse osmosis, chlorination, ozonation,
advanced oxidation, permanganate, hydrogen peroxide, photoly-
sis, semiconductor, and photocatalysis [10, 12]. However, these
methods either cause undesirable side effects or entail higher cost
and treatment complexity. Here, we report a more environmental-
friendly way to achieve high removal efficiency on both M. aerugi-
nosa and its toxin, MC-LR.
In the past, clay has been applied to remove Microcystis spp.
cells and toxins with moderate success. To remove the algal cells,
within 6 h, clay particles at 750 mg/L can settle about 5000 cells
with an initial concentration at ~70,000 cells/mL equivalent to a
7.1% removal [13]. To remove the MC-LR, higher than 833 mg/L
of clay particles achieved 81% removal [14]. A most recent study
showed that clay modified by hexadecyl trimethyl ammonium
­bromide (CTAB) can deactivate 92% of tested M. aeruginosa [15].
We have reported that nanosilicate platelet (NSP) can remove

Glu Mdha

H CO2H CH3 O
N
H-N NH
O CH2 Ala
Adda CH3 H
O
H CH3
H OCH3 H H O

H NH H-N
H CH3 CH3 H H H CH3 H
X
-

N N
O
Y
O H CO2H O
β-Me-Asp

Fig. 1 Structural formula of microcystins [11]

 A Method to Prepare Magnetic Nanosilicate Platelets for Effective Removal… 87

MC-LR as high as 99% at a dosage of 500 ppm [16]. However, up

to now, none of the studies has been performed on magnetic iron
oxide nanoparticle-decorated nanosilicate platelet (MNSP) for
HAB control regarding algal growth inhibition, settling enhance-
ment, and toxin adsorption.
There are several advantages to development of this method by
using such a nanohybrid material. First, most eutrophic water bod-
ies are quite shallow, and a retrievable material (through magnetic
attraction) can minimize the impact of particle deposit at the bot-
tom. Second, MC-LR adsorbed on MNSP can be recovered physi-
cally upon magnetic separation since the uptake of MNSP may
result in accumulation of toxin in predating organisms. Third, iron
oxides themselves have been reported to be able to remove MC-LR
through adsorption [17, 18].

2 Materials

2.1 Culture 1. M. aeruginosa PCC7820 was kindly provided by Prof. Tsair-­

Preparation Fuh Lin at National Cheng Kung University in Taiwan. This
culture was originally obtained from Institut Pasteur (Paris,
France). Growth medium was a BG-11 medium according to
the literature [19].
2. Grow the culture in a 1-liter flask filled with BG-11 medium.
Seal the flask using a sterile silicon stopper and shake it at
100 rpm on a shaker. Keep the temperature at 25 °C, and use
a 23-W full-spectrum light bulb to provide lighting 12 h a day
(e.g., CGB Full Spectrum Lighting, color temperature at
5500 K, Chang Gung Biotechnology, Taipei, Taiwan).
3. Transfer this culture to fresh media every month. Do not per-
form growth inhibition and settling tests until the cell density
is higher than 107 cells/mL.

2.2 Nanosilicate 1. Prepare nanosilicate platelets, NSP, from the natural mineral of
Platelet Preparation sodium form of montmorillonite (MMT) purchased from
Nanocor Co. (Aberdeen, MS, USA). This clay mineral is the
sodium counter ion form of the smectite silicates with
1.20 meq/g cationic exchange capacity. Use poly(oxypropylene)-
triamine (Jeffamine T403, abbreviated as T403) with a molec-
ular mass of 440 g/mol (Huntsman Chemical Co. or Aldrich
Chemical Co.) and diglycidyl ether of bisphenol A (DGEBA,
tradename BE-188) with an epoxy equivalent mass (EEM) of
188 (Nan-Ya Chemicals, Taiwan). Perform the exfoliation or
delamination of the MMT layered structure in water according
to [20, 21], as exemplified below.
2. Firstly, prepare the exfoliating agent, polyamines. Add to a
50-mL three-necked and round-bottomed flask, equipped with
a mechanical stirrer, a thermometer, and a heating mantle,
88 Shu-Chi Chang et al.

T403 (7.15 g) and DGEBA (2.85 g). Mix these reactants by a

mechanical stirrer at room temperature. Monitor the progress of
epoxy/amine reaction by an FT-IR, showing the disappearance
of 910 cm−1 epoxide absorbance during the process. Stop the
reaction after 4 h; the final product is a viscous, transparent liquid.
After extracting out the unreacted T403, analyze the crude prod-
uct by amine titration and gel permeation chromatography. In
our experiment, the total amine content was 4.8 meq/g and gel
permeation chromatography analysis identified two major peaks.
3. Secondly, disperse the clay of sodium MMT (1.0 g) in 100 mL
of deionized water at 80 °C by overnight vigorous stirring to
form a MMT slurry. Treat the prepared polyamines (4.50 g)
separately at room temperature with a designated amount of
hydrochloric acid (35 wt%, 0.71 g) to form the amine-HCl
salts in water, and then add this to the silicate clay slurry. Stir
this clay mixture at 80 °C for 3 h, and then filter, collect, and
wash thoroughly with 40 mL of water/ethanol several times.
Dry the raw NSP material and pulverize it to powder.
4. This raw NSP, as the form of water slurry of the exfoliated
MMT, could be further modified by complexing with three
alkyl R-dimethylbenzyl ammonium chloride surfactants (R
represents 12:0, 14:0, and 16:0 alkyl groups, in a ratio of
63:30:7, respectively).
5. The final NSP represents a complex with the surfactants to
form an organic silicate platelet. Confirm the physical size of
NSP of about 100 × 100 × 1 nm in lateral dimension by atomic
force microscopy and transmission electron microscopy [20,
21]. A typical characterization using a high-resolution X-ray
diffractometer (HRXRD D8 SSS, Bruker, Billerica, MA, USA)
of NSP is shown in Fig. 2. Characteristic peaks are at 27.4,
31.8, 45.5, and 56.6 degrees on a 2θ scale.

2.3 MNSP 1. In order to synthesize magnetite nanoparticles (see Note 1),

Preparation keep the environment anaerobic. To maintain the anaerobic
synthetic conditions, perform all preparation steps for MNSP
in a glove box (Coy Lab, Grass Lake, MI, USA). Subject all
water used in this synthesis step to purging by nitrogen for at
least 30 min, and then keep it in the anaerobic chamber for at
least 1 day before use.
2. For a typical 0.5% of magnetite coverage of the surface area of
the NSP, add 1.371 g of FeCl2·4H2O, 3.731 g of FeCl3·6H2O,
10.0 g of NSP (dry mass), and 2.208 g of NaOH into 70 mL
of the water from the previous step. Seal the bottles to prevent
oxygen aeration and then mix with a magnetic mixer for 10 min.
3. Add 2.208 g of NaOH to 18.4 mL of deoxygenated water to
prepare a 3.0 N NaOH solution. A coprecipitation method
was used to prepare MNSP. Put the mixture prepared in step 2
into a sonication bath, and hold the temperature at 85 °C for a
 A Method to Prepare Magnetic Nanosilicate Platelets for Effective Removal… 89

Fig. 2 XRD characterization of final NSP

Fig. 3 XRD characterization of MNSP

sonication period of 30 min. Then, add the 3.0 N NaOH solu-

tion (see Note 2). Cool the mixture to room temperature and
transfer it into an anaerobic chamber.
4. In the anaerobic chamber, open the bottle, and wash the
product using deoxygenated water three times. Hold all
MNSP in place in the bottle by using a strong rare earth mag-
net. Thus, a clean batch of MNSP is ready for use (see Note 3).
The HR XRD characterization of MNSP is shown in Fig. 3.
90 Shu-Chi Chang et al.

Fig. 4 HRTEM characterization of MNSP (bar = 20 nm, magnification = 150,000×)

The characteristic peaks for magnetite are 30.2, 35.5, 43.2,

57.1, and 62.6 degrees on a 2θ scale. Also, an electron micro-
graph of MNSP can be taken using a high-resolution transmis-
sion electron microscope (HRTEM) at a magnification of
150,000 folds (e.g., JEM-2010, JEOL, Tokyo, Japan. Fig. 4).

3 Methods

3.1 Growth 1. Perform the growth inhibition test of M. aeruginosa

Inhibition Test (PCC7820) by diluting the cells with BG-11 medium to dif-
ferent concentrations in order to see if MNSP could efficiently
inhibit the growth of the M. aeruginosa within a high concen-
tration range often encountered in an extremely eutrophic
water body. Add MNSPs to final concentrations of 0, 10, 100,
and 500 ppm.
2. Sample all test combinations at 0, 0.5, 2, 4, 12, 24, and 48 h.
Enumerate all samples right away in a hemocytometer slide
using a Carl Zeiss microscope (Axiovert 200, Carl Zeiss AG)
and perform all tests in triplicates (see Note 4).
3. A typical result of growth inhibition test on 2.06 × 106 cells/mL
is shown in Fig. 5. The highest dosage at 500 mg/L of MNSP
resulted in the best growth inhibition of M. aeruginosa.
 A Method to Prepare Magnetic Nanosilicate Platelets for Effective Removal… 91

Fig. 5 Growth inhibition test results

Fig. 6 Settling test on M. aeruginosa using MNSP

3.2 Settling 1. Employ a jar test platform (e.g., JT-6S, Sun Great Technology
Enhancement Test Co, Ltd., Taipei, Taiwan) to test the settling-enhancing effect
of NSP. In this test, cell concentration of M. aeruginosa should
be above 1.00 × 107 cell/mL because a higher concentration
is more realistic in an algal bloom situation.
2. Add NSP to final concentrations of 0, 10, 20, 50, 100, and
500 ppm. After a 30-min mixing of cells and MNSP solutions
at 30 rpm, leave all jars undisturbed for 1 h.
3. Then, sample the supernatants at 0.5, 2, 4, 12, 24, and 48 h
to measure their turbidity using a turbidimeter (e.g., Hach
2100 N, Loveland, Colorado, USA).
4. A typical test result is presented in Fig. 6. The dosage of
10 ppm resulted in the best turbidity reduction. This is mean-
ingful because typical eutrophication indices, like Carlson’s
index, include turbidity as an indicator, in terms of Secchi
depth measurement.
92 Shu-Chi Chang et al.

3.3 Toxin 1. The most toxic microcystin, MC-LR, in powder form, was
Adsorption Test purchased from Enzo Life Sciences, Inc. (New York, USA)
with purity higher than 95%. Prepare a stock solution by add-
ing methanol to a concentration of 500 μg/L and store under
−20 °C.
2. Before the test, dilute it further by adding deionized water to
100 μg/L, and add MNSP to each sample to final concentra-
tions of 0, 10, 100, 500, and 1000 ppm. Rotate all samples at
120 rpm in a 30 °C warm room for 12 h. Take the supernatant
after 30 min of settling, and filter it through a filter with
0.2 μm pore size (Isopore™ Membrane Filters 0.2 μm GTBP,
Millipore, Billerica, Massachusetts, USA) (see Note 5).
3. Then, adjust the pH of the filtered samples to 10.5 before apply-
ing solid phase extraction (57030-U, Supelco) using an Oasis
HLB column (Waters Corporation, Milford, Massachusetts,
USA).
4. Then, elute MC-LR by methanol (containing 0.1% acetic
acid), and concentrate it by purging with high-purity nitrogen
(see Note 6).
5. A typical result of this step is shown in Fig. 7. At an initial
100 μg/L MC-LR, the removal of a 100-ppm MNSP dosage
was 99.39 ± 1.07% (mean ± standard deviation, n = 3). This
removal is significantly higher than our previously reported
removal by using NSP without magnetite coating. The
removal of 100 μg/L of MC-LR using NSP (100 ppm dos-
age) is 36.84 ± 4.80% (mean ± standard deviation, n = 3).

Fig. 7 Toxin removal by using MNSP

 A Method to Prepare Magnetic Nanosilicate Platelets for Effective Removal… 93

In addition, the 100-ppm MNSP dosage yielded higher

removal than 500-­ppm NSP. In other words, it saved about
80% of the dosage using MNSP to achieve a similar removal.

3.4 Toxin 1. Then analyze all samples using a Thermo Scientific Accela
Quantification pump liquid chromatography system coupled to a Finnigan
TSQ Ultra EMR mass spectrometer (Thermo Finnigan
Corporation) with a Phenomenex ODS column (150 × 2.1 mm,
5 μm particle size, Phenomenex International).
2. The settings are H-ESI mode with spray voltage at 4.5 kV,
heater temperature at 270 °C, capillary temperature at 270 °C,
and sheath gas flow rate at 25 arbitrary units.
3. Quantitation transition ion was set at m/z from 995.5 to 135,
and collision energy was at 60. The mobile phases were 0.1%
formic acid in water (A) and 0.1% formic acid in methanol (B).
4. Set the elution condition at 90% A and 10% B for the first
minute, and then ramp to 20% A and 80% B at 6 min, main-
tained at this ratio from 6 to 16 min.
5. At the 17th min, change the elution condition back to 90% A and
10% B, and then maintain at this ratio till the end of 25 min.

4 Notes

1. Magnetite synthesis is very sensitive to oxygen. Under aerobic

conditions, magnetite cannot be obtained. Instead, hematite
and maghemite will be formed, which are not as good as mag-
netite for magnetic recovery and phosphate removal.
2. The concentrated NaOH solution is very corrosive to glass-
ware. Be careful when handling this reagent.
3. Magnetite nanoparticles are highly photolabile. They will be
easily oxidized under light. The finished MNSP has to be
stored in the dark before use.
4. For enumeration of algal cells using a hemacytometer slide,
prior to counting, it is better to wait for at least 5.0 min after
transferring the samples into the slide. In this way, the cells are
more stable within the tiny space, and one can obtain more
consistent results.
5. The filtration of samples from toxin adsorption can be acceler-
ated by using a specialized vacuum chamber (Visiprep™ SPE
Vacuum Manifold, #57044 Supelco, Sigma-Aldrich, St. Louis,
MO, USA). The samples need to be evaporated to dryness
before adding any eluent (methanol with 0.1% acetic acid).
6. The HLB columns need to be primed by 5.0 mL of methanol
(twice) and then 5.0 mL deionized water (twice) before use.
The flow rate of eluent needs to be carefully controlled at
about two drops per second.
94 Shu-Chi Chang et al.

Acknowledgments

The authors would like to thank Ministry of Science and Technology

(Taiwan) of Republic of China (ROC) for the partial financial support
of this research through a grant of MOST104-2221-E-005 -008.

References
1. Miller J, Tyler G, Spoolman SE (2010) of microcystins in Microcystis aeruginosa strain
Environmental science, 13th edn. Cengage from lake Thanh Cong, Hanoi, Vietnam.
Learning, Belmint, CA Chromatographia 54:569–575
2. United Nations World Water Assessment 12. Lawton LA, Robertson PKJ (1999) Physico-­
Programme (2015) The United Nations world chemical treatment methods for the removal
water development report 2015: water for a of microcystins (cyanobacterial hepatotoxins)
sustainable world. UNESCO, Paris from potable waters. Royal Soc Chem
3. Qin B, Zhu G, Gao G, Zhang Y, Li W, Paerl 28:217–214
H, Carmichael W (2010) A drinking water cri- 13. Verspagen JMH, Visser PM, Huisman J (2006)
sis in lake Taihu, china: linkage to climatic vari- Aggregation with clay causes sedimentation of
ability and lake management. Environ Manag the buoyant cyanobacteria Microcystis spp. Aq
45:105–112 Microb Ecol 44:10
4. Rinta-Kanto JM, Konopko EA, DeBruyn JM, 14. Morris RJ, Williams DE, Luu HA, Holmes
Bourbonniere RA, Boyer GL, Wilhelm SW CFB, Andersen RJ, Calvert SE (2000) The
(2009) Lake erie microcystis: relationship adsorption of microcystin-LR by natural clay
between microcystin production, dynamics of particles. Toxicon 38:303–308
genotypes and environmental parameters in a 15. Liu G, Fan C, Zhong J, Zhang L, Ding S, Yan
large lake. Harmful Algae 8:665–673 S, Han S (2010) Using hexadecyl trimethyl
5. Hallegraeff GM (1993) A review of harmful ammonium bromide (CTAB) modified clays to
algal blooms and their apparent global increase. clean the Microcystis aeruginosa blooms in
Phycologia 32:21 Lake Taihu, China. Harmful Algae 9:413–418
6. USEPA (2016) Harmful algal blooms. 16. Chang S-C, Li C-H, Lin J-J, Li Y-H, Lee M-R
USEPA, Washington, DC (2014) Effective removal of Microcystis aerugi-
7. Teixera MGLC, Costa MCN, Carvalho VLP, nosa and microcystin-LR using nanosilicate
Pereira MS, Hage E (1993) Gastroenteritis platelets. Chemosphere 99:49–55
epidemic in the area of the Itaparica dam, 17. Lee J, Walker HW (2011) Adsorption of
Bahia, Brazil. Bull Pan Am Health Organ microcystin-Lr onto iron oxide nanoparticles.
27:244–253 Colloids Surf A Physicochem Eng Asp
8. Jervis R, Welch WM, Hall C (2014) Ohio 373:94–100
drinking water emergency 'not over yet'. USA 18. Gao Y-Q, Gao N-Y, Deng Y, Gu J-S, Shen
Today. Gannett Satellite Information Network, Y-C, Wang S-X (2012) Adsoption of microcys-
Inc., McLean, VA tin-­LR from water with iron oxide nanoparti-
9. Trevino-Garrison I, DeMent J, Ahmed FS, cles. Water Environ Res 84:562–568
Haines-Lieber P, Langer T, Ménager H, Neff 19. Rippka R, Deruelles J, Waterbury JB, Herdman
J, van der Merwe D, Carney E (2015) Human M, Stanier RY (1979) Generic assignments,
illnesses and animal deaths associated with strain histories and properties of pure cultures
freshwater harmful algal blooms—Kansas. of cyanobacteria. J Gen Microbiol 111:1–61
Toxins (Basel) 7:353–366 20. Lin J-J, Chu C-C, Chiang M-L, Tsai W-C
10. Westrick JA, Szlag DC, Southwell BJ, Sinclair (2006) First isolation of individual silicate
J (2010) A review of cyanobacteria and platelets from clay exfoliation and their unique
­cyanotoxins removal/inactivation in drinking self-assembly into fibrous arrays. J Phys Chem
water treatment. Anal Bioanal Chem 397: B 110:18115–18120
1705–1714 21. Chiu C-W, Chu C-C, Cheng W-T, Lin J-J
11. Hummert C, Dahlmann J, Reinhardt K, Dang (2008) Exfoliation of smectite clays by
HPH, Dang DK, Luckas B (2001) Liquid chro- branched polyamines consisting of multiple
matography-mass spectrometry identification ionic sites. Eur Polym J 44:628–636
 Chapter 9

An Immunochromatographic Test Strip to Detect

Ochratoxin A and Zearalenone Simultaneously
Xiaofei Hu and Gaiping Zhang

Abstract
Lateral flow immunochromatographic assays are simple, useful tools for the detection of many different
targets. In this format, a liquid sample containing target of interest moves along a strip of polymeric mate-
rial adhered with various lines where molecules have been attached that excert specific interactions with the
target. Here, we describe a colloidal gold-based immunochromatographic test strip for the simultaneous
detection of ochratoxin (OTA) and zearalenone (ZEN) in corn and other cereals qualitatively and
quantitatively.

Key words Immunochromatographic test strip, Colloidal gold, Mycotoxins, Ochratoxin A,

Zearalenone

1 Introduction

Lateral flow immunoassays (LFIA) are simple, useful tools for a

variety of point-of-care and field-use applications [1–3]. One typi-
cal LFIA format consists of a surface layer named sample pad to
carry sample liquid via the conjugate pad along the strip encoun-
tering the detection zone up to the absorbent pad [4].
As with the ELISA method, there are two kinds of assay format
of LFIA, namely, the direct (sandwich) or competitive one [5–8].
The sandwich LFIA is mainly used to detect macromolecular sub-
stance, as proteins, hormones, or enzymes. The competitive LFIA is
mainly used to determine targets containing small molecules.
Moreover, there are many kinds of labels which can be used for con-
jugates that attach on the conjugate pad, such as colloidal gold [9],
latex, liposomes, fluorescent nanoparticles, quantum dots, or up-
converting phosphor [4, 10]. Thus, lateral flow immunochromato-
graphic test strips are currently widely used to detect veterinary drugs,
pesticides, mycotoxins, prohibited additives, allergen, environmental
pollutants, pathogens, hormones, and metabolites, because of their
sensitivity, selectivity, speed, and ease of use [8, 10–13].

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_9, © Springer Science+Business Media LLC 2017

95
96 Xiaofei Hu and Gaiping Zhang

Conventional lateral flow strip only determines a single target

per strip. Thus, we have labeled monoclonal antibodies for OTA
and ZEN with colloidal gold and have subsequently developed a
lateral flow immunoassay strip of the competitive format which can
simultaneously detect OTA and ZEN.

2 Materials

All solutions were prepared using ultrapure water (the water has a
sensitivity of 18.2 MΩ × cm at room temperature) or double-­
distilled water (DDW). All reagents are analytical grade or higher
and prepared and stored at room temperature, unless otherwise
specified. All the waste materials have been collected and disposed
by professional company designated by the government.

2.1 Colloidal Gold 1. Gold chloride trihydrate, store at 4 °C (see Note 1).
Components 2. 1% gold chloride solution: dissolve 1 g gold chloride trihydrate
(see Note 2) in 100 mL ultrapure water (see Note 3).
3. 1% trisodium citrate solution: dissolve 1 g trisodium citrate
dehydrate in 100 mL ultrapure water.

2.2 Immuno­ 1. Antibodies: anti-OTA and anti-ZEN monoclonal antibodies

chromatographic Test (see Notes 4 and 5) are prepared by Henan Key Laboratory,
Strips Henan Academy of Agricultural Sciences.
2. Buffer for sample pads of strips: 100 mM phosphate buffer
(PB) containing 1% casein, 0.3% Tween 20, 5% trehalose, and
0.05% sodium azide (see Notes 6 and 7), store at 4 °C.
3. Buffer for conjugate pads of strips: 20 mM borate buffer (see
Note 8) containing 3% casein, 1% Triton X-100, and 0.03%
sodium azide (see Note 9), store at 4 °C.
4. Standard solutions of OTA and ZEN: diluting stock solutions
(see Note 10) of OTA or ZEN with DDW, respectively.
5. Sodium chloride solution: dissolve 10 g sodium chloride in
100 mL DDW.

2.3 The Structural Nitrocellulose membrane, glass fiber, and absorbent pad were from
Components Millipore, Bedford, MA, USA. The plastic supporting backbone
of the Immuno­ was supplied by Jiening Biological Technology Co., Ltd., Shanghai,
chromatographic Test China.
Strips
2.4 Preparation 1. Synthesize OTA-BSA by the method described by Wang et al.
of OTA-BSA [14] with slight modification (see Note 11).
and ZEN-BSA 2. Prepare ZEN-BSA according to Sun et al. [12] (see Note 12).
 An Immunochromatographic Test Strip to Detect Ochratoxin A and Zearalenone... 97

3 Methods

All the procedures were carried out at room temperature unless

specified otherwise.

3.1 Preparation 1. Mix 1 mL 1% (w/v) gold chloride solution with 99 mL ultra-

of Colloidal Gold pure water in a very clean glassware (see Note 13) to prepare
Nanoparticles 0.01% gold chloride solution.
2. Add 1.6 mL of a 1% solution of trisodium citrate to 100 mL of
boiling 0.01% gold chloride solution (see Note 14). The resulting
solution, initially had a gray color which changed to lavender and
then, with continued boiling, after 1–3 min developed a red hue.
The resulting colloidal gold particle size was about 25 nm [15].

3.2 Preparation 1. Adjust pH of the colloidal gold solution to 8.2 with 0.2 mol/L
of Colloidal Gold K2CO3 (see Note 15).
Conjugates 2. Determine the most appropriate mAb protein concentration
of the Monoclonal for conjugation: prepare twofold serial dilutions of the stock
Antibody antibody solutions, anti-OTA or anti-ZEN (see Note 16). Mix
30 μL of each dilution with 125 μL of colloidal gold solution.
After incubating for 5 min at room temperature, add 125 μL
of 10% NaCl solution. The color of the mixtures changes from
brilliant red to blue as the concentration of mAb decreases.
The optimum concentration of mAb for colloidal gold labeling
is the lowest concentration of mAb solution at which there is a
change in color [16] (see Note 17).
3. Drop the required volume of the optimal concentration of
mAb protein into 100 mL colloidal gold solution (pH 8.2)
(see Note 18). Gently mix for 70 min, then add 10 mL of 5%
(w/v) casein, and incubate further for 10 min. Centrifuge the
mixture at 12,000 × g for 30 min at 4 °C. Discard the super-
natant, and re-suspend the sediment in 50 mL of gold conju-
gate suspension buffer (see Note 19).

3.3 Assembly The immunochromatographic test strip is based on the competi-

of Immuno­ tive immunoassay principle:
chromatographic Test 1. The strip structure contains the following elements: sample
Strips pads (see Note 20), conjugate pads (see Note 21), absorbent
pads (see Note 22), nitrocellulose (NC) membranes (see
Note 23), and plastic supporting backbone (see Note 24).
Attach the sample pads, conjugate pads, NC membranes,
and absorbent pads to the plastic supporting backbone
sequentially (see Note 25) as shown in Fig. 1.
2. Coat the conjugate pads with colloidal gold-labeled anti-OTA
and anti-ZEN mAbs with a XYZ Biostrip Dispenser separately.
The volume of colloidal gold-labeled anti-OTA mAbs is 7 μL/
cm line, and of anti-ZEN 6 μL/cm line. Dry the conjugate
pads at 42 °C for 1 h.
98 Xiaofei Hu and Gaiping Zhang

Fig. 1 Schematic diagram of simultaneous immunochromatographic strip test. C, control line; T1, OTA test line;
T2, ZEN test line; +, mycotoxin positive; −, mycotoxin negative

3. Spot test and control lines onto NC membrane (see Note 26)
using XYZ Biostrip Dispenser at 0.9 μL/cm line. Spot two test
lines (see Note 27) with OTA-BSA at a concentration of
0.25 mg/mL and with ZEN-BSA at a concentration of
0.30 mg/mL, respectively (see Note 28). Spot the control line
with staphylococcal protein A at a concentration of 1.20 mg/
mL. Dilute all protein solutions for spotting with normal
saline. The membrane is dried at 42 °C for 4 h.
4. Cut the assembled semirigid polyethylene sheet into strips
(0.3 cm × 5 cm).

3.4 Sample The samples included corn, barley, oat, rice, wheat, and other
Preparation cereals:
1. Ground the samples before mixing 5 g with 2 mL of 1 M HCl
for 5 min.
2. Add 5 mL of trichloromethane and shake the mixture violently
for 15 min.
3. Filter the mixture and abandon the upper aqueous phase.
4. Add an equal volume of 0.13 M NaHCO3 to the remaining
organic phase. Shake the mixture violently for 15 min.
5. Centrifuge the mixture for 10 min at 3300 × g.
6. The clear upper phase was used directly for analysis without
further cleanup (see Note 29).

3.5 Assay Procedure 1. Drop 100–150 microliters of spiked sample extracts onto the
and Determination sample pad of a test strip.
of the Analytical 2. Observe the color of the test lines and control line 5–10 min
Parameters later.
3. Examine the strip with the naked eye to qualitatively detect
OTA and ZEN presence (Fig. 2) (see Note 30).
 An Immunochromatographic Test Strip to Detect Ochratoxin A and Zearalenone... 99

Fig. 2 Test strip of corn samples. Note: (a) OTA concentrations (from left to right), 0, 2, 3, 4, 5, and 6 ng/mL; (b)
ZEN concentrations (from left to right), 0, 4, 8, 12, 16, and 20 ng/mL; (c) OTA/ZEN concentrations (from left to
right), 0/0, 2/4, 3/8, 4/12, 5/16, 6/20, and 50/50 ng/mL; C line is the control line; T1 and T2 lines are the OTA
and ZEN test lines, respectively

4. Detect OTA and ZEN quantitatively in samples (Figs. 3 and 4)

by scanning the strips with the densitometer (see Note 31).

4 Notes

1. Gold chloride trihydrate readily absorbs water and should be

kept in a sealed container and stored in a dry place. It can also
be aliquoted and sealed under vacuum.
2. Gold chloride trihydrate should be weighed in a dry room, and
the process should take as little time as possible.
3. The ultrapure water is prepared in our laboratory by PURELAB
Classic UF water purifier (Veolia, UK) with purified water.
4. The OTA monoclonal antibody was prepared from ascitic fluid
induced by the hybridoma secreting anti-OTA ­monoclonal anti-
body (cell line, 2A11). The immunological properties of OTA
monoclonal antibody were determined: the 50% inhibitory con-
centration (IC50) for OTA was 0.163 ng/mL; the cross-reactivity
was 16.3% with OTB and no more than 0.03% with compounds
such as aflatoxin B1, fumonisin B, and zearalenone.
5. The ZEN monoclonal antibody was prepared from ascitic fluid
from mice growing the hybridoma cell line 4A3-F9 secreting
anti-ZEN monoclonal antibody. The immunological proper-
ties of ZEN monoclonal antibody were determined: the 50%
100 Xiaofei Hu and Gaiping Zhang

Fig. 3 ROD curves in corn samples for OTA (a) and ZEN (b). Note: The corn extracts spiked with OTA (0, 1, 2, 3,
4, 5, 6 ng/mL) and ZEN (0, 2, 4, 8, 12, 16, 20 ng/mL−1) standard solutions were tested using the test strips. Test
lines were scanned with a TSR3000 membrane strip reader
 An Immunochromatographic Test Strip to Detect Ochratoxin A and Zearalenone... 101

a b
0.9
0.8 y = -1.0916x + 0.7764 0.9
y = -0.7527 x + 0.9807
0.7
R2 = 0.9862 0.8 R2 = 0.9975
IC50=1.7905ng/ml
0.7 IC50=4.3514ng/ml
0.6
0.6
0.5
B/B0

0.5

B/B0
0.4
0.4
0.3 0.3
0.2 0.2
0.1 0.1
0 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.2 0.4 0.6 0.8 1 1.2 1.4
lg[OTA concentration (ng/ml)] lg[ZEN concentration (ng/ml)]

Fig. 4 Standard curves in corn samples for OTA (a) and ZEN (b) using test strip detection. Note: The X-axis
is expressed as log concentration. B/B0 represents the ratio of ROD of standards divided by that of the
ROD at 0 ng/mL

inhibitory concentration (IC50) for OTA was 1.115 ng/mL;

the cross-reactivity with zearalanone was 53.12% and less than
4% with other analogues such as α-zearalenol, β-zearalenol,
α-zearalanol, and β-zearalanol.
6. 100 mM PB: mix 95 mL 0.2 M monosodium orthophosphate
with 405 mL 0.2 M disodium hydrogen phosphate, then add
DDW to the mixed solution to 1000 mL.
7. Sample pad buffer: mix 1 g casein, 5 g trehalose, 0.05 g sodium
azide, and 0.3 mL Tween-20 in 100 mL PB.
8. 20 mM borate buffer: dissolve 7.63 g sodium tetraborate deca-
hydrate in 1000 mL DDW.
9. Conjugate pad buffer: mix 3 g casein, 1 mL Triton X-100, and
0.03 g sodium azide in 100 mL 20 mM borate buffer.
10. Stock solutions: dissolve 1 mg OTA or ZEN in 1 mL methanol
and store at −20 °C.
11. Preparation of OTA-BSA: 5 mg of OTA and 2.6 mg N-(3-
dimethylaminopropyl)-N'-ethyl-carbodiimide (EDC) were
dissolved in 1 mL dry N,N-dimethylformamide (DMF). The
mixture was then added dropwise with mixing to 2 mL of a
solution of 10 mg BSA in 130 mM sodium bicarbonate
(pH 8.1). Mix at room temperature for 5 h, then add a further
1.2 mg EDC again. The final mixture was incubated overnight
at 4 °C with stirring. The conjugates were purified by dialysis
against 0.01 M phosphate-buffered saline (PBS, pH 7.4) and
stored at −20 °C.
12. Preparation of ZEN-BSA: 7 mg of BSA was dissolved in 1 mL
of distilled water, and the pH adjusted to 10.8 using 1 M
sodium hydroxide. Butane-1,4-diol diglycidyl ether (6 μL) was
added, and the solution was incubated for 20 h at ambient
temperature under a nitrogen atmosphere. 5 mg of ZEN dis-
solved in 1.5 mL of 0.5 M sodium hydroxide, degassed using
102 Xiaofei Hu and Gaiping Zhang

nitrogen, was then added to the epoxy-activated BSA solution.

The reaction mixture was incubated for 20 h at ambient tem-
perature under a nitrogen atmosphere. The conjugates were
purified by dialysis against 0.01 M PBS and stored at −20 °C.
13. Scratches on the inner surface of the glassware must be avoided,
and a mild detergent can be used to clean the glassware and a
soft cloth to scrub its inner surface. After cleaning, all deter-
gent should be removed by rinsing thoroughly with ultrapure
water, and then the glassware boiled with ultrapure water two
or three times.
14. It is best to add trisodium citrate solution to the boiling gold
chloride solution. The trisodium citrate solution should be
added as quickly as possible into the boiling gold chloride solu-
tion with strong agitation.
15. The pH of prepared colloidal gold solution is about 6. Add
12 μL 0.2 M K2CO3 to 1 mL of colloidal gold solution gives a
final pH of 8.2.
16. Twofold serially diluted about 8–12 times in a row of wells.
17. In our experience, the lowest concentration of anti-OTA mAb
protein that produced a change in color is 15 μg/mL colloidal
gold solution. For the anti-ZEN mAb protein, the amount is
19.68 μg/mL colloidal gold solution.
18. For our experiment, we rounded up the optimum anti-ZEN
mAb protein to 20 μg/mL of colloidal gold solution. The opti-
mum anti-OTA mAb protein concentration used was 15 μg/mL
of colloidal gold solution. Before conjugating the mAb protein
to colloidal gold, we diluted the initial mAb protein to 1 mg/mL
with DDW; thus, to 100 mL of colloidal gold solution, 1.5 mL
of the diluted anti-OTA mAb protein solution or 2 mL of the
diluted anti-ZEN mAb protein solution was added.
19. Gold conjugate suspension buffer: add 1 g BSA, 0.03 g sodium
azide to 100 mL of 20 mM borate buffer and store at 4 °C.
20. The sample pads are usually made of glass fiber, cross-linked
silica, cellulose, or polyester cotton. In our laboratory, glass
fiber is used, as a 30 cm × 1.5 cm strip previously soaked in
sample pad buffer, then dried at 42 °C for 4 h.
21. Many materials can be used as conjugate pads, such as glass
fiber, polyester film, and cellulose or non-woven fabrics. Our
laboratory uses a glass fiber conjugate pad as a 30 × 0.8 cm
strip previously soaked with conjugate pad buffer, then dried at
42 °C for 1 h.
22. The absorbent pads are made of white filter paper without any
chemical additives as a 30 × 1.8 cm strip.
An Immunochromatographic Test Strip to Detect Ochratoxin A and Zearalenone... 103

23. A 30 × 1.5 cm strip of nitrocellulose is used. Nitrocellulose can

be replaced by cellulose acetate membrane, polyvinylidene
difluoride membrane, or nylon membrane.
24. The plastic supporting backbone is made of semirigid polyeth-
ylene sheet, which specification is 30 × 6.5 cm.
25. The overlap for consecutive pads is 2 mm.
26. In our laboratory, the distance between the adjacent test lines
1 and 2 and control line is 0.45 cm.
27. If the strip is required to detect a single fungal toxin, then only
one test line is put on the NC membrane.
28. The order of the conjugates in test line 1 and test line 2 can be
changed.
29. The upper clear liquid should be diluted before analysis accord-
ing to the detection sensitivity of strip and the limit set for the
target in the foodstuff.
30. When both mycotoxins are absent (sample OTA−/ZEN−), all
of the mAb−gold conjugates are trapped by the capture
reagents resulting in three vivid red lines (T1, T2, and C line).
Whereas if the sample contains both mycotoxins (OTA+/
ZEN+), only one vivid red line appears at the C line. If two
lines appear (C line and one T line), the sample is positive for
either OTA or ZEA (OTA−/ZEN+ or OTA+/ZEN−). The
control line is colored red to ensure that the strip and the test
procedure are correct. If no color appears at the control line,
either the test procedure was improperly conducted or the
strip was faulty and the sample should be tested again using a
new strip [5]. The sensitivity of the test strip to the naked eye
was set at the lowest concentration of OTA/ZEN giving no
color at the corresponding test line. We found that the color of
the OTA test lines and ZEN test lines disappeared at ≥6 ng/
mL and ≥20 ng/mL, respectively. Therefore, the visual sensi-
tivity for the test strip was found to be 6 ng/mL of OTA and
20 ng/mL of ZEN, respectively. In our laboratory, we con-
sider a sample containing less than 6 ng/mL OTA as negative;
for ZEN, this limit is 20 ng/mL.
31. With TSR3000 membrane strip reader (Bio-Dot), the immu-
nochromatographic strip test can be used for a more quantita-
tive analysis [6]. The higher the concentration of analyte in the
sample the more it competes for the limited amount of mAb−
gold conjugate. Hence, the color intensity of the appropriate
test line is inversely related to the toxin concentration in the
sample. A TSR3000 membrane strip reader was used to scan
the strips and obtain the relative optical density (ROD) of the
test line. The standard curve was constructed by plotting the
G/D × A-ROD (pixel) values obtained from the standard sam-
104 Xiaofei Hu and Gaiping Zhang

ples against the log of the concentration. The IC50 and the
lower detection limit (LDL) were calculated from the regres-
sion equation. The linearity was assessed by the coefficient of
determination (R2). The sigmoidal dose−response curve
­produced by the OTA/ZEN standard samples against its loga-
rithm concentrations and the G/D × A (area), which had good
linearity within the range of 0.62–5.14 ng/mL and 0.94–
20.09 ng/mL for OTA and ZEN, respectively, and correlation
coefficients of 0.9862 and 0.9975. The calculated regression
equations were y = −1.0916x + 0.7764 and
y = −0.7527x + 0.9807. From these, the IC50 for OTA and
ZEN were calculated to be 1.79 ng/mL and 4.35 ng/mL, and
the LDL of OTA and ZEN were 0.77 ng/mL and 1.20 ng/
mL, respectively. So the sensitivity of the test strip, using the
Bio-Dot TSR3000 membrane strip reader, for cereal (extracts
prepared by the aforementioned sample preparation method)
was 0.77 μg/kg and 1.20 μg/kg of OTA and ZEN,
respectively.

Acknowledgment

This research was supported by the National Science & Technology

Pillar Program of “12th Five-Year Plan” (2014BAD13B05) and
the Distinguished Youth Fund Project of Henan Academy of
Agricultural Sciences (2013YQ29). Norman Gregson, Professor at
University of London, reviewed this manuscript and gave lots of
good advice.

References

1. Paek SH, Lee SH, Cho JH et al (2000) 5. Zhao Y, Zhang G, Liu Q et al (2008)
Development of rapid one-step immunochro- Development of a lateral flow colloidal gold
matographic assay. Methods 22:53–60 immunoassay strip for the rapid detection of
2. Putalun W, Morinaga O, Tanaka H et al enrofloxacin residues. J Agric Food Chem
(2004) Development of a one-step immuno- 56:12138–12142
chromatographic strip test for the detection of 6. Song C, Liu Q, Zhi A et al (2011) Development
sennosides A and B. Phytochem Anal of a lateral flow colloidal gold immunoassay
15:112–116 strip for the rapid detection of olaquindox resi-
3. Cho YJ, Lee DH, Kim DO et al (2005) dues. J Agric Food Chem 59:9319–9326
Production of a monoclonal antibody against 7. Li X, Sun Y, Yang S et al (2015) Development
ochratoxin A and its application to immuno- of an immunochromatographic strip for anti-
chromatographic assay. J Agric Food Chem body detection of pseudorabies virus in swine.
53:8447–8451 J Vet Diagn Invest 27:739–742
4. Posthuma-Trumpie GA, Korf J, van 8. Wang Y, Deng R, Zhang G et al (2015) Rapid
Amerongen A (2009) Lateral flow (immuno) and sensitive detection of the food allergen gly-
assay: its strengths, weaknesses, opportunities cinin in powdered milk using a lateral flow col-
and threats. A literature survey. Anal Bioanal loidal gold immunoassay strip test. J Agric
Chem 393:569–582 Food Chem 63:2172–2178
 An Immunochromatographic Test Strip to Detect Ochratoxin A and Zearalenone... 105

9. Zhang GP, Li QM, Yang YY et al (2005) 13. Yang X, Zhang G, Wang F et al (2015)
Development of a one-step strip test for the Development of a colloidal gold-based strip test
diagnosis of chicken infectious bursal disease. for the detection of chlorothalonil residues in
Avian Dis 49:177–181 cucumber. Food Agric Immunol 26:729–737
10. Song C, Zhi A, Liu Q et al (2013) Rapid and 14. Wang XH, Liu T, Xu N et al (2007) Enzyme-­
sensitive detection of beta-agonists using a por- linked immunosorbent assay and colloidal gold
table fluorescence biosensor based on fluores- immunoassay for ochratoxin A: investigation of
cent nanosilica and a lateral flow test strip. analytical conditions and sample matrix on assay
Biosens Bioelectron 50:62–65 performance. Anal Bioanal Chem 389:903–911
11. Zhi A, Li B, Liu Q et al (2010) Development 15. Frens G (1973) Controlled nucleation for the
of a lateral-flow immunochromatographic test regulation of the particle size in monodisperse
device for the rapid detection of difloxacin resi- gold suspensions. Nat Phys Sci 241:20–22
dues. Food Agric Immunol 21:335–345 16. Zhang G, Wang X, Zhi A et al (2008)
12. Sun Y, Hu X, Zhang Y et al (2014) Development Development of a lateral flow immunoassay
of an immunochromatographic strip test for strip for screening of sulfamonomethoxine resi-
the rapid detection of zearalenone in corn. dues. Food Addit Contam Part A Chem Anal
J Agric Food Chem 62:11116–11121 Control Expo Risk Assess 25:413–423
 Chapter 10

Endotoxin Removal from Escherichia coli Bacterial Lysate

Using a Biphasic Liquid System
Janusz Boratyński and Bożena Szermer-Olearnik

Abstract
Lipopolysaccharide (LPS, endotoxin, pyrogen) which is a component of the outer membrane of most
Gram-negative bacteria is a troubling contaminant of crude bacteriophage suspension. Therefore, its
removal is important for bacteriophage applications especially in preparations dedicated for use in therapy
with bacterial infections treatment. The method presented here is used for extractive removal of endotox-
ins from bacteriophage preparations with a water immiscible solvent such as 1-octanol. During extraction
most of the phage lytic activity is retained in the aqueous phase, while endotoxin accumulates in the
organic solvent. The levels of endotoxin in the aqueous bacteriophage rich fraction are determine by
Limulus Amebocyte Lysate or EndoLISA assay and are extremely low.

Key words Lipopolysaccharides, Bacteriophage, Extraction procedure

1 Introduction

Bacteriophage preparations are often contaminated by macromol-

ecules derived from the host bacteria and culture medium. In the
case of phages prepared on Gram-negative bacteria, the major
pyrogen is lipopolysaccharide (endotoxin), which can be found in
the outer membrane of these bacteria [1]. The amount of endo-
toxins is defined by an endotoxic unit (EU) which corresponds to
activity of 100 pg of Escherichia coli lipopolysaccharide. Tolerated
endotoxin levels depend on the route of the administration there-
fore the allowed threshold for most intravenous applications is
5 EU/kg/h [2]. Endotoxin is much more tolerated when admin-
istered orally [3, 4], additionally the endotoxin content of distilled
water is ca. 20 EU/mL [5]. There is no general procedure for
endotoxins removal which is the reason why many endotoxin elim-
ination methods are tailored for specific bioproducts. [6]. For
example, ultrafiltration is a common technique for purification bio-
logical macromolecules [7, 8]. Affinity techniques can be used to
effectively remove lipopolysaccharides using immobilized ligands

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_10, © Springer Science+Business Media LLC 2017

107
108 Janusz Boratyński and Bożena Szermer-Olearnik

such as polymyxin B [9, 10], ʟ-histidine, poly-ʟ-lysine,

poly(ethyleneimine) [11], PEGylated polypeptides [12] and
ʟ-serine immobilized on polyvinylidene fiber [13]. A commercially
available Endo-trap kit efficient at endotoxin removal is also applied
to bacteriophage suspensions [14]. Here, we demonstrate a
method for effective endotoxin removal from a bacteriophage
crude suspension by extraction with water immiscible solvent
namely 1-octanol [15]. The presented procedure allows for endo-
toxins removal from bacteriophage preparations with preserving
antibacterial activity.

2 Materials

2.1 Media Nutrient broth (pH 7.2 ± 0.2): 0.4 g beef extract, 5.4 g enzymatic
digest of casein, 1.7 g yeast extract, 4 g peptone, 3.5 g NaCl per
liter. Dissolve the components in distilled water and sterilize in an
autoclave. Add sterile glucose to a final concentration of 1%.

2.2 Bacterial Strains 1. E. coli B strain—e.g., from the Polish Collection of Microorganisms,
Institute of Immunology and Experimental Therapy, Polish
Academy of Sciences (PCM 1630). Keep the strain on MacConkey
agar at 4 °C.
2. Pseudomonas aeruginosa strain—e.g., from the Polish
Collection of Microorganisms, Institute of Immunology and
Experimental Therapy, Polish Academy of Sciences (PCM
2720). Keep the strain on blood agar at 4 °C.

2.3 Bacteriophages 1. Bacteriophage T4—E. coli-specific bacteriophage from the

American Type Culture Collection (Rockville, Maryland,
USA).
2. Bacteriophage HAP1—E. coli-specific bacteriophage received
from the phage collection of Laboratory of Bacteriophages,
Institute of Immunology and Experimental Therapy, Polish
Academy of Sciences [16].
3. Bacteriophage F8—P. aeruginosa-specific bacteriophage
receive from the phage collection of Laboratory of
Bacteriophages, Institute of Immunology and Experimental
Therapy, Polish Academy of Sciences.

2.4 Determination 1. Limulus Amebocyte Lysate test (Charles River Laboratories

of Endotoxins International).
2. EndoLISA (ELISA-based Endotoxin Detection Assay, Hyglos,
Germany).
 Endotoxin Removal from Escherichia coli Bacterial Lysate... 109

3 Methods

3.1 Preparation 1. Grow the bacterial culture in nutrient broth at 37 °C for

of Crude 8–16 h, until the optical density (OD, 565 nm) reaches 1
Bacteriophage McFarland units.
Suspension 2. Infect the culture with bacteriophage (0.1 PFU/bacteria).
3. Grow this culture at 37 °C for further 8 h.
4. Filtrate the crude bacteriophage suspension using the filters
with 0.22 μm pore size.

3.2 Determination 1. Determine the titer of phage, expressed as plaque forming

of Bacteriophage Titer units (PFU), using the double layer agar technique [17].

3.3 General 1. Supplement crude bacteriophage suspension by adding a

Purification Procedure 0.2 M solution of MgCl2 to the final concentration of 0.02 M.
2. Incubate at 4 °C for 3–24 h.
3. Add the 10–20% of octanol (v/v) to the crude phage suspen-
sion and shake gently for 1–3 h at room temperature.
4. Cool the two-phase mixture to 4 °C for 1–3 h and separate
using separatory funnel for liter-scale, or centrifugation at
4000 × g, 10 min for small samples.
5. Dialyze the collected aqueous phase using dialysis membrane
(12–14 kDa), against 15–25% aqueous ethanol (5 × 4 h), and
subsequently against aqueous 0.15 M NaCl sterile solution
(4 × 4 h).
6. Optionally pass crude lysate through a Pellicon membrane
(1000 kDa) EMD Millipore.
7. The fraction that does not pass through the membrane is the
bacteriophage concentrate (Fig. 1).

3.4 Determination 1. Use the end-point chromogenic Limulus Amebocyte Lysate

of Endotoxins (Table 1) test.
3.4.1 LAL Test 2. Quantification of endotoxins is achieved using the chromo-
genic technique—a standard curve, consisting of three mea-
surement points, is perform for each determination.
3. Dilute samples serially with pyrogen-free water (supplied with
the test) until the measurement is within the range of the stan-
dard curve.
4. Perform each data point in duplicate.
5. Maintain reaction conditions according to the manufacturer’s
recommendation [18].
110 Janusz Boratyński and Bożena Szermer-Olearnik

Fig. 1 Schematic illustration of purification protocol. Dotted line represents optional stages used for improved
efficiency of the purification procedure

3.4.2 EndoLISA Test 1. Mark the endotoxin levels of the T4 bacteriophage prepara-
tions after purification using EndoLISA according to the man-
ufacturer’s instructions.
2. Dilute the samples and standards with Binding Buffer prior to
analysis.
3. Incubate samples on an assay plate overnight at room tempera-
ture with shaking.
4. Wash the plate with Washing Solution and add Assay Reagent.
5. Detect the fluorescent signal immediately in a fluorescence
microplate reader (Synergy H4, BioTek, USA) [19, 20].
 Endotoxin Removal from Escherichia coli Bacterial Lysate... 111

Table 1
Representative results of three different bacteriophages purification runs with 1-octanol extraction

Titer of bacteriophage Endotoxin level (EU/ Relative endotoxin

(109 PFU/mL) mL)a content (EU/109 PFU)
Bacteriophage
Entry strain Beforeb Afterc Beforeb Afterc Beforeb Afterc
1 T4 110 50 34,000 0.9 310 0.02
2d
T4 110 70 34,000 50 310 0.7
3 HAP1 6.0 2.0 60,000 14 10,000 7.0
4 F8 1.9 0.9 3800 8.0 2000 8.9
a
Determined by chromogenic LAL test
b
For crude bacterial lysate
c
For bacterial lysate after purification but before ultrafiltration
d
Phage culture not supplemented with MgCl2 before extraction

4 Notes

1. The presence of magnesium ions increases the efficiency of

pyrogen removal.
2. Other organic solvents use for extraction (1-butanol, olive oil,
dichloromethane) also give good results nevertheless, we
found 1-octanol to be the most practical due to easier phase
separation. Another solvents such as olive oil or dichlorometh-
ane allowed for antibacterial activity preservation but they were
not tested for endotoxin removal efficiency. 1-Butanol was
tested for endotoxin removal and its efficiency is comparable
with the use of 1-octanol. However, the solubility in water of
1-butanol is significantly higher than 1-octanol.
3. Determination of the endotoxin level directly after extraction is
impossible, because the residual 1-octanol in the aqueous phase
disable the LAL test. Thus, sequential dialyses against15–25%
ethyl alcohol and 0.15 M aqueous NaCl are employed.
4. The membrane filtration step (1000 kDa cut-off) is aimed to
remove macromolecules originating from the culture medium
and bacteria. The filtration also causes concentration of the
bacteriophages but do not contribute to the removal of resid-
ual endotoxin (optionally).
5. The Limulus Amebocyte Lysate test and EndoLISA detect
endotoxins with different sensitivity. In the obtained puri-
fied bacteriophages, the LAL test showed a higher level of
endotoxin compared to the results received from
EndoLISA.
112 Janusz Boratyński and Bożena Szermer-Olearnik

6. During preparation of crude bacteriophage suspension the alter-

native method could be used to determine the number of bacte-
ria. This method is using absorbance measurement 260/280 nm
to estimate the number of bacteria in suspension [21].
7. The precise separation of the water phase is crucial for achieving
low endotoxin level in the final product.

References

1. Dofferhoff AS, Nijland JH, de Vries-Hospers 12. Guo J, Meng F, Li X, Wang M, Wu Y, Jing X
HG, Mulder PO, Weits J, Bom VJ (1991) et al (2012) PEGylated click polypeptides syn-
Effects of different types and combinations of thesized by copper-free microwave-assisted
antimicrobial agents on endotoxin release from thermal click polymerization for selective
gram-negative bacteria: an in vitro and in vivo endotoxin removal from protein solutions.
study. Scand J Infect Dis 23:745–754 Macromol Biosci 12:533–546
2. Daneshian M, Guenther A, Wendel A, Hartung 13. Gao JP, Huang M, Li N, Wang PF, Chen HL,
T, von Aulock S (2006) In vitro pyrogen test Xu QP (2011) Efficacy of a novel endotoxin
for toxic or immunomodulatory drugs. adsorber polyvinylidene fluoride fiber immobi-
J Immunol Methods 313:169–175 lized with (L)-serine ligand on septic pigs.
3. Bruttin A, Brussow H (2005) Human volun- J Zhejiang Univ Sci B 12:264–272
teers receiving Escherichia coli phage T4 orally: 14. Merabishvili M, Pirnay J-P, Verbeken G,
a safety test of phage therapy. Antimicrob Chanishvili N, Tediashvili M, Lashkhi N et al
Agents Chemother 49:2874–2878 (2009) Quality-controlled small-scale produc-
4. Abedon ST, Kuhl SJ, Blasdel BG, Kutter EM tion of a well-defined bacteriophage cocktail for
(2011) Phage treatment of human infections. use in human clinical trials. PLoS One 4:e4944
Bacteriophage 1:66–85 15. Szermer-Olearnik B, Boratyński J (2015)
5. Gorbet MB, Sefton MV (2005) Endotoxin: the Removal of endotoxins from bacteriophage
uninvited guest. Biomaterials 26:6811–6817 preparations by extraction with organic sol-
6. de Oliveira MP, Lopes AM, Mazzola PG, vents. PLoS One 10(3): e0122672.
Rangel-Yagui C, Penna TC et al (2007) doi:10.1371/journal.pone.0122672
Methods of endotoxin removal from biological 16. Dabrowska K, Zembala M, Boratynski J,
preparations: A review. J Pharm Pharm Sci Switala-Jelen K, Wietrzyk J, Opolski A et al
10:388–404 (2007) Hoc protein regulates the biological
7. Petsch D, Anspach FB (2000) Endotoxin effects of T4 phage in mammals. Arch
removal from protein solutions. J Biotechnol Microbiol 187:489–498
76:97–119 17. Adams MH (1959) Bacteriophages. Inter
8. Jang H, Kim HS, Moon SC, Lee YR, Yu KY, Science Publ, New York, USA, New York
Lee BK et al (2009) Effect of protein concen- 18. Charles River Endosafe, Limulus Amebocyte
tration and detergent on endotoxin reduction Lysate – Endochrome. Available: www.criver.
by ultrafiltration. BMB Rep 42:462–466 com.
9. Sato T, Shoji H, Koga N (2003) Endotoxin 19. Grallert H, Leopoldseder S, Schuett M, Kurze
adsorption by polymyxin B immobilized fiber P, Buchberger B (2011) EndoLISA: a novel
column in patients with systemic inflammatory and reliable method for endotoxin detection.
response syndrome: the Japanese experience. Nature Methods 8:i–v
Ther Apher Dial 7:252–258 20. Endolisa the world’s first endotoxin detection
10. Fiore GB, Soncini M, Vesentini S, Redaelli A system based on ELISA-technology. Available:
(2010) Mechanisms of polymyxin B endotoxin h t t p : / / w w w. h y g l o s . d e / e n / p r o d u c t s -­
removal from extracorporeal blood flow: ser vices/products/elisa-based-endotoxin-­
hydrodynamics of sorption. Contrib Nephrol detection-method.html.
167:55–64 21. Szermer-Olearnik B, Sochocka M, Zwolińska
11. Petsch D, Rantze E, Anspach FB (1998) K, Ciekot J, Czarny A, Szydzik J, Kowalski K,
Selective adsorption of endotoxin inside a poly- Boratyński J (2014) Comparison of microbio-
cationic network of flat-sheet microfiltration logical and physicochemical methods for enu-
membranes. J Mol Recognit 11:222–230 meration of microorganisms. Postepy Hig Med
Dosw 68:1392–1396
 Chapter 11

Fourier Transform Infrared Spectroscopy as a Tool

in Analysis of Proteus mirabilis Endotoxins
Paulina Żarnowiec, Grzegorz Czerwonka, and Wiesław Kaca

Abstract
Fourier transform infrared spectroscopy (FT-IR) was used to scan whole bacterial cells as well as lipopoly-
saccharides (LPSs, endotoxins) isolated from them. Proteus mirabilis cells, with chemically defined LPSs,
served as a model for the ATR FT-IR method. The paper focuses on three steps of infrared spectroscopy:
(1) sample preparation, (2) IR scanning, and (3) multivariate analysis of IR data (principal component
analysis, PCA).

Key words Lipopolysaccharides, Fourier transform infrared spectroscopy, Multivariate analysis, ATR
FT-IR spectroscopy, Principal component analysis

1 Introduction

Gram-negative bacterial endotoxins (lipopolysaccharides, LPS) are

components of the outer membrane of the cell envelope and can be
responsible for over-activation of the human immunological system
and may lead to endotoxemia [1]. The positive outcomes of biologi-
cal activity involve a humoral response and anti-LPS antibody genera-
tion. Anti-O-polysaccharide antibodies are useful for Gram-negative
bacterial serotyping. Complete structures of LPS are important for
understanding the biological activity of endotoxins.
Infrared spectroscopy is based on the principle that upon scan-
ning with an infrared (IR) beam, the vibrational behavior of mol-
ecules in the sample changes energy quanta and modifying
vibrational and rotational modes [2, 3]. Molecular bonds with an
electric dipole moment within the sample will absorb the infrared
radiation and vibrate in one of a number of ways, following pat-
terns of either stretching, bending, deformation, or combination
vibrations [2, 3]. Different chemical bonds absorb IR at different
wavenumber ranges regardless of other structures in the molecule,
and spectral peaks are derived from the absorption of bond vibra-
tional energy changes in the IR region [3]. These vibrations can

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_11, © Springer Science+Business Media LLC 2017

113
114 Paulina Żarnowiec et al.

then be correlated directly to the biochemical makeup of the sam-

ple, and the resultant infrared absorption spectrum provides a spe-
cific fingerprint that reflects the structure and composition of the
sample. Infrared frequencies (4000–600 cm−1) were used due to
the characteristic absorbance frequencies of organic molecules.
Primary molecular vibrations are often divided into five windows,
reflecting the predominant types of recorded class of organic com-
ponents. The fatty acid region is 2800–3000 cm−1, the amide I and
amide II (amide I/II) region is 1500–1700 cm−1, the mixed region
is 1200–1500 cm−1, the wavenumber region 1200–900 cm−1 is
associated with the stretching vibrations of polysaccharides, and
the 600–900 cm−1 region is a true fingerprint region [2]. The prin-
ciples of IR methods are presented in our and another recently
published reviews [3, 4].
IR has been used to define the lamellar structures of lipid A,
interactions with lysozymes and Mg2+, and water content in LPS
spectra [5, 6]. IR spectroscopy offers a direct view of the presence
of LPS on bacterial cells, without preprocessing. This paper pres-
ents the IR spectra of Proteus mirabilis S1959 (O3) as well as of
their two R mutant cells. The Ra type (strain R110) has its
O-specific part deleted, and the Re type (R45) is deprived of its
core oligosaccharide. Thus, they possess lipooligosaccharides
(LOS). The acquisition of spectra and their mathematical analysis
are given for whole P. mirabilis cells as well as for isolated LPS. IR
spectroscopy in ATR mode and analysis were run in the following
steps: (1) acquisition of IR spectra, (2) computer-based prepro-
cessing, (3) mathematical analysis of preprocessed spectra, and (4)
conclusions based on comparison with control samples.

2 Materials

1. Proteus mirabilis strains S1959, R110, and R45 can be obtained

from the Institute of Microbiology and Immunology, University
of Lodz, Lodz, Poland.
2. Ethanol (70% solution).

2.1 Equipment 1. Infrared spectrometer—Spotlight 400 FT-IR Imaging System

(PerkinElmer).
2. Microcentrifuge.
3. Bacterial incubator.

2.2 Software 1. Spectrum One (PerkinElmer, Waltham, MA, USA) (see Note 1).
2. Statistica 12 (StatSoft Inc., USA).
 Fourier Transform Infrared Spectroscopy as a Tool in Analysis of Proteus mirabilis… 115

3 Methods

3.1 Sample 1. Cultivate the bacteria on Lysogeny broth (LB) plates at 37 °C

Preparation for 18 h (see Note 2).
Prepare samples were prepared following the steps listed in
one of the options given below.
2. Transfer a single bacterial colony by a sterile plastic loop onto
the crystal of the ATR device and leave at 18 °C for drying. Or
scrape one loop of bacteria from the agar plate, suspend it in
100 μL of sterile deionized water, place 30 μL of the bacterial
suspension on the crystal, and allow it to dry [7]. In the case of
freeze-dried bacterial mass and LPSs, the water evaporation
step is omitted.
3. For LPS samples, cultivate bacterial strains in nutrient broth
supplemented with 1% glucose.
4. Harvest the bacterial cells at the end of logarithmic phase of
growth, kill them by addition of 1% phenol (final concentra-
tion), and centrifuge (3000 × g, 20 min, 10 °C) three times
with distilled water, and then freeze-dry.
5. Extract smooth P. mirabilis S1959 (O3) LPS from freeze-­dried
bacterial mass by the phenol–water method [8].
6. Mix 20 g of freeze-dried bacterial cells with 600 mL distilled
water, and immerse the glass container in a 65–67 °C water
bath, for 30 min, with mixing till the suspension is
homogenous.
7. Add 300 mL 90% phenol (from freshly prepared distilled crys-
talline phenol) and continue the extraction for 30 min at
65 °C.
8. Cool the mixture in an ice water bath approximately to 10 °C,
and then centrifuge it for 30 min, 300 × g.
9. Separate the aqueous milky upper phase with the crude LPS.
10. Repeat extraction two times with sediment (phenol phase).
11. Combine all three water phases in a dialysis bag, and dialyze it
in a cool room against tap water for 2 days and then against
distilled water for another 2 days.
12. Concentrate dialyzed material by vacuum distillation, and then
freeze-dry. The sample represents the crude LPS which may
contain smaller amounts of protein and nucleic acids, which
can be removed by enzymatic cleavage (see below).
Rough Ra- and Re-type LOS were obtained from strains R110
and R45, respectively, by the Galanos method (using phenol, chlo-
roform, and light petroleum, 2:5:8 by vol., PCP) [9]:
116 Paulina Żarnowiec et al.

1. To 10 g freeze-dried bacterial cells, add 200 mL PCP, and stir

at 5–10 °C for 60 min.
2. Separate the cells as sediment by centrifugation for 15 min
(4800 × g).
3. Repeat extraction of the sediment.
4. Combine the supernatants and evaporate chloroform and light
petroleum by rotary evaporation (fume hood!).
5. Add to the remaining phenol phase dropwise distilled water,
with gentle stirring until white LPS precipitation occurs.
6. Centrifuge in a refrigerated centrifuge at 3000 × g for 30 min,
wash the sediment (= LPS) two times with acetone, and finally
resolve it in distilled water and freeze-dry.
7. Purify aqueous 1% solutions of crude LPS and LOS from
nucleic acids and proteins by ultracentrifugation (three times
at 105,000 × g, 4 °C, 120 min).
8. For final LPS purification, perform treatment with nucleases
[10], i.e., 2 mg/g LPS RNAse and DNAse (Merck) in 10 mM
Tris–HCl buffer (pH 7.6) for 18 h at room temperature. Add
one drop of toluene to prevent bacterial growth.
9. After overnight dialysis against distilled water, precipitate the
LPS by ultracentrifugation (105,000 × g, 4 °C, 120 min) and
freeze-dry. Pure LPS/LOS should contain less than 2% pro-
teins and only traces of nucleic acids) (see Note 3).

3.2 Acquisition 1. Open the instrument-operating software by double-clicking

of ATR FT-IR Spectra the “Spectrum” icon.
2. Open instrumental settings.
3. Check the path to the location where files are to be saved.
4. Clean the ATR crystal with 70% ethanol or isopropanol and
ensure the crystal is dry before background acquisition (see
Notes 4 and 5).
5. Locate and click on “Instrument” in the panel. Select “Scan”
and record a background spectrum (see Note 6) (Fig. 1).
Prepare samples by following the steps listed in one of the options
given below:
1. Place a colony of bacteria (see Notes 7–9) or a freeze-dried
sample on the crystal of the ATR device and press with a pres-
sure applicator [7, 11, 12], or place 30 μL of a bacterial sus-
pension on the crystal and allow to dry [13].
2. Acquire a sample spectrum with a resolution of 4 cm−1 and 32
or 64 co-added scans. It is common to collect multiple techni-
cal and biological replicate samples; three biological replicates
and three technical replicates are recommended (see Note 10)
[11, 14] (Figs. 2 and 8a).
Fourier Transform Infrared Spectroscopy as a Tool in Analysis of Proteus mirabilis… 117

Fig. 1 Background acquisition, screenshot from PerkinElmer’s Spectrum FT-IR

software

Fig. 2 Sample acquisition, screenshot from PerkinElmer’s Spectrum FT-IR

software
118 Paulina Żarnowiec et al.

3.3 Data Preprocessing of spectra involves different options at each step;

Preprocessing however, there are combinations of these steps that may be more
or less appropriate than others, depending on the sample type,
instrumentation setup, noise level, need for visualization of spec-
tra, personal preferences, and classification performance, among
other factors [3, 4].

3.3.1 Preprocessing 1. Start “Spectrum” and click on “File…” and select “Open…”
Using Spectrum One dataset (Fig. 3).
Software 2. Locate and click “Process” in the top panel.
3. Locate and click “Baseline Correction → Automatic…” (see
Note 11) (Fig. 4).
4. Locate and click on “Smooth” in the panel. Select “User
selected smooth…” and choose the Savitzky–Golay algorithm
and 9 points (see Note 12) (Fig. 5).

Fig. 3 Preprocessing acquisition, screenshot from PerkinElmer’s Spectrum FT-IR software

Fig. 4 Baseline correction, screenshot from PerkinElmer’s Spectrum FT-IR

software
Fourier Transform Infrared Spectroscopy as a Tool in Analysis of Proteus mirabilis… 119

Fig. 5 Smooth, screenshot from PerkinElmer’s Spectrum FT-IR software

Fig. 6 Normalization, screenshot from PerkinElmer’s Spectrum FT-IR software

Fig. 7 Derivative, screenshot from PerkinElmer’s Spectrum FT-IR software

5. Locate and click “Normalization” in the panel. Select “Auto

zero” from the “Zero point” pop-up box for ordinate limit 1
(see Note 13) (Fig. 6).
6. Locate “Derivative” and perform Savitzky–Golay differentia-
tion for the first derivative (most often used; second differen-
tiation is also common) for 15 points (Fig. 7).
7. Open “File…” and “Save as…” to save normalized first deriva-
tive spectral data from the experiment to ASCII (Fig. 8b).
120 Paulina Żarnowiec et al.

Fig. 8 IR absorbance spectra for a P. mirabilis S1959 colony in the spectral range of 4000–650 cm−1. (A)
Original absorbance spectrum; (B) first derivative

3.4 Multivariate Choose a data analysis procedure appropriate to your analytical

Analysis goal; this may be, e.g., clustering or principal component analysis
(unsupervised classification) [7]:
1. Start STATISTICA, click on “Load…,” and select your
dataset.
2. Locate and click on “Statistic” in the panel. Select
“Multivariate…” and choose “Principal component analysis
and classification analysis.”
3. Select the spectral range “1200–900 cm−1” in the variables (see
Note 14).
4. Click “OK” to perform initial computations, and then in the
“Results” dialog box set the “Number of factors” to 2 (see
Note 15) (Fig. 9).
5. Next, click “Plot case factor coordinates 2D” to display the
results of observations (cases) (Fig. 10).
The presented PCA diagram of IR spectra (polysaccharide
region 1200–900 cm−1) shows the clustering of freeze-dried cells
 Fourier Transform Infrared Spectroscopy as a Tool in Analysis of Proteus mirabilis… 121

Fig. 9 “Principal component analysis and classification analysis,” screenshot

from STATISTICA, StatSoft software

Fig. 10 Diagram generated by PCA obtained from the 1200–900 cm−1 spectral region (window 4—polysac-
charides) (A) freeze-dried bacterial mass and (B) LPS extracted from three P. mirabilis strains: S1959 (S), R45
(Z), and R110 (R)

as well as of the isolated LPSs based on polysaccharide content.

Smooth P. mirabilis S1959 (O3) are clustered away from the two
R mutants lacking the O-polysaccharide part (Ra type, strain R110)
or the core oligosaccharide (Re, strain R45). Differences inside the
R and S groups may be due to natural preparation heterogeneity
and the content of contaminants (nucleic acids, proteins, cations).
This indicates ATR FT-IR sensitivity to recorded fingerprints of
particular LPS sample preparations.
122 Paulina Żarnowiec et al.

4 Notes

1. Spectral acquisition software is normally provided by the

instrument manufacturer. Most of such programs also provide
a number of preprocessing, and sometimes more advanced,
data analysis options.
2. Throughout the experiment, it is advisable to use strictly con-
trolled bacterial culture conditions, with the same type of
nutrients. ATR FT-IR spectra of analyzed isolates should be
acquired from an overnight subculture on an agar medium rec-
ommended for a given strain exactly after 18 h of growth at
37 °C (unless the analyzed strain has different culture condi-
tions) to avoid spectral interferences due to colony age [7].
3. Even the same LPS IR spectra might differ due to the prepara-
tion method, LPSs heterogeneity, or traces of contaminants.
Highly pure LPS preparation methods were described [15].
4. It is very important that a background spectrum is recorded
for each sample! Also, a background spectrum should be taken
if atmospheric changes occur (e.g., if a door has been suddenly
opened).
5. The crystal surface must not be scratched under any circum-
stances! It is expected that some residual sample will be left.
Isopropanol will remove most of the remaining residual film
before and after each trial.
6. To acquire a background spectrum, the IR source should not
be in contact with the sample, and it should be open to the
surrounding environment.
7. The ATR crystal should be completely covered by the sample,
and the minimum sample thickness should be three to four
times the depth of penetration to ensure that there is no inter-
ference from the substrate.
8. The same pressure should be applied to all samples.
9. Intimate contact of the sample with the crystal is a very impor-
tant parameter for ATR FT-IR analysis.
10. After the spectral acquisition of each sample, the crystal should
be washed with isopropanol or 70% ethanol; the crystal should
be left to dry.
11. Baseline should be corrected to zero in order to eliminate any
differences between spectra due to shifts in baseline [3, 4].
12. Smoothing reduces instrumental noise and increases informa-
tion content in the spectrum using one of the following algo-
rithms: Savitzky–Golay de-noising [7, 16], PCA noise
reduction, or minimum noise fraction [17].
 Fourier Transform Infrared Spectroscopy as a Tool in Analysis of Proteus mirabilis… 123

13. Data can be normalized using MinMax [7] or vector normal-

ization [18]. Scale the variables: this could be done by stan-
dardization (normalization of variables to zero mean and unit
s.d.) or by normalization to a 0–1 range.
14. Wavenumbers from 1200 to 900 cm−1 are characteristic for the
polysaccharide region in the infrared spectrum and are associ-
ated with IR frequencies of O-specific (lipopolysaccharide)
antigens from the bacterial cell envelope [7, 19–21].
15. As a result, “Quality of representation” should be computed as
more than 80%.

Acknowledgments

This study was supported by grant from the National Science

Center, Poland granted on the basis of decision number DEC-
2012/07/N/NZ6/04118 and grant no. 612 427 from the Jan
Kochanowski University. Some of the experiments were run on
apparatus purchased with European Union grant 307 under 2.2
Innovation Industry.

References 7. Zarnowiec P, Mizera A, Chrapek M et al

(2016) Chemometric analysis of attenuated
1. Brunn GJ, Platt JL (2006) The etiology of sep- total reflectance infrared spectra of proteus
sis: turned inside out. Trends Mol Med mirabilis strains with defined structures of
12:10–16 LPS. Innate Immun 22:325–335
2. Naumann D (2000) Infrared spectroscopy in 8. Westphal O, Jann K, Himmelspach K (1983)
microbiology. In: Meyers RA (ed) Encyclopedia Chemistry and immunochemistry of bacterial
of analytical chemistry: applications, theory, and lipopolysaccharides as cell wall antigens and
instrumentation. John Wiley & Sons Ltd., endotoxins. Prog Allergy 33:9–39
Chichester, pp 102–131 9. Galanos C, Luderitc O (1993) Isolation and
3. Żarnowiec P, Lechowicz Ł, Czerwonka G, precipitation of R-form of lipopolysaccharides:
Kaca W (2015) Fourier Transform Infrared further applications of the PCP method. In:
Spectroscopy (FTIR) as a tool for the identifi- BeMiller J, Whistler R, Shaw D (eds) Methods
cation and differentiation of pathogenic bacte- carbohydr. Chem. John Wiley & Sons, Inc.,
ria. Curr Med Chem 14:1710–1718 pp 11–18
4. Davis R, Mauer L (2010) Fourier transform 10. Gmeiner J (1975) The isolation of two differ-
infrared (FT-IR) spectroscopy: a rapid tool for ent lipopolysaccharide fractions from various
detection and analysis of foodborne pathogenic Proteus mirabilis strains. Eur J Biochem
bacteria. Curr Res Technol Educ Top Appl 626:621–626
Microbiol Microb Biotechnol II:1582–1594 11. Sousa C, Silva L, Grosso F et al (2014)
5. Brandenburg K, Seydel U (1988) Orientation Development of a FTIR-ATR based model for
measurements on membrane systems made typing clinically relevant Acinetobacter bau-
from lipopolysaccharides and free lipid A by mannii clones belonging to ST98, ST103,
FT-IR spectroscopy. Eur Biophys J 16:83–94 ST208 and ST218. J Photochem Photobiol B
6. Seydel U, Koch C, Brandenburg K (1997) Biol 133:108–114
Investigations into lysozyme-endotoxin 12. Winder CL, Goodacre R (2004) Comparison
interactions with FT-IR spectroscopy and of diffuse-reflectance absorbance and attenu-
X-ray diffraction. Spectrosc Biol Mol Modem ated total reflectance FT-IR for the discrimina-
Trends:363–364 tion of bacteria. Analyst 129:1118–1122
124 Paulina Żarnowiec et al.

13. Nagib S, Rau J, Sammra O et al (2014) 18. Bosch A, Miñán A, Vescina C et al (2008)
Identification of Trueperella pyogenes isolated Fourier transform infrared spectroscopy for
from bovine mastitis by fourier transform infra- rapid identification of nonfermenting gram-­
red spectroscopy. PLoS One:9–e104654 negative bacteria isolated from sputum samples
14. Muhamadali H, Weaver D, Subaihi A et al from cystic fibrosis patients. J Clin Microbiol
(2015) Chicken, beams, and Campylobacter: 46:2535–2546
rapid differentiation of foodborne bacteria via 19. Baldauf NA, Rodriguez-Romo LA, Männig A
vibrational spectroscopy and MALDI-mass et al (2007) Effect of selective growth media on
spectrometry. Analyst 141:111–122 the differentiation of Salmonella enterica serovars
15. Pupo E (2011) Isolation of smooth-type lipo- by Fourier-transform mid-Infrared spectroscopy.
polysaccharides to electrophoretic homogene- J Microbiol Methods 68:106–114
ity. In: Holst O (ed) Microbial toxins: methods 20. Al-Qadiri HM, Lin M, Cavinato AG, Rasco
and protocols, Methods in molecular biology, BA (2006) Fourier transform infrared spec-
vol 739. Springer New York, Totowa, NJ, pp troscopy, detection and identification of
101–112 Escherichia coli O157:H7 and alicyclobacillus
16. Brandes Ammann A, Brandl H (2011) strains in apple juice. Int J Food Microbiol
Detection and differentiation of bacterial 111:73–80
spores in a mineral matrix by Fourier transform 21. Kim S, Reuhs BL, Mauer LJ (2005) Use of
infrared spectroscopy (FTIR) and chemometri- Fourier transform infrared spectra of crude
cal data treatment. BMC Biophys 4:14 bacterial lipopolysaccharides and chemo-
17. Baker MJ, Trevisan J, Bassan P et al (2014) Using metrics for differentiation of Salmonella
Fourier transform IR spectroscopy to analyze enterica serotypes. J Appl Microbiol 99:
biological materials. Nat Protoc 9:1771–1791 411–417
 Chapter 12

Laser Interferometry Method as a Novel Tool

in Endotoxins Research
Michał Arabski and Sławomir Wąsik

Abstract
Optical properties of chemical substances are widely used at present for assays thereof in a variety of
scientific disciplines. One of the measurement techniques applied in physical sciences, with a potential for
novel applications in biology, is laser interferometry. This method enables to record the diffusion proper-
ties of chemical substances. Here we describe the novel application of laser interferometry in chitosan
interactions with lipopolysaccharide by detection of colistin diffusion. The proposed model could be used
in simple measurements of polymer interactions with endotoxins and/or biological active compounds, like
antibiotics.

Key words Laser interferometry, Lipopolysaccharide, Chitosan

1 Introduction

Laser interferometry, based on the phenomenon of wave interference,

enables quantitative substance assays by means of measurement of
the difference between light refractive indexes for the studied and
control substances. Properties of laser-generated radiation, i.e.,
low beam divergence, narrow spectral band, and high degree of
coherence, positively influence the assay sensitivity.
Figure 1a depicts the interferometric setup used in our studies,
which is basically suited for recording the refractive index modifi-
cation inside the diffusion cell (Fig. 1b). A beam of monochro-
matic light (λ = 632.8 nm) is splited into two beams, one of which
passes through the diffusion cell. The other beam is directed by the
mirror onto a compensator plate. The interferograms, which
appear due to the interference of the two beams are recorded by a
CCD camera and presented on a graphic screen (inserts in Fig. 3).
A computer image-processing system, completed with a dedicated
software, enables a mathematical analysis of interferograms shown
on the system screen. The changes in solution concentration in
the system involve variations Δn(x,t) of their refractive index.

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_12, © Springer Science+Business Media LLC 2017

125
126 Michał Arabski and Sławomir Wąsik

Fig. 1 Setup of the laser interferometry system used in this study (a) with diffusion cell (b)

The correlation coefficient a between concentration and refractive

index of water solutions is refractometrically determined. The
quantity of C(x,t) is calculated according to the formula

C ( x,t ) = C0 + aDn ( x,t ) (1)

in which C0 is the initial substance concentration.

Recording the interferometric images with a given time step
∆t, one can reconstruct the concentration profiles for different
times. Moreover, the transport parameters, such as substance
quantity N, substance flux Js, the membrane permeability coeffi-
cient ωm, and the substance diffusion coefficient D after time t and
in successive time intervals ∆t, can be determined. These parame-
ters are presented in Table 1.
The possibility of real-time diffusion analysis between unmixed
systems and an analysis based on the change of the solution’s
refractive index favors the application of laser interferometry as
analysis of biomolecular interactions based on the phenomenon of
diffusion, for example, lipopolysaccharides (LPS) with antibiotics
 Laser Interferometry Method as a Novel Tool in Endotoxins Research 127

Table 1
Transport parameters which can be determined by laser interferometry

Formula

Transport parameter After time t In the time interval ∆t of number i

d d¢ d
The amount of the N = S òC1 ( x,t ) dx N = S òC ( x,t + iDt ) dx - S òC ( x,t + ( t - 1) Dt ) dx
transported 0 0 0
substance
d
The substance flux d¢ d
òC ( x,t ) dx òC ( x,t + iDt ) dx - òC ( x,t + ( i - 1) Dt ) dx
Js = 0
JS = 0 0
t Dt
d¢ d¢ d
Membrane
permeability òC ( x,t ) dx òC ( x,iDt ) dx - òC ( x, ( i - 1) Dt ) dx
1

wm = 0
wm = 0 0
coefficient RT éëC2 ( x = 0,t ) - C1 ( x = 0,t ) ùû t RT éëC2 ( x = 0,iDt ) - C1 ( x = 0,iDt ) ùû Dt

Diffusion coefficient
D=
d (t )
2
d 2 ( iDt ) - d 2 ( ( i - 1) Dt )
D=
4t ( erfc -1 ( k ) )
2
4Dt ( erfc -1 ( k ) )
2

C1(x=0,t) and C2(x=0,t) are the concentrations of the solutions on the membrane surfaces at time t, erfc is complemen-
tary error function defined as
¥
2
erfc ( k ) = òe
-h 2
dh
Õ k
(5)

or polymers, analysis of antibiotics diffusion trough bacterial biofilm,

as well as biofilm degradation by bacteriophages [1–6].
In this study, the methodology focuses on laser interferometry
measurements of chitosan interaction with LPS of Pseudomonas
aeruginosa O10 in the presence of colistin.

2 Materials

Prepare all solutions using ultrapure water at room temperature by

mixing for 2 h.
1. Laser interferometry system as depicted in Fig. 1.
2. Dissolve 0.5 g of chitosan in 25 mL of aqueous acetic acid (2%
v/v, pH 3.0) and leave overnight at room temperature with
continuous mechanical stirring to obtain a 2% (w/v) solution
(see Note 1).
3. Dissolve 1 mg of LPS (in this study from P. aeruginosa O10)
in 10 mL of water by sonication (20 kHz, 40%, 30 s) and
mechanical stirring to obtain the LPS solution at 100 μg/mL.
128 Michał Arabski and Sławomir Wąsik

3 Methods

3.1 Chitosan 1. Take two glass beakers (each with capacity 250 mL) and pre-
Hydrogel pare chitosan hydrogel particles in each of them: using a
syringe, inject 5 mL of 2% chitosan solution in aqueous acetic
acid drop by drop (10 μL) to 50 mL of 2% NaOH water with
continuous mechanical stirring, and leave for 3 h at room tem-
perature (25 °C).
2. Centrifuge (15 min, 900 × g) the hydrogel particles at room
temperature (see Note 2).
3. Wash the chitosan hydrogen particle precipitates six times in
100 mL of water with centrifugation (15 min, 900 × g) until a
neutral pH is obtained.

3.2 Incubation 1. Incubate the washed chitosan hydrogel particle precipitates

of Chitosan Hydrogel with 100 μg/mL LPS in 10 mL of water for 18 h at 4 °C.
with LPS and/or 2. Wash the chitosan hydrogel particles three times in 20 mL of
Colistin water with centrifugation (15 min, 2100 × g).
3. Dissolve 2 mg of colistin sodium methanesulfonate in 1 mL of
water by mechanical stirring.
4. Incubate the washed precipitates of chitosan hydrogel parti-
cles with 2 mg of colistin in 1 mL of water for 18 h at room
temperature.

3.3 Preparation 1. Mix 1 mL of colistin/hydrogel suspension (pre-incubated

of Agarose Gel with LPS or free from LPS) with 1 mL of 1.5% low melting
with Colistin point agarose gel at 40 °C by pipetting up and down, and add
it to the bottom of the cuvette as element of the diffusion cell
(Fig. 2; cuvette I).

Fig. 2 Images of chitosan hydrogel taken using optical (a) and fluorescence (b) microscopy (100×
magnification)
 Laser Interferometry Method as a Novel Tool in Endotoxins Research 129

2. As control for laser interferometry analysis, prepare a third

mixture 1 mL of colistin (2 mg/mL) with 1 mL of 1.5% low
melting point agarose gel at 40 °C (see Note 3).
3. After 30 min (gelling process at room temperature), the sam-
ples are ready for analysis of free colistin diffusion from LMP
agarose (in cuvette I) to water (in cuvette II, Fig. 2) by laser
interferometry (see Note 4).

3.4 Laser 1. The prepared samples are placed in one of the arms of the
Interferometry interferometer (diffusion cell), and the interferometric images
(time steps ∆t = 120 s) are recorded. Based on a computer
analysis of the recorded interferograms, the spatio-temporal
concentration distribution C(x,t) (i.e., concentration profile)
of colistin is determined.
2. The selected transport parameters of colistin diffused to the
water are calculated based on the obtained concentration pro-
files according the formulas presented in Table 1 (see Note 5).
Figures 3, 4, 5, 6, and 7 show the diffusion parameters calcu-
lated for colistin released from agarose after pre-incubation
with chitosan or chitosan-LPS complex. Generally, a higher
amount of colistin released from agarose with an immobilized
chitosan hydrogel in comparison to chitosan-LPS complexes
indicates that LPS interacts with chitosan.

Fig. 3 The concentration profiles of colistin released from 0.75% agarose gel after 10, 30, and 60 min, calcu-
lated on the base of interferograms using laser interferometry. Near the gel-water interface, the interferometric
fringes are bent from a straight line, which means that the colistin was released from the gel to the water
phase. The region of fringes curvature expands with time indicating that the thickness of concentration bound-
ary layers (CBL) and the amount of released substance increase with time
130 Michał Arabski and Sławomir Wąsik

Fig. 4 The amount (mol) of colistin alone (●), pre-incubated with chitosan (○) or LPS-chitosan complexes
( ), released from 0.75% agarose measured by laser interferometry. The higher values (mol) of released colis-
tin were observed after pre-incubation with chitosan (2.01 × 10−9) than with chitosan-LPS complexes
(7.29 × 10−10) after 60 min

Fig. 5 The amount (mol) of colistin alone (●), pre-incubated with chitosan (○) or LPS-chitosan complexes ( )
released in successive time intervals from 0.75% agarose measured by laser interferometry system
Fig. 6 Time dependency of diffusive flux of colistin alone (●), pre-incubated with chitosan (○) or LPS-chitosan
complexes ( ), released from 0.75% agarose measured by laser interferometry. The curves correspond to the
curves shown in Fig. 3 and illustrate the kinetics of colistin transport form gel into the water phase. These
curves are typical for the diffusion analyzed at the macroscopic level. In the initial stage of diffusion, we
observe a rapid decrease of colistin flux due to increased concentration polarization

Fig. 7 Time characteristics of concentration field evolution distances x0 = 0, 0.5, 1, and 2 mm of the gel-water
interface. The highest colistin concentration value occurs at the gel-water interface (x0 = 0 mm). A concentra-
tion near to zero indicates to be outside the CBL region; for example, the diffusing colistin reached the point
x0 = 2 mm after time t > 20 min. At the same time and at points x0 = 0, 0,5 and 1 mm, the concentrations were
4.65 × 10−4, 2.19 × 10−4, and 1.85 × 10−4 mol/dm3, respectively. From the plots results also that at points situ-
ated near the gel-water interface, the concentration evolves nonlinearly, whereas at points situated at longer
distances from this interface, the concentration changes practically proportionally to the time
132 Michał Arabski and Sławomir Wąsik

4 Notes

1. The chitosan solution can be stored at room temperature for

1 month.
2. Figure 2 shows chitosan hydrogel particles investigated by
optical and fluorescence microscopy. Hydrogel was stained by
100 μg/mL fluorescein for 3 h at room temperature and six
times washed in water.
3. A major advantage of the laser interferometry technique is the
possibility of analyzing chemical compounds which form
quasi-­real solutions, such as amphiphilic colistin.
4. In the system presented here, only colistin is able to diffuse
from agarose to the water phase. Chitosan particles or
chitosan-­LPS complexes are stabilized in agarose matrix.
5. While planning the interferometric experiment, it should be
taken into account to study the transport of only one substance,
due to the lack of identification of a substance in the multicom-
ponent solutions. Moreover, using laser interferometry solutions
can be studied with sufficient optical transparency.

References
1. Arabski M, Wąsik S, Zych M et al (2013) Analysis lin and colistin transfer through cellulose bio-
of ciprofloxacin and gentamicin diffusion in membrane in the presence of Proteus vulgaris
Proteus mirabilis O18 biofilm by laser interferom- O25 lipopolysaccharide. J Membr Sci 299:
etry method. Acta Biochim Pol 60:707–711 268–275
2. Danis-Wlodarczyk K, Olszak T, Arabski M et al 5. Arabski M, Wąsik S, Dworecki K et al (2009)
(2015) Characterization of the newly isolated Laser interferometric and cultivation methods
lytic bacteriophages KTN6 and KT28 and their for measurement of colistin/ampicillin and
efficacy against Pseudomonas aeruginosa biofilm. saponin interactions with smooth and rough of
PLoS One 10(5):e0127603 Proteus mirabilis lipopolysaccharides and cells.
3. Danis-Wlodarczyk K, Vandenheuvel D, Jang J Microbiol Methods 77:179–183
HB et al (2016) A proposed integrated approach 6. Arabski M, Davydova VN, Wąsik S et al (2009)
for the preclinical evaluation of phage therapy in Binding and biological properties of lipopoly-
Pseudomonas infections. Sci Rep 6:28115 saccharide Proteus vulgaris O25 (48/57)-chito-
4. Arabski M, Wąsik S, Dworecki K et al (2007) san complexes. Carbohydr Polym 78:
Laser interferometric determination of ampicil- 481–487
 Chapter 13

Endotoxin Entrapment on Glass via C-18 Self-Assembled

Monolayers and Rapid Detection Using Drug-Nanoparticle
Bioconjugate Probes
Prasanta Kalita*, Anshuman Dasgupta*, and Shalini Gupta

Abstract
Bloodstream bacterial infections are known to illicit a systemic immune response that can lead to multior-
gan failure and septic shock. The current endotoxin identification techniques in serum are expensive and
elaborate requiring bulky benchtop instrumentation. We demonstrate a new route for endotoxin detection
in which lipopolysaccharides (LPS) in solution are entrapped using C-18 silane-functionalized glass slides
and tagged with polymyxin B sulfate (PMB) drug-conjugated gold nanoparticles. The signal from the
particles is further amplified via the silver reduction approach to yield concentration-dependent colorimet-
ric spots visible to the bare eye. The method is rapid, reliable, and cost-effective and fulfills an urgent
unmet need in the healthcare industry for early septicemia diagnosis.

Key words Lipopolysaccharide (LPS), Endotoxin, Gold nanoparticle (GNP), Polymyxin B sulfate
(PMB), Silver enhancement, C-18 silane

1 Introduction

Endotoxin, also referred to as LPS or pyrogen, is a conserved com-

ponent of most Gram-negative bacterial cell envelopes [1].
Structurally, endotoxin can be segregated into three distinct com-
ponents: (1) the repeating glycan polymer units known as O-antigen
that are responsible for molecular immunogenicity, (2) a core oligo-
saccharide section, and (3) the (toxic) lipid-A region consisting of
disaccharides and long fatty acid chains [2, 3]. Circulating endotox-
ins in the bloodstream are known to trigger an immunological cas-
cade through the release of pro- or anti-­inflammatory cytokines
resulting in several pathological disorders such as hypotension,
intravascular coagulation, multiple-organ failure, severe sepsis, and
septic shock depending on the body’s degree of imbalance [4, 5].

Prasanta Kalita and Anshuman Dasgupta these author contributed equally to this chapter.
*

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_13, © Springer Science+Business Media LLC 2017

133
134 Prasanta Kalita et al.

In the other sectors too such as pharmaceuticals, agriculture, tex-

tile, defense, food, and water industry, the omnipresence of endo-
toxins as contaminants raises grave concerns [6–11]. Various
endotoxin detection methods based on enzyme-linked immuno-
sorbent assay (ELISA), clotting, absorbance, and chemilumines-
cence are commercially available to address the concerns of these
sectors that demand rapid and affordable endotoxin monitoring
[12–15]. Although some of these methods are now widely
accepted, they come with their own drawbacks. For instance, the
LAL test cannot be used to assay endotoxin in blood samples due
to the presence of circulating inhibitors that prevent the coagula-
tion reaction. Similarly, the ELISA detection kits are costly and
require trained personnel to handle the microplate reader systems.
Alternate approaches are therefore required that are simple, cost-
effective and allow assaying complex milieu of biological analytes
with high sensitivity and specificity.
Self-assembled monolayers (SAMs) offer a facile way of entrap-
ping biomolecules or covalently attaching targeting moieties for their
subsequent capture and are very popular in the field of bioassays and
biosensors. For endotoxin detection specifically, thiolated organic
molecules have been self-assembled onto gold surfaces for attaching
different targeting ligands such as PMB and divalent cations like Cu2+
in order to identify trace amounts of endotoxin [16, 17]. In another
ingenious example demonstrated by Peri and coworkers, LPS mole-
cules were successfully self-assembled on the surface of hydrophobi-
cally functionalized magnetic nanoparticles in order to stimulate
TLR4-dependent cell activation in vitro [18]. Our group has also
recently reported a simple way to entrap and detect LPS molecules
present in water and blood samples using silanized glass substrates
[19]. The length of the alkane chains in the silane molecules was kept
similar to the carbon atoms present in the fatty acid chains in the LPS
(typically 16–18) in order to entropically drive the interdigitization
and oriented binding of LPS on the hydrophobic surface, an approach
motivated by the chromatographic separation techniques used for
endotoxin purification of biological samples [20].
To detect the bound endotoxin on the glass surface, the mole-
cules were tagged with gold nanoparticles (GNPs) coated with PMB
drug (clinically known as colistin) known to have a high affinity for
LPS molecules [21–23]. The use of drugs as targeting ligands is fast
gaining popularity as it is superior to traditional moieties such as anti-
bodies, peptides, or aptamers in terms of cost, thermal stability, resis-
tance to denaturation, and ease of conjugation. The GNP conjugate
probes were finally silver enhanced in order to amplify the signal mul-
tifold and get visual signal readout by the naked eye in the clinically
relevant concentration range. A schematic of the assay is shown in
Fig. 1, and a step-by-step experimental protocol to carry this out is
discussed below. As one reads on, the simplicity of the assay and its
suitability for clinical diagnostics will become apparent.
 Endotoxin Entrapment and Detection 135

Fig. 1 A schematic of the procedure used for the silver enhanced endotoxin bio-
assay using PMB-conjugated gold nanoparticles

2 Materials

All the items were purchased from Sigma-Aldrich and stored at

room temperature unless otherwise stated. Refrigerated items were
brought to room temperature before use. All the solutions were
prepared using only pyrogen-free water (commercially available)
and analytical grade reagents. For washing and other purposes,
only ultrapure deionized (DI) water of high resistivity (~18 MΩ cm)
was used.

2.1 Glass 1. Glass activation solution: Prepare 60 mL of piranha solution

Functionalization by slowly pouring H2O2 over concentrated H2SO4 in 3:1 v/v
Components in a glass coplin jar. Alternately, use a basic piranha solution
comprising of 1:1:5 v/v of H2O2:NH4OH:H2O.
2. 10 mM silane solution: Take 58 mL of dry toluene (water con-
tent ≤0.001%) in a glass coplin jar and add 248 μL of
trichloro(octadecyl)silane solution to it while maintaining a
dry environment using either a glove box or by purging nitro-
gen through a nozzle jet just above the working area in order
to minimize any water vapor content (see Note 1).

2.2 Sample Analyte 1. Serum: Take approx. 5 mL of whole blood in a pyrogen-free

Components vial, and allow it to stand for 20 min at room temperature in
order to coagulate. Centrifuge the coagulated blood at
10,380 rcf for 20 min. Collect the supernatant containing the
serum, and store it at −20 °C in 95 μL aliquots in pyrogen-free
vials. Endotoxin-free frozen plasma/serum may alternately be
purchased directly from a vendor.
136 Prasanta Kalita et al.

2. LPS samples: Weigh 1 mg of lyophilized LPS powder

(Escherichia coli serotype 055:B5) in depyrogenated glass
vials, and add it to 1 mL of pyrogen-free water (Hyglos
GmBH, Germany). Vortex the solution vigorously for at
least 30 min and sonicate for another 5–10 min. Prepare
multiple successive dilutions of this 1 mg/mL stock solu-
tion using pyrogen-free water until you reach 1 fg/mL
concentration. Store at 4 °C. To prepare LPS samples in
serum, bring the spiked aqueous LPS solutions to room
temperature. Collect 5 μL from each of these samples, and
add them separately to 95 μL serum aliquots to finally
obtain 100 μL volume samples with varying LPS concen-
trations (see Note 2).

2.3 Detection Probe 1. 40 mM HEPES buffer, pH 7.4: Weigh about 4.766 g of

Components HEPES salt, and add to 500 mL of ultrapure DI water kept in
an autoclaved glass bottle. Shake the solution until all the salt
disappears. Adjust the pH to 7.4 using 25 M NaOH solution.
Store at 4 °C (see Note 3).
2. 9.3 mM dithiolalkane aromatic PEG-6 hydrazide (DTH)
linker solution: Bring 50 g of DTH powder (Sensopath, USA)
stored at −20 °C to room temperature, and add 2 mL of 99.9%
(v/v) ethanol to obtain a 35.3 mM clear solution. Vortex vig-
orously for 10 min. Dilute a portion of this stock to 9.3 mM
concentration using 99.9 v/v % ethanol, and prepare 10 μL
aliquots in PCR tubes. Store the aliquots as well as the remain-
ing stock at −20 °C.
3. Glutaraldehyde (GLA) solution: Prepare 10 mL of 2 w/v %
GLA solution by adding 400 μL of 25 w/v % GLA stock to
9600 μL of 40 mM HEPES buffer at pH 7.4.
4. Polymyxin B sulfate (PMB) solution: Prepare 1 mL of 15.3 μM
PMB solution by taking 1 μL of 15.3 mM stock solution and
adding it to 999 μL of ultrapure DI water in a 1.5 mL
Eppendorf tube. Homogenize the mixture using a vortexer.
Store at 4 °C.
5. GNP suspension: Take approx. 16 nm citrate-stabilized
GNPs suspended in ultrapure DI water, and adjust their
concentration to 2 nM either by dilution or by centrifuga-
tion, and suspend the pellet in DI water. The GNPs may be
purchased or synthesized by any of the known methods in
the literature [24, 25]. Final volume of the suspension
should be at least 10 mL. Store at 4 °C. The UV-visible
spectrum of this suspension should peak around 522 nm
(Fig. 2b). The size distributions of the particles are shown
in Fig. 2a and c.
 Endotoxin Entrapment and Detection 137

Fig. 2 (a) Transmission electron microscopy image taken using Tecnai G2 (FEI) shows spherical particles
approximately 16 nm in size. (b) UV-visible spectra and (c) DLS measurements of GNPs before and after
chemical modification with DTH, DTH-GLA and DTH-GLA-PMB, respectively. A gradual right shift in peaks in
both spectra indicates successful surface functionalization at each modification step. (d) Actual bioassay spots
obtained after performing experiments in quadruplicate at each concentration. Images taken in ambient light
using Nikon COOLPIX 310 CCD camera

3 Methods

All experiments are performed at room temperature unless other-

wise stated. The described protocol for drug-modified GNP sus-
pension preparation is sufficient to carry out multiple experiments.
The excess stock can be stored at 4 °C for at least 3 months.

3.1 PMB Conjugation 1. Take 10 mL of 2 nM citrate-capped GNP suspension in a

to GNPs 15 mL conical glass flask. Add 20 μL of 9.3 mM DTH solution
to this suspension, and keep the mixture at 25 °C inside a
shaker incubator at 200 rpm for 12 h.
2. Transfer the ligand-exchanged GNPs to a 15 mL falcon tube
marked “1” and centrifuge it at 1660 rcf for 15 min. Transfer
the supernatant to another falcon tube marked “2” and centri-
fuge it again at 5080 rcf for 15 min. Do not discard the pellet.
Repeat the process at higher centrifugation speeds of 10,380
and 17,540 rcf. Collect all the pellets from the tubes marked
1–4, and mix them together in a single falcon tube. Re-suspend
this final pellet in 40 mM HEPES buffer, pH 7.4 to a final
volume of 10 mL. The UV-visible spectrum of this DTH-
modified GNP suspension should give an approx. red-shift of
4 nm with respect to bare GNPs (Fig. 2b). The hydrodynamic
size of the particles also increases by ~16 nm as measured by
dynamic light scattering (DLS) (Fig. 2c).
138 Prasanta Kalita et al.

3. Add 9 mL of 2% (w/v) GLA solution to an equal volume of

the above DTH-modified GNP suspension in a clean conical
glass flask. Keep the flask inside a shaker incubator for 8 h at
25 °C (see Note 4).
4. Repeat the same centrifugation steps involved during the
removal of excess DTH from the GNP suspension. Finally sus-
pend the pellet in 16 mL of 40 mM HEPES buffer, pH 7.4.
The UV-visible spectrum of the GLA-modified particles should
give a peak around 527.5 nm that is redshifted by 1.5 nm peak
with respect to DTH-modified GNPs (Fig. 2b). The hydrody-
namic size increases to 30 nm (Fig. 2c).
5. Transfer 16 mL of the GLA-modified GNP suspension
(~1 nM) to a 50 mL glass conical flask. Add to this 10 μL of
15.3 μM PMB solution. Keep the solution in a shaker incuba-
tor at 200 rpm for 12 h at 25 °C.
6. Repeat the same centrifugation steps involved during the
removal of excess DTH from the GNP suspension. Collect
the final pellet and resuspend it in 14 mL of 40 mM HEPES
buffer, pH 7.4 to bring the final concentration of the sus-
pension to 1 nM. The UV-visible spectrum of PMB-modified
particles should give a 1 nm peak shift with respect to GLA-
modified GNPs (Fig. 2b). The final size of the particle
becomes approx. 38 nm at this point (Fig. 2c). Store the
particles at 4 °C (see Note 5).

3.2 Glass 1. Clean a standard glass slide using a soft fiberless tissue (e.g.,
Silanization Kimwipes) soaked in acetone.
2. Soak the glass slide overnight in the acidic piranha solution to
remove organic content and to activate the surface with
hydroxyl (–OH) groups for subsequent reaction. Wash the
slide with ample DI water and dry it under a purge of nitrogen
(see Note 6).
3. Dip the dry glass slide in freshly prepared silane solution kept
inside a glass coplin jar. Place the lid on the coplin jar and keep
it inside a vacuum desiccator for at least 12 h. The desiccator
should be pre-filled with nitrogen gas.
4. Take out the slide and dip it in another glass coplin jar filled
only with dry toluene. Sonicate for 10 min to remove the
excess unbound silane molecules. Repeat this step twice.
5. Take out the glass slide and keep it for drying in an oven at 70 °C
for 1 h, and then store inside a desiccator until further use.

3.3 LPS 1. Wash the silanized glass slide with DI water just prior to begin-
Immobilization ning the LPS immobilization experiments (see Note 7).
on Glass Substrates 2. Using a micropipette, adjacently drop cast two 10 μL droplets
of serum/water on the surface of the silanized glass slide – one
 Endotoxin Entrapment and Detection 139

containing and one without LPS (negative control). The dis-

tance between the droplets should be maintained at least 1 cm
such that their contents do not mix. Keep the slide inside a
pre-­humidified water bath at 37 °C for 30 min (see Note 8).
3. Take the slide out of the water bath, and mark the area on
which the LPS is immobilized on the bottom side of the glass
slide using a permanent marker. Gently wash the area with DI
water using a 1 mL micropipette (see Note 9).
4. Allow the slide to dry for 4–5 min under ambient conditions.

3.4 GNP Tagging 1. After LPS immobilization, add 100 μL of the PMB-modified
and Silver GNP suspension above the LPS-immobilized area (marked
Enhancement earlier). Keep the slide inside the pre-humidified water bath at
for Endotoxin 37 °C and incubate for 40 min.
Detection 2. Take the slide out and rinse it with DI water several times using
a micropipette. Allow the slide to dry for 4–5 min under ambi-
ent conditions.
3. From the silver enhancement kit, take out 50 μL each of solu-
tions A and B in Eppendorf cups. Mix and vortex the two solu-
tions together. Apply the entire mixture volume immediately
to the LPS-immobilized area, and wait for 30–50 s for the
spots to develop. Immediately wash the slide with copious
amounts of ultrapure DI water (see Note 10).
4. A black spot should be visible to the naked eye where the
LPS is immobilized. The rest of the glass slide remains clear
(see Note 11) (Fig. 2d).

4 Notes

1. Even trace amounts of water is sufficient to trigger self-polym-

erization of silane molecules leaving a whitish by product and
incomplete functionalization of glass slides hampering the bio-
assay results.
2. Aqueous LPS solution when stored at 4 °C can be used up to
4 weeks without significant loss in activity. LPS samples pre-
pared in serum should be used fresh (within 3 days) since
serum proteins tend to get denatured and are liable to con-
tamination. Avoid using plastic containers (e.g., polypropyl-
ene, polystyrene, polymethyl mythacrylate, etc.) as LPS tends
to stick to them irreversibly.
3. HEPES is a zwitterionic buffer with low ionic strength. It is
more suitable for doing GNP experiments than other common
buffers such as phosphate buffer saline (PBS) or borate in
which the GNPs tend to aggregate.
140 Prasanta Kalita et al.

4. In the GLA activation step, 360 μL of 25% (w/v) GLA stock

can also be added directly to 18 mL of 1 nM of GNP suspen-
sion. Use of six orders of magnitude higher molar concentra-
tion of GLA (than GNPs) is sufficiently high for restricting
particle aggregation due to interparticle cross-linking. Since
GNPs tend to stick to plastic surfaces upon conjugation with
GLA, all vials or tubes used should be made out of glass to
avoid particle loss.
5. The concentration of the GNP suspension may be calculated
from its UV spectrum using the Beer-Lambert’s law, A = ɛCL,
where A is absorbance, ɛ is particle-size-dependent extinction
coefficient, C is sample concentration, and L is path length of
the cuvette through which light travels. For 16 nm citrate-­
capped GNPs, ɛ is reported as 5.26 × 108 M−1 cm−1 [26]. This
ɛ is assumed constant for all the ligand-modified GNPs in our
study.
6. Base piranha solution is a mild reagent for –OH activation. If
this is used, dip the glass slides in the solution at 70 °C for at
least an hour.
7. A silanized glass slide is highly hydrophobic. The success of the
silanization procedure may be checked by putting a drop of
water on the functionalized area; the water droplet should
bead up and roll off without sticking to the underlying
surface.
8. High humidity (>80%) aids LPS entrapment. Aqueous LPS
solutions should always be vortexed vigorously prior to any
experiment to break aggregates. Do not sonicate the serum
samples.
9. The successful immobilization of LPS may be quickly con-
firmed by adding a drop of water to the functionalized area.
When the slide is tilted, the droplet should not fall down since
LPS immobilization makes the hydrophobic surface a hydro-
philic one.
10. The silver enhancement mixture must be prepared fresh. This
mixture is also sensitive to light, temperature, and the presence
of salts.
11. The spot tends to gain more contrast once the slide is washed
and dried.
Special notes on LPS: LPS is a pyrogenic material that requires
strict safety guidelines to be followed inside the laboratory [27].
Experiments should only be performed inside a BSL-I class bio-
safety cabinet or fume hood using designated equipment disallow-
ing contamination. All LPS-contaminated solutions, vials, and
pipette tips should be discarded after dipping in 30% (v/v) sodium
hypochlorite solution for 24 h at room temperature. Other utensils
should be autoclaved for 1 h and discarded in disposable waste
 Endotoxin Entrapment and Detection 141

bags in a designated place. All glassware should be depyrogeneted

by heating at 250 °C for 4 h. All the personnel involved in the
experiments must be trained and should wear gloves, facial masks,
and lab coats at all times. Hands must be washed with soap after
performing the experiments.
LPS occurs naturally in the environment and does not harm
healthy individuals but can exacerbate asthmatic symptoms in sus-
ceptible individuals. In such an event, visit a doctor immediately.
In case of spills, wearing protective clothing and gloves, the area
should be wiped up with an absorbent material and decontami-
nated with a freshly made bleach solution. In case of undue expo-
sure, report to a doctor immediately.

Acknowledgment

This work was supported by the grant received from DBT-BIRAC

(BIG-II scheme) and the fellowship received from CSIR, New
Delhi.

References
1. Pearson FC, Dubczak J, Weary M et al (1985) composting facilities. Int J Hyg Environ Health
Detection of endotoxin in the plasma of 214:376–383
patients with gram-negative bacterial sepsis by 10. Thorne PS, Perry SS, Saito R et al (2010)
the Limulus amoebocyte lysate assay. J Clin Evaluation of the limulus amebocyte lysate and
Microbiol 21:865–868 recombinant factor C assays for assessment of
2. Ganesh V, Bodewits K, Bartholdson SJ et al airborne endotoxin. Appl Environ Microbiol
(2009) Effective binding and sensing of lipo- 76:4988–4995
polysaccharide: combining complementary 11. Todi S, Chatterjee S, Sahu S et al (2010)
pattern recognition receptors. Angew Chem Epidemiology of severe sepsis in India: an update.
Int Ed Engl 48:356–360 Paper presented at the 30th international sympo-
3. Caroff M, Karibian D (2003) Structure of bac- sium on intensive care and emergency medicine,
terial lipopolysaccharides. Carbohydr Res Brussels, Belgium, 9–12 March 2010
338:2431–2447 12. Foster D, Derzko A, Romaschin A (2004) A
4. Raetz CR, Whitfield C (2002) Lipopolysaccharide novel method for the rapid detection of human
endotoxins. Annu Rev Biochem 71:635–700 endotoxaemia. Clin Lab Int 4:10–12
5. David SA, Silverstein R, Amura CR et al (1999) 13. Bates DW, Parsonnet J, Ketchum PA et al
Lipopolyamines: novel antiendotoxin com- (1998) Limulus amebocyte lysate assay for
pounds that reduce mortality in experimental detection of endotoxin in patients with sepsis
sepsis caused by gram-negative bacteria. syndrome. Clin Infect Dis 27:582–591
Antimicrob Agents Chemother 43:912–919 14. Grallert H, Leopoldseder S, Schuett M et al
6. Turner AP (2013) Biosensors: sense and sensi- (2011) EndoLISA: a novel and reliable method
bility. Chem Soc Rev 42:3184–3196 for endotoxin detection. Nat Methods 8:i–v
7. Turner A (2013) Biosensors: then and now. 15. Teng NN, Kaplan HS, Hebert JM et al (1985)
Trends Biotechnol 31:119–120 Protection against gram-negative bacteremia and
8. Garcia J, Bennett DH, Tancredi D et al (2013) endotoxemia with human monoclonal IgM anti-
Occupational exposure to particulate matter bodies. Proc Natl Acad Sci USA 82:1790–1794
and endotoxin for California dairy workers. Int 16. Zuzuarregui A, Arana S, Pérez-Lorenzo E et al
J Hyg Environ Health 216:56–62 (2013) Novel fully-integrated biosensor for
9. Pankhurst LJ, Deacon LJ, Liu J et al (2011) endotoxin detection via polymyxin B immobi-
Spatial variations in airborne microorganism lization onto gold electrodes. J Sens Sens Syst
and endotoxin concentrations at green waste 2:157–164
142 Prasanta Kalita et al.

17. Cho M, Chun L, Lin M et al (2012) Sensitive 22. Mares J, Kumaran S, Gobbo M et al (2009)
electrochemical sensor for detection of lipo- Interactions of lipopolysaccharide and poly-
polysaccharide on metal complex immobi- myxin studied by NMR spectroscopy. J Biol
lized gold electrode. Sens Actuators B Chem Chem 284:11498–11506
174:490–494 23. Bhor VM, Thomas CJ, Surolia N et al (2005)
18. Piazza M, Colombo M, Zanoni I et al (2011) Polymyxin B: an ode to an old antidote for
Uniform lipopolysaccharide (LPS)-loaded mag­ endotoxic shock. Mol Biosyst 1:213–222
netic nanoparticles for the investigation of 24. Slot JW, Geuze HJ (1985) A new method of
LPS–TLR4 signaling. Angew Chem Int Ed preparing gold probes for multiple-labeling
Engl 50:622–626 cytochemistry. Eur J Cell Biol 38:87–93
19. Kalita P, Dasgupta A, Sritharan V et al (2015) 25. Tsai CY, Lee DS, Tsai YH et al (2004)
Nanoparticle–drug bioconjugate as dual func- Shrinking gold nanoparticles: dramatic effect
tional affinity ligand for rapid point-of-care of a cryogenic process on tannic acid/sodium
detection of endotoxin in water and serum. citrate-generated gold nanoparticles. Mater
Anal Chem 87:11007–11012 Lett 58:2023–2026
20. Ongkudon CM, Chew JH, Liu B et al (2012) 26. Liu X, Atwater M, Wang J et al (2007)
Chromatographic removal of endotoxins: a Extinction coefficient of gold nanoparticles
bioprocess engineer's perspective. ISRN with different sizes and different capping
Chromatography 2012:9 ligands. Colloids Surf B Biointerfaces 58:3–7
21. Schindler M, Osborn MJ (1979) Interaction of 27. Chosewood LC, Wilson DE (2007) Biosafety
divalent cations and polymyxin B with lipo- in microbiological and biomedical laboratories.
polysaccharide. Biochemistry 18:4425–4430 Diane Publishing, Washington
 Chapter 14

A Bioassay for the Determination of Lipopolysaccharides

and Lipoproteins
Marcus Peters, Petra Bonowitz, and Albrecht Bufe

Abstract
The availability of convenient assays for the detection and quantification of pathogen-associated molecular
patterns (PAMPs) is limited. In the case of lipopolysaccharide (LPS) the so-called LAL (limulus amebocyte
lysate) test is available, an assay that is performed with the lysate of the blood of the horse shoe crab.
Although a sensitive and convenient assay, it lacks specificity, since it is affected by other endotoxins like,
for instance, fungal cell walls as well. Here, we describe a bioassay that can be used to detect and quantitate
PAMPs in environmental samples. More specific we demonstrate the usage of TLR2 and TLR4/CD14/
MD2 transfected Hek293 cells to quantitatively determine bacterial lipoproteins and LPS, respectively. We
show the usefulness of these assays to measure LPS in tobacco before and after combustion.

Key words Lipopolysaccharide, Lipoprotein, Toll-like receptor, Bioassay, Pathogen-associated molec-

ular patterns

1 Introduction

The sensitive and accurate measurement of pathogen-associated

molecular patterns (PAMPs) is an important tool in many scientific
fields, especially for immunologists, biochemists, and pharmacolo-
gists. However, the availability of high-quality methods is limited
due to the lack of specific and sensitive immunoassays. Only in the
case of lipopolysaccharide (LPS) there is an alternative in vitro
assay with sensitivity comparable to ELISA, the limulus amebocyte
lysate (LAL) test. This assay works well when the complexity of the
sample is low, e.g., detection of LPS contamination in recombi-
nant proteins isolated from bacterial expression systems. However
with increasing complexity of the sample the LAL test suffers from
its low specificity. In particular, fungal glycans can influence the
outcome of the test [1–3]. Due to these limitations a more specific
assay was released by the same company, the recombinant factor C
(rFC) assay [4]. The rFC assay uses only the LPS binding protein
from the clotting cascade of the LAL test as a biosensor. This assay

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_14, © Springer Science+Business Media LLC 2017

143
144 Marcus Peters et al.

combines sensitivity and specificity but needs special equipment,

that is, a fluorescence reader since it uses a fluorogenic substrate for
development.
An alternative assay that is already in use to qualitatively give
evidence for the presence or absence of PAMPs is a bioassay using
TLR-transfected Hek293 cells [5]. HEK293 cells express genes for
TLR1, TLR3, TLR5, TLR6, and NOD1. Therefore, these cells
normally do not respond to LPS or bacterial lipoproteins. However
when these cells are transfected with TLR4 or TLR2, respectively,
they become reactive to these PAMPs. We have previously shown
that transfected Hek293 cells can be used to detect LPS in envi-
ronmental samples.
TLR4/MD2/CD14 transfected HEK293 cells are useful for the
sensitive and specific quantification of LPS in complex samples such
as dust extracts from the environment [6]. The procedure of PAMP
detection is based on the stimulation of these cells with increasing
concentration of highly purified Escherichia coli LPS leading to release
of IL-8 in a dose-dependent manner. The resulting standard curve is
used to determine LPS concentration in unknown samples.
Here, we extend the use of this assay to determine LPS in
tobacco before and after combustion. Moreover, we describe a
similar assay to determine TLR2 ligands by the use of TLR2-­
transfected HEK293 cells.
Tobacco for production of cigarettes is dried after harvesting the
leaves. During the drying process microbes propagate on the drying
tobacco leaves. Therefore, it is not surprising that tobacco products
contain PAMPs. There are some studies determining the LPS con-
tent of tobacco via LAL test [7] and one study that is using the
specific determination of 3-OH fatty acids of lipid A [8]. LPS is a
molecule with relatively high chemical and physical stability; how-
ever, constant heat treatment normally destroys the biologic activity
of the substance. Despite this fact there is one older study proposing
the presence of LPS in the smoke of combusted cigarettes after the
measurement of smoke extracts with the LAL test. Furthermore,
LPS was found in alveolar macrophages of smokers. Therefore, since
the presence of LPS and other PAMPs in cigarette smoke has an
important impact for understanding smoking-­related disease [9], we
aimed to detect LPS and bacterial lipoproteins in cigarette smoke by
using TLR4-MD2-CD14-­transfected HEK293 cells.

2 Materials

293/mTLR4-MD2-CD14 cells are for studying the activation

of TLR4 by LPS. The 293/mTLR2 cells are useful for the detec-
tion of TLR2 ligands. Both cell types are designed by stable
transfection of Hek293 cells with the pUNO vector. Cells can be
purchased at Invivogen (Toulouse, France).
 A Bioassay for PAMPs 145

2.1 Cell Maintenance Use all solutions and culture media for preparing the stimulation
and Stimulation assay at room temperature.
1. DMEM: 4.5 g/L glucose, l-glutamine, sodium pyruvate, h
3.7 g/L NaHCO3.
2. PBS: Dulbecco’s Phosphate-Buffered Saline without MgCl2
and CaCl2 (Pan-Biotech, Aidenbach, Germany), liquid, sterile-­
filtered, suitable for cell culture.
3. Blasticidin S hydrochloride 10 mg/mL solution in HEPES
buffer. It is required to maintain the plasmids coding for
mTLR4 and mTLR2, respectively.
4. Hygromycin B Gold 100 mg/mL solution in HEPES buffer.
It is to maintain the plasmid coding for MD2 and CD14.
5. Normocin 50 mg/ml red aqueous solution. It is a formulation of
three antibiotics active against mycoplasmas, bacteria, and fungi.
6. Penicillin-Streptomycin: 10,000 U/ml Penicillin, 10 mg/mL
Streptomycin.
7. l-Glutamine 200 mM.
8. LPS: E. coli, chromatographically purified [10].
9. FSL-1: Synthetic di-acyl lipopeptide binding to the TLR2/6
heterodimer (EMCmicrocollections, Tübingen, Germany) (see
Notes 1 and 2).
10. Complete Growth Medium: DMEM containing 4.5 g/L glucose,
10% (v/v) fetal bovine serum, 50 U/mL penicillin, 50 μg/mL
streptomycin, 100 μg/mL Normocin™, 2 mM l-glutamine,
10 μg/mL Blasticidin, 50 μg/mL of Hygromycin B Gold.
11. Freezing Medium: DMEM, 4.5 g/L glucose, 20% (v/v) fetal
bovine serum, 50 U/ml penicillin, 50 μg/ml streptomycin,
100 μg/ml Normocin™, 2 mM l-glutamine, 10% (v/v) DMSO.
12. TC Flask, Cell+ with Vented Cap and TC Plate 96-well
Standard F, (Sarstedt, Nümbrecht, Germany).

2.2 Measurement 1. 96-well ELISA plates, High Binding, F (Sarstedt, Nümbrecht,

of Interleukin-8 in Cell Germany).
Culture Supernatants 2. Coating Buffer: 7.13 g NaHCO3, 1.59 g Na2CO3; q.s. to
1.0 L; pH 9.5 with 1 N NaOH.
3. Assay Diluent: PBS with 10% FCS.
4. Washing Buffer: PBS with 0.05% Tween20.
5. Human IL-8 ELISA (OptEIA™ ELISA Development Set,
BDbiosciences, Heidelberg, Germany) for human interleukin-8
(IL-8) containing IL-8 catcher antibody, IL-8 recombinant
protein, biotinylated IL-8 detector antibody, horse raddish per-
oxidase coupled to streptavidine.
6. TMB substrate set (BDbiosciences, Heidelberg, Germany).
7. Stop solution 1 M sulfuric acid.
146 Marcus Peters et al.

2.3 Cigarette and 1. Research cigarettes (3R4F) were purchased from the University of
Smoke Extracts Kentucky (Kentucky Tobacco Research & Development Center).
2. Smoke extract: ten cigarettes were combusted by a cigarette
smoke generator (DSI, Data science international, St. Paul,
Minnesota). Main and side-stream smoke was passed through
70 mL complete cell culture medium in a gas washing bottle.
3. Cigarette extract: the tobacco from ten cigarettes was pulver-
ized in liquid nitrogen with a mortar. Afterward tobacco pow-
der was stirred for 1.5 h in 70 mL 0.05 triethylamine for
extraction, followed by the addition of 7 mL 1 M Tris–HCl
pH 7.5. Subsequently, extracts were passed through filters
with 0.2 μm pore size.

3 Methods

3.1 Cell Maintenance 293/mTLR4-MD2-CD14 cells were cultured in DMEM 10%

FCS with Normocin, Hygromycin, Penicillin, and Streptomycin as
antibiotics (see Note 3). 293/mTLR2 cells were cultured in the
same Medium without Hygromycin. Cells were subcultured before
reaching confluence by PBS treatment.
1. Do not passage cells more than 20 times to remain fully efficient.
2. Maintain and subculture the cells in Growth Medium supple-
mented with Normocin™ (100 μg/mL), Blasticidin (10 μg/mL),
and Hygromycin B Gold (50 μg/mL).
3. Renew Growth Medium two times a week.
4. Passage cells when a 70–80% confluency is reached; therefore,
exchange the cell culture medium against PBS and detach the
cells by tapping the flask or by flushing with serological pipette
(see Notes 4 and 5).
5. Store cells frozen for longer time periods, by exchanging tissue
culture medium to freezing medium followed by freezing the
cells slowly to −80 °C followed by storage in liquid nitrogen.

3.2 Stimulation Figure 1 shows an example of determination of LPS in tobacco

of Transfected 293/ extracts from 3R4F cigarettes.
mTLR4-­MD2-­CD14
1. Prepare the assay not earlier than 48 h after the last passage of
Hek293 Cells with LPS the cells (see Note 6). Dilute cells 1 × 106/mL.
or Cigarette Extracts
2. Use a sterile 96-well tissue culture plate.
3. Generate a twofold serial dilution of LPS starting from the
highest LPS concentration of 4 ng/ml in complete cell culture
medium to prepare a standard curve of a total of eight points
with culture medium (see Note 7).
4. Prepare the unknown samples by diluting them in cell culture
medium twofold on the same plate. In the example shown in Fig. 1,
 A Bioassay for PAMPs 147

the tobacco extract was serial diluted from 1:500 to 1:16,000,

whereas the smoke extract was diluted from 1:4 to 1:128.
5. Add 100 μL culture medium with 1 × 105 293/mTLR4-
MD2-­CD14 cells to the appropriate wells by using a stepper
pipette.
6. Place the cells in an incubator at 37 °C in 5% CO2 for 24 h.
7. Plates were centrifuged (300 × g, 10 min) and supernatants
stored at −80 °C.

3.3 Stimulation Figure 2 shows an example of determination of bacterial lipopro-

of Transfected 293/ teins in tobacco extracts from 3R4F cigarettes.
mTLR2 Cells
1. Prepare the assay not earlier than 48 h after the last passage of
with FSL-1 or the cells (see Note 6).
Cigarette Extracts
2. Use a sterile 96-well cell culture plate.
3. Generate a twofold serial dilution of FSL-1 starting from the
highest concentration of 5 ng/ml in complete cell culture
medium to prepare a standard curve of a total of eight points
with culture medium (see Note 7).
4. Prepare the unknown samples by diluting them twofold on the
same plate. In the example shown in Fig. 2 the tobacco and the
smoke extract were serial diluted from 1:8 to 1:256.
5. Add 100 μL culture medium with 1 × 105 293/mTLR2 cells
to the appropriate wells by using a stepper pipette.
6. Place the cells in an incubator at 37 °C in 5% CO2 for 24 h.
7. Plates were centrifuged (470 × g, 10 min) and supernatants
stored at −20 °C until measurement of IL-8 is performed.

Fig. 1 Detection of LPS in tobacco extract. Tobacco from 3R4F cigarettes was
extracted with TEA buffer. In addition, smoke of ten cigarettes was passed
through complete cell culture medium in a gas-washing bottle. Subsequently,
the diluted extracts were evaluated for the presence of LPS. By using the LPS
standard curve to compute unknown LPS concentration in cigarettes, we calcu-
lated a total amount of 2.5 μg LPS per cigarette. In contrast, the smoke of com-
busted cigarettes contains amounts of LPS near the detection limit of the test
(gray shaded area of the graph, see Note 8)
148 Marcus Peters et al.

Fig. 2 Detection of TLR2 binding lipoproteins in tobacco extract. Tobacco from

3R4F cigarettes was extracted with TEA buffer. In addition, smoke of ten ciga-
rettes was passed through cell culture medium in a gas-washing bottle.
Subsequently, the extracts were evaluated for the presence of lipoproteins. Both
in tobacco and in smoke extracts the lipoprotein concentration is near the detec-
tion limit of the test (see Note 9)

3.4 Determination A commercial available ELISA development kit for the detection
of IL-8 in Cell Culture of IL-8 in supernatants was used.
Supernatants
1. Supernatants were thawed and centrifuged to sediment cellular
debris (470 × g, 10 min).
2. Coat wells with 50 μL per well of capture antibody diluted
1/250 (see Note 10) in coating buffer.
3. Incubate overnight at 4 °C.
4. Remove liquid from the wells and wash three times with
≥200 μL washing buffer. After the last wash, invert plate and
blot on absorbent paper to remove any residual buffer.
5. Block plates with 200 μL assay diluent per well. Incubate for
1 h at room temperature.
6. Repeat step 4.
7. Prepare the top standard at a concentration of 200 pg/mL in
assay diluent. Perform twofold serial dilutions of the top standard
to make the standard curve ranging from 200 to 3.1 pg/mL.
Pipette 50 μL of each dilution.
8. Prepare sample dilutions as well in assay diluent. Pipette 50 μL
of each sample dilution to a separate well.
9. Incubate for 2 h at room temperature.
10. Repeat step 4 with a total of five washing cycles.
11. Mix equal volumes of detection antibody diluted 1/250
(see Note 10) in assay diluent with the enzyme concentrate
diluted 1/250 (see Note 10) in assay diluent.
12. Pipette 50 μL of the detection reagent prepared in step 11 to
each well.
 A Bioassay for PAMPs 149

13. Incubate for 1 h at room temperature.

14. Repeat step 4 with a total of seven washing cycles.
15. Add 50 μL TMB substrate solution to each well and incubate
for 15 min.
16. Add 25 μL stop solution to each well followed by direct mea-
surement of the absorbance at a wavelength of 450 nm in an
ELISA plate reader.

3.5 Computation GraphPad Prism Version 5.01 was used for data analysis.
of Results
1. Concentration of LPS used for stimulation of cells was entered
as x values and transformed by the equation x = log(x).
2. The corresponding concentrations of IL-8 measured in the cell
supernatant by ELISA were entered as y values.
3. The resulting curve: log(concentration agonist) versus
response(concentration IL-8) was fitted to a nonlinear regres-
sion model. From the resulting dose response curve unknown
LPS concentration in test samples is interpolated automatically
by the software.

4 Notes

1. In this paper, the diacyl lipopeptide FSL-1 binding to the

TLR2/TLR6 heterodimer was used; however, triacyl lipopep-
tides binding to the TLR2/TLR1 heterodimer work as well
(data not shown).
2. For the generation of the standard curve use the best quality
PAMPs available. Insufficiently purified PAMPs will greatly
influence the performance of the assay.
3. Do not handle the cells cold since it will affect the vitality of
the cells. It is important that all buffers and cell culture medium
are pre-warmed. Centrifuge the cells at room temperature.
4. The response of the transfected cells can be altered by the
action of trypsin. Do not use trypsin to detach the cells. Cells
are only loosely adherent; therefore, mechanical detachment is
sufficient to harvest them.
5. Do not let the cells grow over 80% confluency. Otherwise, the
vitality of the cells will be strongly influenced.
6. Do not use cells to perform the assay that do not adhere to the
cell culture flask prior to collection, i.e., when cells look spheri-
cal and swim just above the plastic surface.
7. Normally, the detection limit of the test is below 10 pg/mL
LPS and below 40 pg/mL of bacterial lipopeptides.
150 Marcus Peters et al.

8. The TLR4-MD2-CD14 HEK293 cells show background pro-

duction of IL-8. This IL-8 production is a good indicator for
the vitality of these cells. E.g., when working with organic sol-
vents like for instance DMSO it indicates the DMSO dilution
needed to be nontoxic for the cells. The TLR2-transfected
cells do not show production of IL-8 until stimulated with the
ligand.
9. Keep in mind that the amount of biological active PAMPs
measured in solution depends on the extraction procedure.
Here, we used mild methods to extract the PAMPs from
tobacco. However when the total amount of PAMPs in a sam-
ple is of interest harsher methods must be employed to release
PAMPs that are still bound to microbial debris.
10. Please refer to the manufacturer’s protocol since the dilution
of these test compounds may vary from lot to lot.

Acknowledgments

WeliketothankProf.Dr.Karl-HeinzWiesmüller(EMCmicrocollections,
Tübingen, Germany) for providing us with FSL-1. We like to thank
Prof. Dr. Ulrich Zähringer (formerly Forschungszentrum Borstel,
Germany) for providing us with E. coli LPS. We like to thank Dimitri
Kasakovski for g
­ enerating the cigarette and smoke extracts. This work
was supported by Deutsche Forschungsgemeinschaft (PE 1813/2-1).

References

1. Vassallo R, Limper AH (1999) Fungal ß-­glucan environmental samples. Innate Immun 18:
can yield false-positive results with the limulus 694–699
amebocyte lysate endotoxin assay. Chest 116: 7. Hasday JD, Bascom R, Costa JJ, Fitzgerald T,
583–584 Dubin W (1999) Bacterial endotoxin is an
2. Hirano M, Matsumoto T, Kiyohara H, Yamada active component of cigarette smoke. Chest
H (1994) Lipopolysaccharide-independent 115:829–835
limulus amebocyte lysate activating, mitogenic 8. Pomorska D, Larsson L, Skórska C, Sitkowska
and anti-complementary activities of pectic J, Dutkiewicz J (2007) Levels of bacterial
polysaccharides from chinese herbs. Planta endotoxin in air of animal houses determined
Med 60:248–252 with the use of gas chromatography-mass spec-
3. Cooper JF, Weary ME, Jordan FT (1997) The trometry and Limulus test. Ann Agric Environ
impact of non-endotoxin LAL-reactive materi- Med 14:291–298
als on Limulus amebocyte lysate analyses. PDA 9. Knobloch J, Schild K, Jungck D, Urban K,
J Pharm Sci Technol 51:2–6 Müller K, Schweda EK, Rupp J, Koch A (2011)
4. Ding JL, Ho B (2010) Endotoxin detection— The T-helper cell type 1 immune response to
from limulus amebocyte lysate to recombinant gram-negative bacterial infections is impaired
factor C. Subcell Biochem 53:187–208 in COPD. Am J Respir Crit Care Med 183:
5. Lappin DF, Sherrabeh S, Erridge C (2011) 204–214
Stimulants of Toll-like receptors 2 and 4 are 10. Güttsches AK, Löseke S, Zähringer U,
elevated in saliva of periodontitis patients com- Sonnenborn U, Enders C, Gatermann S, Bufe
pared with healthy subjects. J Clin Periodontol A (2012) Anti-inflammatory modulation of
38:318–325 immune response by probiotic Escherichia coli
6. Peters M, Fritz P, Bufe A (2012) A bioassay Nissle 1917 in human blood mononuclear
for determination of lipopolysaccharide in cells. Innate Immun 18:204–216
 Chapter 15

Capillary Electrophoresis Chips for Fingerprinting

Endotoxin Chemotypes and Subclasses
Béla Kocsis, Lilla Makszin, Anikó Kilár, Zoltán Péterfi, and Ferenc Kilár

Abstract
Endotoxins (lipopolysaccharides, LPS; lipooligosaccharides, LOS) are components of the envelope of
Gram-negative bacteria. These molecules, responsible for both advantageous and harmful biological activ-
ity of these microorganisms, are highly immunogenic and directly involved in numerous bacterial diseases
in humans, such as Gram-negative sepsis. The characterization of endotoxins is of importance, since their
physiological and pathophysiological effects depend on their chemical structure. The differences among
the LPS from different bacterial serotypes and their mutants include variations mainly within the composi-
tion and length or missing of their O-polysaccharide chains. Microchip electrophoretic methodology
enables the structural characterization of LPS molecules from several bacteria and the quantitative evalua-
tion of components of endotoxin extracts. The improved microchip electrophoretic method is based on
the direct labeling of endotoxins by covalent binding of a fluorescent dye. The classification of the S-type
LPSs can be done according to their electrophoretic profiles, which are characteristics of the respective
bacterial strains. According to the number, distribution, and the relative amounts of components in an
endotoxin extract, it is possible to differentiate between the S-type endotoxins from different Gram-­
negative bacterial strains. The microchip electrophoresis affords high-resolution separation of pure and
partially purified (e.g., obtained from whole-cell lysate) S and R endotoxins. This microchip technique
provides a new, standardizable, fast, and sensitive method for the detection of endotoxins and for the
quantitative evaluation of components of an endotoxin extract.

Key words Endotoxin, Lipopolysaccharide, LPS, Chemotype, Whole-cell lysate, Fluorescent label-
ing, Microchip electrophoresis, Endotoxin profiles

1 Introduction

Endotoxins (lipopolysaccharides, LPS; lipooligosaccharides, LOS)

are components of the envelope of most Gram-negative bacteria
[1]. These molecules, responsible for both advantageous and
harmful biological activity of these microorganisms, are highly
immunogenic and can be directly involved in numerous bacterial
diseases in humans, such as Gram-negative sepsis [2]. The amphi-
philic LPS compounds consist of a hydrophobic lipid region (lipid
A) covalently substituted by a hydrophilic oligo- or polysaccharide

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_15, © Springer Science+Business Media LLC 2017

151
152 Béla Kocsis et al.

region [3]. Typically, the polysaccharide region of S (smooth) LPS

chemotypes consists of a core and an O-polysaccharide chain,
whereas the R (rough) LPS or LOS chemotypes lack the
O-polysaccharide chain and sometimes even the core components
(see Fig. 1). The differences among the LPSs from different bacte-
rial serotypes and their mutants include variations mainly within
the composition and length or missing of the O-polysaccharide
chains. The characterization of endotoxins is of importance, since
their physiological and pathophysiological effects depend on their
chemical structure.
When a large number of bacterial mutants and their LPS con-
tent are to be compared, for instance, in the preparation of vac-
cines, LPSs are detected directly from bacterial cultures, generally
by the relatively short method of Hitchcock and Brown [4]. The
procedure involves the enzymatic digestion of whole-cell lysates
(by lysozyme and proteinase K enzymes) leading to partially puri-
fied, protein-free total cellular LPS. An important feature of this
method is that only 1 mL culture volume is necessary for the
­characterization of LPSs, besides it has a short overall process-time
(ca. 40 h).
The analysis of intact LPS molecules and the classification of
the chemotypes are routinely checked by sodium dodecyl sulfate-­
polyacrylamide gel electrophoresis (SDS-PAGE) with silver-­
staining detection [5].
The fast analyses of Gram-negative bacterial endotoxins with
microchip electrophoresis have been described either using nonco-
valent binding of a fluorescent dye (see the application of the
Agilent Protein 80 LabChip) [6–8], or using the covalent binding
of a fluorescent dye to the ethanolamine groups of the endotoxins
(see the application of the High Sensitivity Protein 250 LabChip) [8, 9].
The separation is performed with a sieving matrix included in the

Fig. 1 Basic structure of endotoxins (lipopolysaccharides, LPSs) from Gram-negative bacteria. n: the number
of repeating units, R-type endotoxins (LOS): n = 0; SR-type endotoxins: n = 1, S-type endotoxins (LPS): n > 1
Capillary Electrophoresis Chips for Fingerprinting Endotoxin Chemotypes and Subclasses 153

kits, which provides optimal resolution for the separation of LPS

molecules with different compositions, chain-lengths, and concen-
tration. With the use of these microchip analyses 11 endotoxin
assignations can be performed within ca. 1 h. Optimization of the
experimental conditions (the sample pretreatment, the concentra-
tion of the solutions, and the separation time), along with the clas-
sification, molecular mass estimation, and quantitative analyses of
endotoxin components from purified LPS or partially purified sam-
ples from whole-cell lysates, has been made systematically.
The classification of LPS can be done according to the electro-
phoretic profiles, which are characteristics of the respective bacte-
rial strains. The LPS components appear as multiple “waves” of
peaks. According to the number, distribution, and the relative
amounts of components, differentiation of two main groups
(groups 1 and 2), and within group 1, three subgroups (groups 1a,
1b, 1c) (Fig. 2a–c) can be made. Components with 13 or 15 repeat-
ing units, appearing with relatively high amounts, possess molecu-
lar masses of ca. 15,500 Da in group 1a, or ca. 17,000–18,000 Da
in group 1b, respectively. In group 1c, the relative “maximum” in
the quantity distribution appears at components with ca.
23–24,000 Da, corresponding to 18–29 repeating units. The defi-
nite decrease in the relative amounts of the LPS components in
group 2 (Fig. 3) up to a molecular mass of ca. 23 kDa is a quite
characteristic, while the first endotoxin component (with a molec-
ular mass of 4400 Da) is produced in about 50 times higher
amount, compared to the other LPS components.
It is highly important to estimate the molecular masses of
endotoxin components. However, the determination of the exact
molecular masses in SDS-PAGE is not possible, because molecular
mass standards in this technique are available only for proteins. In
contrast, the microchip electrophoresis method may be used for
molecular mass estimation based on the calculated molecular
masses of LPS components with known structures. Constructing
the log M versus 1/t diagrams, using mass data previously known
(e.g., with MS or NMR studies) the molecular masses of the com-
ponents can be estimated [9] (Fig. 4). It should be noted that
endotoxins classified into group 1a have a different log M-1/t line
compared to the others showing the significant difference in the
migration properties of the endotoxin components in group 1a as
compared to those in the other groups. For instance, components
with 24,000 Da in group 1a are detected at ca. 38 s, while compo-
nents in the other groups with the same mass migrate much slower
(detection time is ca. 58 s) (Fig. 4).
The electrophoretic profiles of purified LOS samples extracted
from Shigella sonnei 4303 (phase II) and Salmonella enterica sv.
Minnesota (Salmonella Minnesota) R595 show only one or two,
low-molecular-mass components (Fig. 5). These low molecular
Fig. 2 Microchip electrophoretic profiles of LPS components prepared from bacteria of (a) Escherichia coli O83
(group 1a); (b) Escherichia coli O21 (group 1b); and (c) Salmonella Adelaide O35 (group 1c). The LPS molecules
are visualized by covalently bound fluorescent dye applying the High Sensitivity Protein 250 LabChip kit.
Experimental conditions: HSP 250 Protein microchip; running buffer, polydimethyl-acrylamide-­based linear
polymer solution at pH 8 (its adsorption to the capillary wall reduces electroendosmotic flow to almost zero,
thus the negatively charged LPS-dye-SDS complexes migrate toward the anode); fluorescence detection with
630 nm excitation and 680 emission wavelengths; injection: 80 s at 1000 V; separation: 60 or 90 s at 900 V. The
sample concentration applied in the chip wells is 0.1 μg/μL. The first peak corresponds to the system peak
(S.P.), i.e., the signal of the nonbound dye, while the subsequent peaks in the electropherogram correspond to
LPS components with increasing numbers (n) of repeating units in their O-polysaccharide chain (i.e., the first
endotoxin peak corresponds to a component built from the lipid A and the core parts). The estimated molecular
masses of the first and last sample peaks are indicated. Δ: molecular mass of a repeating unit (Da)
Capillary Electrophoresis Chips for Fingerprinting Endotoxin Chemotypes and Subclasses 155

Fig. 2 (continued)

Fig. 3 Microchip electrophoretic profiles of LPS from Proteus morganii O34 prepared from bacteria (group 2).
Experimental conditions are the same as in Fig. 2
156 Béla Kocsis et al.

Fig. 4 The log M versus 1/t calibration curves for the endotoxin components. The molecular masses of the
LPS known from other studies are used to construct the lines. The calibration lines for group 1a (square),
group 1b (circle), and group 1c (triangle) are calculated from the migration data in the electropherograms
in Figs. 2 and 3 [9]

components sometimes partially overlap with the system peak (see

Fig. 5b).
The limits of detection values (at S/N = 3), obtained with the
covalent labeling of LPS or LOS, are considerably lower (0.43 and
1.13 ng/μL, applying the High Sensitivity Protein 250 LabChip
analyses for the LPS and LOS, respectively) than those found in
the previously developed microchip method using noncovalent
labeling of the LPSs (ca. 5 ng/μL LPS with the Protein 80
LabChip).
Protein-free LPS samples, obtained from bacterial lysates, can
be directly analyzed (without further purification) with the micro-
chip electrophoresis method. Fig. 6a shows the electrophoretic
profile of the partially purified LPS in the E. coli O83 whole-cell
lysate. The “wave-like” pattern of 20 peaks is consistent with the
profile in Fig. 2a, although the relative peak area ratios are slightly
different in the two patterns. The analysis of an R-type partially
purified endotoxin sample prepared from the whole-cell lysate of
Salmonella Minnesota R595 bacteria is shown in Fig. 6b. The elec-
tropherogram shows a single-component profile analogue to the
purified R-chemotype endotoxin in Fig. 5b.
Capillary Electrophoresis Chips for Fingerprinting Endotoxin Chemotypes and Subclasses 157

Fig. 5 Microchip electrophoretic profiles of R-type (a) Sh. sonnei 4303 and (b) S. Minnesota R595 endotoxin
components prepared from bacteria. The experimental conditions are the same as in Fig. 2, except for that the
S. Minnesota R595 sample was diluted 1:200. One (or two) component(s) is (are) found in the LOS. The major
sample peak appears as fast-migrating, small-molecular-mass substance after (and in some cases partially
overlapping with) the system peak (S.P.)
158 Béla Kocsis et al.

Fig. 6 Microchip electrophoretic profiles of partially purified endotoxin samples from whole-cell lysates of (a)
S-type E. coli O83 and (b) rough R-type S. Minnesota R595 bacteria. The experimental conditions are the same
as in Fig. 2. Although the electrophoretic profiles of these partially purified LPS are very similar to those of the
pure LPS of the same strains (Figs. 2a and 5b, respectively), the relative peak area ratios are slightly different
in the two patterns
Capillary Electrophoresis Chips for Fingerprinting Endotoxin Chemotypes and Subclasses 159

2 Materials

2.1 Bacterial Cell 1. Bacterial stock is stored at −80 °C.

Culture and Lysis 2. 5 mL Mueller-Hinton broth (Oxoid Ltd., UK): 3.0 g beef
extract, 17.5 g acid hydrolysate of casein, 1.5 g starch in 1 L
deionized water. Store the broth at 4 °C.
3. Dissolve lysozyme in water at 100 mg/mL and store at −20 °C.
4. LPS lysing buffer: 2% (w/v) sodium dodecyl sulfate (SDS), 4%
(v/v) β-mercaptoethanol, 10% (w/v) glycerol, 1 M Tris–HCl
buffer, pH 6.8, 0.05% (w/v) bromophenol blue. Store at
25 °C.
5. Proteinase K enzyme. Store at −20 °C.
6. Ethanol/magnesium-chloride solution: Dissolve magnesium-­
chloride (3.8 mg) in 50 mL ethanol and store at 25 °C. The
reagents are of analytical grade.

2.2 Preparation 1. Gram-negative bacterial stock is stored at −80 °C.

of Pure LPS 2. Extraction of pure S-type LPSs is carried out with the hot phe-
nol–water extraction method of Westphal et al. [10]: 110 g
phenol and 65 mL distilled water, 100 mg of sodium chloride
(per 100 g of wet cells), cold acetone, sodium hydroxide.
3. The R-type pure LPSs are obtained by the phenol-chloroform-­
light petroleum method of Galanos et al. [11]. Extraction was
made in liquid phenol (90 g dry phenol and 11 mL water),
chloroform, and light petroleum in a volume ratio of 2:5:8.

2.3 LPS Covalently 1. 30 mM Tris/HCl buffer, pH 8.5 (store at 4 °C).

Labeled 2. Sieving matrix: 4.5% polydimethyl-acrylamide-based linear
with Fluorescent Dye polymer in an SDS-containing buffer, pH 8.0 (this is included
in the Agilent High Sensitivity Protein 250 LabChip kit).
3. Destaining solution: this solution is included in the Agilent
High Sensitivity Protein 250 LabChip kit.
4. Sample buffer/DTT solution: Mix 100 μL sample buffer
(included in the Agilent High Sensitivity Protein 250 LabChip
kit) with 3.5% (v/v) 1 M 1,4-dithiothreitol (DTT). Store at
−20 °C.
5. Mix the labeling dye (a fluorescent dye included in the Agilent
High Sensitivity Protein 250 LabChip kit) with DMSO: Add
54 μL DMSO onto the pellet of 1 vial labeling dye and mix
until all solid components are completely dissolved (the fluo-
rescent dye reagent needs reconstitution in DMSO, which is
known to facilitate the entry of organic molecules into tissues,
so it should be used with appropriate care). Prepare a diluted
dye solution (2 μL stock solution diluted ten times with 18 μL
160 Béla Kocsis et al.

distilled water) daily for the labeling experiments. Store the

dye in the dark (see Note 1).
6. Ethanolamine (this is corrosive) is stored at −20 °C.

2.4 Electrophoresis 1. Microchip developed for protein sizing [12] (e.g., Agilent
on Microchip High Sensitivity Protein 250 chip from Agilent Technologies).
2. Agilent 2100 Bioanalyzer system equipped with a diode laser
for fluorescence detection (Agilent Technologies) and with
2100 Expert software.
3. Syringe priming station for filling the microchip (Agilent
Technologies).
4. Electrode cleaner (Agilent Technologies). This is a microchip
without capillaries; thus, the space connecting the wells can be
filled with ca. 350 μL aqueous solution.

3 Methods

These instructions describe the analyses of pure and partially puri-

fied endotoxin samples from whole cell lysates of bacteria and
assume the use of the Agilent 2100 Bioanalyzer microchip electro-
phoresis system. The methods are easily adaptable to other micro-
chip formats. It is critical that the molecular mass range of the
sieving matrix used for the separation of LPS components is opti-
mal and fluorescent dye is used for labeling and detection.

3.1 Preparation 1. Grow the bacteria at 37 °C in Mueller–Hinton broth in a fer-

of Pure LPS mentor (e.g., Biostat U30D, Braun Melsungen, Germany)
until the late logarithmic phase (about 10 h) and then collect
the cells by centrifugation.
2. Prepare pure LPS by applying the hot phenol–water extraction
method [10], by extracting the hydrophilic LPSs from smooth
Gram-negative bacteria. Suspend 20 g acetone-dried bacteria in
distilled water, mix this bacterial suspension and 90% w/v phe-
nol/water 1:1 (v/v), and heat it up to 65 °C for 10 min. After
the LPS are extracted, let the mixture cool down and centrifuge
it at 4 °C. Collect the water phase that contains the LPS. Repeat
the extraction step twice. Dialyze the combined water phases in
a tubing (Kalle GmbH, Wiesbaden, Germany) against tap water
for 3 days, and against distillated water for another 3 days.
Concentrate the dialyzate in a rotary evaporator and ultracen-
trifuge the crude extract three times for 4 h at 100,000 × g.
Re-suspend the sediment in distilled water and freeze-dry it.
3. Store the endotoxins in lyophilized form, and prepare 2 mg/mL
suspensions in phosphate-buffered isotonic saline solution
before use.
Capillary Electrophoresis Chips for Fingerprinting Endotoxin Chemotypes and Subclasses 161

3.2 Preparation 1. Grow the bacteria at 37 °C in Mueller–Hinton broth in a fer-

of Pure LOS mentor until the late logarithmic phase (about 10 h) and then
collect the cells by centrifugation.
2. Extract the LOS from acetone-dried bacteria according to
[11], which is an efficient extraction method for the hydro-
phobic LOS (lacking the hydrophilic O-side chains) from
rough bacteria. Suspend the bacteria (20 g) in 100 mL phenol-
chloroform-­light petroleum (2:5:8, v:v:v) and stir for about
20 min. Precipitate the LOS dropwise with distilled water from
the phenol extract after evaporation of chloroform and light
petroleum in a rotary evaporator. Wash the precipitated LOS
with diethyl ether, dry it in a fume hood, dissolve it in water,
and centrifuge it at 100,000 × g for 4 h. Re-suspend the sedi-
ment in distilled water and freeze-dry it.
3. Store the LOS in a lyophilized form, and prepare 2 mg/mL solu-
tions in phosphate-buffered isotonic saline solution before use.

3.3 Preparation 1. Streak bacterial stocks on a Mueller-Hinton agar plate and

of LPS Samples grow overnight at 37 °C in a bacterial incubator.
from Whole-Cell 2. Transfer one colony into 5 mL of Mueller-Hinton broth using
Lysate a sterile inoculating loop and incubate at 37 °C on a shaker
overnight.
3. Collect 1 mL of the cell culture in an Eppendorf tube and wash
with 1 mL water once by centrifugation at 6000 × g for 3 min.
4. Re-suspend the pellets in 1 mL water and heat at 100 °C for
30 min.
5. Transfer 200 μL of the cooled suspension to an Eppendorf
tube and add 4 μL of the lysozyme solution. Incubate the mix-
ture at 37 °C for 30 min. During this process the peptidogly-
can layers of the bacterial cell wall are disintegrated.
6. Add 200 μL of the LPS lysing buffer to the cooled mixture and
incubate the lysates at 100 °C for 10 min. During this step cel-
lular materials, such as proteins, LPS, nucleic acids, cell debris,
are released from the bacterial cells.
7. For proteolytic digestion, dissolve Proteinase K at 20 μg/μL in
water, and add 20 μL of this in two portions (10–10 μL) to the
cell lysate and incubate at 65 °C for 4 h (the second portion is
added in the middle of the incubation time, i.e., after 2 h).
8. Precipitate the LPS content by adding 800 μL of the ethanol/
magnesium-chloride solution and store the mixture at −20 °C
overnight.
9. Centrifuge the mixture at 13,000 × g for 15 min, and suspend
the sediment (containing LPS) in 30 μL deionized water and
sonicate in an ultrasonic bath.
162 Béla Kocsis et al.

3.4 Preparation of 1. Add the LPS samples to the Tris/HCl buffer in 9:1 ratio (see
LPS Covalently Note 2), e.g., 4.5 μL LPS sample and 0.5 μL Tris/HCl
Labeled buffer.
with Fluorescent Dye 2. Mix 5 μL of LPS samples in Tris/HCl buffer with 0.5 μL of
the fluorescent dye/DMSO solution and incubate for 10 min
at room temperature (see Note 2).
3. Add 0.5 μL of ethanolamine and incubate for 10 min at room
temperature. During this step the excess dye is quenched after
reaction with ethanolamine.
4. Dilute the LPS sample mixtures tenfold with deionized water.
5. Combine 4 μL of the labeled and diluted LPS sample mixtures
with 2 μL of the sample buffer/DTT solution and incubate at
100 °C for 5 min, then centrifuge (6000 × g, 15 s).

3.5 Loading 1. Fill the capillaries of the microchip with sieving matrix. Pipette
the Microchip 12 μL of the sieving matrix (see Note 3) in the well marked
for the Electrophoresis with and fill the chip channels hydrodynamically applying
the special syringe priming station.
2. Adjust the base-plate of the syringe priming station to position
“A” and the syringe clip to its middle position.
3. Put the microchip in the priming station and the plunger at
1 mL.
4. Close the syringe priming station and press the plunger until
held by the clip.
5. Release the clip after 90 s and slowly pull back the plunger to
1 mL position after 5 s, the priming station is then opened.
6. After the filling, make sure that no bubbles are present in the
capillaries. The filling is proper if the capillaries cannot be seen
on the backside of the chip.
7. Pipette 12 μL of the sieving matrix (see Note 3) in the other
three wells marked with “G” and 12 μL of the destaining solu-
tion in the well marked with “DS” (see Note 4).
8. Load 6 μL of LPS samples onto the sample wells in the micro-
chip, and in the Ladder well marked with (see Note 5). Fill
all sample wells and the Ladder well of the microchip before
the electrophoretic analysis.
9. Place the loaded microchip in the Agilent 2100 Bioanalyzer
equipment (see Note 6).

3.6 Electrophoresis 1. Place the loaded chip in the receptacle of the Agilent 2100
on Microchip Using Bioanalyzer and carefully close the lid, so the electrodes in the
Agilent 2100 cartridge fit into the wells of the chip.
Bioanalyzer 2. Select the High Sensitivity Protein 250.xsy assay (for the analy-
sis of the covalently labeled LPS samples) from the Assay menu
Capillary Electrophoresis Chips for Fingerprinting Endotoxin Chemotypes and Subclasses 163

and start the assay immediately by clicking on the start button

(see Note 7).
3. The analyses of all samples in the microchip are accomplished
within 30 min (see Note 8). The data procession by the soft-
ware includes baseline correction, alignment of separated runs
in one chip, and the display of the peak data in the microchip
electropherograms as gel-like images on a gray scale (this is not
shown here). In all samples the peaks corresponding to the
LPS components appear directly after the first (relatively big)
system peak (see Note 9). The LPS appear as multiple “waves”
of peaks (LPS components) with a broad molecular mass dis-
tribution (containing both, high and low mobility compo-
nents), according to the number of repeating units in the
O-polysaccharide chain, while the LOS contain only one (or
two) peak(s) with high mobility (low molecular mass)
component(s) since they lack the O-polysaccharide chain.
Examples of results are shown in Figs. 2 and 3.
4. Immediately remove the chip from the Agilent 2100
Bioanalyzer when the assay is finished (see Note 10).
5. After removing the microchips, apply the electrode cleaner to
remove possible contaminants from the electrodes. Slowly fill
one of the wells of the electrode cleaner with 350 μL deionized
water and then place it in the Agilent 2100 Bioanalyzer. Close
the lid and wait for about 10–30 s before the removal of the
electrode cleaner.

3.7 Determination 1. Calculate the molecular masses of LPS components (i.e., lipid
of the Molecular A, core region, and repeating units) with known structures
Masses of the LPS from other experiments, such as MS or NMR.
Components 2. For the endotoxins, however, of which all three data are not
known, it is possible to construct the log M versus 1/t
diagrams.
3. Apply the available calibration curves (see Fig. 4) for endotoxin
components, of which the molecular structures (masses) of
one or more constituents are not known, and estimate the
molecular masses.

4 Notes

1. Remove light covers only when pipetting. The dye contained

in the reagents decomposes when exposed to light and this
reduces the signal intensity.
2. Use 0.5 mL vials. Using larger vials may lead to poor results,
caused by evaporation.
164 Béla Kocsis et al.

3. Always insert the pipette tip to the bottom of the well when
dispensing the liquid. Placing the pipette at the edge of the
well may lead to poor results.
4. The unbound fluorescent dye present in the capillary channels
is diluted before the detection. The diluting solution consist-
ing of Tris/HCl buffer and SDS without sieving matrix is
introduced via a cross-section just before the detection point.
This helps to get a better signal-to-noise ratio.
5. The original protein chip kit uses the ladder well for the deter-
mination of the molecular masses of proteins. A molecular
mass standard mixture would be injected into this ladder well,
but—since the protein standards have different migration
properties than LPS—the determination of the molecular
masses of LPS components cannot be done with the help of
those standards. See the molecular mass estimation in
Subheading 3.6. The ladder well, hence, this well can be used
for the analysis of an eleventh sample.
6. Loaded chips should be used within 5 min. Reagents and the
sample solutions may undergo evaporation, leading to poor
results.
7. The Agilent 2100 Bioanalyzer should not be touched during
analysis and should never be placed on a vibrating surface.
8. The analysis starts with the electrophoretic injection into the
capillaries for 80 s (the injected volume is ca. 40 pL) and the
run times are 60 or 90 s (the collection of the data starts at ca.
10 s after the voltage is applied).
9. The complex of the fluorescent dye and ethanolamine gener-
ally appears as a system peak. However, this does not disturb the
separation of the LPS components. The sensitivity of this
method is high, since satisfactory patterns are obtained from
1 mL bacterial cell cultures, which contain ca. 108 cells and the
endotoxin content is less than 1 ng.
10. Leaving the chip for a period longer than 1 h in the Bioanalyzer
may cause contamination of the electrodes.

Acknowledgments

The work was supported by the grants TÁMOP 4.2.1/B-10/2/

KONV-2010&#8211;0002, 4.2.2/B-10/1&#8211;2010-0029,
OTKA K-100667, and the UNKP-16-4 New National Excellence
Program of the Ministry of Human Capacities. Lilla Makszin
acknowledges the financial support of the PTE ÁOK-Post-Doc
grant.
Capillary Electrophoresis Chips for Fingerprinting Endotoxin Chemotypes and Subclasses 165

References
1. Caroff M, Karibian D (2003) Structure of bac- 7. Kilár A, Farkas V, Kovács K, Kocsis B, Kilár F
terial lipopolysaccharides. Carbohydr Res (2008) Novel quantitative electrophoretic
338:2431–2447 analysis of endotoxins on microchips.
2. Holst O, Ulmer AJ, Brade H, Flad H-D, Electrophoresis 29:1713–1722
Rietschel ET (1996) Biochemistry and cell 8. Kocsis B, Kilár A, Makszin L, Kovács K, Kilár F
biology of bacterial endotoxins. FEMS (2011) Capillary electrophoresis chips for finger-
Immunol Med Microbiol 16:83–104 printing endotoxin chemotypes from whole-cell
3. Magalhaes PO, Lopes AM, Mazzola PG, lysates. In: Holst O (ed) Microbial toxins: meth-
Rangel-Yagui C, Penna TCV, Pessoa A (2007) ods and protocols. Springer, NY, pp 89–99
Methods of endotoxin removal from biological 9. Makszin L, Kilár A, Felső P, Péterfi Z, Kocsis B,
preparations: a review. J Pharm Pharm Sci Kilár F (2012) Quantitative microfluidic analy-
10:388–404 sis of S- and R-type endotoxin components
4. Hitchcock PJ, Brown TM (1983) Morphological with chip capillary electrophoresis.
heterogeneity among Salmonella lipopolysac- Electrophoresis 33:3351–3360
charide chemotypes in silver-stained polyacryl- 10. Westphal O, Lüderitz O, Bister F (1952) Über
amide gels. J Bacteriol 154:269–277 die Extraktion von Bakterien mit Phenol
5. Tsai CM, Frasch CE (1982) A sensitive silver Wasser. Z Naturforsch B 7:148–155
stain for detecting lipopolysaccharides in poly- 11. Galanos C, Lüderitz O, Westphal O (1969) A
acrilamide gels. Anal Biochem 119:115–119 new method for extraction of R-lipopoly­
6. Kilár A, Péterfi Z, Csorba E, Kilár F, Kocsis B saccharides. Eur J Biochem 9:245–249
(2008) Capillary electrophoresis chips for 12. Bousse L, Mouradian S, Minalla A, Yee H,
screening of endotoxin chemotypes from whole- Williams K, Dubrow R (2001) Protein sizing
cell lysates. J Chromatogr A 1206:21–25 on a microchip. Anal Chem 73:1207–1212
 Chapter 16

Micromethods for Isolation and Structural

Characterization of Lipid A, and Polysaccharide Regions
of Bacterial Lipopolysaccharides
Alexey Novikov, Aude Breton, and Martine Caroff

Abstract
Lipopolysaccharides (LPS) are major components of the external membrane of most Gram-negative bac-
teria, providing them with an effective permeability barrier. They are essentially composed of a hydrophilic
polysaccharide region (PS) linked to a hydrophobic one, termed lipid A. The LPS polysaccharide moiety
is divided into the core oligosaccharide (OS) and O-chain repetitive elements. Depending on their indi-
vidual variable fine structures, LPS may be potent immunomodulators. The lipid A structure is a key
determinant for LPS activity. However, the presence of the core region, or at least of the highly charged
3-deoxy-d-manno-oct-2-ulosonic acid molecules, is also important for preserving the native lipid A con-
formation within individual LPS molecules. We describe herein four rapid and practical micromethods for
LPS, lipid A, and core OS structural analyses. The first method allows the direct isolation of lipid A from
whole bacteria cell mass; the second describes conditions for the sequential release of fatty acids enabling
the characterization of their substitution position in the lipid A backbone, to be determined by matrix-­
assisted laser desorption/ionization mass spectrometry (MALDI-MS). The third one is a microscale pro-
cedure for the mass spectra screening of LPS, lipid A, and PS using triethylamine and citric acid. The
fourth method is a chromatography procedure for Rough-type LPS on thin-layer-chromatography. These
methods were developed to be coupled to mass-spectrometry (e.g., MALDI-MS) but can also be used
with other analytical techniques (e.g., chromatography). Examples are given with reference to two major
human pathogens: Bordetella pertussis and Pseudomonas aeruginosa; to one porcine pathogen: Actinobacillus
pleuropneumoniae; and to commercial samples of Salmonella Minnesota Re595 LPS.

Key words Gas chromatography, 3-Deoxy-d-manno-oct-2-ulosonic acid, Lipooligosaccharide,

Lipopolysaccharide, Matrix-assisted laser desorption/ionization mass spectrometry, Polyacrylamide
gel electrophoresis, Sodium dodecyl sulfate, Triethylamine, Thin-layer chromatography

1 Introduction

Endotoxins are major components of the Gram-negative bacterial

outer membrane and occur either as lipooligosaccharides (LOS)
composed of a lipid A region covalently linked to a core oligosac-
charide, or as lipopolysaccharide (LPS) composed of lipid A linked
through a core oligosaccharide to a polysaccharide (O-chain),

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_16, © Springer Science+Business Media LLC 2017

167
168 Alexey Novikov et al.

made of repetitive oligosaccharide subunits. Lipid A is anchored

in the outer layer of the external membrane. In some bacteria, it
is a powerful immunomodulator, responsible for most of the
­biological activities of the LPS. Small LPS amounts from such bac-
teria can have beneficial effects, but larger amounts may cause
endotoxic shock.
LPS and lipid A preparations are variably heterogeneous. This
is seen in their ladder-like profile on SDS-PAGE or on thin-layer
chromatograms (TLC) [1] and in LPS, lipid A, and PS mass spec-
tra [2, 3]. This heterogeneity is due to different levels of biosyn-
thesis related to different numbers of O-chain repetitive units, as
well as to variability in other structural elements like fatty acids,
phosphate groups, and other specific structural modifications (e.g.,
methylation, acetylation, substitution with different amino deriva-
tives, etc.). LPS molecular masses range from about 2 to 20 kDa.
Analysis and comparison of the separated structural regions of
LPS are usually performed on samples from preparative-scale
extractions [2]. However, most commonly used methods for
extracting both LPS and LOS are time-consuming and require an
appreciable amount of bacteria.
We present here a method for the direct selective separation
and extraction of lipids A from bacterial cells, and for their analy-
sis by mass spectrometry (MS) [4]. The method, which is fast
and easy, can be applied to μg-mg quantities of lyophilized bac-
teria, and is extremely efficient for analysis of bacterial isolates or
samples recovered in vivo, as well as for checking modifications
as a function of culture conditions for example. Use of the direct
lipid A isolation method will be illustrated by MS analysis of
Pseudomonas aeruginosa lipid A structural modifications induced
by transitions between planktonic and biofilm growth styles of
these bacteria [3].
Alkaline treatments of lipid A selectively remove ester-linked
fatty-acids [5, 6]. Following the kinetics of fatty-acids release, spe-
cific to each type (primary or secondary-linked), by MS we obtained
more detailed structural information with respect to lipid A acyla-
tion patterns. This second method is perfectly compatible with the
first one, and can be applied to micro amounts of lipid A directly
isolated from bacteria, as illustrated using lipid A and LPS of
Bordetella pertussis, the agent of whooping cough [7, 8].
A third one, among our methods, consists in the rapid cleavage
of LPS and isolation of the lipid A together with the PS moieties
for MS screening [9]. The combination of triethylamine (TEA) a
well-known lipid A solubilizing agent, and of citric acid, that we
described for disaggregation of LPS aggregates [9] was used at a
pH value of about 3.4 to 4.4, suitable for LPS mild hydrolysis at
100 °C. The potency of the two components was shown to increase
the solubility, to improve desorption/ionization, as well as to pre-
serve LPS labile residues during hydrolysis (amino acids,
 Micromethods for LPS Structural Analyses 169

­ hosphates). The screening of lipid A and PS mildly liberated from

p
the crude LPS of different Actinobacillus pleuropneumoniae field
isolates, a porcine pathogen, together with the comparison of dif-
ferent commercial LOS from Salmonella Minnesota Re595 mutant
are illustrating here the potency of this method. The S. Minnesota
deep-rough mutant LOS were selected for being repeatedly
described in the literature as the simplest and shortest LOS. This
quick description led many biologists to using these LOS as the
more common standards for testing biological activities. We show
herein by matrix-assisted laser desorption/ionization (MALDI)-MS
that the structures of these LOS are not as simple and reproducible
as thought, and recommend their characterization by MS before
use. TLC analysis also demonstrates the level of heterogeneity and
nonreproducibility.

2 Materials and Techniques

2.1 Bacterial Strains 1. Grow Gram-negative bacterial cells in required experimental

and Cultures conditions.
2. Before harvesting, kill bacteria by a method appropriate for the
species under study, e.g., in cold 2% phenol or by incubation
for 40 min at 56 °C and then examine the growth for the
absence of viable bacteria.

2.2 Chemicals 1. LPS: The three different commercial batches of LPS from
Re595 S. Minnesota, SmRe595-1: L-9764 55F-4023,
SmRe595-2: L-9764 121H4026, and SmRe595-3: L9764
075M4033V were all from Sigma-Aldrich and had been
extracted by the phenol-chloroform-light petroleum method.
2. Solvents: Ultra-pure water was obtained with a Millipore
Milli-­Q® system (resistivity >18 MΩ cm). Chloroform and
methanol were of Normapur grade.
3. Acids: Isobutyric acid (synthesis grade) and citric acid
(Normapur grade).
4. Bases: Ammonia 28% solution (Normapur grade), methyl-
amine 41% solution (purum grade), triethylamine.
5. MALDI-MS matrix: 2,5-Dihydroxybenzoic acid (DHB,
purum grade).

2.3 Mass Matrix-assisted laser desorption/ionization mass spectrometry

Spectrometry (MALDI-MS) instructions can be followed using appropriate
MALDI-MS system.
The presented MS experiments were performed in the linear
mode, with delayed extraction, using a Perseptive Voyager STR
(PE Biosystem, France) and/or a Shimadzu Axima Performance
170 Alexey Novikov et al.

time-of-flight mass spectrometer. The reflectron mode can be used

to obtain a better resolution, but it could also lead to underestima-
tion, or even loss, of some fragile molecular species (e.g., lipids A
with substituted phosphate groups) due to their fragmentation
between the ion-source and the reflector. Lipid A, cores, and LPS
mass-spectra are recorded in the negative-ion mode at 20 kV with
the adjustment of the extraction delay time to optimized resolu-
tion and signal-to-noise ratio. Spectra obtained in the positive-ion
mode give important additional fragmentation data, but they will
not be presented here [10] (see Note 1).
Analyte ions are desorbed from the DHB matrix by pulses
from a 337 nm nitrogen laser. Addition of citric acid at a 0.1 M
concentration to the matrix solution was found to greatly improve
the resolution and signal-to-noise ratio of LPS, cores, and lipid A
mass spectra [11]. Citric acid chelates cations, and by insertion
between the molecules causes disaggregation of LPS and lipid A
micelles, thus improving solubility and cocrystallization of the ana-
lyte with the matrix resulting in a better desorption/ionization
process (see Note 2).

3 Methods

Isolation of lipid A is effected by mild hydrolysis of the acid-labile

link between the proximal Kdo residue of the oligo- or poly-­
saccharide and the lipid moiety of LPS. The most widely used
acetic-­acid hydrolysis method to liberate free lipid A can also cleave
other acid-labile lipid A bonds such as those linking glycosidic
phosphate, ethanolamine pyrophosphate, amino sugars {e.g.,
4-amino-4-deoxy-l-arabinopyranose (Arap4N) ] or 2-amino-­ 2-
deoxy-d-glucopyranose (GlcpN) [8] esterifying lipid A phosphate
groups}. A later developed mild hydrolytic method using sodium-
acetate at pH 4.4 [12] was modified by the addition of 1% SDS, in
order to disrupt micelles formed by amphiphilic LPS molecules
[13]. The modified method is presented here and used for the
comparison with the new micromethod (also presented below)
[4]. The latter procedure was found to be as mild as the former
SDS-promoted hydrolysis and did not modify, or eliminate, any
lipid A native element substitutions.
With the biological activities of LPS being intimately related
to their structures [14, 15], it was a priority to find methods giv-
ing quick information on their structures, their stability and
reproducibility with various batches, and their variability with
their production under different growth conditions. As lipid A is
responsible for most of the biological properties of the LPS mol-
ecule, it was necessary to find a new lipid A isolation method that
was fast, mild, and applicable to microscale bacterial samples.
 Micromethods for LPS Structural Analyses 171

These conditions were achieved by hydrolyzing lipid A directly

from the bacterial surface and using MALDI-MS for its charac-
terization, i.e., establishing its molecular heterogeneity, degree
of acylation, presence of some fatty acids, and other substituents
[4]. Our aim was to develop methods for the production of suf-
ficient material for several purposes in addition to direct MS
analysis, e.g., analyses by chromatography and application of
selective mild-alkaline treatments for the sequential liberation of
fatty acids to establish lipid A acylation patterns. The latter is
illustrated here with micro quantities of B. pertussis lipids A
directly isolated from bacterial cells. The lipids A, which were
isolated by micro-hydrolysis of the bacteria, can thus be further
characterized in a few hours as schematized in Fig. 1 with B.
pertussis lipid A structure.
The triethylamine-citrate method gives quick data for the
structural analysis of lipid A and PS moieties after mild hydrolysis
of the native LPS molecules. The method is illustrated with three
A. pleuropneumoniae (App) field isolates and underlines the vari-
ability of core structures for LPS in a single species and even sero-
type. This method allows, at the same time, the screening of lipids
A. In LPS commercial batches, it is indispensable to check the LPS
structures present in a given sample as they could vary from one
batch to another. The method, illustrated here with S. Minnesota
Re595 LOS confirms its important role and reliability.

3.1 Isolation of Lipid 1. Suspend 5 mg of LPS or LOS in 500 μL of a 10 mM sodium-­

A from LPS/LOS acetate buffer (pH 4.4) containing 1% SDS in a screw-cap
on a Normal Analytical tube. Place in an ultrasonic bath to obtain a clear solution (see
Scale [13] (See Note 3) Notes 4 and 5).
2. Heat the sample at 100 °C for 1 h in an Eppendorf®-
Thermomixer system or similar, under stirring at 1000 rpm.
The duration can be adjusted depending on the solubility of
the samples.
3. Cool in an ice-water bath and lyophilize the sample with an
excess of water to avoid foam formation.
4. Remove SDS by washing with a mixture of 100 μL of water
and 500 μL of acidified ethanol (prepared by adding 10 μL of
4 M HCl to 2 mL of 95% ethanol). Centrifuge at 2000 × g for
10 min (see Note 6).
5. Wash the sample twice again with 500 μL of nonacidified etha-
nol and centrifuge.
6. Extract the lipid A from the dried sample by 500 μL of a mix-
ture of chloroform:methanol:water (3:1.5:0.25; by vol.).
7. Dry the lipid A extract, disperse it in 500 μL of water by soni-
cation, and lyophilize to yield the lipid A as a powder.
172 Alexey Novikov et al.

Fig. 1 Schematic representation of the analytical steps used for the analysis of lipid A isolated after hydrolysis
of freeze-dried bacteria. The latter were hydrolyzed in a mixture of isobutyric acid-1 M ammonium hydroxide
(5:3; by vol.) for 2 h at 100 °C. The doted arrow indicates alkaline treatment for 5 h at 50 °C with 28% ammo-
nium hydroxide leading to partial O-deacylation, and the black arrows indicate alkaline treatment for 5 h at
37 °C with 41% methylamine, leading to complete O-deacylation. The structure displayed is that of B. pertus-
sis lipid A

3.2 Isolation of Lipid This micromethod was aimed at using the TEA-citrate reagent for
A and PS/OS facilitating MS analysis thanks to its ability to disaggregate LPS
from LPS/LOS: A Rapid molecules as well as to retaining bivalent cations. The method is
MicroScale Method, thus used first for LPS analysis without heating, then for obtaining
Using Triethylamine-­ spectra of the separated lipid A and core constituents. MALDI-­
Citrate, and Preserving TOF is a semiquantitative MS method and it can be used to
Labile Constituents [9] ­determine the composition and the predominant LPS species pres-
ent in each sample (see Notes 3 and 7).
 Micromethods for LPS Structural Analyses 173

1. Suspend LPS/LOS samples in an Eppendorf® tube (see Note 8)

at a concentration of 5 μg/μL for LOS and 10 μg/μL for LPS
in 0.01 M TEA-citrate (1:1 molar ratio, pH 3.6). Disperse the
suspension by agitation (vortex) and use of ultrasonic bath if
available.
2. Dilute the 5–10 μg/μL stock solution to 1–2 μg/μL concen-
tration with 0.01 M TEA-citrate. Sonicate it again. The sample
is ready for MALDI-MS analysis as described in Subheading
3.4.
3. Split the lipid A and PS/OS regions by heating the initial sam-
ple (5–10 μg/μL) at 100 °C for 1 h in an Eppendorf®-
Thermomixer system or similar, under stirring at 1000 rpm
(see Note 9). The minimal final volume should be kept around
40 μL to avoid evaporation. Cool the samples for 5 min at 4 °C
or in an ice-water bath and then perform a short 20 s spin to
recover the drops from the top and vortex.
4. Analyze directly the hydrolyzed micro-samples by MS: Dilute
the hydrolysate to 1–2 μg/μL concentration with 0.01 M
TEA-­citrate and sonicate. The sample is ready for MS analysis.
Use DHB in 0.1 M aqueous citric acid as the matrix solution.
Both lipid A and PS-related molecular ions are expected to be
observed (see Notes 1, 7 and 10).
5. Lyophilize or desiccate the hydrolyzed samples.
6. Extract the excess of salt with methanol: add methanol at a
concentration of 10 μg/μL and suspend the sample by sonica-
tion using an ultrasonic bath. Centrifuge the tube at 4 °C,
7000 × g for 10 min. Discard the methanol supernatant. Good
practice would be to test the methanol supernatant by
MALDI-MS to be sure that no specific molecular species are
lost (see Note 6).
7. Extract the lipid A from the pellet with a mixture of
chloroform:methanol:water (3:1.5:0.25; by vol.) at a concen-
tration of 10 μg/μL. Suspend the pellet by sonication.
Centrifuge the tube at 4 °C, 7000 × g for 10 min. Recover the
supernatant.
8. Repeat step 7 with half a volume of the chloroform:methanol:
water mixture. Pool both chloroform:methanol:water superna-
tants. The sample is ready for lipid A analysis as described in
Subheading 3.4.
9. Suspend the pellet in water at 10 μg/μL and homogenize by
using vortex and sonication. Centrifuge the tube at 4 °C,
7000 × g, during 20 min. Dilute the supernatant containing
the PS moiety if necessary, then deposit it on the MALDI sam-
ple plate, cover with the matrix solution (10 μg/μL DHB in
0.1 M aqueous citric acid), and let it dry as described in
Subheading 3.4.
174 Alexey Novikov et al.

A quick version of the method also gives good results by test-

ing native LPS samples suspended in TEA-Citrate as
described. Then compare it to a sample of the same mixture
directly after treatment at 100 °C. In this case, there is no
need to remove salts; they were selected for direct use. The
longer version of the method presented above is only neces-
sary when samples have never been studied, when masses are
not already known, and when superposed lipid A and core
peaks cannot be sorted out. The quick method is very conve-
nient for rapid comparison between different batches, growth
conditions, species, or kinetics of hydrolysis (see Notes 3, 4,
5, and 9).

3.3 Isolation of Lipid 1. After collecting a culture pellet, wash the cells with a buffer or
A from Whole Cells: water, only if necessary as it could fragilize LPS molecules, and
A Micro Method [4] lyophilized it.
(See Note 3) 2. Suspend 10 mg of lyophilized bacteria in 400 μL of a mixture
of isobutyric acid:1 M ammonium hydroxide (5:3; by vol.)
under the hood, in a screw-cap tube and heat for 2 h at 100 °C
with stirring (see Notes 4 and 8).
3. Cool the sample to 4 °C, and centrifuge at 2000 × g for 15 min.
4. Dilute the supernatant with water (1:1; by vol.) and lyophilize
it.
5. Wash the material obtained twice with 400 μL of methanol and
centrifuge (2000 × g for 15 min).
6. Extract the insoluble lipid A twice with 100 μL of
chloroform:methanol:water (3:1.5:0.25; by vol.). This extract
can be used directly for MALDI-MS (see Notes 1 and 7).

3.4 MALDI-Mass 1. Take up LPS at 1 μg/μL in water. Desalt the solution with a
Spectrometry few grains of Dowex 50 W-X8 (H+) (see Note 1).
of Isolated LPS, Lipid 2. Take up lipid A in a mixture of chloroform:methanol:water
A, and Core moieties (3:1.5:0.25; by vol.) at 1 μg/μL or use directly the lipid A
[11] extract obtained in Subheading 3.2. Desalt the solution with a
few grains of Dowex 50 W-X8 (H+) (see Notes 1 and 10).
3. Take up core and/or PS moieties at 1 μg/μL in water. Desalt
the solution with a few grains of Dowex 50 W-X8 (H+) (see
Notes 1 and 10).
4. Drop 0.5–1 μL of the different solutions on the target.
5. Add 0.5–1 μL of the matrix solution (DHB dissolved at
10 μg/μL in the same solvent for lipid A, or in 0.1 M citric
acid in water for LPS and PS) and dry. Different ratios
between sample and matrix have to be tested for defining the
best concentrations (0.5:1/1:1/1:0.5).
6. Submit the sample on the target to MALDI-MS analysis, as
described in Subheading 2.3 (see Notes 1 and 7).
 Micromethods for LPS Structural Analyses 175

3.5 Sequential Sequential fatty-acid release by mild-alkali treatment was found to

Liberation of Ester- be useful in establishing lipid A acylation patterns [5, 6]. Fatty
Linked Fatty Acids acids being ester-linked in direct acylation of the di-glucosamine
of LPS, or Lipid A, backbone (primary acylation) were released more readily than
by Mild Alkali those bound to the hydroxyl groups of other fatty acids (secondary
Treatment for Analysis acylation). We set up experimental conditions for breaking: (1)
by Mass Spectrometry only primary ester fatty acid bonds and (2) all fatty acid ester bonds.
The conditions were established from our experience with numer-
ous samples and proved that the technique can be scaled down if
LPS or lipid A supply is limiting (see Note 3).
1. Dry under a stream of nitrogen, lipid A solutions (50 μL),
obtained in Subheadings 3.2 and 3.3, or the equivalent amount
of LPS in an Eppendorf® tube (see Note 8).
2. In order to only liberate primary ester-linked fatty acids, add 50 μL
of 28% ammonium hydroxide, sonicate, then close the tube, and
stir for 5 h at 50 °C in a Thermomixer-system or in a simple ther-
mostated bath using magnetic stirring (see Notes 5 and 9).
3. Dry the sample under a stream of nitrogen.
4. Take up the modified lipid A into 50 μL of a mixture of
chloroform:methanol:water (3:1.5:0.25; by vol.) or into 50 μL
of water when a LPS sample is treated. This sample is ready to
be analyzed by MS, as described in Subheading 3.4 (see
Notes 1, 7 and 10).
5. In order to liberate both the primary and secondary ester-linked
fatty acids, repeat the above steps 1–4 replacing ammonium
hydroxide by 41% methylamine and keeping the mixture for
5 h at 37 °C under stirring.
6. If necessary, the liberated fatty acids can be recovered by
extraction and tested by GC/MS.

4 Examples of Analysis

4.1 Illustration The micro-method for lipid A isolation, directly from bacterial
of P. aeruginosa cells, is particularly useful for comparing the lipid A structures of
Lipid A Structural different species and strains of a genus, as well as for the same
Modifications ­bacteria grown under different culture conditions (e.g., media,
Induced by the temperature, oxygen tension, salt stress, bivalent ions concentra-
Switch Between tion, etc.), or colonizing different niches (e.g., environmental iso-
Planktonic and Biofilm lates versus clinical isolates) (see Note 3).
Lifestyles [3] Here, we give the example of lipids A isolated from P. aeruginosa
strain PAO1 grown in planktonic culture, in biofilm, then in plank-
tonic culture inoculated by biofilm-isolated bacteria (hereafter
assigned as “B → P” culture). Mass spectra of the lipids A, extracted
by the micro-method from ~4 mg samples of dried bacteria cultured
in these three ways, are presented in Fig. 2a–c, the corresponding
structures are presented in Fig. 2d. The observed modifications con-
tribute to higher inflammatory response in human monocytes [2].
Fig. 2 Negative-ion MALDI mass-spectra of lipids A isolated (a) from a planktonic culture of PAO1 bacteria, (b)
from biofilm ones, and (c) from Biofilm to planktonic cultures. 4 FA four fatty acids, 5 FA five fatty acids, 6 FA
six fatty acids, 7 FA seven fatty acids. (d) Structures of the major hexa-acyl lipid A species appearing at m/z
1617 and of the different penta-acyl lipid A species appearing at m/z 1431, 1447, 1463, and 1578. Comparison
of MALDI mass-spectra shows an important decrease of the 12:0(2-OH) content, and an increase of 12:0
content in the lipid A of biofilm cultured bacteria
 Micromethods for LPS Structural Analyses 177

4.2 Fatty Acids The kinetics of esterified fatty acid liberation from lipid A samples
Positioning was shown to be dependent on the substitution pattern, and could
on the Di-GlcN be used to determine the fatty-acid linkage. We first showed that
Backbone, liberation of primary ester-linked fatty acid (at C3 and C3′) was
by Sequential Alkaline much faster than that of the fatty acids in secondary ester linkages
Treatment [6]: (at C2 and C2′ acyloxyacyls), thus allowing a distinction between
Example them by the application of the two-steps protocol described in
of B. pertussis Lipid A Subheading 3.5 [6]. Furthermore, we observed that following
kinetics of primary fatty-acid liberation, we could discriminate
between the fatty acids located at the C3 and C3′ positions.
The kinetics presented in Fig. 3a was obtained from B. pertussis
Tohama I LPS sample [8] selected for its basically single lipid A
molecular-species composition. In the initial mass spectrum
(T = 0), only one lipid A major peak is observed at m/z 1559, cor-
responding to the structure displayed in Fig. 1. We followed the
evolution of the mass spectra as a function of time, during the
alkaline treatment (28% ammonium hydroxide at 50 °C). Mass
spectra were recorded after treatment times of 8 min, 15 min,
30 min, 1 h, and 2 h. Drastic differences in the fatty-acid-releasing
times were observed that related to their linkage position. Thus, at
C3, the 10:0(3-OH) is completely liberated only after 8–15 min,
transforming the initial penta-acyl lipid A into a tetra-acyl one (m/z
1389), while at C3′, the 14:0(3-OH) is only released after 1–2 h,
leading to disappearance of the peak at m/z 1389, and to the con-
secutive appearance of the one at m/z 1163. The time of release of
the acyloxyacyl 14:0 at C2′ is again much longer. In fact, to com-
pletely cleave this bond requires the use of stronger hydrolysis con-
ditions, as described in Subheading 3.5. Only after this second
treatment, the peak at m/z 953 remains in the spectrum, corre-
sponding to a di-acyl lipid A with amide-linked 14:0(3-OH) fatty
acids at C2 and C2′ (data not shown). Similar kinetics was obtained
from a lipid A sample (Fig. 3b). In this case, we used lipid A from
B. pertussis 1414 strain that was obtained by direct hydrolysis of
dried bacterial cells. Two major molecular species were present in
the initial spectrum at m/z 1559 (penta-acyl lipid A) and at m/z
1333 (tetra-acyl lipid A with a free position at C3′). In conclusion,
twice as long treatments were necessary to liberate the same fatty
acids from the lipid A sample compared to LPS, probably owing to
solubility differences between the two samples.

4.3 Illustration Cultures of three field isolates studied in this work were obtained
of the TEA-­Citrate from Professor Gottschalk laboratory, Montreal University,
Method with Faculty of Veterinary Medicine. The isolates were denoted as
A. pleuropneumoniae: FMV94-­ 9216, 00-7025 and 13-007 by the furnisher. All the
Core and Lipid three samples were of serotype 1 according to the structure of
A Screening their capsular polysaccharides (see Note 3). Preliminary analyses
Among Three Field of LPS by SDS-PAGE confirmed Smooth-type phenotypes to be
Isolates similar to that of a reference serotype 1 strain [16, 17].
178 Alexey Novikov et al.

Fig. 3 Negative-ion MALDI mass spectra of lipid A obtained from: (a) B. pertussis strain Tohama I LPS, and (b) B.
pertussis strain BP1414 lipid A. The corresponding time of hydrolysis with 28% ammonium hydroxide at 50 °C
is given in each spectrum. A clear difference is observed for fatty-acids release between the two samples, due
to differences in solubility. Both strains also display different natural degrees of heterogeneity
 Micromethods for LPS Structural Analyses 179

Crude LPS were extracted from lyophilized bacteria [18].

TEA-citrate hydrolysis was applied to these samples as described in
Subheading 3.2. MALDI-MS was performed as described in
Subheading 3.4 (see Notes 1, 7, 8, and 10).
The lipid A profiles of the three isolates are very similar
(Fig. 4a–c), with a major hexa-acyl molecular species appearing at

Fig. 4 MALDI-MS analysis of lipid A and core OS regions of LPS from three A. pleuropneumoniae serotype 1
field isolates. (a, b, c) Lipid A negative-ion mass-spectra obtained respectively for isolates FMV94-9216,
00-7025, and 13-007. (d, e, f) Core OS negative-ion mass-spectra obtained respectively for isolates
FMV94-­9216, 00-7025 and 13-007. (g) Structure of the major hexa-acyl lipid A molecular species from A.
pleuropneumoniae. (f) Complete and incomplete core OS structures of A. pleuropneumoniae serotype 1,
observed for the three field isolates under study. S Sugar
180 Alexey Novikov et al.

m/z 1825 (Fig. 4g). Minor penta- and tetra-acyl molecular species
are observed at m/z 1615 and at m/z 1389 respectively.
First, it should be noted that all core-OS peaks correspond to
their anhydrous form. This is consistent with the fact reported ear-
lier [17] that the only Kdo residue present is phosphorylated at the
C4 position. As shown earlier [19], this substitution is lost during
acid hydrolysis, by beta-elimination. Absence of satellite peaks at
+18 mass units corroborates the fact that the Kdo molecule is 100%
phosphorylated (see Note 5).
The core-oligosaccharide structure described for serotype 1
App [17] represents a deca-saccharide (Kdo + 4 Hep + 2 Glc + 2
Gal + GalNAc). The anhydrous form of this classical structure is
observed at m/z 1840 (see Fig. 4e and f). It is observed for the
field isolates 13-007 and 00-7025 (Fig. 4f and e respectively).
Isolate 00-7025 reveals the presence of nona-saccharides in which
the terminal GalNAc residue is missing (m/z 1636, Fig. 4e), and
isolate FMV94-9216 reveals major octa-saccharide missing the ter-
minal GalNAc-Gal disaccharide (m/z 1474, Fig. 4d).
For all the tree isolates core-OS molecular species are observed
substituted with one or two Gly residues at +57 or +114 mass units
[9]. For isolate 13-007 (Fig. 4f) major molecular species corre-
spond to nonsubstituted deca-saccharide with a smaller peak cor-
responding to one Gly substitution. For 00-7025 and FMV94-9216
isolates, the major peaks correspond to species substituted with
one Gly residue (m/z 1897, 1693, and 1531) as well as with two
Gly residues (m/z 1954, 1750, and 1588).
The LPS core moieties of this porcine pathogen were shown to
be responsible for adhesion to respiratory tract cells and mucus
[20]. The presence of Gly was reported to influence the vaccine
capacities of the corresponding structures [21].

4.4 Lipid It has been described in the early 90s [1] that different proportions
A Screening of Three of isobuyric acid-1 M ammonium hydroxide could be used as
Commercial LOS mobile phase for TLC analysis of LPS, lipid A, and core oligosac-
Batches from S. charides. Here, a new ratio between the two components was
Minnesota Re595 shown to be efficient for comparing the three Re595 deep-rough
mutants by HPTLC.
4.4.1 TLC
1. Deposit the solutions (2 μL) of the three Re595 mutant LPS
batches (25 μg/μL in water) in lanes of 0.5 cm on the baseline
of HPTLC Silica-gel-coated glass-plates.
2. Develop for 1.5 h for separating the different molecular species
in the solvent mixture made of isobutyric acid:1 M ammonium
hydroxide, (5:2; by vol.) (see Note 4).
3. Reveal the spots by charring (10% sulfuric acid in ethanol).
As shown in Fig. 5, the three LPS HPTLC profiles are fairly dif-
ferent and too much complicated for representing the well-­known
simplest LPS structure with lipid A and two Kdo molecules, they are
often referred to. Sample A contains at least 8–10 molecular species,
 Micromethods for LPS Structural Analyses 181

Fig. 5 HPTLC migration of the three samples (a) 50 μg SmRe595-1, (b) 60 μg

SmRe595-2, (c) 30 μg SmRe595-3 in isobutyric acid: 1 M ammonia (5:2; by vol.).
The three samples display different profiles of migration indicating structural
diversity and heterogeneity between them

sample B at least about the same number but with different types,
and with differences in the relative intensity of each type. Sample C
revealed to be the less heterogeneous, with 4–6 molecular species.
To better characterize the structural heterogeneity of each sample,
we performed a comparative MALDI-MS analysis of the three sam-
ples, as presented below.

4.4.2 TEA-Citrate The TEA-citrate treatment was used to quickly compare the struc-
Hydrolysis of the Three tures present in the three commercial LPS samples (see Note 3).
TLC-Tested Commercial S. The lipid A is generally composed of six fatty acids: 14:0(3-­
Minnesota Re595 LPS OH) linked at C-2 and C-3, 14:0[3-O(12:0)] linked at C-2′, and
Followed by MALDI-MS 14:0[3-O(14:0)] at C-3′. Posttranslational modifications often
Analysis of Their Lipids A appear with, for example, esterification of the 14:0(3-OH) at C2
by a 16:0. The phosphate groups on GlcpN I or II could be substi-
tuted by phosphoethanolamine (PEA) or Arap4N residues [22].
As shown in Fig. 6, and expected after the TLC results, the three
lipid A spectra showed differences and behaved differently upon
hydrolysis (see Notes 1, 7, 8, and 10):
–– For the three samples, the major peak appears at m/z 1797 and
corresponds to a hexa-acylated molecule. The dephosphory-
lated associated molecule (m/z) is present in different amounts
182 Alexey Novikov et al.

1797,4
1360,2
100
14:0(C3-14:0) 16:0
1280,1

1570,0
1587,2

1928,3

2035,3
H2PO3 ArapN4 14:0 ArapN4

1490,8
0
B

1797,6
100

1928,6 1920,9
Intensity

2035,2
1718,0
H2PO3 EtNPO3 EtNPO3

2158,5
1360,3

2052,1
1587,2
1507,2

2167,1
1280,4

H2PO3

1955,8
Arap4N
Arap4N

0
C
100 1797,4

2035,8
1920,8
1717,4

2158,3
1928,6
1587,2

1955,9

2052,1
1974,9
0
1500 2000 2500
m/z

Fig. 6 Negative-ion MALDI mass-spectra of lipid A after TEA-Citrate treatment. (a) SmRe595-1, 45 min of
treatment; (b) SmRe595-2, 20 min of treatment; (c) SmRe595-3, 45 min of treatment. EtNPO3 ethanolamine
phosphate, H2PO3 phosphate group

depending on the batch. Batch 2, which was submitted to a

shorter time of hydrolysis (20 min. vs 45 min. For the two
other samples), already presents the highest level of dephos-
phorylation, followed by batch 3 at 45 min. Batch 1 only dis-
plays a negligible peak of monophosphoryl lipid A. These
results demonstrate that some important dephosphorylation
can take place, even under very mild hydrolysis conditions.
From the huge dephosphorylation level observed just after
20 min of treatment in sample 2, we have to consider that the
preparation of this very sample leads to important physico-
chemical differences, and this impacts the experiments, at the
structural, and therefore biological level (see Note 9).
–– The peak at m/z 1920 corresponds to a hexa-acylated molecu-
lar species substituted with a PEA group. It is hardly detected
in batch 1.
–– The peak at m/z 1928 corresponds to the 1797 molecular spe-
cies substituted with an Arap4N residue. This substitution is
present in the three batches.
–– Lipid A molecular species containing 16:0 (corresponding to
peaks at m/z 2035) vary in intensity from one to another
batch, and batches 2 and 3 display this palmytoylated form,
also substituted with PEA (m/z 2158).
 Micromethods for LPS Structural Analyses 183

–– The penta-acylated lipid A molecular species are observed in

the three batches at m/z 1587 (with their dephosphorylated
equivalent molecular species at m/z 1507 for batch 2).
Proportions vary greatly depending on each batch.
–– The tetra-acylated lipid A molecular species appears as a major
peak in the batch 1 spectrum, and such structures are known
for their antagonist activity. It is present at a much smaller pro-
portion in Batch 2 and absent from batch 3.
The corresponding described lipid A structures are given in
Fig. 7. The influence of the number and nature of fatty acids in
biological studies has already been stressed [23].
Different commercial batches from a single supplier, or from
different ones, cannot be considered identical, just by referring to
their name.

HO HO
O O
O O
HO P O HO P O O
O
OH HO NH O OH O NH O
HO HO
O O O
O O
NH NH O P OH
O O P OH O O O
O O O
O OH O O O OH
O
OH OH
OH OH

MW=1361,65 MW=1798,37
HO
O
O
HO P O O
H2N
H HO OH O NH O
HO
O O O O
HO O O
O P O O NH
O O P OH
OH OH O NH O O O O
HO O OH
O O O O O
NH O P OH O
O O O O OH O
O O OH
O
OH
OH

MW=1929,49 MW=2036,77

Fig. 7 Structure of the predominant lipid A molecular species present in each sample. (a) SmRe595-1,
(b) SmRe595-2, (c) SmRe595-3
184 Alexey Novikov et al.

H2N
H HO HO
O O
O O O
HO O P O HO P O
O O
OH OH O NH HO O NH2 OH O NH HO O NH2
O O O O O O O O
NH NH O P O P
O O P O P O O O
O O O O O O
O OH OH O O OH OH
O O O
OH O
OH OH O

MW=2052,54 MW=2159,82

Fig. 7 (continued)

5 Notes

1. MALDI-MS is a very convenient MS process; however,

depending on the temperature, humidity, or other variables
the crystallization can be a critical step and has to be repeated
if a spectrum is weak or absent.
2. Other MS methods can be successfully performed, but they are
often more dependent on the presence of salt and other con-
taminants, thus requiring further purification steps.
3. The methods presented here are rapid, facile, and effective
procedures for the analysis of microscale samples and for fol-
lowing the evolution and biosynthesis of lipid A structures
under different growth conditions, as illustrated. The method
can also be readily adapted to the verification of quality and
reproducibility of the bacterial production.
4. Each method can be experimentally adapted to the starting
material and to the amount of bacterial sample.
5. A phosphorylated terminal Kdo unit can render the glycosidic
bond less labile [24]. Depending on the mass ratio between the
glycosidic and lipid moieties, the degree of acylation of the lipid
A, as well as the presence of polar groups, the solubility of the
samples can be highly variable. As with all methods, adjustments
in the experimental conditions are sometimes required.
6. In the large-scale LPS hydrolysis method (Subheading 3.1),
the ethanol supernatant should remain clear during the SDS
extraction step; otherwise, some lipid A might be lost in this
fraction. If this is the case, ethanol should be concentrated and
 Micromethods for LPS Structural Analyses 185

added back to the sample before re-extraction, and the cen-

trifugation step should be performed at a higher speed and
lower temperature. The same applies to any “washing steps.”
7. During MS analysis, different ratios between sample and matrix
volumes should be tested to obtain the best-quality spectra
with respect to signal-to-noise ratio and to the resolution of
the peaks.
8. Polypropylene screw cap Eppendorf® tubes can be used for all
stages of microscale sample preparation. Preliminary tested
and certified by the manufacturer tubes should be used. Use of
tubes of unknown origin and poor plastic quality can lead
to appearance in mass-spectra of plastic related artifact peaks
(a polymer with 44 mass units repeating pattern is putatively
attributed to polyethylene glycol). It is therefore recom-
mended to test all new batches.
9. Each hydrolysis time is given as the optimum for the large
amount of different LPS samples tested; however if some
dephosphorylation is observed in the spectra, the acid hydroly-
sis time should be reduced and adjusted for a given LPS.
10. The desalting process performed with the Dowex resin before
MS analysis is a crucial step. It can be easily effected by the
deposition of a drop of water on a piece of Parafilm® on which
a few grains of resin are added and mixed. Mixtures with chlo-
roform are desalted in an Eppendorf® tube.

Acknowledgments

Aude Breton is a recipient of a CIFRE grant in collaboration with

the LPS-BioSciences company. We thank Pr. Gottschalk for his
kind exchanges and expertise with App strains and isolates.

References
1. Caroff MGL, Karibian D (1990) Several uses 4. El Hamidi A, Tirsoaga A, Novikov A, Hussein
for isobutyric acid-ammonium hydroxide sol- A, Caroff M (2005) Microextraction of bacte-
vent in endotoxin analysis. Appl Environ rial lipid A: easy and rapid method for mass
Microbiol 56:1957–1959 spectrometric characterization. J Lipid Res
2. Caroff M, Brisson JR, Martin A, Karibian D 46:1773–1778
(2000) Structure of the Bordetella pertussis 5. Aussel L, Therisod H, Karibian D, Perry MB,
1414 endotoxin. FEBS Lett 477:8–14 Bruneteau M, Caroff M (2000) Novel varia-
3. Ciornei CD, Novikov A, Beloin C, Fitting tion of lipid A structures in strains of different
C, Caroff M, Ghigo JM, Cavaillon JM, Yersinia species. FEBS Lett 465:87–92
Adib-­ C onquy M (2010) Biofilm-forming 6. Tirsoaga A, El Hamidi A, Perry MB, Caroff M,
Pseudomonas aeruginosa bacteria undergo Novikov A (2007) A rapid, small-scale p­ rocedure
lipopolysaccharide structural modifications for the structural characterization of lipid A
and induce enhanced inflammatory cyto- applied to Citrobacter and Bordetella strains: dis-
kine response in human monocytes. Innate covery of a new structural element. J Lipid Res
Immun 16:288–301 48:2419–2427
186 Alexey Novikov et al.

7. Caroff M, Deprun C, Richards JC, Karibian D 16. Altman E, Brisson JR, Perry MB (1986)
(1994) Structural characterization of the lipid Structure of the O-chain of the lipopolysaccha-
A of Bordetella pertussis 1414 endotoxin. ride of Haemophilus pleuropneumoniae sero-
J Bacteriol 176:5156–5159 type 1. Biochem Cell Biol 64:1317–1325
8. Marr N, Tirsoaga A, Blanot D, Fernandez R, 17. Caroff M (2004) Novel method for isolating
Caroff M (2008) Glucosamine found as a sub- endotoxins. PCT WO04062690, 2005
stituent of both phosphate groups in Bordetella 18. Michael FS, Brisson JR, Larocque S, Monteiro
lipid A backbones: role of a BvgAS-activated M, Li J, Jacques M, Perry MB, Cox AD
ArnT ortholog. J Bacteriol 190:4281–4290 (2004) Structural analysis of the lipopolysac-
9. Chafchaouni-Moussaoui I, Novikov A, Bhrada charide derived core oligosaccharides of
F, Perry MB, Filali-Maltouf A, Caroff M (2011) Actinobacillus pleuropneumoniae serotypes 1,
A new rapid and micro-scale hydrolysis, using 2, 5a and the genome strain 5b. Carbohydr
triethylamine citrate, for lipopolysaccharide Res 339:1973–1984
characterization by mass spectrometry. Rapid 19. Lebbar S, Haeffner-Cavaillon N, Karibian D,
Commun Mass Spectrom 25:2043–2048 Le Beyec Y, Caroff M (1995) 252Cf-plasma
10. Karibian D, Brunelle A, Aussel L, Caroff M desorption mass spectrometry analysis of lip-
(1999) 252Cf-plasma desorption mass spec- ids A obtained by an elimination reaction
trometry of unmodified lipid A: fragmentation under mild conditions. Rapid Commun Mass
patterns and localization of fatty acids. Rapid Spectrom 9:693–696
Commun Mass Spectrom 13:2252–2259 20. Ramjeet M, Deslandes V, St Michael F, Cox
11. Therisod H, Labas V, Caroff M (2001) Direct AD, Kobisch M, Gottschalk M, Jacques M
microextraction and analysis of rough-type (2005) Truncation of the lipopolysaccharide
lipopolysaccharides by combined thin-layer outer core affects susceptibility to antimicro-
chromatography and MALDI mass spectrom- bial peptides and virulence of Actinobacillus
etry. Anal Chem 73:3804–3807 pleuropneumoniae serotype 1. J Biol Chem
12. Rosner MR, Tang J, Barzilay I, Khorana HG 280:39104–39114
(1979) Structure of the lipopolysaccharide 21. Gamian A, Mieszała M, Lipiński T, Zielińska-­
from an Escherichia coli heptose-less mutant: Kuźniarz K, Gawlik-Jędrysiak M, Dzierzba K,
I. Chemical degradations and identification of Pietkiewicz J, Szeja W (2011) O-aminoacylation
products. J Biol Chem 254:5906–5917 of bacterial glycoconjugates: from native struc-
13. Caroff M, Tacken A, Szabo L (1988) Detergent- ture to vaccine design. Curr Pharm Biotechnol
accelerated hydrolysis of bacterial endotoxins 12:1781–1791
and determination of the anomeric configura- 22. Karibian D, Deprun C, Caroff M (1993)
tion of the glycosyl phosphate present in the Comparison of lipids A of several Salmonella and
“isolated lipid A” fragment of the Bordetella per- Escherichia strains by 252Cf plasma desorption
tussis endotoxin. Carbohydr Res 175:273–282 mass spectrometry. J Bacteriol 175:2988–2993
14. Caroff M, Karibian D (2003) Structure of bac- 23. Heine H, Rietschel ET, Ulmer AJ (2001) The
terial lipopolysaccharides. Carbohydr Res biology of endotoxin. Mol Biotechnol
338:2431–2447 19:279–296
15. Caroff M, Karibian D, Cavaillon JM, Haeffner-­ 24. Caroff M, Lebbar S, Szabó L (1987) Do endo-
Cavaillon N (2002) Structural and functional toxins devoid of 3-deoxy-d-manno-2-­­octulosonic
analyses of bacterial lipopolysaccharides. acid exist? Biochem Biophys Res Commun
Microbes Infect 4:915–926 143:845–847
 Chapter 17

Mass Spectrometry for Profiling LOS and Lipid

A Structures from Whole-Cell Lysates: Directly from a Few
Bacterial Colonies or from Liquid Broth Cultures
Béla Kocsis, Anikó Kilár, Szandra Péter, Ágnes Dörnyei, Viktor Sándor,
and Ferenc Kilár

Abstract
Lipopolysaccharides (LPSs, endotoxins) are components of the outer cell membrane of most Gram-­
negative bacteria and can play an important role in a number of diseases of bacteria, including Gram-­
negative sepsis. The hydrophilic carbohydrate part of LPSs consists of a core oligosaccharide (in the case
of an R-type LPS or lipooligosaccharide, LOS) linked to an O-polysaccharide chain (in the case of an
S-type LPS), which is responsible for O-specific immunogenicity. The hydrophobic lipid A anchor is com-
posed of a phosphorylated diglucosamine backbone to which varying numbers of ester- and amide-linked
fatty acids are attached and this part of the LPSs is associated with endotoxicity. The detailed chemical
characterization of endotoxins requires long-lasting large-scale isolation procedures, by which high-purity
LPSs can be obtained. However, when a large number of bacterial samples and their LPS content are to be
compared prompt, small-scale isolation methods are used for the preparation of endotoxins directly from
bacterial cell cultures. The purity of the endotoxins extracted by these methods may not be high, but it is
sufficient for analysis.
Here, we describe a fast and easy micromethod suitable for extracting small quantities of LOS and a
slightly modified micromethod for the detection of the lipid A constituents of the LPSs from bacteria
grown in different culture media and evaluate the structures with mass spectrometry. The cellular LOS and
lipid A were obtained from crude isolates of heat-killed cells, which were then subjected to matrix-assisted
laser desorption/ionization mass spectrometry analysis. The observed ions in the 10-colony samples were
similar to those detected for purified samples. The total time for the sample preparation and the MS analy-
sis is less than 3 h.

Key words Crude cell lysate, Lipid A isolation, Matrix-assisted laser desorption/ionization mass
spectrometry, Microextraction, LOS

1 Introduction

Lipopolysaccharides (LPSs) are components of the outer cell mem-

brane of most Gram-negative bacteria. They are also called endo-
toxins, as they may possess a distinctive range of biological effects
including lethal toxicity, typically by the intense activation of the

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6_17, © Springer Science+Business Media LLC 2017

187
188 Béla Kocsis et al.

complement and cytokines system, leading to septic shock [1]. The

basic structure of a smooth or S-type LPS molecule comprises three
covalently linked regions: the “lipid A” phosphoglycolipid, the
“oligosaccharide core,” and the “O-polysaccharide chain” (or
“O-antigen”). A second type of LPSs, called rough or R-type LPS,
or more correctly lipooligosaccharide, LOS, consists of only the
lipid A and core moieties. Toxicity is associated with the lipid A
moiety of LPSs and strongly depends on the acylation and phos-
phorylation patterns.
The development of techniques for the isolation of LPSs from
the bacterial outer membrane is important with respect to investi-
gations of LPS structure and function. Highly purified LPSs can be
obtained by large-scale extraction and long-lasting (5–10 days)
purification processes, such as the phenol–chloroform–light petro-
leum [2] or the hot phenol–water [3] methods. However, during
the isolation procedure the LPSs might undergo degradation
through loss of phosphate (P), phosphoethanolamine (PEtN),
and/or fatty acyl chains, which are important in the host-pathogen
interactions [4]. Several scaled-down analytical procedures have
been developed in an effort to determine native LPS [5, 6] and
particularly lipid A structures [7–9] as quick as possible from small
numbers of bacterial cells. Routine and easy methods are essential,
for instance, to determine structural variations resulting from var-
ied cell growth conditions, or to compare LPSs from laboratory
strains and clinical isolates, or in the quality control of endotoxins
prepared for human vaccines.
Matrix-assisted laser desorption ionization (MALDI) mass
spectrometry is widely used for the structural characterization of
native LOS and chemically hydrolyzed lipid A from different bac-
teria, whereas the elucidation of intact LPS (containing several
O-repeating units) with MS is often difficult [10].
Here, we present two extremely simple, small-scale isolation
methods: one for LOS and another for the lipid A part of either
LOS or LPS obtained directly from crude lysates of whole bacterial
cells, followed by their MALDI-TOF MS analysis [11]. The micro-
methods are based on the disruption of the intact cells—taken
directly from agar plate cell cultures or liquid culture media—by
the application of heat (100 °C) to release LPS or LOS from the
cell membrane, and the partial purification by washing the culture
pellets with pure water to remove intracellular components (nucleic
acid, proteins) and other water-soluble cell wall materials. The lipid
A constituents of the LPS/LOS are hydrolyzed directly from the
heat-killed cells of R- or S-type bacteria with aqueous citric acid
solution at elevated temperature (100 °C). The resulting LOS and
lipid As are found in partially purified forms in the culture pellet,
which can be isolated with centrifugation and directly subjected to
MALDI-MS analysis without further purification or freeze-drying.
The entire procedure, including sample preparation and MS analy-
 Mass Spectrometry of R-LPS and Lipid A from Bacterial Cell Lysates 189

sis, can be completed in less than 3 h, and it is probably the easiest

procedure for preparing LOS or lipid A, practically free from pro-
tein and nucleic acid, although some bacterial contaminants (e.g.,
lipid co-extractives) may be present.
This quick and simple approach could provide a rapid and effi-
cient way to give preliminary information on the LOS and lipid A
structures from small quantities of cells, as illustrated analyzing the
LOS of two model bacteria, Salmonella Minnesota R595 (Fig. 1)
and Shigella sonnei R41 (Fig. 2), and the lipid A parts of two
Escherichia coli bacteria (Fig. 3). The observed ions in the whole-­
cell lysate LOS samples were essentially similar to those detected
for the purified samples. Due to the mild conditions of this method,
biologically important but labile substituents (such as PEtN),
which were not previously reported, have been detected (Fig. 2).
Also, some dephosphorylation was observed in the purified R-LPS
samples, as shown in their mass spectra.

2 Materials

2.1 Preparation 1. Gram-negative bacterial stock is stored at −80 °C.

of Pure LOS 2. Distilled water and diethyl ether.
3. Rotary evaporator and flask.
4. Ultracentifuge and rotors.
5. Lyophilizator (freeze-dryer).
6. Extraction of pure LOS is carried out with the phenol-­
chloroform-­ light petroleum [2]. For extraction, mix liquid
phenol (90 g dry phenol and 11 mL water), chloroform, and
light petroleum in a volume ratio of 2:5:8.

2.2 Bacterial Cell 1. Store Gram-negative bacterial stocks at −80 °C. The following
Culture and Lysis strains are used for the analyses of LOS with known structures:
S. Minnesota R595, S. sonnei R41; and for the lipid A analysis:
E. coli O55 and ATCC 25922.
2. Mueller-Hinton agar plates.
3. 3 mL Mueller-Hinton broth: 3.0 g Beef extract, 17.5 g casein
acid hydrolysate, 1.5 g starch in 1 L deionized water. Store the
broth at 4 °C.
4. High-purity water (LC-MS Chromasolv grade), citric acid.
5. Shaker incubator for maintaining cultures at 37 °C.
6. Microcentrifuge containing rotor for 1.5–2.0 mL microcentri-
fuge tubes.
7. Vortex mixer.
Fig. 1 Negative-ion, linear mode MALDI-TOF mass spectra of purified LOS from S. Minnesota R595 (a) and
R-type LPSs isolated from crude cell lysates of the S. Minnesota R595 bacterium grown in liquid medium (b)
or on agar plate (c). DHB was used in (a) and (c), and THAP matrix in (b). +Na indicates a sodium adduct. The
predicted structures for the detected [M−H]− ions are shown. The full structure of the R-LPS from S. Minnesota
R595 has been reported earlier [12]
Fig. 2 Negative-ion, linear mode MALDI-TOF mass spectra of purified LOS from S. sonnei R41 (a) and LOS isolated
from crude cell lysates of the S. sonnei R41 bacterium grown in liquid medium (b) or on agar plate (c). As MALDI
matrix THAP was used. +Na and +K indicate a sodium and a potassium adduct, respectively. The predicted struc-
tures for the detected [M−H]− ions are shown. The full structure of the LOS from S. sonnei R41 has been reported
earlier [13]. The phosphate group (for instance at C1) could provide an attachment site for the PEtN moiety
192 Béla Kocsis et al.

Fig. 3 Negative-ion, reflectron mode MALDI-TOF mass spectra of lipid A isolated from crude cell lysates of E.
coli O55 (a) and E. coli ATCC 25922 bacteria (b) grown on agar plate. As MALDI matrix DHB was used. The
predicted structures for the detected [M−H]− ions are shown. Fatty acid alkane chain length heterogeneities
for the major hexa-acylated ion at m/z 1796 are indicated by the mass differences of ±14 u (–CH2– group)
 Mass Spectrometry of R-LPS and Lipid A from Bacterial Cell Lysates 193

2.3 MS 1. Dowex 50WX8-200 (H+) cation-exchange resin that has been

converted into its ammonium form (see Note 1).
2. Ammonium-hydroxide solution, high-purity water.
3. Ultrasonic bath sonicator.
4. THAP matrix: saturated (ca. 20 mg/mL) solution of
2′,4′,6′-trihydroxyacetophenone monohydrate in
acetonitrile:water (1:1, v/v) prepared in a microcentrifuge
tube. Mix and bath-sonicate for 20 min. Vortex for 1 min and
centrifuge at 12,000 × g for 5 min before use.
5. DHB matrix: saturated (ca. 10 mg/mL) solution of
2,5-­dihydroxybenzoic acid in 0.1 M citric acid aqueous solu-
tion prepared in a microcentrifuge tube. Mix and bath-sonicate
for 20 min. Vortex for 1 min and centrifuge at 12,000 × g for
5 min before use.
6. Peptide Calibration standard mixture (Bruker Daltonics,
Bremen, Germany).
7. MALDI plate (stainless-steel target).
8. Mass spectrometer: e.g., Autoflex II MALDI-TOF MS instru-
ment (Bruker Daltonics, Bremen, Germany) equipped with a
1.2 m drift tube and a nitrogen laser (337 nm) using a 50 Hz
firing rate.
9. Flex Analysis 2.4 software packages (Bruker Daltonics, Bremen,
Germany).

3 Methods

3.1 Preparation 1. Grow the bacteria at 37 °C in Mueller–Hinton broth in a fer-

of Pure LOS mentor (e.g., Biostat U30D, Braun Melsungen, Germany)
until the late logarithmic phase (about 10 h) and then collect
the cells by centrifugation.
2. Extract LOS in pure form from acetone-dried bacteria accord-
ing to the method of Galanos et al. [2], which is an efficient
extraction method for LOS from rough bacteria. Suspend the
bacteria (20 g) in 100 mL phenol-chloroform-light petroleum
(2:5:8, v/v/v) and stir for 20 min. Precipitate the LOS by add-
ing dropwise distilled water from the phenol phase which is
obtained after the evaporation of chloroform and light petro-
leum in a rotary evaporator. Wash the precipitated LOS with
diethyl ether, then dry in a fume hood, dissolve in distilled
water, and centrifuge at 100,000 × g ultracentrifugation for
4 h. Re-suspend the sediment in distilled water and freeze-dry.
3. Store the LOS in lyophilized form, and prepare 1 mg/mL
aqueous suspensions before use.
194 Béla Kocsis et al.

3.2 Isolation of LOS 1. Streak bacterial stocks on a Mueller-Hinton agar plate and
from Whole Cells grow overnight at 37 °C in a bacterial incubator.
3.2.1 Isolation of LOS 2. Transfer one colony into 3 mL of culture medium (Mueller-­
from Cells Grown in Liquid Hinton broth) using a sterile inoculating loop and incubate at
Medium 37 °C on a shaker overnight.
3. Transfer the cell culture into an Eppendorf tube in two por-
tions (1.5–1.5 mL) and centrifuge at 6000 × g for 3 min (trans-
fer the second portion into the same Eppendorf tube after the
supernatant from the first portion has been discarded). Discard
the supernatant.
4. Re-suspend the pellets with 1 mL of water with a vortex and
centrifuge at 6000 × g for 3 min. Discard the supernatant.
Repeat this step two times.
5. Re-suspend the pellets with 1 mL of water and incubate at
100 °C for 30 min.
6. After cooling at room temperature, centrifuge the sample at
12,000 × g for 5 min. Discard the supernatant, and the sedi-
ment (ca. 10 μL) containing LOS is ready to use.

3.2.2 Isolation of LOS 1. Streak bacterial stocks on a Mueller-Hinton agar plate and
from Cells Grown on Agar grow overnight at 37 °C in a bacterial incubator.
Plate 2. Transfer ten colonies (about 1.5 mg) with a sterile inoculating
loop into an Eppendorf tube containing 150 μL of water and
mix by repeatedly pipetting the suspension up and down.
3. Incubate the cell suspension at 100 °C for 30 min.
4. After cooling at room temperature, centrifuge the sample at
12,000 × g for 5 min. Discard the supernatant, and the sedi-
ment (ca. 10 μL) containing the LOS is ready to use.

3.3 Isolation of Lipid 1. Streak bacterial stocks on Mueller-Hinton agar plate and grow
A from Whole Cells overnight at 37 °C in a bacterial incubator.
2. Transfer ten colonies (about 1.5 mg) with a sterile inoculating
loop into an Eppendorf tube containing 1 mL of 0.1 M citric
acid aqueous solution (see Note 2).
3. Mix the cell suspension by vortexing and incubate at 100 °C
for 90 min.
4. After cooling at room temperature, centrifuge the samples at
12,000 × g for 5 min. Discard the supernatant, and the sedi-
ment (ca. 10 μL) containing free lipid A is ready to use.

3.4 Sample 1. Mix 10 μL of pure LPS or LPS or lipid A from cell lysate sus-
Preparation for Mass pensions with 10 μL of 0.1 M citric acid solution in an
Spectrometry Eppendorf tube and sonicate for 10 min (see Note 3).
2. Desalt 5 μL of the sample suspension with some grains (ca.
5 μL suspension) of Dowex 50WX8-200 (NH4+) cation-
exchange beads (see Note 4).
 Mass Spectrometry of R-LPS and Lipid A from Bacterial Cell Lysates 195

3. Deposit 0.5–1 μL from this sample suspension onto a spot of

the MALDI plate (see Note 5) and mix gently (see Note 6)
with 0.5–1 μL of the saturated solution of the DHB matrix
(dissolved in 0.1 M citric acid solution) or the THAP matrix
(dissolved in acetonitrile–water mixture (1:1, v/v)) (see Note
7) and analyze immediately after drying (see Notes 8–10).
4. Submit the sample on the target to MALDI–MS analysis and
acquire spectra by scanning the sample for optimal ion signals.

3.5 Determination The presented MS experiments were done in linear or reflectron

of Purified and Whole- modes (see Note 11). Acquire mass spectrometry data in the nega-
Cell Lysate LOS tive ionization mode at 19 kV, using pulsed ion extraction (allow-
and Lipid A Samples ing 120 ns delay between generation and extraction/acceleration
with MALDI-TOF Mass of the ions). Adjust the laser power between 65% and 85% of its
Spectrometry maximal intensity (see Note 12). Record the mass spectra of the
LOS over the m/z range 900–5000, and of the lipid A over the
m/z range 900–2500. One spectrum was the sum of 500 laser
shots on a sample spot. Perform the calibration of the instrument
externally using a peptide calibration standard. Identify the LOS
and lipid A components according to the molecular mass of their
quasimolecular [M−H]− ions.
Examples of analysis are given for two LOS samples from S.
Minnesota R595 (Fig. 1) and S. sonnei R41 (Fig. 2) with known
structures [12, 13], and for two lipid A samples from E. coli O55
and ATCC 25922 (Fig. 3). In all cases, a complex pattern of ions
was obtained in the mass spectra due to natural sample heteroge-
neity and in-source fragmentation [10]. Fine structural variations
were attributable to different degrees of lipid A acylation and phos-
phorylation, as well as the presence of phosphate substituents [e.g.,
4-amino-4-deoxy-l-arabinopyranose (Arap4N) or ethanolamine].
The high-intensity signals for the individual LOS components
from the cell lysates were of the same m/z, as observed in the spec-
tra of the high-purity LPS analogs. However, some differences
could be seen in the relative signal intensities, the appearance of
dephosphorylation products in the purified samples, and the
­detection of an additional PEtN substituent in the whole-cell lysate
LOS sample of S. sonnei R41 (Fig. 2).
The mass spectra of the two E. coli lipid As (Fig. 3) isolated
from whole cells displayed all the expected mass spectral peaks—
with high signal-to-noise ratio—of an E. coli-type lipid A.

4 Notes

1. The Dowex 50WX8-200 (H+) cation-exchange resin has to be

converted into its ammonium form suitable for sample desalt-
ing. Wash about 5 mL of the protonated form of the commer-
cial resin two times with 25 mL of 1 M NH4OH solution for
196 Béla Kocsis et al.

10 min, and once with 10 mL of 0.1 M NH4OH solution for

10 min, with continuous stirring. Then, wash the resin five to
six times with 10 mL of distilled water, until the supernatant is
neutral and store in distilled water for usage.
2. By boiling the sample in 0.1 M citric acid (pH 2.9), the lipid
part is cleaved from the terminal Kdo (3-deoxy-d-manno-oct-­­
2-ulosonic acid) residue of the oligo- or polysaccharide part.
We have obtained similar results by applying 1% (v/v) acetic
acid (pH 3.5), the most widely used mild-acid hydrolysis
method to liberate free lipid A from purified LPS [14]. However,
with both procedures there is a risk of partially losing other
acid-labile linkages, such as phosphate, phosphoethanolamine,
amino sugars (e.g., Arap4N), or ester-linked fatty acid.
3. Citric acid chelates the divalent cations (e.g., Ca2+ or Mg2+)
present in the sample, and promotes disaggregation of LPS
and lipid A micelles by insertion between the molecules, thus
improving solubility.
4. The presence of contaminating alkali metal ions (primarily Na+ and
K+) can lead to adduct ion formation {[M + nNa + mK − (n + m + 1)
H]−, n, m = 0, 1, 2,…}. Multiple peaks (e.g., M + 23 − 2,
M + 39 − 2) provide complicated mass spectra and decrease sensi-
tivity. The effect of alkali metal cations can be substantially reduced
by rapid desalting with a cation exchange resin in the NH4+ form.
This desalting process can be carried out by the deposition of the
resin suspension on the top of a small piece of Parafilm on which
the same volume of sample suspension is added and mixed by
repeatedly pipetting the suspension up and down. However, if the
salts are not removed completely, small adduct signals could still be
observed in the MALDI mass spectra.
5. Care must be taken not to pipette any Dowex beads along with
the sample suspension onto the MALDI target, as they could
disturb the matrix crystallization and/or analyte incorporation.
6. The matrix-sample suspension is mixed on the sample plate by
repeatedly pipetting the suspension up and down. Putting the
sample first, followed by the matrix layer onto the MALDI
plate, does not change the quality of the MALDI spectra.
However, the premixing of the sample and matrix solutions in
a tube should be avoided as—in our case—this resulted in poor
quality of the spectra.
7. In general, the selection of matrix is empirical. Even though
DHB or THAP have proven to be the most efficient for the ion-
ization of LPS-derived samples, other matrices such as ATT
(6-aza-2-thiothymine) or CMBT (5-chloro-2-mercapto-benzo-
thiazole)—this latter applied only for lipid A samples—are also
commonly used for MALDI-TOF MS analysis in endotoxin
research.
 Mass Spectrometry of R-LPS and Lipid A from Bacterial Cell Lysates 197

8. Allow the plate to air dry in an area where it will not be exposed
to contaminants. The solvent evaporation using the DHB
matrix from aqueous solution takes usually less than 5 min,
while for the THAP matrix prepared in a volatile solvent takes
usually less than 2 min.
9. After the rapid solvent evaporation, the matrix-sample mixture
forms a crystalline precipitate. The DHB preparations result in
a finely dispersed homogeneous matrix-sample layer with very
dense, thick crystals, occasionally showing a white crust ring.
By contrast, the THAP preparations show a heterogeneous
spatial distribution with long crystals.
10. The crystallization can be a critical step and has to be repeated
if a spectrum is not obtained. The best ion formation occurs
from particular spots, called “hot spots,” a factor strongly
increasing measurement times. It is therefore important to
acquire and average many single-shot spectra from several
positions (usually 10–20) within a given sample spot to gain
representative sample data. One reason for hot spot formation
could be the heterogeneous incorporation of the analyte into
the co-crystallized matrix-sample complex.
11. The linear mode gives higher sensitivity but the resolution is
more improved in the reflectron mode. Masses obtained
with linear mode conditions are average masses, whereas in
reflectron mode, the isotopic patterns of the molecules are
resolved, so the mass values determined correspond to
monoisotopic masses. Because of the lower resolution
power of the TOF analyzer in linear mode, the peaks are
broader compared to the measurements with reflectron
mode conditions. The linear mode is used for the intact
R-type LPS endotoxins and the reflectron mode is used for
the lipid A samples.
12. As fluctuation of the laser power also contributes to the vari-
able signal intensities of the samples, it should be adjusted to
limit the acceptable analyte signal intensity above background
noise before the spectra are averaged.

Acknowledgments

The work was supported by the grants OTKA K-100667, OTKA

K-106044 and UNKP-16-4-III New National Excellence Program
of the Ministry of Human Capacities. Á.D. acknowledges the
financial support of the János Bolyai Research Scholarship
(Hungarian Academy of Sciences).
198 Béla Kocsis et al.

References
1. Rietschel ET, Brade H (1992) Bacterial endo- spectrometric characterization. J Lipid Res
toxins. Sci Am 267:54–61 46:1773–1778
2. Galanos C, Lüderitz O, Westphal O (1969) A 9. Zhou P, Chandan V, Liu X, Chan K, Altman E,
new method for extraction of R-lipopoly­ Li J (2009) Microwave assisted sample prepa-
saccharides. Eur J Biochem 9:245–249 ration for rapid and sensitive analysis of H.
3. Westphal O, Lüderitz O, Bister F (1952) Über pylori lipid A applicable to a single colony.
Die Extraktion Von Bakterien Mit Phenol J Lipid Res 50:1936–1944
Wasser. Z Naturforsch B 7:148–155 10. Kilár A, Dörnyei Á, Kocsis B (2013) Structural
4. Nummila K, Kilpelainen I, Zähringer U, Vaara characterization of bacterial lipopolysaccha-
M, Helander IM (1995) Lipopolysaccharides rides with mass spectrometry and on- and off-­
of polymyxin B-resistant mutants of Escherichia line separation techniques. Mass Spectrom Rev
coli are extensively substituted by 2-aminoethyl 32:90–117
pyrophosphate and contain aminoarabinose in 11. Kilár A, Péter Sz, Dörnyei Á, Sándor V, Kocsis
lipid A. Mol Microbiol 16:271–278 B, Kilár F (2016) Structural analysis of endo-
5. Hitchcock PJ, Brown TM (1983) Morphological toxins from bacterial culture suspensions.
heterogeneity among Salmonella lipopolysaccha- J Mass Spectrom—submitted for publication
ride chemotypes in silver-stained polyacrylamide 12. Brandenburg K, Wagner F, Müller M, Heine
gels. J Bacteriol 154:269–277 H, Andrä J, Koch MHJ, Zähringer U, Seydel
6. Thérisod H, Labas V, Caroff M (2001) Direct U (2003) Physicochemical characterization
microextraction and analysis of rough-type and biological activity of a glycoglycerolipid
lipopolysaccharides by combined thin-layer from Mycoplasma fermentans. Eur J Biochem
chromatography and MALDI mass spectrom- 270:3271–3279
etry. Anal Chem 73:3804–3807 13. Kilár A, Dörnyei Á, Bui A, Szabó Z, Kocsis B,
7. Yi EC, Hackett M (2000) Rapid isolation Kilár F (2011) Structural variability of endo-
method for lipopolysaccharide and lipid A from toxins from R-type isogenic mutant of Shigella
Gram-negative bacteria. Analyst 125:651–656 sonnei. J Mass Spectrom 46:61–70
8. El Hamidi A, Tirsoaga A, Novikov A, Hussein 14. Wilkinson SG (1996) Bacterial lipopolysaccha-
A, Caroff M (2005) Microextraction of bacte- rides—themes and variations. Prog Lipid Res
rial lipid A: easy and rapid method for mass 35:283–343
 Index

A E
Actinobacillus pleuropneumoniae����������������� 169, 171, 177–180 Electrochemical DNA (E-DNA)�����������������������������������9–11
Agarose gel��������������������������������������������� 41, 43–45, 128–129 Electropolymerization�������������������������������������������� 51, 54, 58
Analysis EndoLysa test��������������������������������������������������� 108, 110, 111
multivariate�����������������������������������������������������������������120 Endotoxin
principal component������������������������������������������� 120, 121 profiles������������������������������������������������ 153–156, 158, 168
Aptamer����������������������������������������������������� 9–21, 61–67, 134 removal���������������������������������������������65, 85–93, 107–112,
Assay 138, 163
fluorescence������������������������������������������������������������������31 unit (EU)��������������������������������������������������������������������107
toxicity�������������������������������������������������10, 29, 33, 38, 187 Enzyme-linked immunosorbent assay
(ELISA)����������1, 5, 50, 95, 108, 134, 143, 145, 148
B Escherichia coli������������������������������� 30, 33, 107–112, 135, 144,
Bacillus anthracis�����������������������������������������������������������61, 81 145, 154, 156, 158, 189, 192, 195
Bacillus cereus����������������������������������������������������������� 61–67, 81
F
Bacillus thuringiensis������������������������������������������������ 61, 67, 81
Bacteriophage Fatty acids
F8�������������������������������������������������������������������������������108 ester-linked�������������������������������������������������������� 168, 175,
HAP1�������������������������������������������������������������������������108 177, 196
T4����������������������������������������������������������������������� 108, 110 positioning������������������������������������������������������������������177
Bioconjugate probes�������������������������������������������������133–141 sequential liberation�������������������������������������������� 171, 175
Biosensing Fealden software����������������������������������������������� 12, 14, 15, 20
electrochemical�������������������������������������������������������17–19 Food
Biosensors������������������������������������������������������� 9–21, 134, 143 pathogen�����������������������������������������������������������������������61
Bloom���������������������������������������������������������������������������85, 91 poisoning����������������������������������������������������������������������61
Bordetella pertussis�������������������������������������������� 168, 171, 172, Fourier transform infrared spectroscopy (FTIR)
177, 178 acquisition�������������������������������������������������� 116–118, 122
Botulism amide I/II region��������������������������������������������������������114
protein����������������������������������������������������������������������9–21 fatty acid region����������������������������������������������������������114
fingerprint��������������������������������������������������������� 114, 121,
C 151–164
Chitosan frequencies���������������������������������������������� 18, 19, 114, 123
hydrogel������������������������������������������������������ 128, 129, 132 mixed region���������������������������������������������������������������114
LPS complex���������������������������������������������� 129, 130, 132 pre-processing����������������������������������������������������� 118, 119
Cholera�����������������������������������������������������������������������������1–7 wavenumber region������������������������������������� 113, 114, 123
Cigarette extract����������������������������������������������� 144, 146–148
Clostridia�����������������������������������������������������������������������������38 G
Clostridium tetani����������������������������������������������������������37–45 Gold
Colistin����������������������������������������������128–129, 131, 132, 134 colloidal���������������������������������������������������������� 95–97, 102
C-18 silane���������������������������������������������������������������133–141 electrode����������������������������������������������������� 10, 12, 14, 15
Gram-negative���������������������������������107, 113, 133, 151, 152,
D 159, 160, 167, 169, 187, 189
3-Deoxy-D-manno-oct-2-ulosonic acid (Kdo)�����������������196 Gram-positive��������������������������������������������������������������37, 61

Otto Holst (ed.), Microbial Toxins: Methods and Protocols, Methods in Molecular Biology, vol. 1600,
DOI 10.1007/978-1-4939-6958-6, © Springer Science+Business Media LLC 2017

199
 Microbial Toxins: Methods and Protocols
200 Index
  
H Microcystis aeruginosa
growth inhibition test��������������������������������������� 87, 90–91
Hek293 cells settling enhancement test���������������������������������������85–93
TLR-transfected���������������������������������������������������������144 Microscopy
Hemocytometer������������������������������������������������������������31, 90 cytofluorometry������������������������������������������������������������27
I Monolayers
Immunoaffinity������������������������������������������������������������������70 self-assembled (SAM)����������������������������������������133–141
Immuno-sensors�����������������������������������������������������������49–58 MTS assay��������������������������������������������������������������������25, 28
Interleukin-8��������������������������������������������������������������������145 Mycotoxins������������������������������������������������� 5, 65, 92–93, 144
microcystin LR
L
adsorption test��������������������������������������������������92–93
Laser interferometry�������������������������������������������������125–132 quantification���������������������������������������� 5, 65, 93, 144
Limulus amebocyte lysate (LAL)�������������������� 108–111, 134, ochratoxin A���������������������������������������������������������95–104
143, 144 zearalenone�����������������������������������������������������������95–104
Lipid A
isolation������������������������������������������������������������� 168, 170, N
172, 175, 192 Nanoparticles
screening�������������������������������������������������������������177–183 gold��������������������������������������������������������� 95, 97, 134, 135
Lipooligosaccharide (LOS)���������������������� 114, 151, 167, 188 magnetic labeling����������������������������������������������������������50
Lipopolysaccharide (LPS)������������������������ 152, 158, 159, 188 Neurotoxin
aggregates����������������������������������������������������� 50, 140, 168 botulinum��������������������������������������������������������� 10, 12, 13
chemotype����������������������������������������������������������151–164
immobilization���������������������������������������������������138–140 P
molecular mass determination������������ 153, 156, 163, 168
Pathogen-associated molecular patterns
R-type preparation�����������������������������152, 156, 157, 159,
(PAMP)�������������������������������������������������� 143, 144
188, 190, 197
Piranha solution����������������������������������������������� 135, 138, 140
S-type preparation������������������������������ 152, 158, 159, 188
Polymerase chain reaction (PCR)
Lipoprotein
amplification������������������������������������������� 1, 38–44, 50, 66
FSL-1��������������������������������������������������������� 145, 147–148
FLASH based��������������������������������������������������������38, 44
Liquid chromatography������������������������������������������ 29, 30, 93
real-time����������������������������������������10, 50, 52, 61–67, 126
Listeria monocytogenes
sequencing�������������������������������������������������� 40–41, 43, 72
detection����������������������������������������������� 49, 51–53, 55–56
Polymyxin B������������������������������������������������������������� 108, 136
Listeriosis���������������������������������������������������������������������������49
Proteomics�������������������������������������������������������������� 70, 75, 80
Loop-mediated isothermal amplification assay
Proteus mirabilis��������������������������������������������������������113–123
(LAMP)������������������������������������� 1, 38, 41–42, 45
Pseudomonas aeruginosa
M biofilm������������������������������������������������ 127, 168, 175–177
exotoxin A��������������������������������������������������������������29, 34
Magnetic planktonic��������������������������������������������������� 168, 175–177
beads Pyrrole-NHS���������������������������������������������� 50–51, 53–54, 56
antibody-coated������������������������������������������������75–76
aptamer-linked������������������������������������������� 62, 64, 67 Q
nanosilicate platelets (MNSP)
Quickfold module��������������������������������������������������������������20
preparation��������������������������17, 62, 64, 87–90, 96–98,
109, 159–162, 168, 185, 189 R
Mass spectrometry
MALDI-TOF��������������������������������������������� 70, 172, 188, Refractive index�������������������������������������������������� 58, 125, 126
190, 191, 193, 195 Ricin protein
tandem�������������������������������������������������������� 65, 72, 77, 78 preparation��������������������������������������������������������������13, 17
Microarray
S
antibody������������������������������������������������������������������������50
Microchip electrophoresis Salmonella
covalent binding��������������������������������������������������� 80, 152 enterica sv. Minnesota��������������������������������������������������153
fluorescent dye���������������������152, 154, 159–160, 162, 164 Minnesota������������������������������������153, 156–158, 169, 171,
noncovalent binding���������������������������������������������������152 180–183, 189, 190, 195
 Microbial Toxins: Methods and Protocols
Index   
201
SDS-PAGE Toxin
silver staining��������������������������������������������������������������152 AB��������������������������������������������������������������������������25–35
Selected reaction monitoring (SRM)���������������������������������70 cholera������������������������������������������������������������������������1–7
Shigella sonnei�������������������������������������153, 157, 189, 191, 195 detection���������������������������������������������������������������������1–7
Smoke extract��������������������������������������������������� 144, 146–148 diphtheria���������������������������������������������������������������29, 34
Spore inhibitors���������������������������������������������������� 26, 29–32, 34
lysis�������������������������������������������������������������������������62, 65 shiga�����������������������������������������������������������������������������27
trapping������������������������������������������������������������������61–67
Surface enhance raman scattering (SERS)�������������������������50 V
Surface plasmon resonance (SPR) Vero cell������������������������������������������������������������ 26, 27, 31–33
imaging (SPRI)������������������������������������������������������50–58 Vibrio cholerae��������������������������������������������������������������������1–6
Voltammetry������������������������������������������������12, 15, 16, 18, 19
T
Test strip W
immunochromatographic����������������������������� 1–7, 95–104
Whole-cell lysate��������������� 152, 153, 156, 158, 161, 187–197
Tetanospasmin��������������������������������������������������������������������39
Tetanus������������������������������������������������������������������� 37–39, 45 Y
TetR gene���������������������������������������������������������������������������39
TetX gene���������������������������������������������������������������������38–45 Yersinia pestis
Thin-layer chromatography (TLC)����������������������������������168 inactivation�������������������������������������������������������������71, 77
Toll-like receptor (TLR)�������������������������� 134, 144, 149, 150 protein extraction����������������������������������������������������71, 77