Fanconi Anemia

Monographs in Human Genetics

Vol. 15

Series Editor

Michael Schmid, Würzburg

Fanconi Anemia
A Paradigmatic Disease for the
Understanding of Cancer and Aging

Volume Editors

Detlev Schindler, Würzburg

Holger Hoehn, Würzburg

59 figures, 34 in color, and 14 tables, 2007

Basel · Freiburg · Paris · London · New York ·

Bangalore · Bangkok · Singapore · Tokyo · Sydney
 Holger Hoehn, M.D. Detlev Schindler, M.D.
Professor of Human Genetics Professor of Human Genetics
Department of Human Genetics Department of Human Genetics
University of Würzburg University of Würzburg
Am Hubland Am Hubland
D–97074 Würzburg D–97074 Würzburg

Library of Congress Cataloging-in-Publication Data

Fanconi anemia : a paradigmatic disease for the understanding of cancer and

aging / volume editors, Detlev Schindler, Holger Hoehn.
p. ; cm. – (Monographs in human genetics, ISSN 0077-0876 ; v. 15)
Includes bibliographical references and indexes.
ISBN-13: 978-3-8055-8277-3 (hard cover : alk. paper)
1. Fanconi’s anemia–Genetic aspects. I. Schindler, Detlev. II. Hoehn,
Holger. III. Series.
[DNLM: 1. Fanconi Anemia–genetics. W1 M0567P v. 15 2007 / WH 175 F1995
2007]
RC641.7.F36F34 2007
616.1⬘52042–dc22
2007011109

Bibliographic Indices. This publication is listed in bibliographic services, including Current Contents® and
Index Medicus.

Disclaimer. The statements, options and data contained in this publication are solely those of the individ-
ual authors and contributors and not of the publisher and the editor(s). The appearance of advertisements in the
book is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness,
quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property
resulting from any ideas, methods, instructions or products referred to in the content or advertisements.

Drug Dosage. The authors and the publisher have exerted every effort to ensure that drug selection and
dosage set forth in this text are in accord with current recommendations and practice at the time of publication.
However, in view of ongoing research, changes in government regulations, and the constant flow of information
relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for
any change in indications and dosage and for added warnings and precautions. This is particularly important when
the recommended agent is a new and/or infrequently employed drug.

All rights reserved. No part of this publication may be translated into other languages, reproduced or
utilized in any form or by any means electronic or mechanical, including photocopying, recording, microcopying,
or by any information storage and retrieval system, without permission in writing from the publisher.

© Copyright 2007 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland)

www.karger.com
Printed in Switzerland on acid-free paper by Reinhardt Druck, Basel
ISSN 0077–0876
ISBN 978–3–8055–8277–3
Dedicated to George M. Martin
 Contents

IX Editorial
Schmid, M. (Würzburg)

XI Preface
Schindler, D.; Hoehn, H. (Würzburg)

Introductory Remarks

1 Why, What and How Can We Learn

from a Rare Disease Like Fanconi Anemia?
Schroeder-Kurth, T. (Würzburg)

9 Fanconi Anemia: A Disease with Many Faces

Dietrich, R. (Unna); Velleuer, E. (Düsseldorf)

23 Milestones in Fanconi Anemia Research

Digweed, M. (Berlin); Hoehn, H. (Würzburg); Sperling, K. (Berlin)

39 Fanconi Anemia Genes: Structure, Mutations, and

Genotype-Phenotype Correlations
Kalb, R.; Neveling, K.; Herterich, S.; Schindler, D. (Würzburg)

59 Cancer in Fanconi Anemia and Fanconi Anemia Genes in Cancer

Neveling, K.; Kalb, R.; Schindler, D. (Würzburg)

79 Clonal Chromosomal Aberrations in Bone Marrow Cells of

Fanconi Anemia Patients: Results and Implications
Neitzel, H.; Kühl, J.-S.; Gerlach, A.; Ebell, W.; Tönnies, H. (Berlin)

VI
95 Interphase FISH-Assay for the Detection of MDS- and
AML-Associated Chromosomal Imbalances in Native
Bone Marrow and Peripheral Blood Cells
Tönnies, H.; Huber, S.; Volarikova, E.; Gerlach, A.; Neitzel, H. (Berlin)

110 Applications of Cell Cycle Testing in Fanconi Anemia

Schindler, D.; Friedl, R.; Gavvovidis, I.; Kalb, R.; Neveling, K. (Würzburg);
Linka, Y.; Hanenberg, H. (Düsseldorf); Kubbies, M. (Penzberg);
Hoehn, H. (Würzburg)

131 Prenatal Diagnosis of Fanconi Anemia: Functional and

Molecular Testing
Bechtold, A.; Kalb, R.; Neveling, K.; Friedl, R.; Gottwald, B.; Herterich, S.
(Würzburg); Schmugge Liner, M. (Zürich); Heilmann, C. (Copenhagen);
Hanenberg, H. (Düsseldorf); Schindler, D. (Würzburg)

149 Revertant Mosaicism in Fanconi Anemia: Natural

Gene Therapy at Work
Hoehn, H.; Kalb, R.; Neveling, K.; Friedl, R.; Bechtold, A.;
Herterich, S. (Würzburg); Sun, Y. (Nanjing, P.R. China); Gruhn, B. (Jena);
Hanenberg, H. (Düsseldorf); Schindler, D. (Würzburg)

173 Stem Cell Transplantation in Fanconi Anemia – Recent

Advances with Alternative Donors
Eyrich, M.; Winkler, B.; Schlegel, P.G. (Würzburg)

183 Fanconi Anemia Genes in Vertebrates: Evolutionary Conservation,

Sex-Linkage, and Embryonic Expression of FANCC and FANCG
in Avian Cells
Nanda, I.; Buwe, A. (Würzburg); Wizenman, A. (Neuherberg/Munich);
Takata, M. (Hiroshima); Haaf, T. (Mainz); Schartl, M.; Schmid, M. (Würzburg)

200 Studying Homologous Recombination in Fanconi Anemia

Demuth, I.; Digweed, M. (Berlin)

211 Functional Knock-Down of Human RAD51 for Testing the

Fanconi Anemia-BRCA Connection
Rio, P. (Düesseldorf, Madrid); Hanenberg, H. (Düsseldorf, Indianapolis, Ind.)

226 Author Index

227 Subject Index

Contents VII
 Editorial

Fourteen volumes of Monographs in Human Genetics were published

between 1966 and 1992. Since then a plethora of experimental data obtained by
the new molecular techniques has led to paradigmatic changes in understanding
the mechanisms operating in human heredity. Currently human genetics pre-
sents as a highly diversified field covering an ever increasing number of topics,
ranging from basic research to medical practice. This calls for a specialized
forum where the rapid advances in human genetics are reviewed by experts of
different fields. In response to these developments, the traditional book series
Monographs in Human Genetics will be revived with two volumes per year,
focusing on important hereditary diseases, their molecular basis, their clinical
impact and their eventual treatment. With its concise but highly informative
reviews, Monographs in Human Genetics provides essential reading not only
for researchers but also for physicians and students interested in specific
genetic diseases. All articles published in this book series are reviewed accord-
ing to classical standards.
The present volume is devoted to ‘Fanconi Anemia’, a chromosome insta-
bility disorder whose molecular basis has been all but elucidated in recent
years. By their very nature, monograph-types of publications are more compre-
hensive and synthetic rather than hot-off-the-press reports. A case in point is the
topic ‘Fanconi anemia’ where the pace of gene discovery has accelerated during
recent years. While this volume was in preparation three new disease causing
genes have been discovered, the most recent of which (FANCI) was too recent
to be included. Even though monographs may not cover the very last develop-
ments in a given field, their undisputed value arises from their unhurried and in

IX
depth treatment of a given subject that can hardly be achieved with the usual
publish-or-perish types of publications. As such, monographs fulfill the impor-
tant task to remind the reader that any type of scientific progress has its roots in
a multifaceted landscape of prior insights and achievements that need to be col-
lected and preserved in order to provide fertile ground for future growth.
I would like to thank all the authors for their interesting contributions, the
Editors Holger Hoehn and Detlev Schindler for their invaluable support and
help in the organization of this book, and the Publisher Thomas Karger for hav-
ing offered the opportunity to reinitiate this book series.

Michael Schmid
Würzburg, January 2007

Editorial X
 Preface

The revival of the series ‘Monographs in Human Genetics’ starts with a

volume on Fanconi anemia, a rare inherited disease that was discovered 80 years
ago by an eminent Swiss pediatrician, Guido Fanconi, then working in Geneva.
The Editors of the volume gratefully acknowledge the initiative by Karger
Publishers (Basel, Switzerland), and by Michael Schmid of the University of
Würzburg, Germany, in getting this new series started. The volume not only
pays tribute to Guido Fanconi’s discovery of a disease that teaches us much
about the connection between genetic instability, cancer and premature aging,
but also pays tribute to Traute Schroeder-Kurth who in 1964 discovered the
chromosome instability in Fanconi anemia, and who has contributed numerous
important observations on the clinical and genetic aspects of the disease. On the
occasion of Traute Schroeder-Kurth’s 75th birthday we convened a small meet-
ing in Würzburg in order to honor her seminal contributions to the diagnosis
and pathophysiology of Fanconi anemia. The talks presented at the Schroeder-
Kurth symposium provide the basis for this volume. We are very grateful to the
contributing authors for complementing and updating their previous oral pre-
sentations. Following introductory chapters that include a historical account,
exemplatory case reports, and a summary of the current status of FA genes and
their mutations, there are two chapters on neoplasia in FA, three chapters
reviewing diagnostic approaches in FA, including prenatal diagnosis, and one
chapter each on the phenomenon of revertant mosaicism or ‘natural gene ther-
apy’ and on hematopoietic stem cell transplantation as the only curative
approach in FA. The final three chapters deal with evolutionary aspects of the
FA genes with special emphasis on the avian genome, with recombinational

XI
types of DNA repair in FA, and with the establishment of a test system for the
Rad51 recombinase in homology-directed DNA repair. Even though it is
impossible to cover all aspects of Fanconi anemia within the space available for
such a volume, we hope that it may serve as useful introduction to a disease that
provides unique insights into the complex network of genomic maintenance
systems which protect us from cancer and premature aging.
We dedicate this volume to George M. Martin, M.D., Professor Emeritus
of Pathology and Genetics at the University of Washington, Seattle, USA, on
the occasion of his 80th birthday. Like Traute Schroeder-Kurth with respect to
Fanconi anemia, George Martin was one of the first scientists to recognize the
intimate relationship between genetic instability, cancer and aging by studying
the Werner progeria syndrome. Very early on he was convinced of the model
character and unique value of the rare human chromosomal breakage syn-
dromes. He emphasized that these rare experiments of nature would provide
unprecedented insights into the complex mechanisms of DNA damage recogni-
tion and repair which are essential for long lived, warm-blooded species such as
ours. George Martin encouraged us to study the human caretaker gene syn-
dromes, including Fanconi anemia, and we owe him much in terms of motiva-
tion, mentorship, and guidance.

Detlev Schindler
Holger Hoehn
Würzburg, January 2007

Preface XII
 Introductory Remarks

Schindler D, Hoehn H (eds): Fanconi Anemia. A Paradigmatic Disease for the Understanding of
Cancer and Aging. Monogr Hum Genet. Basel, Karger, 2007, vol 15, pp 1–8

Why, What and How Can We

Learn from a Rare Disease Like
Fanconi Anemia?
Traute Schroeder-Kurth
Department of Human Genetics, University of Würzburg, Würzburg, Germany

Abstract
In a field that embraces multiple aspects of both clinical and basic research and that
moves impressively fast, any answers to the questions why, what and how can we learn from
a rare disease like Fanconi anemia (FA) must remain tentative and preliminary. However,
there are very encouraging advances, most notably at the level of understanding the molecu-
lar basis of FA and at the level of treatment via hematopoietic stem cell transplantation. For
the sake of our patients we clearly need to arrive at meaningful genotype-phenotype correla-
tions and individualized risk profiles. This requires prospective and longterm studies carried
out in close cooperation among patients, clinicians and basic scientists. There are a number
of open questions, for example relating to the mechanisms of chromosomal breakage and
DNA repair, to the spectrum of genetic changes that herald and promote the emergence of
leukaemia and solid tumors, and to the emergence of genetically reverted cells in blood and
bone marrow of FA patients. Researchers in the fields of cancer and aging should be encour-
aged to and are likely to benefit from the study of Fanconi anemia, as are our patients from
the welcome results of such studies.
Copyright © 2007 S. Karger AG, Basel

The year 2007 marks the 80th anniversary of the original description by
Guido Fanconi of ‘Familial infantile pernicious-like anemia’, a rare genetic dis-
ease that since carries the name of this eminent physician-scientist [1]. Why
should we study such a rare disease, what can we learn from its myriad clinical
and molecular manifestations, and how should we go about it? I will try to give
some brief answers to these questions even though by necessity these answers
are tentative and preliminary. Despite undoubted progress, there still are many
aspects of this enigmatic disease that we do not fully understand.
 Why Study Fanconi Anemia?

First of all, for the sake of the affected patients. Even though there is no
definite cure for FA on the horizon, there is encouraging progress regarding
diagnosis and treatment. Timely diagnosis is crucial both for the institution of
adequate treatment and for the prevention of inadequate medical management
that may result from failure to recognize the hereditary and specific nature of
the disease. If one looks back at the last monograph on Fanconi anemia that I
had the pleasure to edit with my colleagues Arleen Auerbach and Gunter Obe in
the year 1989 [2], the progress that has since been made is impressive and
encouraging. Two major areas stand out: molecular genetics, and hematopoietic
stem cell transplantation. Progress in the understanding of the molecular basis
of Fanconi anemia is astounding indeed: even though genetic heterogeneity of
FA had been convincingly documented prior to 1989 [3, 4], it was not until
1992 that the first FA gene had been cloned and, as of this writing, 12 FA caus-
ing genes have been identified. Several more are in the offing, as there are a
number of patients who apparently do not belong to any of the known comple-
mentation groups [5]. There is also increasing understanding of how the FA
proteins work, how they work together and how they might interact with other
proteins in maintaining a stable genome [6, 7].
Although less spectacular than the gene discovery itself, the virtual explo-
sion of knowledge concerning the molecular basis of Fanconi anemia during the
past 10–15 years translates into immediate benefits to FA patients and their fami-
lies. For example, precise knowledge of the affected gene and the specific type of
mutation in a given patient will increasingly be of prognostic value and serve as a
rational guide for optimized medical management. We clearly need to collect
many more data in order to arrive at clinically useful genotype-phenotype corre-
lations. We should strongly oppose arguments that deem the collection of such
additional data unnecessary once the clinical diagnosis of FA has been established
and confirmed by a chromosome breakage or cell cycle test. The ultimate goal of
such efforts is to arrive at individualized risk profiles for FA patients similar to
what has already been achieved with congenital abnormality scores [8]. Biallelic
mutations in two of the FA genes (FANCD1/BRCA2 and FANCN/PALB2) have so
far been associated with very early childhood cancers [9, 10], and awareness of
such correlations are of undisputed value in the clinical management of these
young tumor patients. Precise knowledge of the disease causing mutations also
improves and facilitates prenatal diagnosis that remains an unwelcome but com-
promise option for some of the FA families. Last but not least, participation in
gene therapy trials requires knowledge of underlying genes and mutations.
Although one may question the extent to which the vastly improved
molecular knowledge has so far been of immediate benefit to the majority of

Schroeder-Kurth 2
FA patients, such reservations do least apply to the progress in the field of
hematopoietic stem cell transplantation (HSCT). The outcomes of matching
donor sibling transplantations have continuously improved such that today early
HSCT is highly recommended in such families [11]. Unrelated donor trans-
plantation still carries a higher risk, but modified conditioning protocols using
fludarabine and T-cell depletion, reduction or complete absence of whole body
irradiation, the introduction of minitransplants and improved post-transplant
care have contributed to much better longterm survival [12]. The serious prob-
lem of posttransplant squamous cell carcinomas remains to be solved [13]. The
message from all this progress is to consider the option of HSCT in each patient
at a much earlier stage of the disease and with much more optimism than we
could have had in the year 1989.

What Can We Learn from Fanconi Anemia?

We have learned that introducing a gene into something as accessible as a

bone marrow stem cell does not necessarily result in improvement [14]. There
are many reasons for the slow progress in gene therapy, but there is no reason to
give up. Improvements in stem cell collection and improved gene transfer pro-
tocols are on the horizon [15, 16]. Notwithstanding all previous failures and
disappointments, the replacement of a defective gene copy by an intact gene
remains an attractive option which should and needs to be further explored.
Looking back at the recent progress of FA research it seems to me a kind of
miracle that at first chromosome instability showed up and initiated the search
for possible molecular mechanisms causing breaks, gaps and reunions which
we saw in the chromosome preparations. Our very first observations reported in
1964 taught us that chromosomal instability was present not only in patients
presenting with FA symptoms, but also in their siblings without any clinical
manifestation of the disease [17]. We also quickly learned that chromosome
instability occurs in vivo as it was observed in direct preparations of bone mar-
row cells [18]. If dividing bone marrow cells exhibit chromosome aberrations,
this surely must reflect a fundamental and intrinsic defect of genomic mainte-
nance with severe consequences for the renewal of blood cell lineages. The
more we learned about the complex structure that makes up a chromosome, the
more naive appeared the simple notion of a DNA lesion that was not properly
repaired and therefore manifests as a chromosome break. Why do breaks
reunite if there is a repair defect, and what are the prerequisites for a reunion to
take place? Why do crosslink-induced radials preferentially form between non-
homologous autosomes and are virtually absent from gonosomes? [19]. What is
the role of low copy repeats that are increasingly recognized as focal points for

Why, What and How Can We Learn from a Rare Disease Like Fanconi Anemia? 3
non-allelic homologous recombination? [20]. Does the chromosomal position
within the nucleus influence the opportunity for the type of radial formation?
There are many unsolved questions, there still is much to be learned.
One of the most impressive lessons we learned from patients with Fanconi
anemia is the intimate relationship between genetic instability and cancer. FA
patients carry a high risk for acute leukaemia, squamous cell carcinomas and
other tumors [21]. Young FA patients with certain gene mutations have a higher
risk to develop leukaemia. The older a FA patient gets the higher is the risk to
develop a solid tumor. There are impressive case histories describing adult
patients in whom a leukaemia or a solid tumor were the first manifestations of
FA [22, 23]. With the exception of FANCF that is frequently inactivated by an
epigenetic mechanism in various types of tumors, the other FA genes seem to be
relatively intact in most neoplasias as if they were needed for tumor cell growth
[24]. However, monoallelic truncating mutations in at least three of the FA genes
(FANCD1/BRCA2, FANCJ/BRIP1, and FANCN/PALB2) have been shown to
confer variant degrees of breast cancer susceptibility in non-FA patients [25, 26].
Those are important new insights, connecting a rare with a common disease.
In 1971 we postulated that in vivo chromosomal instability leads to multi-
ple somatic mutations which finally may confer a selective advantage to a sin-
gle mutated cell, giving rise to malignant cell growth [27]. By careful bone
marrow studies, this process has been directly observed in FA patients. Specific
chromosomal changes in bone marrow cells appear to herald progression to
malignancy even prior to any clinical manifestation of leukaemia [28]. It was
quite clear in the 1970s that aberrant clones may also appear in peripheral blood
lymphocyte cultures. Over time, these cytogenetically aberrant clones were
subject to clonal attenuation and clonal succession, but they undoubtedly were
bad news for the patient. One of these patients I followed for more than six
years with repeated chromosome studies [29]. He died at age 35 of bronchial
cancer (unpublished). These types of longterm observations are important for
our patients and should be supported as much as possible. There is much to be
learned about the close relationship between FA and cancer for the sake of our
patients, and for our understanding of the origin of cancer [30].
Even though a fundamental change of the high cancer risk in FA is not in
sight, we have already learned that predictive and preventive measures are benefi-
cial for our patients. As already mentioned, regular monitoring for aberrant bone
marrow clones, particularly those involving monosomy 7 and 3q duplications, is
of great value for both physicians and patient in their difficult decision as to
whether and when to proceed with HSCT [28]. Likewise, non-invasive measures
for the early detection of oral cavity and genital area lesions as currently devel-
oped by the group of Ruud Brakenhoff in Amsterdam [31] will gain even greater
importance with more and more patients undergoing and surviving HSCT.

Schroeder-Kurth 4
 Another lesson we are beginning to appreciate from Fanconi anemia is the
discovery of reverse mutations in blood cells [32]. One in four or five patients
displays MMC resistant cells among the original MMC sensitive cells. These
patients are mosaic FA patients who have a chance to escape BM failure and,
possibly, the development of leukaemia. Much research has to be done to clar-
ify the underlying mechanisms. The likelihood of ‘natural gene therapy’ leading
to revertant mosaicism appears to be higher in patients belonging to certain
complementation groups and in the presence of compound heterozygosity.
Longterm observations of these mosaic FA patients will ultimately teach us
whether there is more than a temporary benefit of this type of ‘natural gene
therapy’ to the individual, and what the prerequisites are for somatic reversions
to occur. And, of course, could there be ways to enhance the occurrence of such
reversions for the benefit of our patients?

How Should We Go About It?

First and foremost, FA research needs the cooperation of the affected fam-
ilies and patients as well as the unrelenting interest and commitment of their
doctors. A well established (and well funded!) FA registry such as pioneered by
Arleen Auerbach in New York is essential for both clinical care and basic
research [33, 34]. It should include regular follow-ups with complete informa-
tion on the natural history of the disease in each individual family. For practical
purposes, such registries should be organized at the regional or national levels,
but they should and must communicate with each other. Today we have every
reason to believe that most FA families are aware of the importance of FA
research for the sake of their children and thus are willing to cooperate with
such registries. Many research projects require the participation and involve-
ment of the patients themselves, and there is little doubt that the affected
patients and families can still teach us a lot about their disease. On the technical
side, new tools such as retroviral complementation, knockdown of FA genes via
RNA interference, or MLPA for the detection of large deletions have con-
tributed both to diagnosis and research. With such technical advances at hand
the time might have come to tackle the old question of what really causes the
oxygen sensitivity of FA cells [35, 36].
We now know that at least some of the FA genes are highly conserved dur-
ing evolution and that they play a major role during DNA replication and
recombination in premeitotic stages and meiosis. Some of the FA genes have
been recognized as very ancient ‘caretaker’ genes that have emerged even prior
to the vertebrate lineage [37]. In addition to valuable insights gained from ver-
tebrate model organisms such as zebrafish, mouse and birds, scientists working

Why, What and How Can We Learn from a Rare Disease Like Fanconi Anemia? 5
on ancient model organisms like yeast, C. elegans and Drosophila should be
invited and encouraged to participate in FA-oriented basic research. Last but
not least, Fanconi anemia should be introduced as a human model system into
basic cancer research. It has already been shown that targeted inactivation of
FA genes in cancer cells may offer specific therapeutic options [38, 39]. I am
convinced that both, our patients and our colleagues involved in cancer
research will benefit from inclusion of Fanconi anemia into current research
paradigms.
Concerning the premature aging phenotype of Fanconi anemia, it is obvi-
ous that truly progeroid features (such as in the Hutchinson-Gilford or Werner
syndromes) are far less conspicuous in FA. The decline of bone marrow func-
tion in FA clearly is a much more progressive and devastating event compared
to what happens during normal aging. This is the likely reason why Martin and
Oshima did not include FA in their presentation of human genetic instability
syndromes with progeroid features [40]. However, squamous cell carcinomas of
the oral cavity and genital area are typical tumors of advanced age that occur
very prematurely in FA patients. Likewise, endocrine abnormalities including
hyperinsulinemia, growth hormone deficiency and hypothyroidism as typical
endocrine abnormalities of older individuals affect more than 80% of FA
patients during young adulthood [41]. Impaired gametogenesis and premature
reproductive aging are additional features of FA patients that are reminiscent of
an accelerated aging process. Collectively, the premature manifestation, in FA
patients, of clinical features also encountered during normative aging seems to
indicate that genetic instability per se may be one of the most significant factors
that contributes to and promotes aging. From this point of view, scientists
involved in aging research might benefit from the study of Fanconi anemia.
Conversely, Fanconi anemia patients may ultimately benefit from the insights
gained from the ongoing efforts of serious and scientifically sound anti-aging
research.

References

1 Lobitz S, Velleuer E: Guido Fanconi (1892–1979): a jack of all trades. Nat Rev Cancer 2006;6:
893–898.
2 Schroeder-Kurth TM, Auerbach AD, Obe G (eds): Fanconi Anemia. Clinical, Cytogenetic and
Experimental Aspects. Berlin, Heidelberg, New York, Springer-Verlag, 1989.
3 Zakrzewski S, Sperling K: Genetic heterogeneity of Fanconi’s anemia demonstrated by somatic
cell hybrids. Hum Genet 1980;56:81–84.
4 Duckworth-Rysiecki G, Cornish K, Clarke CA, Buchwald M: Identification of two complementa-
tion groups in Fanconi anemia. Somat Cell Mol Genet 1985;11:35–41.
5 Hanenberg H, Batish SD, Pollok KE, Vieten L, Verlander PC, et al: Phenotypic correction of pri-
mary Fanconi anemia T cells with retroviral vectors as a diagnostic tool. Exp Hematol 2002;30:
410–420.

Schroeder-Kurth 6
 6 Taniguchi T, D’Andrea AD: Molecular pathogenesis of Fanconi anemia: recent progress. Blood
2006;107:4223–4233.
7 Gurtan AM, D’Andrea AD: Dedicated to the core: understanding the Fanconi anemia complex.
DNA Repair (Amst) 2006;5:1119–1125.
8 Rosenberg PS, Huang Y, Alter BP: Individualized risks of first adverse events in patients with
Fanconi anemia. Blood 2004;104:350–355.
9 Alter BP, Rosenberg PS, Brody LC: Clinical and molecular features associated with biallelic
mutations in FANCD1/BRCA2. J Med Genet 2007;44:1–9.
10 Reid S, Schindler D, Hanenberg H, Barker K, Hanks S, et al: Biallelic mutations in PALB2 cause
Fanconi anemia subtype FA-N and predispose to childhood cancer. Nat Genet 2006;39:162–164.
11 Rosenberg PS, Alter BP, Socie G, Gluckman E: Secular trends in outcomes for Fanconi anemia
patients who receive transplants: implications for future studies. Biol Blood Marrow Transplant
2005;11:672–679.
12 Wagner JE, Eapen M, Macmillan ML, Harris RE, Pasquini R, et al: Unrelated donor bone marrow
transplantation for the treatment of Fanconi anemia. Blood 2007;109:2256–2262.
13 Alter BP: Fanconi’s anemia, transplantation, and cancer. Pediatr Transplant 2005;9:81–86.
14 Williams DA, Croop J, Kelly P: Gene therapy in the treatment of Fanconi anemia, a progressive
bone marrow failure syndrome. Curr Opin Mol Ther 2005;7:461–466.
15 Jacome A, Navarro S, Casado JA, Rio P, Madero L, et al: A simplified approach to improve the
efficiency and safety of ex vivo hematopoietic gene therapy in Fanconi anemia patients. Hum
Gene Ther 2006;17:245–250.
16 Kelly PF, Radtke S, Kalle C, Balcik B, Bohn K, et al: Stem cell collection and gene transfer in
Fanconi anemia. Mol Ther 2007;15:211–219.
17 Schroeder TM, Anschütz F, Knopp A: Spontane Chromosomenaberrationen bei familiärer
Panmyelopathie. Humangenetik 1964;1:194–196.
18 Schroeder TM: Cytogenetische und cytologische Befunde bei enzymopenischen Panmyelopathien
und Pancytopenien. Humangenetik 1966;2:287–316.
19 Newell AE, Akkari YM, Torimaru Y, Rosenthal A, Reifsteck CA, et al: Interstrand crosslink-
induced radials form between non-homologous chromosomes, but are absent in sex chromosomes.
DNA Repair (Amst) 2004;3:535–542.
20 Erdogan F, Chen W, Kirchhoff M, Kalscheuer VM, Hultschig C, et al: Impact of low copy repeats
on the generation of balanced and unbalanced chromosomal aberrations in mental retardation.
Cytogenet Genome Res 2006;115:247–253.
21 Rosenberg PS, Greene MH, Alter BP: Cancer incidence in persons with Fanconi anemia. Blood
2003;101:822–826.
22 Liu JM, Auerbach AD, Young NS: Fanconi anemia presenting unexpectedly in an adult kindred
with no dysmorphic features. Am J Med 1991;91:555–557.
23 Huck K, Hanenberg H, Gudowius S, Fenk R, et al: Delayed diagnosis and complications of
Fanconi anaemia at advanced age – a paradigm. Br J Haematol 2006;133:188–197.
24 Neveling K, Kalb R, Schindler D: Cancer in Fanconi anemia and Fanconi anemia genes in cancer;
in Schindler D, Hoehn H (eds): Fanconi Anemia, Basel, Karger-Verlag, 2007, pp 59–78.
25 Seal S, Thompson D, Renwick A, Elliott A, Kelly P, et al: Truncating mutations in the Fanconi
anemia J gene BRIP1 are low-penetrance breast cancer susceptibility alleles. Nat Genet 2006;38:
1239–1241.
26 Rahman N, Seal S, Thompson D, Kelly P, Renwick A, et al: PALB2, which encodes a BRCA2-
interacting protein, is a breast cancer susceptibility gene. Nat Genet 2006;39:165–167.
27 Schroeder TM, Kurth R: Spontaneous chromosomal breakage and high incidence of leukemia in
inherited disease. Blood 1971;37:96–112.
28 Tonnies H, Huber S, Kuhl JS, Gerlach A, Ebell W, Neitzel H: Clonal chromosomal aberrations in
bone marrow cells of Fanconi anemia patients: gains of the chromosomal segment 3q26q29 as an
adverse risk factor. Blood 2003;101:3872–3874.
29 Schroeder TM, Drings P, Beilner P, Buchinger G: Clinical and cytogenetic observations during a
six-year period in an adult with Fanconi⬘s anemia. Blut 1976;34:119–132.
30 Mathew CG: Fanconi anaemia genes and susceptibility to cancer. Oncogene 2006;25:5875–5884.

Why, What and How Can We Learn from a Rare Disease Like Fanconi Anemia? 7
31 Bremmer JF, Braakhuis BJ, Ruijter-Schippers HJ, Brink A, Duarte HM, et al: A noninvasive
genetic screening test to detect oral preneoplastic lesions. Lab Invest 2005;85:1481–1488.
32 Lo Ten Foe JR, Kwee ML, Rooimans MA, Oostra AB, Veerman AJ, et al: Somatic mosaicism in
Fanconi anemia: molecular basis and clinical significance. Eur J Hum Genet 1997;5:137–148.
33 Auerbach AD, Schroeder TM: First announcement of the Fanconi Anemia International Registry.
Blood 1982;60:1054.
34 Kutler DI, Singh B, Satagopan J, Batish SD, Berwick M, et al: A 20-year perspective on the
International Fanconi Anemia Registry (IFAR). Blood 2003;101:1249–1256.
35 Pagano G, Degan P, d’Ischia M, Kelly FJ, Nobili B, et al: Oxidative stress as a multiple effector in
Fanconi anaemia clinical phenotype. Eur J Haematol 2005;75:93–100.
36 Mukhopadhyay SS, Leung KS, Hicks MJ, Hastings PJ, Youssoufian H, Plon SE: Defective mito-
chondrial peroxiredoxin-3 results in sensitivity to oxidative stress in Fanconi anemia. J Cell Biol
2006;175:225–235.
37 Nanda I, Buwe A, Wizenman A, Takata M, Haaf T, Schartl M, Schmid M: Fanconi anemia genes
in vertebrates: evolutionary conservation, sex-linkage, and embryonic expression of FANCC and
FANCG in avian cells; in Schindler D, Hoehn H (eds): Fanconi Anemia, pp 183–199. Basel,
Karger-Verlag, 2007.
38 van der Heijden MS, Brody JR, Dezentje DA, Gallmeier E, Cunningham SC, et al: In vivo thera-
peutic responses contingent on Fanconi anemia/BRCA2 status of the tumor. Clin Cancer Res
2005;11:7508–7515.
39 Gallmeier E, Calhoun ES, Rago C, Brody JR, Cunningham SC, et al: Targeted disruption of
FANCC and FANCG in human cancer provides a preclinical model for specific therapeutic
options. Gastroenterology 2006;130:2145–2154.
40 Martin GM, Oshima J: Lessons from the human progeroid syndromes. Nature 2000;408:263–266.
41 Wajnrajch MP, Gertner JM, Huma Z, Popvich J, Lin K, et al: Evaluation of growth and hormonal
status in patients referred to the International Fanconi Anemia Registry. Pediatrics 2001;107:
744–754.

Prof. Dr. med. Traute Schroeder-Kurth

Wilhelm-Doles-Str. 7
D–97246 Eibelstadt (Germany)
Tel./Fax ⫹49 9303 8762

Schroeder-Kurth 8
 Schindler D, Hoehn H (eds): Fanconi Anemia. A Paradigmatic Disease for the Understanding of
Cancer and Aging. Monogr Hum Genet. Basel, Karger, 2007, vol 15, pp 9–22

Fanconi Anemia: A Disease with

Many Faces
R. Dietricha, E. Velleuerb
a
Deutsche Fanconi-Anämie-Hilfe e.V., Unna; bZentrum für Kinder- und
Jugendmedizin, Universitätsklinikum Düsseldorf, Germany

Abstract
We present the clinical case histories of 7 patients with Fanconi Anemia (FA) in order
to illustrate the widely divergent phenotypes and clinical courses of the disease. Moreover,
these case histories demonstrate that androgen therapy and hematopoietic stem cell trans-
plantation are therapeutic options that have proven their worth despite side effects that can be
severe or even fatal. We show that mild forms and severe forms of the disease can occur in
individuals with identical mutations, as in siblings from a single family. We wish to convince
the reader that there is no ‘typical’ FA patient, even though there are a number of congenital
malformations and clinical manifestations whose patterns and combinations should alert the
physician to the diagnosis of Fanconi anemia. Each individual case history teaches us certain
facets but never the entire spectrum of a disease that has as many faces as affected individu-
als. Fanconi anemia is a disease physicians must think of. We will soon reach a point where
parents and patients need be less discouraged and less concerned by the diagnosis of Fanconi
anemia since there is tremendous progress in understanding the causes and treating the con-
sequences of the disease. There is no definitive cure on the horizon, but there is continuous
improvement of the quality of life for many patients. Early diagnosis, up-to-date information,
closely knit surveillance, participation in family support groups, and a mutually trustful
alliance with expert physicians and scientists are the cornerstones of optimal care.
Copyright © 2007 S. Karger AG, Basel

Since FA is a rare entity, family support groups play an important role in

the collection and dissemination of information concerning the disease. In addi-
tion to the families themselves, physicians and researchers benefit from the vast
amount of knowledge and personal experience that accumulate within the realm
of daily family life. Much of this experience has been summarized and is
updated regularly in special publications and newsletters edited by the Fanconi
Anemia Research Fund (FARF) and its German equivalent, the Deutsche
Fanconi-Anämie-Hilfe e.V. [1–3]. Family support groups organize annual meet-
ings where families have the opportunity to meet each other and to establish
personal contacts to specialized physicians and FA researchers. Family support
groups explain the aims and importance of research projects to their lay mem-
bers and encourage participation. As pointed out repeatedly by one of the lead-
ing researchers in the field, Professor Hans Joenje of the Free University
Amsterdam, the close interaction and mutual trust between family support
groups and FA researchers has been instrumental in the deciphering of the
genetic, cellular and molecular basis of the disease.
Members of academic institutions frequently approach family support
groups asking whether they could provide photographs of a ‘typical’ FA patient
for the purpose of teaching or publication. No one knows better than the
affected families themselves that there is no ‘typical’ FA patient but rather a
highly variable spectrum of phenotypes and clinical courses. There are patients
born with multiple congenital anomalies and severe growth retardation, and
patients without any congenital anomalies and normal growth. There are
patients with neonatal onset of thrombocytopenia and early bone marrow fail-
ure, and there are adult patients with close to normal or even completely normal
blood counts. There are patients who succumb to solid tumors or leukemia as
newborns or toddlers, and patients whose first manifestations of the disease are
malignancies arising during adulthood. Recent anecdotal evidence may illus-
trate this point: The authors showed photographs of 30 proven FA patients to a
group of experienced FA specialists who came to the conclusion that they could
not recognize any features of FA in 8 of these 30 pictures, and only minimal
signs in further five. Thus, there were no ‘typical’ clinical features in almost
half of the 30 patients.
In non-consanguineous, random populations there are only between 2 and
5 affected individuals per million births such that the average physician has no
personal experience with the manifold manifestations and highly variable clin-
ical course of the disease. Given this pronounced variability of phenotypes and
variable onset, severity and course of hematopoietic failure, the correct diagno-
sis may be missed or delayed. However, in the interest of optimal patient care,
there is urgent need for a timely and correct diagnosis. This will become even
more important in future years as the outcome of alternative donor hematopoi-
etic stem cell transplantation (HSCT) continues to improve. Since FA is a reces-
sive disorder with a 25% risk to siblings, early diagnosis is also mandated by
considerations of family planning.
There are numerous clinical problems which require special attention and
special treatments in FA patients. For example, conventional chemotherapy
containing DNA crosslinking agents is poorly tolerated by these highly sensi-
tive individuals. In case of solid tumors, surgery and subsequent radiotherapy

Dietrich/Velleuer 10
are the treatments of choice. For FA patients with incipient leukemia, the only
valid option is immediate HSCT. In order to detect leukemic changes as early as
possible, FA patients should be monitored at regular intervals for clonal chro-
mosome changes arising in their bone marrow cells. There is solid evidence that
a specific pattern of chromosome changes emerges prior to the onset of clini-
cally overt leukemia (see chapter by Neitzel et al.). The emergence of such aber-
rant cells is a warning sign that requires immediate action in terms of
preparation for HSCT. Four German patients have recently been transplanted
after positive bone marrow findings thus sparing them the additional (and often
fatal) complications of overt leukemia.
The following case reports are meant to illustrate the high degree of vari-
ability of phenotype and clinical course encountered among FA patients. Since
by now there are 12 and possibly more genes whose mutational changes lead to
the development of FA, much of the clinical variability can be attributed to this
tremendous amount of genetic and, at the single gene level, mutational hetero-
geneity. In the absence of prospective studies we still are fairly ignorant about
genotype-phenotype correlations, even though tentative evidence for such cor-
relations exists for some of the FA genes (see chapter by Kalb et al.). Despite
recent progress in our understanding of the molecular basis of the disease, there
are many open questions relating to gene and protein function that need to be
answered before we can explain, for example, why manifestations of the disease
can be so varied and different, even among siblings carrying identical muta-
tions. In presenting these case reports, the authors wish to point out that even
though these descriptions are kept in the usual neutral and medical style, these
individuals are exceptional human beings by instructing us not only about med-
ical facts, but also by teaching us humility towards the manifold facets, pleasant
and unpleasant, of human life.

Philipp

Philipp (fig. 1) was delivered via C-section in the 37th week of pregnancy
after sonography at 32 weeks had revealed a hydrocephalus. He weighed
2125 g, his height was 44.5 cm, and his head circumference was 32 cm. In addi-
tion, there were multiple structural and functional defects, including persisting
ductus Botalli, cleft lip and palate (Pierre-Robin sequence), bilateral radius
aplasia, bilateral thumb aplasia, anal stenosis, aqueduct stenosis with hydro-
cephalus, hypospadia, cryptorchism, pulmonary hypertension due to atelecta-
sis, strabism, multiple café-au-lait spots, horseshoe kidney, rib anomalies.
Intubation and assisted ventilation were required during the first week of
life. Subsequent pneumonia was controlled with antibiotics. Philipp had to be

Fanconi Anemia: A Disease with Many Faces 11

 Fig. 1. (from left to right) Philipp at 6, Dominik at 7 and Michael at 25 years.

fed via gastric tube, and the anal stenosis had to be corrected surgically. From
day 3 of his life, Philipp received Vojta-type of physical therapy which was very
exhausting for him, possibly due to the at that time still unknown persistence of
the duct of Botalli. His parents felt that Philipp’s declining health was due to
these stressful exercises. Because of increasing intracranial pressure, Philipp
had to be re-intubated at the age of 2 weeks, and a shunting procedure was per-
formed. There was an episode of bloody liquor four days after shunting, but
Philipp’s condition gradually improved. After four weeks the shunt became
infected and had to be removed. An attempt to close the persisting ductus
Botalli by drug therapy failed, but surgical closure was finally successful. At the
age of 4 months, a plate was installed in his palate in preparation for the later
surgical correction of the large cleft. Except for low thyroid function (for which
he was treated with thyroid hormone) laboratory parameters were within nor-
mal limits, including peripheral blood counts. He was discharged from the hos-
pital at age 5 months with the preliminary diagnosis of VACTERL association.
There was an additional episode of increased intracranial pressure at age 8 months
which required treatment. Growth retardation became apparent around age 1.
Because of the bilateral atresia of his auditory canals, Philipp received a hear-
ing device, and glasses were prescribed to improve strabism and poor eye sight.
Also at age 1 his cleft palate was closed.
During the second and third years of his life, his pronounced statomotoric
retardation was treated with physiotherapy. Developmental milestones were
severely retarded. When he entered kindergarten at age 3 Philipp could neither
speak, nor feed himself, nor walk, and he required assistance for upright sitting
and standing. During the course of recurrent infections his thrombocyte counts
were noted to decrease such that at age 5 the possibility of FA was considered

Dietrich/Velleuer 12
and quickly confirmed by cell cycle and chromosome breakage studies. Philipp
belongs to the rare X-linked complementation group FA-B.
At the time of his diagnosis hemoglobin was 5.7 g/dl, thrombocytes were
20,000/␮l, leukocytes were 3,300/␮l. His blood counts continued to decrease,
requiring regular thrombocyte and erythrocyte transfusions. A venous port
catheter was implanted, but became infected and had to be replaced. Until age
6, Philipp received a total of 15 erythrocyte concentrates and 35 units of throm-
bocytes. Even though his thrombocyte counts returned to normal shortly after
the transfusions, a week later his counts could be as low as 8,000/␮l. Further
complications were esophagitis and esophageal bleeding between ages 5 and 6.
Because of his multiple impairments and poor health status, neither his parents
nor the attending physicians considered HSCT as a viable option for Philipp,
who requires full time care. In a recent issue of the German FA family news let-
ter his parents write: ‘If he feels well, Philipp likes to laugh and conveys the
impression of alertness, happiness and interest in his environment. Despite all
the physical, mental and emotional energy it takes to care for him, we get so
much back from this very special child and we would not want to miss him’.

Dominik

Dominik (fig. 1) was born at 36 weeks of an uneventful pregnancy to

healthy, non-consanguineous parents. Following spontaneous delivery, numer-
ous congenital abnormalities were noted: partial right radial ray aplasia, aplasia
of the right thumb, hypoplasia of the left thumb, membranous duodenal atresia,
hypospadia, hyperopia and strabismus, renal dysplasia with vesico-urethral
reflux, severely stunted growth (weight and length under the 3rd percentile).
Because of normal peripheral blood counts, the child was thought to be affected
by a sporadic developmental disturbance (VATER-association). During the
following years, Dominik was repeatedly hospitalized for the surgical repair of
his duodenum, thumbs, kidneys and urethra. Despite recurrent urinary tract
infections and progressive renal insufficiency, he continued to grow along the
3rd percentile. Low thrombocyte and erythrocyte counts were first noted at age 3,
and the clinical suspicion of FA was confirmed by cell cycle and chromosome
breakage studies. Dominik carries the mutation c.1650delT (exon 8) in FANCB
on his X-chromosome.
He was started on oxandrolone therapy (25 mg/day) with excellent response
of his blood counts. Side effects were signs of virilization and sonographic
changes of liver parenchyma, but liver enzymes and AFP remained within nor-
mal limits. After one year of androgen therapy his blood counts began to decline
such that, starting at age 6, he required weekly thrombocyte transfusions.

Fanconi Anemia: A Disease with Many Faces 13

Erythrocyte transfusions were required every third week. Because of obvious
non-response, oxandrolone was discontinued and briefly replaced by erythro-
poietin, however without any success. Bone marrow biopsy showed high grade
hypoplasia with myelodysplastic changes indicative of MDS. There were 4–5%
blasts, and bone marrow cytogenetics revealed monosomy 7.
Since his older sister was available as HLA-compatible donor, Dominik
underwent bone marrow transplantation. Because of his renal insufficiency he
had to be continuously dialyzed, but there were no other complications. Two years
after HSCT he received a renal transplant that was likewise successful. With the
help of a tutor, Dominik was able to complete 4 years of elementary school. Since
he always was last in his class, he transferred to a school for the handicapped
where he belonged to the top students. At 6 years after HSCT and 4 years after
renal transplant, Dominik had stable renal function and normal blood counts.
Growth hormone therapy was started at age 12. Today, at age 13, Dominik still is
on immunosuppressive medication but leads a nearly normal life.

Michael

Even though Michael (fig. 1) was born at term, he was small for date,
weighing 2,250 g and measuring 45 cm. He displayed a number of birth defects,
including bilateral radial ray and thumb aplasia, coloboma of the right eye, and
rightsided inguinal hernia. Feeding was difficult with frequent regurgitation
requiring a gastric tube with individual meals lasting up to 2 h. A membranous
narrowing of his esophagus was finally diagnosed as cause of his problems
which improved after surgical correction. But because of continuous difficul-
ties to swallow, Michael required mainly liquid or semi-solid foods until the age
of 23. Failure to thrive prompted an investigation of his growth hormone levels at
age 3, but results were normal. The diagnosis of FA was considered at age 4 and
confirmed by chromosome breakage studies but his complementation group
was never determined. Because of poor vision he required glasses from age 6. A
tendency to bleed and cytopenia became apparent at age 9. Michael was treated
with cortisol for two weeks which he did not tolerate well. There were multiple
episodes of esophageal bleedings which were thought to be due to his low
thrombocyte counts, but esophageal varices were diagnosed at age 23 together
with ‘fibrosis’ of his liver. Attempts to stop the bleeding episodes by sclerosing
the varices were only partly successful. Despite these recurrent medical prob-
lems, Michael did well in school and obtained his baccalaureate and his driving
license. At age 24 Michael was diagnosed with pituitary dysfunction, hypo-
gonadism and hypothyroidism. From there on, he was treated with anabolic
steroids, growth hormone and thyroxin. At age 28 his height was 152 cm and he

Dietrich/Velleuer 14
weighed 46 kg. He developed diabetes mellitus which was controlled with oral
medication. Under continuous androgen therapy his blood counts remained
rather stable, with normocytic anemia (hemoglobin 10.9 g/dl) and thrombocytes
the order of 93,000/␮l. He required no more than 12 erythrocyte transfusions
during his entire lifetime.
At age 31, Michael had to undergo bilateral cataract extraction, laser coagu-
lation of the retina of his right eye, and surgical resection of a malignant bone
tumor of his right jaw. He nevertheless finished his apprenticeship, obtained a
license as a motor boat pilot, and joined the administrative staff of the local
state opera. A year later he experienced increasing pains of the neck and was
diagnosed with advanced squamous cell carcinoma of the lateral and posterior
oropharyngeal wall extending to the base of his skull that had escaped detection
on prior routine examinations. He underwent a heroic 14 h operation with sub-
sequent radiotherapy. After he had received a total dose of only 8 Gy, the radio-
therapy had to be discontinued because of severe thrombocytopenia which
could not be controlled. Michael died at age 32 from the complications of
surgery, radiotherapy and pneumonia.

Marleen B.

Marleen (fig. 2) was born two weeks prior to her due date with hypoplastic
thumbs and a number of fairly large café-au-lait spots. After her first feeding,
she suffered from intestinal tract obstruction due to annular pancreas requiring
immediate surgical correction. Because of her anomalies, a chromosome analy-
sis was ordered yielding a normal female karyotype. All additional investiga-
tions were non-contributory. A first low blood count was observed at age 6, but,
because of a recent episode of infection, was considered transitory and not fol-
lowed up with additional tests. However, upper respiratory tract infections
increased in frequency and severity over the next year, and repeated blood tests
showed a decrease in all three blood cell lineages such that the diagnosis of FA
was finally considered and confirmed by the appropriate laboratory tests.
Marleen B. belongs to complementation group FA-A and is heterozygous for
the FANCA mutation c.4010⫹(1-18)del. The other mutation is unknown.
Marleen quickly became transfusion dependent, and attempts to prevent
her bone marrow failure by androgen and cortisol treatments were unsuccess-
ful. To make things worse, AML was diagnosed at age 8.5. Although treatment
with thioguanin was successful in eradicating her blast cells, it led to rapid dete-
rioration of her bone marrow function. Aspergillosis of her lungs was treated
with granulocyte transfusions (collected from her relatives), but her condition
remained critical. A matching donor could not be found within the available

Fanconi Anemia: A Disease with Many Faces 15

 Fig. 2. (left) Marleen B. at age 13; (center) her hands after surgery (left thumb and
index finger); (right) Marleen S. at age 15.

time such that her physicians were forced to use her father despite 2 mismatches
of his HLA surface antigens. Surprisingly, the bone marrow of her father
engrafted rapidly and without any complication. Much sooner than anticipated
she could be discharged with stable blood counts and in good general health.
Today, seven years after transplant, she is still healthy and lives a full life. She
opted for surgical correction of one of her thumbs (fig. 2) and trained the other
so well that she has excellent manual function. Because of proven growth hor-
mone deficiency, she was started on growth hormone therapy at age 12 which
led to a regular adolescent growth spurt. A very satisfying story for all involved.

Marleen S.

Marleen (fig. 2) was delivered three weeks past her due date via C-section.
She weighed 2,750 g. During her second day of life she underwent surgery for
esophageal atresia type IIIb that was successfully corrected. There were no
complications, and there was no evidence for any other congenital defect. Her
developmental milestones were normal, and she started to use her first words at
the age of 18 months. However, she was difficult to understand and further
investigations revealed a hearing impairment due to middle ear malfunction.
A tympanoplastic procedure as unilateral correction (on her left side) was
attempted, but without notable success. She therefore received bilateral hearing
aids which she still uses today. Supportive logopedic training improved her
speech and hearing.

Dietrich/Velleuer 16
 Somewhat lower than normal blood counts were first noted during prepar-
ation for ear surgery at age 4. A control investigation at age 5 yielded the fol-
lowing results: hemoglobin 10.4 g/dl, leukocytes 3,300/␮l, thrombocytes
70,000/␮l. Neutrophils were low with 10%. During the following months,
hemoglobin rose to 12 g/dl and neutrophils to 35%. There were no infections
such that no measures were taken. Chromosome analysis at that time revealed a
normal female karyotype, but there were no chromosome breakage studies.
In the meantime a horseshoe kidney and low-grade microcephaly had been
additionally diagnosed, but it took until age 6 before the possibility of FA was
considered and confirmed by cell cycle studies. Marleen turned out to belong to
group FA-A, carrying the FANCA mutation c.1567-1G⬎C in a homozygous or
heterozygous state. In addition to the findings described above, a thorough phys-
ical examination at age 6.75 years noted a brownish skin color of her trunk and
freckling of her left axilla. Except for slightly decreased blood counts, all other
investigations were within normal limits, including bone marrow cytology. At
that time her height was 119 cm (50th percentile), her weight was 23 kg (between
the 50th and 75th percentile), and the final report emphasizes her excellent
health. A heart murmur was noted at age 12 which turned out to be due to a per-
sisting ductus arteriosus Botalli. Angioplastic closure was uneventful.
Contrary to most other FA patients, Marleen’s blood counts remained sta-
ble over the years such that she needed no further intervention. Annual bone
marrow biopsies gave only normal results without evidence for cytogenetic
changes. At ages 13 and 14 three hundred ml of bone marrow were obtained
under general anesthesia, frozen and deposited as reserve for future autologous
transplantation and/or gene therapy. Interestingly, there was further improve-
ment of her blood parameters in the wake of these procedures. Neither cell
cycle nor chromosome breakage studies provided evidence for somatic
mosaicism. Now at age 15, Marleen leads a normal teenage life. Her only handi-
cap is her hearing impairment which she controls with her hearing aids. She is
an example of a very mild clinical course despite several congenital anomalies.

Sarah Ninja and Valeska

One of the unsolved questions in FA relates to the observation of discord-

ance of phenotype and clinical course among siblings with identical mutations.
In this context we summarize the clinical histories of two sisters belonging to
complementation group FA-A, both compound heterozygous carriers of the
mutations: c.1360_1826del467 pat (genomic deletion including exons 15–20),
and c3788_3790delTCT mat (p.F1263del). Table 1 compares phenotypes and
clinical course of the two sisters who were born at term with similar birth

Fanconi Anemia: A Disease with Many Faces 17

 Table 1. Divergent course of the disease among siblings with identical mutations

Clinical parameter Sarah Ninja Valeska

Adult height 138 cm 154 cm

Adult weight 41 kg 65 kg
Radial ray anomalies both thumbs, moderately severe single thumb, mild
Café au lait spots ⫹⫹⫹ ⫹⫹
Diagnosis of FA age 7.5 age 2 (because of sister)
Start of androgen treatment age 8 age 8 (following an episode
of osteomyelitis)
Duration of androgen treatment 12 years (due to large size adenomas 12 years (2 mg/kg/day for 3 years,
at age 9 stepwise reduction 0.5 mg/kg/day for 2 years and
to 0.2 mg/kg/day after age 15) 0.1 mg/kg/day for 7 years)
Liver adenomas ⫹⫹⫹⫹⫹ none
Transfusion dependency age 5.5 age 19
(erythrocytes)
Transfusion dependency age 9 none
(thrombocytes)
Duration of transfusion dependency 13 years ⬍1 year
Lifetime total of transfusions ⬎700 (300 ery./400 thrombo.) ⬍15 (erythrocytes only)
Profuse bleeding episodes multiple, for many years few
(nose and gingiva)
Episodes of physical weakness, multiple and severe none
forced inactivity
Infections multiple and severe few
G-CSF treatments starting at age 15 (neutrophils none
below 500)
Clonal bone marrow changes from age 15, multiple clones, from age 19, single
chromosome 3q affected monosomy 7 clone
Age at and cause of death 21 years, multiorgan failure 20 years, severe infections and
due to irreversible brain hemorrhage after alternate
pancytopenia donor HSCT

weights (2.5 kg Sarah Ninja, 2.2 kg Valeska), similar low head circumference
(32.5 and 31 cm, respectively) and similar body length (46 and 47 cm, respec-
tively). Overall, the sisters’ phenotypes were relatively mild with short stature,
moderate microcephaly, thumb anomalies, and pigment spots as the only mani-
festations pointing towards FA (fig. 3).
Sarah Ninja experienced a severe clinical course with repeated life-threatening
episodes of epistaxis, very severe anemia, thrombocytopenia and neutropenia,
leading to early and life-long transfusion dependency and recurrent infections
requiring antibiotics and, from age 15, regular G-CSF treatments. She received

Dietrich/Velleuer 18
 Fig. 3. Sarah Ninja at ages 6 (left) and 19 (center); Valeska at age 17 (right).

the diagnosis of aplastic anemia at age five, and the diagnosis of FA at age 7.5.
Her younger sister Valeska showed a much milder course. Other than Sarah Ninja,
who had been treated with transfusions for 2.5 years prior to starting her on
androgens, Valeska received androgen therapy prior to the onset of transfusion
dependency. After two years, her androgens (oxymetholon, 2 mg/kg/day) were
reduced stepwise over a period of several years to a maintenance dose of
0.1mg/kg/day. Although there were signs of virilization, Valeska never experienced
any severe complications of androgen therapy (large liver adenomas) that forced a
stepwise reduction of androgen therapy to a minimum dose of 0.1 mg/kg/day in
Sarah Ninja. Until her final year of life when the appearance of a monosomy 7
clone in her bone marrow mandated preparation for HSCT, Valeska had sufficient
bone marrow function to lead a normal and very active teenage life. In contrast,
from her childhood years on, Sarah Ninja’s quality of life was impaired, with
many intermittent episodes of severe distress and near-fatal medical complica-
tions due to severe bleedings and/or infections. Without the dedicated care of her
family and her physicians she would have fallen victim to her refractory bone
marrow failure at a much younger age. It is very tragic indeed that the younger
sister Valeska, having been much less affected by the disease throughout her life-
time, succumbed to the complications of alternate donor HSCT and thus died at
an even younger age than her more severely affected older sister.
The medical histories of these two siblings illustrate several points:
(1) clinical course and quality of life can be very different among affected indi-
viduals, even among siblings carrying identical mutations, (2) in the case of the
two sisters, the quality of life depended above all on the degree of residual bone
marrow function and/or responsiveness to androgen therapy, (3) whereas bone
marrow failure is, within obvious limits, amenable to dedicated medical and
family care, the appearance of cytogenetically aberrant bone marrow clones, in
particular trisomy 3q and/or monosomy 7, may herald leukemia which leaves
HSCT as the only but decidedly high risk option.

Fanconi Anemia: A Disease with Many Faces 19

 Concluding Remarks

The first patient among our case reports illustrates the most severe form of
Fanconi anemia, with severe pre- and postnatal growth retardation, failure to
thrive, a multitude of malformations and malfunctions, including severe cogni-
tive impairment which otherwise is rare. Such a child requires 24 h medical and
family care, and his or her life expectancy is severely limited by the multi-
system disease. Despite these severe handicaps, his family felt that caring for
Philipp is a demanding but likewise rewarding experience. The other patients
were far less affected, but the majority nevertheless required life-saving surgery
as newborns or at some point later in their lives. Surgery included correction of
esophageal atresia, esophageal membranes, pancreas annulare, duodenal atre-
sia, urethral obstruction, anal atresia, heart, eyes (cataract extraction), middle
ear, and thumbs. Without these surgical interventions, either these patients
would not have survived or their quality of life would have been severely
impaired.
Pharmacological intervention consisted mainly of androgen therapy which
stabilized blood counts in half of the patients, in the case of Valeska as long as
12 years. Growth hormone was helpful in three of the patients with proven
growth hormone deficiency, and G-CSF was evidently useful in cases of severe
granulocytopenia. With the exception of Marleen S. who never required trans-
fusions, blood product substitution became mandatory in the majority of
patients at some point, amounting to more than 700 transfusions in the case of
Sarah Ninja or to only 13 transfusions in the case of Michael who died at age 32
from squamous cell carcinoma, the most feared solid tumor in FA patients.
Programs and methods for early detection and prevention of SCC must be one
of the highest priorities in FA medical research. Michael is also an example of
endocrine deficiencies which seem to develop much sooner in FA patients than
in ‘normal’ aging, including pituitary dysfunction, hypothyroidism, hyper-
gonadotrophic hypogonadism and diabetes mellitus.
The case history of Dominik illustrates that HSCT is the therapy of choice
if a matching sibling donor is available, and that additional medical complica-
tion, like kidney failure, can be overcome once bone marrow function is
restored. AML as the most frequent hematopoietic malignancy in FA struck
Marleen B. at age 8.5, and despite many concerns a partially mismatched trans-
plant from her father engrafted well, a perfect example of the beneficial nature
of a ‘graft versus leukemia reaction’. With additional growth hormone therapy
and hand surgery her quality of life is excellent. This is also the case for
Marleen S. despite the fact that she had to undergo surgery for esophageal atre-
sia, for middle ear problems and for persisting ductus Botalli. Marleen S. is an
example of FA patients whose low blood counts remain stable over many years

Dietrich/Velleuer 20
without any specific therapy, and without evidence for somatic reversion of one
of her constitutional mutations.
That siblings with identical mutations can show divergent phenotypes
and a divergent course of their disease has been noted and reported before
[e.g. reference 4]. In the case of Sarah Ninja and Valeska this divergence may
in part have been due to the protracted diagnosis of Sarah Ninja who had to be
treated for aplastic anemia with blood products very early in life, 2.5 years
prior to starting her own androgen therapy. Being aware of the diagnosis in the
family when the younger sister Valeska began to show signs of incipient bone
marrow failure, she was treated with androgens right away and she did not
respond with the massive adenomatous growth in the liver that forced drastic
reduction of androgen therapy in Sarah Ninja’s case. We still are largely igno-
rant why adult height and weight, bleeding episodes, infections, medical
emergencies and overall quality of life were so different between the two
sisters carrying identical mutations. As postulated for the divergence of
ages-of-onset of chorea Huntington [5] polymorphisms in genes that interact
or are associated with the affected FA gene might provide some explanation
but remain to be identified in the case of FA. Finally, the case history of
Valeska reminds us that despite undoubted progress (see Chapter by Eyrich
et al.) alternate donor HSCT remains a high risk procedure, particularly if
coincident with clonal bone marrow changes, preexisting chronic infections,
and adulthood.
Altogether, we hope this brief presentation of case histories will intro-
duce the non-specialist reader to the impressive spectrum of phenotypic
manifestations and the widely divergent clinical courses of FA. There is no
doubt that medical interventions and devoted parental care have improved the
quality of life for the great majority of FA patients. There is a clear bias in the
literature and in physicians’ minds in favor of severe cases such as repre-
sented by Philipp among our case histories, because children with severe
growth deficits and malformations are easily recognized and diagnosed.
Likewise, our sample of case histories by necessity lacks the description of
cases that lead a normal or close to normal life as adults, and who may be
diagnosed only in case of malignancy, late onset bone marrow failure, or a
sibling affected with FA [e.g. references 6 and 7]. We personally know of at
least two adult individuals without any congenital malformations and com-
pletely or close to normal blood counts who carry the diagnosis of FA (based
upon family history and chromosome breakage studies) but lead a normal
working and family life. Working out genotype-phenotype correlations in
such extreme cases, either mild or severe, possibly by studying polymor-
phisms in FA-related genes, will remain one of the great challenges in FA
research in the years to come.

Fanconi Anemia: A Disease with Many Faces 21

 Acknowledgements and Dedication

The authors and editors are deeply grateful to the patients and their families who con-
sented to the publication of their first names, medical histories and photographs with the
explicit wish to inform the readers of this volume about the many faces of Fanconi anemia.
They would like to dedicate this paper to the memory of Michael, Sarah Ninja, Valeska and
all the other FA patients who lost their heroic struggle against a deadly disease but who
remain alive in our hearts.

References

1 Joyce O: Fanconi Anemia. Standards for Clinical Care. Fanconi Anemia Research Fund Inc.,
Eugene, Oregon, 1999.
2 Frohnmayer L, Frohnmeyer D: Fanconi Anemia. A Handbook for Families and Their Physicians.
ed 3. Eugene, Oregon, Fanconi Anemia Research Fund Inc., 2000.
3 Fanconi Anämie. Ein Handbuch für Eltern, Patienten und ihre Ärzte. Translated version and exten-
sion of reference 2. Deutsche Fanconi-Anämie-Hilfe e.V., Unna, 2005.
4 Schoof E, Beck JD, Joenje H, Doerr HG: Growth hormone deficiency in one of two siblings with
Fanconi’s anemia complementation group FA-D. Growth Horm IGF Res 2000;10:290–294.
5 Metzger S, Bauer P, Tomiuk J, Laccone F, Didonato S, et al: The S18Y polymorphism in the
UCHL1 gene is a genetic modifier in Huntington’s disease. Neurogenetics 2006;7:27–30.
6 Kwee ML, van der Kleij JM, van Essen AJ, Begeer JH, Joenje H, et al: An atypical case of Fanconi
anemia in elderly sibs. Am J Med Genet 1997;68:62–66.
7 Huck K, Hanenberg H, Gudowius S, Fenk R, Kalb R, et al: Delayed diagnosis and complications
of Fanconi anemia at advanced age – a paradigm. Br J Haematol 2006;133:188–197.

Ralf Dietrich
Deutsche Fanconi-Anämie-Hilfe e.V.
Böckenweg 4
D–59427 Unna (Germany)
Tel. ⫹49 23 08 21 11 or 23 24, E-Mail [email protected]

Dietrich/Velleuer 22
 Schindler D, Hoehn H (eds): Fanconi Anemia. A Paradigmatic Disease for the Understanding of
Cancer and Aging. Monogr Hum Genet. Basel, Karger, 2007, vol 15, pp 23–38

Milestones in Fanconi Anemia Research

M. Digweeda, H. Hoehnb, K. Sperlinga
a
Institute of Human Genetics, Charité – Universitätsmedizin Berlin, Berlin,
b
Department of Human Genetics, University of Würzburg, Würzburg, Germany

Abstract
The recessive disease Fanconi anemia (FA) is a prototype chromosome instability syn-
drome which shows a high level of spontaneous and induced chromosomal aberrations in
combination with a significantly increased cancer risk. Thus, the underlying defect must be
directly or indirectly involved in a fundamental cellular task of long-lived mammalian cells,
the maintenance of genomic integrity. In addition to the delineation of the FA clinical and
cellular phenotypes, prominent milestones in FA research include proof of extensive genetic
heterogeneity, identification of (so far) 12 disease genes, and elucidation of the FA pathway
involved in the repair of crosslinks at arrested replication forks. What is referred to as the
FA/BRCA pathway represents only part of a complex network of protein-protein interactions
which is far from being understood. The ultimate milestone in FA research would be the
achievement of an individualized and curative therapy. Currently, hematopoietic stem cell
transplantation is a promising but still high-risk therapy, and gene therapy is still at the exper-
imental stage. Nonetheless, in a significant proportion of patients, a kind of ‘natural gene
therapy’ can be observed which results from intragenic recombination or from compensating
second site mutations. The elucidation of these somatic events and their underlying mecha-
nisms can be considered a milestone in genetic research. The remarkable progress in FA
research during the last 10–15 years was fostered by the foundation of FA patient support
groups in many countries. These support groups are gratefully acknowledged as important
motivational milestones in FA research.
Copyright © 2007 S. Karger AG, Basel

‘Many phenomena that basic research tries to explain would simply be

unknown had they not been uncovered by the study of diseases. Phenomena such
as spontaneously enhanced chromosome instability in Fanconi anemia or Bloom
syndrome with all their consequences for somatic mutation and cancer formation
were discovered accidentally in the process of examining certain patients for
diagnostic reasons’. This statement in the introductory chapter of Vogel and
Motulsky’s famous textbook, ‘Human Genetics – Problems and Approaches’ [1]
is still valid today. Currently, following the identification of 12 different genes
leading to the Fanconi anemia (FA) phenotype, the statement can be extended to
include the understanding of basic biological principles related to the mainte-
nance of genome stability, the elucidation of relevant pathways and their inter-
connection in genetic networks. Here, a brief and necessarily subjective review is
given of what most researchers in the field might consider milestones in FA
research. More details are presented in subsequent articles of this volume and in a
number of excellent reviews [2–9].

The Foundations of FA Research: Clinical Observations

In 1927, the Swiss pediatrician Guido Fanconi (1892–1979) first described

3 brothers with a specific combination of bone marrow failure (pancytopenia)
and various physical anomalies, such as short stature, hypogonadism and
hyperpigmentation, pointing to a new autosomal recessive disorder [10].
Generally, the substantiation of a new genetic disease is more likely to be based
on particularly severely affected individuals, as these individuals receive prefer-
ential medical attention. As a consequence, the disease in FA index cases is reg-
ularly more severe than in their affected sibs [11]. Only around 70–75% of FA
patients display the ‘classical’ FA phenotype which consists of short stature in
combination with a typical spectrum of congenital defects, including radial ray
defects, pigmentary changes, urogenital malformations, atresia of the external
ear canal, and many others [12]. Endocrine abnormalities reminiscent of pre-
mature aging occur in nearly all FA patients with developmental defects. They
include hyperinsulinemia, growth hormone insufficiency and hypothyroidism
[13]. Reproduction is severely impaired due to failure of spermatogenesis in
males and a low pregnancy rate in females [14].
Following the initial and follow-up descriptions by Fanconi himself
[10, 15], a landmark paper in FA research was the publication by Gmyrek and
Syllm-Rappoport in 1964 who reviewed phenotypic features and clinical course
of 129 patients [16]. Much of what we know today about the spectrum of con-
genital malformations and course of the disease is already mentioned in this
paper. From a European perspective, another landmark publication was the
1976 paper by Traute Schroeder and colleagues, entitled ‘Formal Genetics of
Fanconi’s Anemia’ [17]. Based upon an extensive collection of pedigrees, the
authors provided unequivocal proof of autosomal recessive inheritance despite
the surprising fact that none of the families examined showed evidence for con-
sanguinity. There was a slight preponderance of affected males which retro-
spectively would be compatible with a small proportion of X-linked recessive
cases which we now know are due to mutations in the FANCB gene, the only FA

Digweed/Hoehn/Sperling 24
gene located on the X-chromosome. Because the authors noted a high intrafa-
milial correlation for age of onset and for the severity of malformations, they
correctly predicted genetic heterogeneity and stated that ‘apart from the stan-
dard type, an especially mild type with late onset, few malformations, and a rel-
atively benign course seems to exist’ [17]. They could not have been more
correct, since today we are well aware that ‘mild’ mutations exist in some of the
FA genes, and that the phenomenon of revertant mosaicism may mitigate the
clinical course. Around 1982, Arleen Auerbach started the Fanconi Anemia
International Registry in the United States which has made a number of land-
mark contributions to our understanding of the clinical and genetic heterogene-
ity of FA so far [18–22].
A definitive clinical milestone in FA research was the realization that
patients exhibit a high risk for myelodysplastic syndrome (MDS), and for
developing overt malignancies [23]. Leukemia is found in about 10% of
younger patients (especially acute myeloblastic leukemia, AML), and solid
tumors, particularly squamous cell carcinomas of the oral cavity and genital
area have been reported in almost 10% of older patients. This, in combination
with bone marrow failure, explains the significantly reduced life expectancy of
FA patients. The cumulative incidence of solid tumors reaches 30% by the age
of 45 years and remains the major threat to older FA patients [24, 25]. Since
malignant growth is one of the possible endpoints of somatic mutations, the
high cancer risk reflects the genetic defect in FA, which is related to the defec-
tive maintenance of genomic integrity. The discovery of spontaneous chromo-
somal instability as cytogenetic expression of the (at that time still elusive)
genetic defect marks the first of the many experimental milestones of FA
research.

Chromosomal Instability and Sensitivity to

DNA Crosslinking Agents

On the time scale shown in figure 1, the first milestone following

Fanconi’s original description of three patients was the discovery of chromoso-
mal instability. In 1964, Traute Schroeder and colleagues, then working at the
University of Heidelberg, reported increased spontaneous chromosomal insta-
bility in blood lymphocytes of two affected brothers [26]. Less than a year later,
this key observation was confirmed by the Zurich group of Werner Schmid,
with Guido Fanconi himself as senior author of the paper [27]. The chromoso-
mal aberrations described by the Schroeder and Schmid groups were mainly of
the chromatid type and manifested at metaphase as chromatid breaks or chro-
matid interchanges. Obviously, these originate during the S-phase of the cell

Milestones in Fanconi Anemia Research 25

 Sensitivity towards DNA crosslinkers
Risk for MDS and malignancies

FANCA Revertant mosaicism

Chromosomal instability

FA complex
First description of FA

Genetic heterogeneity

FANCD1
FANCD2
Oxygen sensitivity

FANCL
FANCF

FANCB
FANCJ
FANCE

FANCM
FANCC
1920 1930 1960 1970 1980 1990 2000 2010

HSCT

Fig. 1. Milestones in FA-research. The figure does not yet include FANCN as the most
recently identified FA gene whose product proved identical to PALB2, a protein that functions
downstream of FANCD2 as ‘partner and localizer of BRCA2’ [26, 77].

cycle and involve newly replicated DNA. Since each chromatid break is the vis-
ible manifestation of a DNA double-strand break, the characteristic chromoso-
mal instability of FA indicates that the underlying defect in this genetic disease
is directly or indirectly involved in a fundamental cellular process, the repair of
DNA double-strand breaks. Almost at the same time, spontaneous chromoso-
mal instability was also observed by James German and colleagues in Bloom
syndrome [28]. While in FA cells the breakpoints of the interchanges are almost
randomly distributed, in Bloom syndrome they preferably involve homologous
regions, indicating the impairment of different DNA repair pathways [29].
Looking back at these early data from today’s vantage point it is quite remark-
able how careful observations at the cytogenetic level were able to predict mol-
ecular differences.
About 10 years after the discovery of spontaneous chromosomal instability
as a cytogenetic hallmark of FA, Sasaki and Tonomura showed that the sponta-
neous chromosome instability in FA is associated with a high rate of induced
chromosomal aberrations after treatment with DNA crosslinking agents [30].
The determination of cellular sensitivity towards crosslinking agents such as
diepoxybutane, mitomycin C, cisplatinum or nitrogen mustard still serves as gold
standard for the confirmation of the clinical diagnosis of FA [31]. Because of
the highly variable clinical phenotype, each putative patient requires confirma-
tion via the demonstration of increased in vitro sensitivity towards crosslinking

Digweed/Hoehn/Sperling 26
agents. As an alternative to chromosome breakage studies, crosslink sensitivity
can also be assessed via cell cycle analysis [32].
The term chromosome instability syndrome now covers an increasing
number of recessive disorders which share a high level of spontaneous and
induced chromosomal aberrations. However, the type of chromosome anom-
alies and the specificity of the clastogens vary considerably among the different
disorders. For instance, in Ataxia telangiectasia and Nijmegen breakage syn-
drome, the break points of the spontaneous translocations and inversions in
lymphocytes preferentially involve the T-cell receptor and immunoglobulin
gene loci. Other than FA, these patients are highly sensitive to ionizing radia-
tion, but they share a common characteristic of all chromosome instability syn-
dromes which is a sharply increased risk of malignancy [8].

Analysis of Genetic Heterogeneity

In most organisms, the existence of (intergenic) heterogeneity in recessive

traits can easily be studied by crossing pairs of mutants and analyzing their off-
spring. In the case of heterogeneity the progeny of such crossings has a normal
phenotype due to complementation of the parental defects. In our species, fam-
ily studies have only rarely provided evidence for the presence of distinct genes.
The arrival of somatic cell genetics allowed researchers to fuse different human
cell lines and thus perform experimental complementation studies in vitro. In
FA, normalization of the characteristic hypersensitivity to crosslinkers in prolif-
erating cell hybrids should indicate complementation, and thus prove the exis-
tence of genetic heterogeneity. This was first demonstrated by Zakrzewski and
Sperling in 1980 after fusion of an SV40 transformed FA cell line with diploid
fibroblasts derived from another patient. In contrast to the parental cells, the
resulting cell hybrids did not show hypersensitivity to DNA cross-linkers prov-
ing mutual complementation of distinctive gene defects [33]. However, due to
the inherent chromosomal instability and frequent chromosome loss caused by
the SV40 genome, using SV40 transformed cell lines has severe shortcomings.
These problems were overcome by using Epstein-Barr virus immortalized B
lymphoblasts from different FA patients. EBV transformed lymphoid cell lines
remain mostly diploid and can be sustained with different selectable markers
for the isolation of cell hybrids [34]. Such cell fusion studies using lymphoid
cell lines were first performed in the laboratory of Manuel Buchwald in Toronto
and later on, most extensively, in the laboratory of Hans Joenje in Amsterdam.
These tedious cell fusion studies were of fundamental importance for the ensu-
ing progress of FA research, and the Amsterdam laboratory of Hans Joenje
deserves credit for having provided evidence for at least 12 and possibly even

Milestones in Fanconi Anemia Research 27

more FA complementation groups [9]. Reference cells lines for each newly
described complementation group were generously provided by the Amsterdam
group to laboratories all over the world, boosting FA research.
The genes underlying 12 complementation groups have been identified so
far. Their cDNAs are extremely useful for assigning unclassified patients to a
specific complementation group. For this, the relevant cDNAs are inserted into
retroviral or episomal expression vectors, which are then transferred into the
patient’s cell line and tested for complementation [35]. This was first exempli-
fied in 1997 by using a recombinant retroviral vector which stably integrates into
the host genome [36]. In the following years, vector design and transfection effi-
ciencies were substantially improved by the laboratories of David Williams and
Helmut Hanenberg such that assignment of FA patients to the respective com-
plementation groups has become part of diagnostic routine [37, 38]. Interestingly,
a number of patient cell lines are not complemented by any of the known FA
cDNAs rendering the actual figure of complementation groups higher than 12.
Clearly, experimental demonstration and molecular proof of extensive genetic
heterogeneity has been another major milestone in FA research.

Identification of FA Genes

As illustrated on the time scale of figure 1, the early nineties of the 20th
century marked the beginning of a new, molecular area in FA research. The first
FA gene identified in 1992 belonged to complementation group C and was
termed FACC (FA complementation group C complementing). Since then, the
nomenclature has changed and today is based on a truncation of the name
Fanconi and the letter of the respective complementation group. Therefore the
12 FA genes are: FANCA, FANCB (syn. FAAP95), FANCC, FANCD1 (syn.
BRCA2), FANCD2, FANCE, FANCF, FANCG (syn. XRCC9), FANCJ (syn.
BRIP1), FANCL (syn. FAAP43, PHF9), FANCM (syn. FAAP250, KIAA1596)
and FANCN (syn. PALB2). Figure 2 shows the human chromosome map with
the location of the 12 known FA genes. With the exception of FANCC and
FANCG, which are both located on chromosome 9, and FANCA and FANCN,
which are both located on chromosome 16, the other FA genes are spread over
different chromosomes, including FANCB on the X.
The advent of the Human Genome project paved the way for the identifi-
cation of human disease genes by the technique of positional cloning, reported
for the first time in 1986 [39]. Positional cloning is based on the knowledge of
a gene’s location in the genome, the isolation of candidate genes from the crit-
ical region and the identification of mutations in a candidate gene in patient
DNA. A prerequisite for the success of this approach is the availability of many

Digweed/Hoehn/Sperling 28
 FANCL FANCD2
(2p16.1) (3p25.3)
FANCE
(6p21–22) FANCC FANCG FANCF
(9q22.3) (9p13) (11p15)

1 2 3 4 5 6 7 8 9 10 11 12
FANCN/
FANCD1/ PALB2 FANCJ/
BRCA2 FANCM (16p12) FANCA BRIP1
(13q12.3) (14q21.3) (16q24.3) (17q22)
FANCB
(Xp22.32)

13 14 15 16 17 18 19 20 21 22 X Y

Fig. 2. Human karyotype map with locations of the 12 known Fanconi anemia genes.
Figure taken from reference [75], with permission.

affected individuals, usually from different families, but belonging to the same
complementation group. Thus, due to the extensive genetic heterogeneity, this
approach has obvious limitations in FA. In contrast, the older technique of
‘functional cloning’ of a disease gene depends on fundamental information
about the basic biochemical defect and – due to our ignorance – is rarely applic-
able. In the pre-genomic era, many different biochemical lesions have been
attributed to FA cells, including alterations affecting DNA ligase activity, the
intracellular distribution of topoisomerase, UV excision repair or cellular
NAD⫹ levels. Today we know that these were only secondary phenomena.
Expression cloning is a form of functional cloning which does not require
knowledge of the primary defect. In the case of FA it is based on the transfec-
tion of cells with a normal cDNA library or, alternatively, fusing patient cells with
so called mini cells containing different human chromosomes/chromosomal
regions. Successfully complemented cells are not longer sensitive to cross-
linkers since they now contain a functional FA cDNA. The next step is the iden-
tification of the complementing cDNA/genomic DNA. This approach led to the
cloning of the first gene underlying Fanconi anemia (FANCC) in 1992 by
Manuel Buchwald’s group in Toronto, opening up the molecular avenues in FA
research [40]. Most of the other FA genes were identified by the expression
cloning approach (FANCA, FANCE, FANCF, FANCG, and FANCD2). FANCA
was also independently identified by positional cloning, as was FANCJ.

Milestones in Fanconi Anemia Research 29

 Based on the observation that many FA proteins form a complex together
with other unidentified proteins, FAAPs (Fanconi Anemia Associated Proteins)
that could be isolated by immunoprecipitation, three additional FA genes were
identified. Using this biochemical approach, Weidong Wang collaborating with
the Joenje group showed that FAAP43, a protein with a molecular weight of
43 kDa, was absent in the only patient belonging to complementation group L,
thereby defining FAAP43 as the elusive FANCL gene [41]. Subsequently, the
FA-associated proteins FAAP95 and FAAP250 were shown to be defective in
patients belonging to complementation groups B (X-linked) and M [9].
Altogether, these studies confirmed that each complementation group cor-
responds to a distinct FA gene – with one exception. The exception is group D,
which comprises the FANCD1 and the FANCD2 genes. This heterogeneity is not
well understood, but might be due to the fact that FANCD1, in contrast to almost
all other FA genes, acts downstream of FANCD2 which might have affected the
results of the initial complementation studies. Most importantly, FANCD1 was
subsequently shown to be identical to BRCA2 by systematic screenings of FA
patients for mutations in BRCA2 [42]. It came as a great surprise, of course, that
FANCD1 corresponds to BRCA2, a prominent cancer gene known to be
involved in homology directed DNA repair via its association with the RAD51
recombinase. A single patient initially assigned to complementation group H
was later shown to carry biallelic mutations in FANCA such that the reference
cell line had to be reassigned to group FA-A [43].
In many of the FA genes no functional domains are apparent in the protein
sequences and no strong homologies exist in nonvertebrate species. Conse-
quently, their biochemical function has remained obscure. One exception was
the FANCG gene which was shown to be identical to XRCC9, a gene known to
be involved in DNA repair [44]. The situation changed completely when in
2001 the groups of Markus Grompe and Alan D’Andrea discovered the
FANCD2 gene which is highly conserved in plants (A. thaliana), nematodes
(C. elegans), and insects (Drosophila), indicating the possible conservation of a
‘basic’ FA pathway in lower organisms [45]. In addition, two of the most
recently identified FA genes, FANCJ and FANCM, are identical to conserved
genes with known DNA maintenance functions. FANCJ is identical to the
BRIP1/BACH1, a DNA helicase, which interacts with BRCA1. FANCM has
homology with both helicases and endonucleases, including the archaeal Hef
protein. Another conserved FA gene, FANCL, has a ring finger motif that is typ-
ical for E3 ubiquitin ligases [9].
Most FA patients belong to the complementation groups A, C and G.
However, there are also ethnic differences, which are mainly due to founder
mutations. Thus, most FA patients in the Afrikaans-speaking population in South
Africa belong to group A [46], whereas in the Ashkenazi-Jewish population,

Digweed/Hoehn/Sperling 30
group C is most frequent [47]. Interestingly, there is no strong correlation
between a given complementation group and a given clinical phenotype, simply
because different mutations in the same FA gene can lead to strikingly different
phenotypic consequences. As a rule, null mutations lead to more severe and ear-
lier disease manifestation than mutations that result in a partially functional
gene product [5]. There is no doubt that identification of each single FA gene
represents a milestone in FA research, both with respect to the practice of med-
ical genetics and the understanding of FA gene and protein function.

Elucidation of FA Pathways and Networks

Elucidation of the FA pathway should link the clinical and cellular FA phe-
notype to the mutations in the affected genes. However, recent insights into the
molecular pathophysiology of FA have indicated an enormous degree of com-
plexity in which individual pathways are only parts of a network of protein-
protein interactions. Nonetheless, the most important advances in medical
research are often obtained by reducing complexity to basic principles. In the
case of FA, one major unanswered question is how chromosome instability is
linked to the susceptibility of FA cells to DNA crosslinking agents.
Some important milestones in this research were:
• The proof in 1997 that the FA proteins, A and C, form a nuclear complex
and the subsequent realization that the FA proteins B, E, F, G, L, and M are
also part of this core complex [48, 49].
• The discovery that the FANCD2 protein takes a central position in the FA
pathway. The FANCD2 protein is activated by the addition of a ubiquitin
molecule in response to DNA damage [50].
• The finding that the FANCD2 protein directly interacts with known DNA
repair proteins such as BRCA1 and BRCA2/FANCD1, linking DNA
double-strand-break (DSB) repair with DNA cross-link repair [50, 51].
It is now well established that each of the proteins that are assembled to the
FA core complex is needed for monoubiquitination of FANCD2. The catalytic
function is obviously provided by the FANCL protein with its E3 ubiquitin lig-
ase domain. Thus, all these 8 proteins act upstream of the evolutionarily con-
served FANCD2 protein. The posttranslational modification of FANCD2
appears to be crucial for its activity. Proficient monoubiquitination is easily rec-
ognized by the appearance of a larger FANCD2 isoform (FANCD2-L). Western
blotting thus permits convenient subtyping of FA cell lines, since patients with
defects in any of the core complex genes fail to monoubiquitinate FANCD2 and
display only the small (FANCD2-S) isoform [52]. Activated, intact FANCD2
protein is targeted to discrete nuclear foci. The combination of proteins

Milestones in Fanconi Anemia Research 31

comprising these foci varies with the cell cycle and the type of DNA damage.
FANCD2 co-localizes in these foci with FANCD1/BRCA2 and RAD51 (pro-
teins involved in error-free repair) or BRCA1 and RAD50 (proteins involved in
error prone repair). FA foci are found in S-phase cells and after treatment with
crosslinkers and ionizing radiation. Thus, the association with proteins involved
in DNA repair indicates that the FA pathway is directly involved in the DNA-
damage-response during S-phase [5, 9].
Following DNA damage, the FA core complex directly binds to chromatin,
whereby the FANCM protein with its helicase motif seems to be essential for
recognizing crosslinked DNA [53]. It has been speculated that the activated
core complex stabilizes arrested replication forks blocked by crosslinks,
thereby promoting repair via DNA double-strand-break intermediates, either by
translesion synthesis or by homologous recombination. FANCD1/BRCA2,
FANCJ, and FANCN are involved in these repair processes and act downstream
of FANCD2 in the FA pathway(s). Clearly, we are only just beginning to under-
stand the complexity of FA mediated crosslink repair, and confirmation of the-
oretical models by biochemical studies is still very limited.
Moreover, it is still unknown how the high sensitivity of FA cells to chro-
mosomal breakage by atmospheric oxygen is linked to the FA pathway [54].
Reducing oxygen tension in tissue culture incubators from 20% to 5% normal-
izes the cellular FA phenotype with respect to chromosomal breakage and
G2-phase delay [55]. The molecular explanation of the obvious oxygen sensi-
tivity of FA cells will represent another milestone in FA research. It also needs
to be clarified whether, in the patients themselves, oxidative stress is a primary
or secondary phenomenon, and to which extent oxidative stress contributes to
the features of the FA clinical phenotype [56, 57].
What is the reason for the excessive apoptosis of bone marrow stem cells in
FA patients resulting in hematopoietic failure? What is the functional relevance
of the interaction of FANCC with STAT1, hsp70, NADPH cytochrome p450
reductase, FAZF, GRP94 and cdc2 [7], and in how many additional pathways are
the other FA proteins involved? Clearly, the complexity of protein-protein inter-
actions is bewildering and, in addition, depends on the specific stage of develop-
ment and the individual tissue. The combination of reductionistic approaches
with systems biology might lead to insight into these networks. It is safe to pre-
dict that the way to this goal will be paved with future milestones in FA research.

The Difficult but Promising Road to Therapy

Deeper insight into the molecular pathogenesis of FA will hopefully result

in a better understanding of its pathophysiology, which is the prerequisite for a

Digweed/Hoehn/Sperling 32
rational, individualized therapy. From the patient’s point of view, this would
clearly be the ultimate milestone in FA research. Currently, there is no causal
therapy for the genetic instability underlying the clinical manifestations of FA.
In this situation, prenatal diagnosis is requested by couples at risk [58]. Both
functional studies, e.g. determination of sensitivity towards DNA crosslinking
agents, and direct FA gene mutation analysis can be performed with fetal cells.
Although prenatal diagnosis is an important by-product of FA research,
it merely represents a makeshift solution rather than a genuine therapeutic
milestone.
A number of conventional treatments are available to combat bone marrow
failure, including steroid and cytokine medications, but the first line of therapy
is hematopoietic stem cell transplantation (HSCT), pioneered in FA among oth-
ers by Elaine Gluckman and her coworkers [59]. As indicated graphically in
figure 1, both the number of patients transplanted and the success of HSCT
have greatly increased during recent years, mostly due to improvements of con-
ditioning regimens and graft T-cell depletion [60]. Even with suitable donors
HSCT still carries a relatively high risk. In this context, a key observation made
in the laboratory of Heidemarie Neitzel has great impact on the difficult deci-
sion whether and when to subject a patient to transplantation. Neitzel and col-
leagues found that specific clonal chromosomal aberrations, most significantly
gains of 3q, precede the onset of myelodysplasia and/or acute myelocytic
leukemia [61]. Knowing whether a potentially devastating conversion to malig-
nant cell growth is in the offing is an important step forward in the medical
management of FA patients. In combination with the establishment of a rapid
and sensitive FISH-assay [62], this has important consequences for early inter-
vention and treatment of FA patients. Given a low a priori chance of 1 in 4, and
given the small sizes of present day families, matching sibling donors are avail-
able in only a minority of affected families. In such a precarious situation, pre-
implantation genetic diagnosis (PGD) for the selection of an HLA identical
embryo as potential stem cell donor has been proposed and performed [63, 64].
The first case involving an FA family made headlines worldwide, both positive
and negative (‘designer baby’). Clearly, PGD is neither a desirable milestone in
FA research nor a breakthrough for patient care but rather an expression of the
desperate situation of parents faced with limited therapeutical possibilities.
Both the economical and psychological burden of PGD are tremendous and the
success rate is, for biological reasons, very low such that PGD is unlikely to
have a major impact on the medical care of FA families.
Despite considerable efforts, gene therapy is still at the experimental
stage [65]. Nonetheless, in a significant proportion of FA patients a kind of
‘natural gene therapy’ can be observed. This is based on the observation that
some patients have two sets of peripheral blood cells: MMC sensitive and

Milestones in Fanconi Anemia Research 33

MMC insensitive [66, 67]. In some cases, the wildtype cells completely
replace the defective ones. Skin fibroblasts, however, remain MMC sensitive.
This self-correction, which ideally should take place in a hematopoietic stem
cell [68] may more often than not lead to improvement of the hematological
status [67, 69, 70]. In recessive diseases, such as FA, different molecular
mechanisms have been shown to be instrumental in somatic reversion, includ-
ing intragenic recombination, gene conversion, or back-mutation. In addition,
partial restoration of protein function can be due to compensating mutations
in cis of one of the affected alleles [66, 71]. The elucidation of these mecha-
nisms can surely be considered a milestone in genetic research. A dogma of
classical genetics was that recombination occurred only between genes but not
within a gene. In the fifties of the last century, this dogma was rejected for
prokaryotes and fungi, in the sixties, for Drosophila, and, in the nineties, for
our species. Intragenic recombination is highly increased in compound het-
erozygotes with the chromosomal instability syndrome Bloom syndrome [72].
In FA the occurrence of reverted cells in peripheral blood indicates a likely
proliferative advantage of reversed cells within the FA bone marrow. In addi-
tion, analysis of the clonal expansion of these stem or progenitor cells is
expected to give insight into their proliferating potential, which is of relevance
for somatic gene therapy in general. Unfortunately, in the murine FA knockout
mouse models, hematopoiesis is almost unaffected and tumor incidence is not
increased [73, 74]. Thus, with respect to the development of somatic gene
therapy and the understanding of FA pathophysiology, these models are only
of limited value.
The transition from the cellular to the (molecular) genetic approach about
15 years ago opened completely new perspectives for FA research and gave
exciting new insights into basic problems in cell biology. We now know that the
FA family of genes plays an important role in the maintenance of genomic sta-
bility, and thereby in the prevention of cancer and premature aging. This
progress was accompanied and promoted by the foundation of FA patient sup-
port groups in many countries. Their goal is to catalyze and support scientific
research on Fanconi anemia. In the US, under the guidance of Lynn and Dave
Frohnmayer, the Fanconi Anemia Research Fund has raised considerable sums
for FA research. In addition, the FARF, as it is known, has edited several
brochures for FA patients and their physicians, provides newsletters and infor-
mation on research projects, and organizes annual meetings both for scientists
and FA families. Parallel to this development in the USA, Ralf Dietrich founded
a patient support group in Germany with aims and activities similar to those of
FARF. The enormous contribution of the Deutsche Fanconi-Anämie-Hilfe
e.V. to progress in the understanding of FA can hardly be overstated. As scien-
tists we greatly appreciate the support and motivation provided by the family

Digweed/Hoehn/Sperling 34
support groups. As such, each of these groups clearly represents a most
welcome milestone in the common quest to improve the life expectancy and
quality of life of FA patients.

Note Added in Proof

The elusive gene underlying complementation group FA-I (Levitus et al., 2004) has
been identified as a paralog and binding partner of FANCD2, bringing the total number of
identified FA genes to 13. FANCI was detected as an ATM/ATR kinase target protein
required for resistance to mitomycin C. Like FANCD2, the newly detected FANCI protein is
monoubiquitinated at a highly conserved lysine residue. In the chain of events leading to
DNA crosslink repair, FANCI and FANCD2 appear to form an interdependent complex (‘ID-
complex’) required for mutual ubiquitination, activation and association with chromatin
(Smorgorzewska et al., 2007).

References

Levitus M, Rooimans MA, Steltenpool J, Cool NF, Oostra AB, et al: Heterogeneity in Fanconi anemia:
evidence for 2 new genetic subtypes. Blood 2004;103:2498–2503.
Smogorzewska A, Matsuoka S, Vinciguerra P, McDonald ER III, Hurov KE, et al: Identification of the
FANCI Protein, a monoubiquitinated FANCD2 Paralog required for DNA repair. Cell 2007;
doi:10.1016/j.cell.2007.03.009.

References

1 Vogel F, Motulsky AG: Human genetics; in Vogel F, Motulsky AG (eds): Human Genetics
‘Problems and Approaches’, ed 3. Heidelberg-Berlin, Springer, 2002.
2 Digweed M, Sperling K: Molecular analysis of Fanconi anemia. Bioessays 1998;18:579–585.
3 Joenje H, Ketan JP: The emerging genetic and molecular basis of Fanconi Anemia. Nat Rev Genet
2001;2:446–457.
4 D’Andrea AD, Grompe M: The Fanconi Anemia/BRCA pathway. Nature 2002;3:23–34.
5 Kennedy RD, D’Andrea AD: The Fanconi Anemia/BRCA pathway: new faces in the crowd. Genes
Dev 2005;19:2925–2940.
6 Niedernhofer LJ, Lalai AS, Hoeijmakers HJ: Fanconi Anemia (Cross) linked to DNA repair. Cell
2005;123:1191–1198.
7 Bagby GC, Alter BP: Fanconi Anemia. Semin Hematol 2006;43:147–156.
8 Digweed M: Genomic instability in Fanconi anemia and Nijmegen Breakage Syndrome. Genome
Dyn Stab 2007;1:363–380.
9 Levitus M, Joenje H, de Winter JP: The Fanconi anemia pathway of genomic maintenance. Cell
Oncol 2006;28:3–29.
10 Fanconi G: Familiäre infantile perniziosartige Anämie (perniziöses Blutbild und Konstitution).
Jahrb Kinderh 1927;117:257–280.
11 Glanz A, Fraser FC: Spectrum of anomalies in Fanconi anemia. J Med Genet 1982;19:412–416.
12 Gillio AP, Verlander PC, Batish SD, Giampietro PF, Auerbach AD: Phenotypic consequences of
mutations in the Fanconi anemia FAC gene: an International Fanconi Anemia Registry study.
Blood 1997;90:105–110.

Milestones in Fanconi Anemia Research 35

13 Wajnrajch MP, Gertner JM, Huma Z, Popovic J, Lin K, et al: Evaluation of growth and hormonal sta-
tus in patients referred to the International Fanconi Anemia Registry. Pediatrics 2001;107: 744–754.
14 Alter BP, Frissora CL, Halperin DS, Freedman MH, Chitkara U, et al: Fanconi’s anemia and preg-
nancy. Br J Haematol 1991;77:410–418.
15 Fanconi G: Familial constitutional panmyelocytopathy, Fanconi’s anemia (F.A.). I. Clinical aspects.
Semin Hematol 1967;4:233–240.
16 Gmyrek D, Syllm-Rapoport I: On Fanconi’s Anemia. Analysis of 129 Cases. Z Kinderheilkd
1964;91:297–337.
17 Schroeder TM, Tilgen D, Kruger J, Vogel F: Formal genetics of Fanconi’s anemia. Hum Genet
1976;32:257–288.
18 Auerbach AD, Rogatko A, Schroeder-Kurth TM: International Fanconi Anemia Registry: relation
of clinical symptoms to diepoxybutane sensitivity. Blood 1989;73:391–396.
19 Auerbach AD, Allen RG: Leukemia and preleukemia in Fanconi anemia patients. A review of the
literature and report of the International Fanconi Anemia Registry. Cancer Genet Cytogenet 1991;51:
1–12.
20 Giampietro PF, Adler-Brecher B, Verlander PC, Pavlakis SG, Davis JG, Auerbach AD: The need
for more accurate and timely diagnosis in Fanconi anemia: a report from the International Fanconi
Anemia Registry. Pediatrics 1993;91:1116–1120.
21 Butturini A, Gale RP, Verlander PC, Adler-Brecher B, Gillio AP, Auerbach AD: Hematologic
abnormalities in Fanconi anemia: an International Fanconi Anemia Registry study. Blood 1994;84:
1650–1655.
22 Kutler DI, Singh B, Satagopan J, Batish SD, Berwick M, et al: A 20-year perspective on the
International Fanconi Anemia Registry (IFAR). Blood 2003;101:1249–1256.
23 Schroeder TM, Kurth R: Spontaneous chromosomal breakage and high incidence of leukemia in
inherited disease. Blood 1971;37:96–112.
24 Alter BP, Caruso JP, Drachtman RA, Uchida T, Velagaleti GV, Elghetany MT: Fanconi anemia:
myelodysplasia as a predictor of outcome. Cancer Genet Cytogenet 2000;117:125–131.
25 Rosenberg PS, Greene MH, Alter BP: Cancer incidence in persons with Fanconi anemia. Blood
2003;101:822–826.
26 Schroeder TM, Anschütz F, Knopp A: Spontane Chromosomenaberrationen bei familiärer
Panmyelopathie. Humangenetik 1964;1:194–196.
27 Schmid W, Scharer K, Baumann T, Fanconi G: Chromosomal fragility in familial panmyelopathy
(Fanconi type). Schweiz Med Wochenschr 1965;95:1461–1464.
28 German J, Archibald R, Bloom D: Chromosomal breakage in a rare and probably genetically
determined syndrome of man. Science 1965;148:506–507.
29 Schroeder TM, German J: Bloom’s syndrome and Fanconi’s anemia: demonstration of two distinc-
tive patterns of chromosome disruption and rearrangement. Humangenetik 1974;25:299–306.
30 Sasaki MS, Tonomura A: A high susceptibility of Fanconi’s anemia to chromosome breakage by
DNA cross-linking agents. Cancer Res 1973;33:1829–1836.
31 Auerbach AD: Fanconi anemia diagnosis and the diepoxybutane (DEB) test. Exp Hematol 1993;21:
731–733.
32 Seyschab H, Friedl R, Sun Y, Schindler D, Hoehn H, et al: Comparative evaluation of diepoxybu-
tane sensitivity and cell cycle blockage in the diagnosis of Fanconi anemia. Blood 1995;85:
2233–2237.
33 Zakrzewski S, Sperling K: Genetic heterogeneity of Fanconi’s anemia demonstrated by somatic
cell-hybrids. Hum Genet 1980;56:81–84.
34 Strathdee CA, Duncan AM, Buchwald M: Evidence for at least four Fanconi anemia genes includ-
ing FACC on chromosome 9. Nat Genet 1992;1:196–198.
35 Pulsipher M, Kupfer GM, Naf D, Suliman A, Lee JS, et al: Subtyping analysis of Fanconi anemia
by immunoblotting and retroviral gene transfer. Mol Med 1998;4:468–479.
36 Fu KL, Foe JR, Joenje H, Rao KW, Lui JM, Walsh CE: Functional correction of Fanconi anemia
group A hematopoietic cells by retroviral gene transfer. Blood 1997;90:3296–3303.
37 Hanenberg H, Batish SD, Pollok KE, Vieten L, Verlander PC, Leurs C, et al: Phenotypic correction
of primary Fanconi anemia T cells with retroviral vectors as a diagnostic tool. Exp Hematol 2002;30:
410–420.

Digweed/Hoehn/Sperling 36
38 Chandra S, Levran O, Jurickova I, Maas C, Kapur R, et al: A rapid method for retrovirus-mediated
identification of complementation groups in Fanconi anemia patients. Mol Ther 2005;12:
976–984.
39 Royer-Pokora B, Kunkel LM, Monaco AP, Goff SC, Newburger PE, et al: Cloning the gene for an
inherited human disorder – chronic granulomatous disease – on the basis of its chromosomal location.
Nature 1986;322:32–38.
40 Strathdee CA, Gavish H, Shannon W, Buchwald M: Cloning of cDNAs for Fanconi’s anemia by
functional complementation. Nature 1992;356:763–767.
41 Meetei AR, Levitus M, Xue Y, Medhurst AL, Zwaan M, et al: X-linked inheritance of Fanconi ane-
mia complementation group B. Nat Genet 2004;36:1219–1224.
42 Howlett NG, Taniguchi T, Olson S, Cox B, Waisfisz Q, et al: Biallelic inactivation of BRCA2 in
Fanconi anemia. Science 2002;297:606–609.
43 Joenje H, Levitus M, Waisfisz O, D’Andrea A, Garcia-Higuera I, et al: Complementation analysis
in Fanconi anemia: assignment of the reference FA-H patient to group A. Am J Hum Genet 2000;67:
759–762.
44 De Winter JP, Waisfisz Q, Rooimans MA, van Berkel CGM, Bosnoyan-Collins L, et al:
The Fanconi anemia group G gene FANCG is identical with XRCC9. Nat Genet 1998;20:
281–283.
45 Timmers C, Taniguchi T, Hejna J, Reifsteck C, Lucas L, et al: Positional cloning of a novel Fanconi
anemia gene FANCD2. Mol Cell 2001;7:241–248.
46 Tipping AJ, Pearson T, Morgan NV, Gibson RA, Kuyt LP, et al: Molecular and genealogical evi-
dence for a founder effect in Fanconi anemia families of the Afrikaner population of South Africa.
Proc Natl Acad Sci USA 2001;98:5734–5739.
47 Whitney MA, Saito H, Gibson RA, Moses RE, Grompe M: A common mutation in the FACC gene
causes Fanconi anemia in Ashkenazi-Jewish individuals. Nat Genet 1993;4:202–205.
48 Kupfer GM, Naf D, Suliman A, Pulsipher M, D’Andrea AD: The Fanconi anemia proteins, FAA
and FAC, interact to form a nuclear complex. Nat Genet 1997;17:487–490.
49 De Winter JP, van der Weel L, de Groot J, Stone S, Waisfisz O, et al: The Fanconi anemia protein
FANCF forms a nuclear complex with FANCA, FANCC and FANCG. Hum Mol Genet 2000;9:
2665–2674.
50 Garcia-Higuera I, Taniguchi T, Ganesan S, Meyn MS, Timmers C, et al: Interaction of the Fanconi
anemia proteins and BRCA1 in a common pathway. Mol Cell 2001;7:249–262.
51 Taniguchi T, Garcia-Higuera I, Andreassen PR, Gregory RC, Grompe M, D’Andrea AD: S-phase-
specific interaction of the Fanconi anemia protein, FANCD2, with BRCA1 and RAD51. Blood
2002;100:2414–2420.
52 Shimamura A, D’Andrea AD: Subtyping of Fanconi anemia patients: implications for clinical
management. Blood 2003;102:3459.
53 Wang X, Andreassen PR, D’Andrea AD: Functional interaction of monoubiquitinated FANCD2
and BRCA2/FANCD1 in chromatin. Mol Cell Biol 2004;24:5850–5862.
54 Joenje H, Arwert F, Eriksson AW, de Koning H, Oostra AB: Oxygen-dependence of chromosomal
aberrations in Fanconi’s anemia. Nature 1981;290:143–144.
55 Schindler D, Hoehn H: Fanconi anemia mutation causes cellular susceptibility to ambient oxygen.
Am J Hum Genet 1988;43:429–435.
56 Pagano G, Youssoufian H: Fanconi anemia proteins: major roles in cell protection against oxida-
tive damage. Bioessays 2003;25:589–595.
57 Pagano G, Degan P, d’Ischia M, Kelly FJ, Nobili B, et al: Oxidative stress as a multiple effector in
Fanconi anemia clinical phenotype. Eur J Haematol 2005;75:93–100.
58 Auerbach AD, Adler B, Chaganti RS: Prenatal and postnatal diagnosis and carrier detection of
Fanconi anemia by a cytogenetic method. Pediatrics 1981;67:128–135.
59 Gluckman E, Auerbach AD, Horowitz MM, Sobocinski KA, Ash RC, et al: Bone marrow trans-
plantation in Fanconi anemia. Blood 1995;86:2856–2862.
60 Boulad F, Gillio A, Small TN, George D, Prasad V, et al: Stem cell transplantation for the
treatment of Fanconi anemia using a fludarabine-based cytoreductive regimen and T-cell-
depleted related HLA-mismatched peripheral blood stem cell grafts. Br J Haematol 2000;111:
1153–1157.

Milestones in Fanconi Anemia Research 37

61 Tönnies H, Huber S, Kühl JS, Gerlach A, Ebell W, Neitzel H: Clonal chromosomal aberrations in
bone marrow cells of Fanconi anemia patients: gains of the chromosomal segment 3q26q29 as an
adverse risk factor. Blood 2003;101:3872–3874.
62 Tönnies H, Huber S, Volarikova E, Gerlach A, Neitzel H: Interphase FISH-assay for the detection
of MDS- and AML-associated chromosomal imbalances on uncultivated bone marrow and periph-
eral blood cells; in Schindler D, Hoehn H (eds): Fanconi Anemia, pp 95–109. Karger, Basel, 2007.
63 Grewal SS, Kahn JP, MacMillan ML, Ramsay NK, Wagner JE: Successful hematopoietic stem cell
transplantation for Fanconi anemia from an unaffected HLA-genotype-identical sibling selected
using preimplantation genetic diagnosis. Blood 2004;103:1147–1151.
64 Rechitzky S, Kuliev A, Tur-Kaspa I, Morris R, Verlinsky Y: Preimplantation genetic diagnosis
with HLA matching. Reprod Biomed Online 2004;9:210–221.
65 Croop JM: Gene therapy for Fanconi anemia. Curr Hematol Rep 2003;2:335–340.
66 Kwee ML, Poll EH, van de Kamp JJ, de Koning H, Eriksson AW, Joenje H: Unusual response to
bifunctional alkylating agents in a case of Fanconi anemia. Hum Genet 1983;64:384–387.
67 Lo Ten Foe JR, Kwee ML, Rooimans MA, Oostra AB, Veerman AJ, et al: Somatic mosaicism
in Fanconi anemia: molecular basis and clinical significance. Eur J Hum Genet 1997;5:
137–148.
68 Gregory JJ Jr, Wagner JE, Verlander PC, Levran O, Batish SD, et al: Somatic mosaicism in
Fanconi anemia: evidence of genotypic reversion in lymphohematopoietic stem cells. Proc Natl
Acad Sci USA 2001;98:2532–2537.
69 Gross M, Hanenberg H, Lobitz S, Friedl R, Herterich S, et al: Reverse mosaicism in Fanconi
anemia: natural gene therapy via molecular self-correction. Cytogenet Genome Res 2002;98:
126–135.
70 Soulier J, Leblanc T, Larghero J, Dastot H, Shimamura A, et al: Detection of somatic mosaicism
and classification of Fanconi anemia patients by analysis of the FA/BRCA pathway. Blood
2005;105:1329–1336.
71 Waisfisz Q, Morgan NV, Savino M, de Winter JP, van Berkel CG, et al: Spontaneous functional
correction of homozygous Fanconi anemia alleles reveals novel mechanistic basis for reverse
mosaicism. Nat Genet 1999;22:379–383.
72 Ellis NA, Lennon DJ, Proytcheva M, Alhadeff B, Henderson EE, German J: Somatic intragenic
recombination within the mutated locus BLM can correct the high sister-chromatid exchange phe-
notype of Bloom syndrome cells. Am J Hum Genet 1995;7:1019–1027.
73 Chen M, Tomkins DJ, Auerbach W, McKerlie C, Youssoufian H, et al: Inactivation of Fac in mice
produces inducible chromosomal instability and reduced fertility reminiscent of Fanconi anemia.
Nat Genet 1996;12:448–451.
74 Cheng NC, van de Vrugt HJ, van der Valk MA, Oostra AB, Krimpenfort P, et al: Mice with a tar-
geted disruption of the Fanconi anemia homolog Fanca. Hum Mol Genet 2000;9:1805–1811.
75 Kalb R, Neveling K, Nanda I, Schindler D, Hoehn H: Fanconi anemia: causes and consequences of
genetic instability; in Volff J-N (ed): Genome and Disease. Genome Dynamics. Karger, Basel,
2006, vol 1, pp 218–242.
76 Xia B, Dorsman JL, Ameziane N, de Vries Y, Rooimans MA et al: Fanconi anemia is associated
with a defect in the BRCA2 partner PALB2. Nat Genet 2007;39:159–161.
77 Reid S, Schindler D, Hanenberg H, Barker K, Hanks S, et al: Biallelic mutations in PALB2 cause
Fanconi anemia subtype FA-N and predispose to childhood cancer. Nat Genet 2007;39:162–164.

Karl Sperling
Institute of Human Genetics, Charité – Universitätsmedizin Berlin
Augustenburger Platz 1
D–13353 Berlin (Germany)
Tel. ⫹49 30 450566081, Fax ⫹49 30 450566904, E-Mail [email protected]

Digweed/Hoehn/Sperling 38
 Schindler D, Hoehn H (eds): Fanconi Anemia. A Paradigmatic Disease for the Understanding of
Cancer and Aging. Monogr Hum Genet. Basel, Karger, 2007, vol 15, pp 39–58

Fanconi Anemia Genes: Structure,

Mutations, and Genotype-Phenotype
Correlations
R. Kalba, K. Nevelinga, S. Herterichb, D. Schindlera
a
Department of Human Genetics, and bDepartment of Clinical Chemistry and
Pathobiochemistry, University of Würzburg, Würzburg, Germany

Abstract
This article is meant to introduce the reader to the FA/BRCA pathway, the currently
known human FA genes, and their mutations. We review structure and mutational profiles of
the 12 FA genes identified to date. The FA genes display striking variability of size and muta-
tional load which partly reflect their genomic composition and functional constraints.
Biallelic mutations in three genes (FANCA, FANCC and FANCG) account for over 80% of all
FA patients worldwide, and there are only single or very few patients with mutations in the
more recently discovered FA genes. FANCA as the most frequently affected gene displays the
entire spectrum of genetic alterations, including at least 32% large deletions correlated to
Alu-mediated recombination. Viability of FA-D2 patients appears to depend on the presence
of hypomorphic mutations and residual protein. Gene frequencies in excess of 1:100 have
been observed, mainly for mutations in FANCA and FANCC, in isolated populations with
defined founder mutations. With the exception of complementation groups FA-D1 and FA-N
whose patients overwhelmingly present with malignancies in early childhood, most FA genes
lack strict genotype-phenotype correlations. Rather, the particular type of mutation in a given
gene appears to determine the severity of congenital malformations, the age of onset of bone
marrow failure, and the length of survival. FA patients who reach adulthood may carry mild
mutations or may have developed somatic reversions. Prospective studies are needed for a
more definitive exploration of genotype-phenotype correlations.
Copyright © 2007 S. Karger AG, Basel

Around 80% of all patients suffering from Fanconi anemia (FA) belong to
one of three prevalent complementation groups (FA-A, FA-C and FA-G).
According to the International Fanconi Anemia Registry (IFAR), 57% of the
FA patients have mutations in FANCA, 15% in FANCC and 9% in FANCG, fol-
lowed by 4% in FANCD1 and 3% in FANCD2 [1]. Mutations in the other FA
genes are rare, and only single patients have been described for FANCL and
FANCM. The estimated overall incidence of FA is 1 in 200,000 with a carrier
frequency of 1 in 300 [2]. However, much higher patient and carrier frequen-
cies are encountered in certain populations due to founder and inbreeding
effects (see below).
The initial FA-complementation groups were defined via somatic cell fusion
[3, 4], but a variety of strategies were applied for the identification of the under-
lying genes. Expression cloning was used for the identification of FANCA [5],
FANCC [6], FANCE [7], FANCF [8] and FANCG [9], while positional cloning
independently led to the identification of FANCA [10], but also to the identifica-
tion of FANCD2 [11] and FANCJ [12, 13]. FANCD1 was identified by a classical
candidate gene approach [14], as was FANCN [15]. Biochemical isolation of the
respective proteins led to the identification of FANCB [16], FANCL [17] and
FANCM [18]. The identification of the first FA genes (FANCC, FANCA and
FANCG) provided only few hints to their molecular function. Except for FANCG,
which is identical with XRCC9, the initially discovered FA genes shared neither
significant homologies to other proteins nor to known functional domains. This
changed when Howlett and coworkers reported that the breast cancer associated
protein 2 (BRCA2) carries biallelic mutations in FA patients belonging to com-
plementation group D1 [14]. This finding implicated the likely involvement of at
least some of the FA proteins in DNA repair via homologous recombination. In
addition, some of the subsequently discovered genes contain protein domains
with defined biochemical functions. For example, FANCL or PHF9/Pog has a
RING domain with ubiquitin E3 ligase activity [17], FANCJ (identical to
BRIP1/BACH1) has a DNA helicase domain [13], and FANCM (homolog to Hef)
shows homologies to DNA translocases [18]. Specific functional properties can
also be deduced for two of the most recently discovered FA proteins, FANCJ and
FANCN, since these proteins interact with and contribute to the functions of the
well known breast cancer proteins BRCA1 and BRCA2 [12, 19]. The current
view is that FA proteins participate in a complex molecular crosstalk with a large
number of other proteins involved in the regulation and execution of DNA main-
tenance functions, including homology directed DNA repair [20, 21].

The FA/BRCA Pathway

As a specific but not exclusive function, FA genes are thought to play an

important role in the removal of DNA interstrand crosslinks (ICL). Cells of all
FA subgroups display a characteristic hypersensitivity against DNA damage
induced by bifunctional alkylating agents implying cooperation of FA proteins
in a common pathway, which is referred to as FA/BRCA pathway. The current

Kalb/Neveling/Herterich/Schindler 40
 DNA damage

FA core complex
Interstrand crosslinks
F Stalled replication forks
Double strand breaks
E C
G
A M IR
B ? ATM NBS1
BLM FAAP MRE11
100
L ATR RAD50
RPA TOPIIIa
NMR complex

ub
ub P P
D2 D2 D2

ub
USP1

Nuclear foci

Cell cycle control

ATR
BLM RAD51 XRCC3
RPA BRCA2/ NBS1
FANCN FANCD1
ub BRCA1 MRE11 DNA repair
D2 RAD50 – homologous recombination
H2AX
H2AX
G BRIP1/
H2AX
– translesion synthesis
H2AX H2AX
H2AX FANCJ
E C

Fig. 1. Current concept for the involvement of FA proteins in the response to DNA
damage. For explanations see text.

model of FA protein function (cf. fig. 1) includes direct links to recombinational

types of DNA repair and translesion synthesis [22, 23]. As shown in figure 1, spe-
cific types of DNA damage result in the assembly of at least 8 FA proteins
(FANC-A, B, C, E, F, G, L and M) into a nuclear multisubunit complex (FA
core complex) that is required for FANCD2 monoubiquitination at lysine 561
[24]. A member of the complex, the RING finger containing protein FANCL,

Fanconi Anemia Genes: Structure, Mutations, and Genotype-Phenotype Correlations 41

provides the presumptive ubiquitin E3 ligase activity [18]. An extended form of
the FA core complex is the so called ‘BRAFT complex’ (BLM, RPA, FA, Topo
III) in which other proteins like BLM helicase, topoisomerase III alpha and
replication protein A (RPA) participate [25], underlining the close association if
not partial overlap of DNA replication and DNA repair functions.
A central step in the FA/BRCA pathway is the monoubiquitination of
FANCD2. This post-translational modification causes a dynamic nuclear relo-
calization of FANCD2 from a state of diffuse nuclear distribution into discrete
nuclear foci containing many other proteins involved in DNA repair [24, 26].
Colocalization with FANCD2 was found for a number of proteins involved in
DNA damage signaling including ATR kinase [27], components of homologous
recombination (RAD51, BRCA1, BRCA2) [28], proteins of the non-homolo-
gous end joining pathway such as NBS1 [29, 30], the BLM helicase, and par-
tially other FA proteins (FANCE, FANCC, FANCJ) [1, 31]. FANCM (together
with FAAP24) and FANCD2 are able to bind directly to DNA with a structure
specific affinity [32, 33] similar to that observed for FANCJ/BRIP1 and
FANCD1/BRCA2, the only known FA proteins that are not required for FANCD2
monoubiquitination.
In addition to the proteins of the FA core complex, the ATR kinase is
required for DNA damage induced monoubiquitination and relocation of
FANCD2 [27]. ATR-mediated activation of FANCD2 takes place in response to
DNA-ICLs or stalled replication forks. It appears to depend on the intact FA
core complex but also on NBS1, linking FA to the non-homologous end-joining
(NHEJ) pathway of DNA repair [29, 34]. Depletion of the ATR-interacting pro-
tein, ATRIP, in Xenopus egg extracts results in defective chromatin binding of
xFANCD2 [35]. ATR functions either by direct phosphorylation of FANCD2,
as indicated by in vitro experiments, or indirectly by phosphorylation of com-
ponents participating or regulating the FA/BRCA pathway [27].
As shown in figure 1, following exposure to ionizing radiation, FANCD2 is
phosphorylated at serine 222 in an ATM-kinase dependent manner which acti-
vates an intra S phase checkpoint [36]. ATM-mediated phosphorylation depends
on NBS1, implying a regulatory connection between the NBS1/MRE11/
RAD50 complex and the FA proteins [30]. More detailed descriptions and dis-
cussions of the increasingly complex and manifold interactions of the Fanconi
anemia proteins can be found in series of recent reviews [20, 22, 23, 37].

Fanconi Anemia (FA) Genes

The FA genes are scattered widely throughout the human genome and vary
considerably in both size and structure, as summarized in table 1. Sequence

Kalb/Neveling/Herterich/Schindler 42
 Table 1. Summary of Fanconi anemia (FA) complementation groups, corresponding genes and proteins
Fanconi Anemia Genes: Structure, Mutations, and Genotype-Phenotype Correlations

Complementation Gene Chromosome Number Protein Protein domainsb Reference

group location of exons size, kDa

Name Frequency, %a

FA-A 57 FANCA 16q24.3 43 163 partial leucine zipper, [5]

NES, NLS
FA-B 0.3 FANCB/FAAP95 Xp22.3 10 95 NLS [16]
FA-C 15 FANCC 9q22.3 15 63 – [6]
FA-D1 4 FANCD1/BRCA2 13q12.3 27 384 BRC repeats, RAD51-/ [14]
DSS1-/ssDNA
binding, NLS
FA-D2 3 FANCD2 3p25.3 44 155 (162) – [11]
FA-E 1 FANCE 6p21–22 10 58 – [7]
FA-F 2 FANCF 11p15 1 42 homology to ROM [8]
FA-G 9 FANCG/XRCC9 9p13 14 68 tetratricopeptide repeat [9]
FA-I rare – – – – – –
FA-J 1.6 FANCJ/BRIP1/BACH1 17q22 20 140 DEAH-helicase, [12, 13]
BRCA1-binding
FA-L 0.1 FANCL/PHF9 2p16.1 14 43 WD40 repeats, PHD finger [17]
FA-M rare FANCM/Hef 14q21.3 23 250 homology to helicase and [18]
endonuclease domains
FA-N rare FANCN/PALB2 16p12 13 131 C-terminal WD40-like [15]
repeats

a
Based on the International Fanconi Anemia Registry (IFAR), reviewed in reference [1].
b
NES: nuclear export sequences; NLS: nuclear localization signals.
43
comparisons indicate that some of the FA genes must have arisen fairly recently
in evolution, lacking homologs beyond the vertebrate kingdom. However, there
are also FA genes/proteins that are more ancient and highly conserved. For
example, FANCM is closely related to the archael protein Hef (helicase-associated
endonuclease for fork-structured DNA) [19], and FANCJ belongs to the group
of recQ-like helicases that share highly conserved domain structures. In addi-
tion, homologs have been found for FANCD2 in D. melanogaster, C. elegans
and A. thaliana [11], and for FANCL in D. melanogaster and A. gambiae [18].
This might reflect a basic function of the conserved FA proteins, most likely in
the area of DNA repair [38], which has been complemented by the recruitment
of the other FA genes later in evolution. Since oxidative stress and oxidative
damage appear to activate the FA/BRCA pathway, an attractive hypothesis
posits that warm-blooded and long-lived species require the extended FA fam-
ily of genes as additional protection against DNA damage caused by reactive
oxygen species.
Figure 2 provides a graphic summary of the known FA genes, depicting
relative size and intron/exon structure. In addition, numbers and types of muta-
tions are indicated above and below the respective genes. As such, figure 2
illustrates the impressive degree of structural and mutational variation among
the known FA genes, both at the genomic and the cDNA levels. One of the
smallest FA-specific genomic regions, comprising less than 5 kb and only a sin-
gle exon, contains the entire FANCF gene, whereas FANCC and FANCJ are rel-
atively large genes spanning 218 kb and 180 kb, albeit with widely differing
intron/exon structures.
The occurrence of genetic alterations in a given gene depends on a variety
of factors related to gene structure, including the presence of repetitive ele-
ments and base pair composition such as CpG dinucleotides. Physical size by
itself clearly cannot be correlated with the number of mutations in a given FA
gene. A multitude of mutations have been described in relatively small or
medium sized genes like FANCG and FANCA, while genes with large tran-
scripts like FANCM and FANCD1 appear to be affected by relatively few alter-
ations. With respect to type, quantity and distribution of mutations, FANCA
clearly stands out. It is so far the only FA gene with a high proportion of large
deletions, amounting to at least 32% [39, 40] of all observed mutations in
FANCA. The prevalence of large deletions has been correlated with the pres-
ence of numerous Alu-type sequences, promoting Alu-mediated recombination
[41]. With the advent of the MLPA (multiplex ligation dependent probe ampli-
fication) technique, it is very likely that the proportion of deletions detected in
FANCA and perhaps in other FA genes will increase in the future. As illustrated
by the example shown in figure 3, MLPA uncovers deletions that escape detec-
tion with standard PCR-based techniques of mutation analysis.

Kalb/Neveling/Herterich/Schindler 44
 Fig. 2. Relative size, exon/intron structure and mutational spectra of the known FA
genes. In order to accommodate and compare the distribution of mutations, exons were
drawn 50-fold larger than corresponding introns. Mutational changes are shown above and
below the respective genes. Key to symbols: 䊊 missense mutation, 䊉 nonsense mutation, 䉭
small deletion/insertion or duplication, aberrant splicing/splice site mutation, 䉲 large
insertion, large deletion, X translation initiation (unpublished data and references for:
FANCA [45], FANCB [16, 49], FANCC [2, 6, 88], FANCD1 [51, 52], FANCD2 [11], FANCE [7],
FANCF [8], FANCG [86], FANCJ [12, 13], FANCL [17], FANCM [18] and IFAR (Rockefeller
University)).

Fanconi Anemia Genes: Structure, Mutations, and Genotype-Phenotype Correlations 45

 Ratio sample/control 1.0

0.5

0.0
1 4 7 10 13 16 19 22 25 28 31 34 37 40
Exon

Fig. 3. Multiplex ligation dependent probe amplification (MLPA) assay for the detec-
tion of deletions in FANCA. A reduction of the sample to control ratio to approximately 0.5
indicates a deleted region. The example shown is from a compound heterozygous FA-A
patient. One allele carries a large deletion involving exons 15 to 20; the other allele carries a
small deletion in exon 38 that involves only 3 bp (c.3788-3790delTCT) and obviously affects
the primer binding site.

FANCD2, a gene nearly of the same size and structure as FANCA, also con-
tains a high number of repetitive elements including many Alu repeats but so far
there are only two deletions involving a single exon. The mutational record of
the FA genes indeed suggests that evolutionarily conserved genes like
FANCD2, FANCL or FANCM may be more important than the more recent
FANCA or FANCG genes which seem to be highly tolerant of mutations. In
addition to serving as catalytic subunit of the FA core complex and as such
responsible for the monoubiquitination of FANCD2 there may be other, as yet
unknown function(s) of FANCL that might explain the relative paucity of
respective patients and mutations. Much more in depth analysis of FA gene
structure and function will be required to clarify whether the less frequently
affected genes truly have more important functions, or whether their underrep-
resentation among FA patients has other, possibly structural reasons. A case in
point is the highly conserved FANCD2 gene. All FA-D2 patients examined to
date have at least one hypomorphic or leaky mutation resulting in residual lev-
els of protein which implicates lethality of biallelic null mutations [42].
However, patients with mutations in some of the evolutionarily novel genes like
FANCE and FANCB also are exceedingly rare, suggesting that mutational inac-
tivation of these genes might not be as permissive as mutational inactivation of
FANCA, FANCC or FANCG. To add to the complexity, several functions other
than its participation in the FA core complex have been suggested for FANCA,
and yet biallelic null mutations in FANCA are compatible with viability and
defects in this gene are impressively numerous.

Kalb/Neveling/Herterich/Schindler 46
 Annotations of Individual FA Genes (in Alphabetical Order)
FANCA codes for a protein of 163 kDa which contains a leucine zipper-like
motif, two overlapping bipartite nuclear localization signals (NLS) [43] and
five functional leucine-rich nuclear export sequences (NES) that contribute to
CRM1 (chromosome region maintenance 1)-dependent nuclear export [44].
Since the first identification in 1996 more than 200 different mutations of all
possible types have been described (see fig. 2) (reviewed in reference [45]).
Large deletions are notably frequent in FANCA with a striking overlap of the
affected exons. Even though the exact breakpoints have not been determined in
all cases, a number of reports indicate a direct correlation between the presence
of Alu repeats and deletion breakpoints [39, 46, 47]. In addition to large dele-
tions, small deletions and insertions/duplications are also prevalent [48]. These
changes often are associated with short direct repeats, homonucleotide tracts,
and hotspot consensus sequences like CpG dinucleotides. Base substitutions
account for less than half of the mutations in FANCA and cause premature ter-
mination, amino acid exchanges, or affecting translation initiation or normal
exon recognition.
FANCB is the only X-linked FA gene. Approximately 60% of the genes
located at Xp22.3 exhibit biallelic expression in females, which however is not
the case for FANCB which undergoes X-chromosomal inactivation (NCBI,
OMIM, *300515). To date, mutations in FANCB have been found in five male
FA patients [16, 49] including a deletion affecting the promotor region and por-
tions of the 5⬘ untranslated regions (exon 1), three microdeletions/-insertions,
and a single splice site mutation. The analysis of three healthy female FANCB
carriers revealed the preferential inactivation of the X-chromosome carrying
the mutated allele in both peripheral blood cells and fibroblasts, suggesting a
proliferative advantage for cells expressing the wildtype allele early in embryo-
genesis [16].
FANCC was the first FA gene identified in 1992 [6]. It gives rise to two
alternative transcripts differing in usage of the first untranslated exon, denoted
as –1 or –1a [41]. At the genomic level, FANCC is one of the larger FA genes
spanning more than 218 kb, but coding for a protein with a molecular weight of
63 kDa. Two putative p53 binding sites of unknown functional significance
have been described in FANCC [50]. Mutations in FANCC are responsible for
10–15% of all FA cases, but surprisingly few private mutations have been
reported. As shown in figure 2, the mutational spectrum is heterogeneous, and
alterations are scattered widely throughout the gene with the exception of a pos-
sible cluster in the C-terminal region.
FANCD1 is identical to the gene coding for the breast cancer associated
protein 2 (BRCA2) [14] whose monoallelic mutations predispose to breast,
ovarian, pancreatic and possibly other cancers. Biallelic mutations in

Fanconi Anemia Genes: Structure, Mutations, and Genotype-Phenotype Correlations 47

FANCD1/BRCA2 result in a severe form of FA with a high and very early risk
for the development of malignancy [51–53]. At the mRNA level, FANCD1/
BRCA2 is the largest of the FA genes with an open reading frame of 10254 nt.
There are numerous functional domains which have been extensively reviewed
[54, 55]. Crucial for FANCD1/BRCA2 function are 8 BRC repeats, a helicase
domain followed by 3 oligonucleotide/-saccharide binding (OB1, OB2, OB3)
folds, and a tower domain inserted in OB2. The FANCD1/BRCA2 protein
shows structure-specific DNA binding activity and interacts directly with sev-
eral other proteins some of which have been implicated in DNA repair like
RAD51, FANCG and FANCD2 [56–58]. FANCD1/BRCA2 appears to function
in multiple and diverse cellular processes including stabilization of stalled
replication forks, homologous recombination (via interaction with the RAD51
recombinase), and regulation of cytokinesis [59]. Mutations in FANCD1/
BRCA2 are distributed over the entire gene, but monoallelic preservation of a
functional BRC-repeat domain has been observed for the majority of mutations
in FA-D1 patients. Compared to the prevalence of gene carriers in different
populations, the frequency of FA-D1 patients appears low, suggesting non-
viability of most biallelic types of mutations [52, 60].
FANCD2 is expressed in alternatively spliced variants, one with a large
continuous exon 43, and one with an additional exon 44, the latter giving rise to
a functional protein [11]. FANCD2 is a highly conserved gene that is flanked by
two distinct pseudogene regions FANCD2-P1 and FANCD2-P2 showing
homologies to the middle portion of the gene [42]. FANCD2 protein is recruited
to chromatin after DNA damage-induced monoubiquitination at lysine 561 [24,
61] where it colocalizes with other proteins involved in DNA repair including
BRCA2, RAD51 and BRCA1 [56, 57]. Purified FANCD2 has DNA-binding
activity with structure-specific affinity to branch points and free DNA ends
such as Holliday junctions and DNA double strand breaks [33], supporting its
active participation in DNA repair pathways. Based on the analysis of more
than 30 patients and disregarding recurrent mutations, there is an almost ran-
dom distribution of mutations (cf. fig. 2). In fact, more than 50% of the muta-
tions observed in FANCD2 affect splicing, a figure that is much higher than in
any of the other FA genes. Residual levels of FANCD2 protein were detected in
all patient-derived cell lines, which underline the functional importance of
FANCD2 [42].
FANCE was identified by expression cloning and mutation analysis of
three families assigned to complementation group FA-E [7]. Except for the
original families, no further mutations of FA-E patients have been reported.
FANCF consists of only a single exon. The encoded protein shares homolo-
gies to the prokaryotic RNA-binding protein (ROM) [8], which is without
known significance [62]. To date, only 5 nonsense mutations and microdeletions

Kalb/Neveling/Herterich/Schindler 48
have been found in FA-F patients [8]. These alterations appear to cluster in the
5⬘ portion of the gene. However, FANCF has gained special prominence as a
gene that is frequently silenced by hypermethylation of its promotor region in
various types of malignancies [63–65] (see chapter by Neveling et al.).
FANCG is identical to XRCC9 which was identified in Chinese hamster
cells as a gene involved in DNA repair [9, 66]. FANCG has seven tetratricopep-
tide repeat motifs (TRPs) [67] thought to function as a scaffold for mediating
protein-protein interaction. The molecular function of FANCG is not restricted
to the FA core complex, since it has been shown to interact with BRCA2 and
XRCC3 directly and independently of other FA proteins [68]. Mutations in
FANCG are of all types, with the exception of deletions. They are scattered
throughout the gene, with possible preference for the N- and C-terminal regions
(cf. fig. 2).
FANCJ is identical to the BRCA1-interacting protein 1 (BRIP1), also
denoted as BRCA1-associated C-terminal helicase 1 (BACH1) [12, 13].
FANCJ/BRIP1 was originally found as a BRCA1 interacting protein containing
two functional domains, a DNA-dependent ATPase and a 5⬘- to 3⬘-DNA heli-
case [69]. In vitro, BRIP1 has DNA unwinding activity with a high preference
for forked duplex substrates and apparently executes efficient strand displace-
ment from a D loop substrate representing a likely intermediate in DNA repair
via homologous recombination [70]. On the genomic level, FANCJ belongs to
the larger FA genes, but only 8 different mutations have been described to date
with most of the affected patients being homozygous or heterozygous for
c.2392T⬎C (R798X) [12, 13].
FANCL was identified by mass spectroscopy of a previously isolated 43-
kDa FA-associated polypeptide [17]. Sequence analysis revealed identity to the
WD40-repeat containing PHD finger protein-9 (PHF9) which is highly con-
served. FANCL is thought to serve as the catalytic subunit of the FA core com-
plex during monoubiquitination of FANCD2. Consistent with this important
role in the FA/BRCA pathway, there is only a single patient known to date
whose disease is caused by a homo- or hemizygous insertion of 177 bp into a
pyrimidine-rich sequence between intron 10 and exon 11, leading to skipping of
exon 11.
FANCM also belongs to the highly conserved FA genes [18]. It shows sim-
ilarities to the archael protein Hef which displays a helicase and an endonucle-
ase domain. However, the biological relevance of these domains for the current
function of FANCM has yet to be established. Using specific antibodies lack of
FANCM was discovered in a single cell line from a patient previously excluded
from most of the known complementation groups [18]. Genetic work up
revealed compound heterozygous mutations, a maternal nonsense mutation in
exon 13 and a paternal deletion within exon 15.

Fanconi Anemia Genes: Structure, Mutations, and Genotype-Phenotype Correlations 49

 FANCN was recently found to be identical to PALB2 (partner and localizer of
BRCA2) [15]. Originally, PALB2 was identified as a BRCA2-interacting protein
that promotes BRCA2 stability and localization in nuclear structures [19]. Using
a candidate gene approach, biallelic mutations were found in the PALB2 gene in a
cohort of FA patients that could not be assigned to any of the other known com-
plementation groups. Cells from FA-N patients show intact FANCD2 monoubiq-
uitination which localizes the function of FANCN to the group of proteins acting
downstream of FANCD2, including FANCD1/BRCA2 and FANCJ/BRIP1. Early
childhood cancer appears to be a common denominator of mutational inactiva-
tion of the downstream genes FANCD1/BRCA2 and FANCN/PALB2 [15].

Founder Mutations and Ethnically Associated Mutations

A number of FA causing mutations are recurrently prevalent in certain

populations due to founder or inbreeding effects. Examples of such ‘marker’
mutations are shown in table 2. A prominent example is the FANCA mutation
c.295C⬎T in the Spanish Gypsy population where it is encountered with a car-
rier frequency of 1/64 to 1/70 [71]. The occurrence of the mutation is restricted
to the Iberian peninsula where it presumably emerged less than 600 years ago in
a Gypsy family that migrated to Spain. Carrier frequencies higher than 1/100
have also been observed in the Ashkenazi Jewish population for the FANCC
mutation c.456⫹4A⬎T (previously IVS4⫹4A⬎T or c.711⫹4A⬎T) [2]. This
mutation must have originated in the Israelite population that left Palestine dur-
ing the Roman Empire and settled in Europe, since it is not encountered among
Sephardic Jews [72].
A classical founder effect was demonstrated for the large intragenic dele-
tion of exons 12–31 in the FANCA gene which has been shown to account for
60% of FA chromosomes in 46 unrelated Afrikaner FA patients [73]. This muta-
tion was apparently introduced to the Cape region of South Africa at the end of
the 17th century by a French Huguenot couple. In the black South African pop-
ulation, including individuals from Swaziland, Mozambique, and Malawi, 82%
of all FA patients were found to carry the deletion c.637_643delTACCGCC in
FANCG [74]. The distribution of this mutation in diverse geographic regions
and tribes was used to date the origin to the arrival of the Bantu speakers in
South Africa around 400 AD.
In regions of high historical migration like Europe or North America some
mutations were found recurrently among patients of the same ethnic origin (cf.
table 2). However, due to the obvious heterogeneity of the populations there is
little or no increased allele frequency, even if haplotype analysis suggests a
common ancestry. Notable exceptions are mutations found in geographically

Kalb/Neveling/Herterich/Schindler 50
 Table 2. Recurrent/founder mutations in Fanconi anemia genes

Gene Mutation Ethnicity Allele frequencya Other populations Reference

FANCA c.295C⬎T Spanish Gypsy 1/64–1/70 Portuguese [71]

Fanconi Anemia Genes: Structure, Mutations, and Genotype-Phenotype Correlations

c.890_893del Tunisian Jewish rare – [82]

c.894_1359del Irish n.d. – [76]
c.894_1626del South African n.d. – [73]
c.1007_3066del South African n.d. German (Western Ruhr), [73]
French Huguenot ancestry
c.1115_1118del variable n.d. – [48]
c.2574C⬎G Indian Jewish 1/53 – [82]
c.2172_2173dupG Moroccan Jewish 1/200 – [82]
c.3398delA South African n.d. – [73]
c.3788_3790del Brazilian n.d. – [48, 83]
c.4275delT Moroccan Jewish rare – [82]
FANCC c.67delG (c.322delG) Northern European n.d. Dutch [2]
c.456⫹4A⬎T (c.711⫹4A⬎T) Ashkenazi Jewish 1/89 Sephardic Jews (with [2, 84]
Ashkenazi ancestry)
c.456⫹4A⬎T (c.711⫹4A⬎T) Japanese n.d. – [81]
FANCD2 c.1948-16T⬎G Turkish n.d. Turkish North Americans, Czech [42]
c.1948-6C⬎A German n.d. North American [42]
c.2444G⬎A Spanish n.d. Italian, North American [42]
FANCG c.307⫹1G⬎C Japanese n.d. Korean [85, 86]
c.313G⬎T German n.d. – [87]
c.637_643del black South African 1/35–1/475 Bantu-speakers: Swaziland, [74]
Mozambique, Malawi
c.1066C⬎T Japanese/Korean ⬍1/156 Korean [85]
c.1077-2A⬎G Portuguese-Brazilian n.d. Portuguese [86]
c.1183_1192del German n.d. – [87]
c.1480⫹1G⬎C French-Acadian n.d. – [86]
c.1649delC Turkish n.d. – [87]
51

c.1794_1803del Northern European n.d. – [86]

FANCJ c.2392C⬎T Inuit n.d. different ethnic groups [12]

a
n.d.: not determined.
isolated regions. In Italy, for example, two mutations in FANCA, c.790C⬎T and
c.3559insG were detected in otherwise unrelated families within the Campania
region [75]. Founder mutations or mutations sharing the same haplotype have
been reported for several FA complementation groups (FA-A, C, D2, G and J)
among different ethnic backgrounds, and independent of the overall mutation
frequencies of the FA genes.
Not all recurrent mutations can be sufficiently explained by founder effects,
but may represent examples of gene regions vulnerable to genetic change
(‘mutational hotspots’). A case in point is the mutation c.2392C⬎T in FANCJ
which was detected in a few unrelated Inuit patients that share the same haplo-
type. However, the same mutation was also found in patients from diverse popu-
lations and ethnical backgrounds where it was correlated with at least three
different haplotypes [12]. The combination of different haplotypes and the rarity
of FA-J patients imply an independent recurrence of the mutation c.2392C⬎T.

Genotype-Phenotype Correlations

Clinical course and severity of FA vary strongly between and even within
families, suggesting relatively weak, if any, genotype-phenotype correlations.
In particular, the onset of hematological abnormalities and longterm survival
appear to depend on additional factors such as number and severity of infec-
tions, nutritional status, and access to and quality of medical care. Any definite
conclusion about genotype-phenotype correlations must await the results of
prospective studies which are not available to date. A retrospective study com-
paring the haematological profiles of complementation groups FA-A, -C and
-G failed to reveal statistically significant differences in the onset of aplastic
anemia [76]. However, the manifestation of cytopenia was more severe in
patients belonging to complementation group FA-G. In addition, FA-G patients
may develop AML or MDS earlier than patients of the other complementation
groups. Congenital abnormalities were comparable among FA-A and FA-G
patients, but less severe than in certain FA-C patients. The clinical data of
patients belonging to complementation group D2 suggests certain differences
to other FA groups [42]. Compared to other patients listed in the IFAR database,
FA-D2 patients exhibit a higher average frequency of congenital abnormalities.
Their clinical course is additionally marked by early average onset of bone mar-
row failure and early dependency on hematopoietic stem cell transplantation. It
should be noted, however, that this FA-D2 patient cohort contained a number of
patients with identical mutations which complicates the statistical comparison.
An undoubtedly severe phenotype is caused by biallelic mutations in
FANCD1/BRCA2. Many of these patients succumb to their malignancies

Kalb/Neveling/Herterich/Schindler 52
(Wilms tumor, medulloblastoma and leukemias) prior to the onset of pancy-
topenia and thus may never be diagnosed as FA [51, 52]. Because of the unusu-
ally severe and early onset of neoplasia in FA-D1 patients, their phenotype
appears to be distinctive and may be more appropriately labeled FA-like rather
than genuine FA [51, 77].
Genotype-phenotype correlations are further complicated by the obvious
heterogeneity of the mutational spectrum within each FA gene, the private char-
acter of mutations and, of course, the high prevalence of compound heterozygos-
ity increasing the possibility of somatic reversion. Any missense change
involving a non-conserved position within the gene should be formally tested for
its functional significance. The availability of site-directed mutagenesis, retrovi-
ral vectors with high transduction efficiencies, and loss of MMC sensitivity as a
functional marker of complementation facilitate the performance of such ‘con-
firmatory’ tests. Adachi and coworkers [78] have published an exemplary study
in which the functional significance of a large number of FANCA mutations has
been tested. In terms of cost and labor such confirmatory studies tend to exceed
the capacity of the average diagnostic laboratory and might best be performed by
specialized centers. Notwithstanding these difficulties, empirical studies have
already shown that some mutations can be correlated with distinct phenotypes.
The mutation c.1007_3066del in FANCA, for example, is associated with an
increased rate of AML or MDS and a higher frequency of severe congenital
abnormalities [76]. In contrast, the FANCC c.67delG alteration is associated
with milder hematologic symptoms and a lower rate of somatic abnormalities
when compared to the majority of other FA patients. Another example of a mild
mutation is the missense mutation c.2444G⬎A in FANCD2 which seems to be
associated with a relatively mild clinical course [42].
A mild clinical course may either be due to mutations that retain some
degree of protein function or to reverse mosaicism. Mild mutations and/or
reverse mosaicism will be preferentially found among older FA patients or
patients who have escaped the diagnosis of FA because of lack of clinical symp-
toms. A prominent example of such a cryptic clinical course has recently been
described by Huck et al. [79] whose patient lacked somatic abnormalities and
was diagnosed as FA, at the remarkable age of 49 years, only because of an
unusually severe response to chemotherapy for bilateral breast cancer. Concerning
somatic reversion, there are numerous examples of reversions in the FANCA,
FANCC and FANCD2 genes. Somatic reversions preferentially involve com-
pound heterozygous patients and, depending on the cell lineages affected, may
lead to improved blood counts and a mild or protracted clinical course (see
chapter by Hoehn et al.).
An additional level of complexity is introduced by the genetic background
of the respective patients or patient group. A case in point is the mutation

Fanconi Anemia Genes: Structure, Mutations, and Genotype-Phenotype Correlations 53

c.456⫹4A⬎T in FANCC which shows a high prevalence in patients of both
Ashkenazi Jewish and Japanese origin [80, 81]. The Ashkenazi Jewish patients
are severely affected whereas the Japanese patients have a much milder pheno-
type both in terms of hematologic onset and pattern of congenital abnormalities.
In summary, there is wide variation of phenotypes and clinical course
among the most prevalent FA subgroups FA-A, -C and -G. The cumulative
record of genotype-phenotype correlations clearly indicates that the nature of
the underlying mutation is more important than the underlying complementa-
tion group. An exception to this rule are biallelic mutations in FANCD1/BRCA2
and FANCN/PALB2 which cause unusually severe phenotypes with early child-
hood leukemia and solid tumors, setting FA-D1 and FA-N patients apart from
the majority of FA patients.

References

1 Taniguchi T, D’Andrea AD: The molecular pathogenesis of Fanconi anemia: recent progress.
Blood 2006;107:4223–4233.
2 Verlander PC, Kaporis A, Liu Q, Zhang Q, Seligsohn U, Auerbach AD: Carrier frequency of the
IVS4 ⫹ 4 A→T mutation of the Fanconi anemia gene FAC in the Ashkenazi Jewish population.
Blood 1995;86:4034–4038.
3 Joenje H, Oostra AB, Wijker M, di Summa FM, van Berkel CG, et al: Evidence for at least eight
Fanconi anemia genes. Am J Hum Genet 1997;61:940–944.
4 Levitus M, Rooimans MA, Steltenpool J, Cool NF, Oostra AB, et al: Heterogeneity in Fanconi ane-
mia: evidence for 2 new genetic subtypes. Blood 2004;103:2498–2503.
5 Lo Ten Foe JR, Rooimans MA, Bosnoyan-Collins L, Alon N, Wijker M, et al: Expression cloning
of a cDNA for the major Fanconi anaemia gene, FAA. Nat Genet 1996;14:320–323.
6 Strathdee CA, Gavish H, Shannon WR, Buchwald M: Cloning of cDNAs for Fanconi’s anaemia by
functional complementation. Nature 1992;356:763–767.
7 de Winter JP, Leveille F, van Berkel CG, Rooimans MA, van Der Weel L, et al: Isolation of a
cDNA representing the Fanconi anemia complementation group E gene. Am J Hum Genet
2000;67:1306–1308.
8 de Winter JP, Rooimans MA, van Der Weel L, van Berkel CG, Alon N, et al: The Fanconi anaemia
gene FANCF encodes a novel protein with homology to ROM. Nat Genet 2000;24:15–16.
9 de Winter JP, Waisfisz Q, Rooimans MA, van Berkel CG, Bosnoyan-Collins L, et al: The Fanconi
anaemia group G gene FANCG is identical with XRCC9. Nat Genet 1998;20:281–283.
10 The Fanconi anemia/breast cancer consortium: Positional cloning of the Fanconi anaemia group A
gene. Nat Genet 1996;14:324–328.
11 Timmers C, Taniguchi T, Hejna J, Reifsteck C, Lucas L, et al: Positional cloning of a novel Fanconi
anemia gene, FANCD2. Mol Cell 2001;7:241–248.
12 Levran O, Attwooll C, Henry RT, Milton KL, Neveling K, et al: The BRCA1-interacting helicase
BRIP1 is deficient in Fanconi anemia. Nat Genet 2005;37:931–933.
13 Levitus M, Waisfisz Q, Godthelp BC, de Vries Y, Hussain S, et al: The DNA helicase BRIP1 is
defective in Fanconi anemia complementation group J. Nat Genet 2005;37:934–935.
14 Howlett NG, Taniguchi T, Olson S, Cox B, Waisfisz Q, et al: Biallelic inactivation of BRCA2 in
Fanconi anemia. Science 2002;297:606–609.
15 Reid S, Schindler D, Hanenberg H, Barker K, Hanks S, et al: Biallelic mutations in PALB2
cause Fanconi anemia subtype FA-N and predispose to childhood cancer. Nat Genet 2007;39:
162–164.

Kalb/Neveling/Herterich/Schindler 54
16 Meetei AR, Levitus M, Xue Y, Medhurst AL, Zwaan M, et al: X-linked inheritance of Fanconi
anemia complementation group B. Nat Genet 2004;36:1219–1224.
17 Meetei AR, de Winter JP, Medhurst AL, Wallisch M, Waisfisz Q, et al: A novel ubiquitin ligase is
deficient in Fanconi anemia. Nat Genet 2003;35:165–170.
18 Meetei AR, Medhurst AL, Ling C, Xue Y, Singh TR, et al: A human ortholog of archaeal DNA
repair protein Hef is defective in Fanconi anemia complementation group M. Nat Genet
2005;37:958–963.
19 Xia B, Sheng Q, Nakanashi K, Ohashi A, Wu J, et al: Control of BRCA2 cellular and clinical func-
tions by a nuclear partner, PALB2. Mol Cell 2006;22:719–729.
20 Bagby GC, Alter BP: Fanconi anemia. Semin Hematol 2006;43:147–156.
21 Surralles J, Jackson SP, Jasin M, Kastan MB, West SC, Joenje H: Molecular cross-talk among
chromosome fragility syndromes. Genes Dev 2004;18:1359–1370.
22 Thompson LH, Hinz JM, Yamada NA, Jones NJ: How Fanconi anemia proteins promote the four
Rs: Replication, recombination, repair, and recovery. Environ Mol Mutagen 2005;45:128–142.
23 Kennedy RD, D’Andrea AD: The Fanconi Anemia/BRCA pathway: new faces in the crowd. Genes
Dev 2005;19:2925–2940.
24 Garcia-Higuera I, Taniguchi T, Ganesan S, Meyn MS, Timmers C, et al: Interaction of the Fanconi
anemia proteins and BRCA1 in a common pathway. Mol Cell 2001;7:249–262.
25 Meetei AR, Sechi S, Wallisch M, Yang D, Young MK, et al: A multiprotein nuclear complex con-
nects Fanconi anemia and Bloom syndrome. Mol Cell Biol 2003;23:3417–3426.
26 Wang X, Andreassen PR, D’Andrea AD: Functional interaction of monoubiquitinated FANCD2
and BRCA2/FANCD1 in chromatin. Mol Cell Biol 2004;24:5850–5862.
27 Andreassen PR, D’Andrea AD, Taniguchi T: ATR couples FANCD2 monoubiquitination to the
DNA-damage response. Genes Dev 2004;18:1958–1963.
28 Taniguchi T, Garcia-Higuera I, Andreassen PR, Gregory RC, Grompe M, D’Andrea AD: S-phase-
specific interaction of the Fanconi anemia protein, FANCD2, with BRCA1 and RAD51. Blood
2002;100:2414–2420.
29 Pichierri P, Rosselli F: The DNA crosslink-induced S-phase checkpoint depends on ATR-CHK1
and ATR-NBS1-FANCD2 pathways. EMBO J 2004;23:1178–1187.
30 Nakanishi K, Taniguchi T, Ranganathan V, New HV, Moreau LA, et al: Interaction of FANCD2 and
NBS1 in the DNA damage response. Nat Cell Biol 2002;4:913–920.
31 Pace P, Johnson M, Tan WM, Mosedale G, Sng C, et al: FANCE: the link between Fanconi
anaemia complex assembly and activity. EMBO J 2002;21:3414–3423.
32 Ciccia A, Ling C, Coulthard R, Yan Z, Xue Y, et al: Identification of FAAP24, a Fanconi anemia
core complex protein that interacts with FANCM. Mol Cell 2007;25:331–343.
33 Park WH, Margossian S, Horwitz AA, Simons AM, D’Andrea AD, Parvin JD: Direct DNA bind-
ing activity of the Fanconi anemia D2 protein. J Biol Chem 2005;280:23593–23598.
34 Stiff T, Reis C, Alderton GK, Woodbine L, O’Driscoll M, Jeggo PA: Nbs1 is required for ATR-
dependent phosphorylation events. EMBO J 2005;24:199–208.
35 Sobeck A, Stone S, Costanzo V, de Graaf B, Reuter T, et al: Fanconi anemia proteins are required
to prevent accumulation of replication-associated DNA double-strand breaks. Mol Cell Biol
2006;26:425–437.
36 Taniguchi T, Garcia-Higuera I, Xu B, Andreassen PR, Gregory RC, et al: Convergence of the fan-
coni anemia and ataxia telangiectasia signaling pathways. Cell 2002;109:459–472.
37 Cantor SB, Andreassen PR: Assessing the link between BACH1 and BRCA1 in the FA pathway.
Cell Cycle 2006;5:164–167.
38 Marek LR, Bale AE: Drosophila homologs of FANCD2 and FANCL function in DNA repair.
DNA Repair (Amst) 2006;5:1317–1326.
39 Morgan NV, Tipping AJ, Joenje H, Mathew CG: High frequency of large intragenic deletions in
the Fanconi anemia group A gene. Am J Hum Genet 1999;65:1330–1341.
40 Callen E, Tischkowitz MD, Creus A, Marcos R, Bueren JA, et al: Quantitative PCR analysis
reveals a high incidence of large intragenic deletions in the FANCA gene in Spanish Fanconi ane-
mia patients. Cytogenet Genome Res 2004;104:341–345.
41 Savoia A, Centra M, Ianzano L, de Cillis GP, Zelante L, Buchwald M: Characterization of the 5⬘
region of the Fanconi anaemia group C (FACC) gene. Hum Mol Genet 1995;4:1321–1326.

Fanconi Anemia Genes: Structure, Mutations, and Genotype-Phenotype Correlations 55

42 Kalb R, Neveling K, Hoehn H, Schneider H, Linka Y, et al: Hypomorphic mutations in the gene
encoding a key Fanconi anemia protein, FANCD2, sustain a significant group of FA-D2 patients
with severe phenotype. Am J Hum Genet 2007, in press.
43 Lightfoot J, Alon N, Bosnoyan-Collins L, Buchwald M: Characterization of regions functional
in the nuclear localization of the Fanconi anemia group A protein. Hum Mol Genet 1999;8:
1007–1015.
44 Ferrer M, Rodriguez JA, Spierings EA, de Winter JP, Giaccone G, Kruyt FA: Identification of
multiple nuclear export sequences in Fanconi anemia group A protein that contribute to CRM1-
dependent nuclear export. Hum Mol Genet 2005;14:1271–1281.
45 Levran O, Diotti R, Pujara K, Batish SD, Hanenberg H, Auerbach AD: Spectrum of sequence vari-
ations in the FANCA gene: an International Fanconi Anemia Registry (IFAR) study. Hum Mutat
2005;25:142–149.
46 Levran O, Doggett NA, Auerbach AD: Identification of Alu-mediated deletions in the Fanconi
anemia gene FAA. Hum Mutat 1998;12:145–152.
47 Centra M, Memeo E, d’Apolito M, Savino M, Ianzano L, et al: Fine exon-intron structure of the
Fanconi anemia group A (FAA) gene and characterization of two genomic deletions. Genomics
1998;51:463–467.
48 Levran O, Erlich T, Magdalena N, Gregory JJ, Batish SD, et al: Sequence variation in the Fanconi
anemia gene FAA. Proc Natl Acad Sci USA 1997;94:13051–13056.
49 Holden ST, Cox JJ, Kesterton I, Thomas NS, Carr C, Woods CG: Fanconi anaemia complementa-
tion group B presenting as X-linked VACTERL with hydrocephalus syndrome. J Med Genet
2006;43:750–754.
50 Liebetrau W, Buhner M, Hoehn H: Prototype sequence clues within the Fanconi anaemia group C
gene. J Med Genet 1995;32:669–670.
51 Wagner JE, Tolar J, Levran O, Scholl T, Deffenbaugh A, et al: Germline mutations in BRCA2:
shared genetic susceptibility to breast cancer, early onset leukemia, and Fanconi anemia. Blood
2004;103:3226–3229.
52 Reid S, Renwick A, Seal S, Baskcomb L, Barfoot R, et al: Biallelic BRCA2 mutations are associ-
ated with multiple malignancies in childhood including familial Wilms tumour. J Med Genet
2005;42:147–151.
53 Alter BP: The association between FANCD1/BRCA2 mutations and leukaemia. Br J Haematol
2006;133:446–448; author reply 448.
54 Shivji MK, Venkitaraman AR: DNA recombination, chromosomal stability and carcinogenesis:
insights into the role of BRCA2. DNA Repair (Amst) 2004;3:835–843.
55 West SC: Molecular views of recombination proteins and their control. Nat Rev Mol Cell Biol
2003;4:435–445.
56 Hussain S, Wilson JB, Medhurst AL, Hejna J, Witt E, et al: Direct interaction of FANCD2 with
BRCA2 in DNA damage response pathways. Hum Mol Genet 2004;13:1241–1248.
57 Hussain S, Witt E, Huber PA, Medhurst AL, Ashworth A, Mathew CG: Direct interaction of
the Fanconi anaemia protein FANCG with BRCA2/FANCD1. Hum Mol Genet 2003;12:
2503–2510.
58 Tarsounas M, Davies D, West SC: BRCA2-dependent and independent formation of RAD51
nuclear foci. Oncogene 2003;22:1115–1123.
59 Daniels MJ, Wang Y, Lee M, Venkitaraman AR: Abnormal cytokinesis in cells deficient in the
breast cancer susceptibility protein BRCA2. Science 2004;306:876–879.
60 Popp H, Kalb R, Fischer A, Lobitz S, Kokemohr I, et al: Screening Fanconi anemia lymphoid cell
lines of non-A, C, D2, E, F, G subtypes for defects in BRCA2/FANCD1. Cytogenet Genome Res
2003;103:54–57.
61 Montes de Oca R, Andreassen PR, Margossian SP, Gregory RC, Taniguchi T, et al: Regulated
interaction of the Fanconi anemia protein, FANCD2, with chromatin. Blood 2005;105:1003–1009.
62 Leveille F, Blom E, Medhurst AL, Bier P, Laghmani el H, et al: The Fanconi anemia gene product
FANCF is a flexible adaptor protein. J Biol Chem 2004;279:39421–39430.
63 Tischkowitz M, Ameziane N, Waisfisz Q, De Winter JP, Harris R, et al: Bi-allelic silencing of the
Fanconi anaemia gene FANCF in acute myeloid leukaemia. Br J Haematol 2003;123:469–471.

Kalb/Neveling/Herterich/Schindler 56
64 Narayan G, Arias-Pulido H, Nandula SV, Basso K, Sugirtharaj DD, et al: Promoter hypermethyla-
tion of FANCF: disruption of Fanconi Anemia-BRCA pathway in cervical cancer. Cancer Res
2004;64:2994–2997.
65 Taniguchi T, Tischkowitz M, Ameziane N, Hodgson SV, Mathew CG, et al: Disruption of the
Fanconi anemia-BRCA pathway in cisplatin-sensitive ovarian tumors. Nat Med 2003;9:568–574.
66 Liu N, Lamerdin JE, Tucker JD, Zhou ZQ, Walter CA, et al: The human XRCC9 gene corrects
chromosomal instability and mutagen sensitivities in CHO UV40 cells. Proc Natl Acad Sci USA
1997;94:9232–9237.
67 Blom E, van de Vrugt HJ, de Vries Y, de Winter JP, Arwert F, Joenje H: Multiple TPR motifs char-
acterize the Fanconi anemia FANCG protein. DNA Repair (Amst) 2004;3:77–84.
68 Hussain S, Wilson JB, Blom E, Thompson LH, Sung P, et al: Tetratricopeptide-motif-mediated
interaction of FANCG with recombination proteins XRCC3 and BRCA2. DNA Repair (Amst)
2006;5:629–640.
69 Cantor SB, Bell DW, Ganesan S, Kass EM, Drapkin R, et al: BACH1, a novel helicase-like pro-
tein, interacts directly with BRCA1 and contributes to its DNA repair function. Cell
2001;105:149–160.
70 Gupta R, Sharma S, Sommers JA, Jin Z, Cantor SB, Brosh RM Jr: Analysis of the DNA substrate
specificity of the human BACH1 helicase associated with breast cancer. J Biol Chem
2005;280:25450–25460.
71 Callen E, Casado JA, Tischkowitz MD, Bueren JA, Creus A, et al: A common founder mutation in
FANCA underlies the world’s highest prevalence of Fanconi anemia in Gypsy families from Spain.
Blood 2005;105:1946–1949.
72 Peleg L, Pesso R, Goldman B, Dotan K, Omer M, et al: Bloom syndrome and Fanconi’s anemia:
rate and ethnic origin of mutation carriers in Israel. Isr Med Assoc J 2002;4:95–97.
73 Tipping AJ, Pearson T, Morgan NV, Gibson RA, Kuyt LP, et al: Molecular and genealogical evi-
dence for a founder effect in Fanconi anemia families of the Afrikaner population of South Africa.
Proc Natl Acad Sci USA 2001;98:5734–5739.
74 Morgan NV, Essop F, Demuth I, de Ravel T, Jansen S, et al: A common Fanconi anemia mutation
in black populations of sub-Saharan Africa. Blood 2005;105:3542–3544.
75 Savino M, Borriello A, D’Apolito M, Criscuolo M, Del Vecchio M, et al: Spectrum of FANCA
mutations in Italian Fanconi anemia patients: identification of six novel alleles and phenotypic
characterization of the S858R variant. Hum Mutat 2003;22:338–339.
76 Faivre L, Guardiola P, Lewis C, Dokal I, Ebell W, et al: Association of complementation group and
mutation type with clinical outcome in Fanconi anemia. European Fanconi Anemia Research
Group. Blood 2000;96:4064–4070.
77 Hirsch B, Shimamura A, Moreau L, Baldinger S, Hag-alshiekh M, et al: Association of biallelic
BRCA2/FANCD1 mutations with spontaneous chromosomal instability and solid tumors of child-
hood. Blood 2004;103:2554–2559.
78 Adachi D, Oda T, Yagasaki H, Nakasato K, Taniguchi T, et al: Heterogeneous activation of the
Fanconi anemia pathway by patient-derived FANCA mutants. Hum Mol Genet 2002;11:
3125–3134.
79 Huck K, Hanenberg H, Gudowius S, Fenk R, Kalb R, et al: Delayed diagnosis and complications
of Fanconi anaemia at advanced age – a paradigm. Br J Haematol 2006;133:188–197.
80 Kutler DI, Auerbach AD: Fanconi anemia in Ashkenazi Jews. Fam Cancer 2004;3:241–248.
81 Futaki M, Yamashita T, Yagasaki H, Toda T, Yabe M, et al: The IVS4 ⫹ 4 A to T mutation of the
fanconi anemia gene FANCC is not associated with a severe phenotype in Japanese patients. Blood
2000;95:1493–1498.
82 Tamary H, Bar-Yam R, Shalmon L, Rachavi G, Krostichevsky M, et al: Fanconi anaemia group A
(FANCA) mutations in Israeli non-Ashkenazi Jewish patients. Br J Haematol 2000;111:338–343.
83 Magdalena N, Pilonetto DV, Bitencourt MA, Pereira NF, Ribeiro RC, et al: Frequency of Fanconi
anemia in Brazil and efficacy of screening for the FANCA 3788-3790del mutation. Braz J Med
Biol Res 2005;38:669–673.
84 Whitney MA, Saito H, Jakobs PM, Gibson RA, Moses RE, Grompe M: A common mutation in the
FACC gene causes Fanconi anaemia in Ashkenazi Jews. Nat Genet 1993;4:202–205.

Fanconi Anemia Genes: Structure, Mutations, and Genotype-Phenotype Correlations 57

85 Yagasaki H, Oda T, Adachi D, Nakajima T, Nakahata T, et al: Two common founder mutations of
the fanconi anemia group G gene FANCG/XRCC9 in the Japanese population. Hum Mutat
2003;21:555.
86 Auerbach AD, Greenbaum J, Pujara K, Batish SD, Bitencourt MA, et al: Spectrum of sequence
variation in the FANCG gene: an International Fanconi Anemia Registry (IFAR) study. Hum Mutat
2003;21:158–168.
87 Demuth I, Wlodarski M, Tipping AJ, Morgan NV, de Winter JP, et al: Spectrum of mutations in the
Fanconi anaemia group G gene, FANCG/XRCC9. Eur J Hum Genet 2000;8:861–868.
88 Gibson RA, Morgan NV, Goldstein LH, Pearson IC, Kesterton IP, et al: Novel mutations and poly-
morphisms in the Fanconi anemia group C gene. Hum Mutat 1996;8:140–148.

Reinhard Kalb
Department of Human Genetics, University of Würzburg
Biozentrum, Am Hubland
D–97074 Würzburg (Germany)
Tel. ⫹49 931 8884089, Fax ⫹49 931 8884069, E-Mail [email protected]

Kalb/Neveling/Herterich/Schindler 58
 Schindler D, Hoehn H (eds): Fanconi Anemia. A Paradigmatic Disease for the Understanding of
Cancer and Aging. Monogr Hum Genet. Basel, Karger, 2007, vol 15, pp 59–78

Cancer in Fanconi Anemia and Fanconi

Anemia Genes in Cancer
K. Neveling, R. Kalb, D. Schindler
Department of Human Genetics, University of Würzburg, Biocenter,
Würzburg, Germany

Abstract
At the cellular level, defects in Fanconi anemia (FA) genes manifest themselves as
hypersensitivity to DNA damaging agents which can be assessed by increased chromosome
breakage and cell cycle changes. As long-term manifestations of cellular genetic instability,
FA patients are at high risk for the development of a rather narrow spectrum of malignancies.
Neoplasia occurs at a much younger age than in non-FA patients, and chiefly includes acute
myeloid leukemia (AML) and squamous cell carcinomas (SCC). Given the role of FA genes in
the maintenance of genetic stability, one would expect that FA genes are frequently altered in
various kinds of tumors arising in non-FA patients. Much to our surprise and with the possible
exception of pancreatic carcinoma, there appears to be no convincing evidence for a frequent
association between either germline or somatic alterations in FA genes and malignancies aris-
ing in non-FA patients. Another notable exception is the FANCF gene which features pro-
minently among the many genes that are silenced in cancer via promoter CpG island
hypermethylation. However, the map location of FANCF adjacent to the 11p hotspot region of
hypermethylation raises concerns about a strictly causal relationship between FANCF inacti-
vation and tumorigenesis. Altogether, the available evidence suggests that tumor cells just like
normal cells benefit from intact FA genes, possibly by keeping their replication machinery
intact and by optimizing their defense against the adverse effects of reactive oxygen species.
Copyright © 2007 S. Karger AG, Basel

The risk of neoplasia is increased in most genetic instability disorders caused

by mutations in human caretaker genes. It is thought that this increased risk
results from defects in the cellular response to DNA damage, including DNA
damage recognition and DNA repair. Cell genetic evidence suggests that defec-
tive caretaker genes render the genome unstable and thereby promote the error
rate in somatic cells. This would of course have the most devastating conse-
quences in a somatic stem cell [1]. Since only a limited number of genetic
 Chemicals, ionizing radiation, reactive O2-species

DNA-damage

Intact Caretaker genes Defective

Damage ATM, ATR Genetic

recognition BLM, NBS1, MRE11 instabiliy
FANC genes,
etc.
Checkpoints Mutation

DNA-repair Cell death Cancer

Fig. 1. Hypothetical role of human caretaker genes in the DNA damage response. For
explanation see text.

changes appear to be required for the emergence of a clinically expressed malig-

nancy, genetic instability due to defective caretaker genes is likely to speed up the
process of malignant transformation [2]. Certain cellular proteins may represent
crucial targets for the tumor initiating and tumor promoting effects of genetic
instability. To these belong DNA polymerases, proteins which repair DNA, pro-
teins which affect chromatin structure, kinetochore proteins, spindle proteins, and
proteins which regulate apoptosis and cell cycle events in response to DNA dam-
age [2]. Since FA proteins seem to be involved in various aspects of DNA main-
tenance including replication, recombination, repair and recovery [3], they
themselves feature prominently among the ‘at risk’ proteins of genetic instability.
A current concept depicting the role of caretaker genes in the DNA damage
response is shown in figure 1. DNA damage can be induced in different ways, for
example by chemicals or ionizing radiation or, endogenously, by reactive oxygen
species. As long as caretaker genes are intact, DNA damage is recognized and,
dependent on the kind of damage, checkpoints are activated, and either DNA
repair or elimination via apoptosis will occur. In the case of defective caretaker
genes, a regulated and orderly cell response is not guaranteed. DNA damage
recognition and repair might be impaired, not occur at all, or be error-prone,
resulting in genetic instability. As a result, mutations accumulate, ultimately lead-
ing either to cell death or to the emergence of malignant cell growth.
From a clinical point of view, there are two striking observations in support
of a causal connection between the genetic instability phenotype of FA cells and

Neveling/Kalb/Schindler 60
 Table 1. Types of malignancies in human caretaker syndromes

Sensitivity Syndrome Gene Malignancy

Chemicals like 4-NQO, Werner syndrome WRN helicase soft tissue sarcomas
BrdU Bloom syndrome BLM helicase all types of leukemias
and solid tumors
Rothmund-Thompson recQ-IV helicase soft tissue sarcomas
syndrome
Ionizing radiation Ataxia telangiectasia ATM kinase lymphoma
protection Nijmegen breakage NBS1 lymphoma
syndrome
AT-like syndromes MRE11, RAD50, LIG4 lymphoreticular neoplasia
Reactive oxygen species, Fanconi anemia FANCA, B, C, D1, D2, AML, squamous cell
crosslinking agents E, F, G, I, J, L, M, N carcinoma
(?) Reactive oxygen species breast cancer BRCA1, BRCA2 breast cancer
ovarian cancer ovarian cancer

the emergence of neoplasia. The first observation concerns the age of onset of
the neoplastic change which in contrast to the general population occurs in the
teens [4] or, in the case of biallelic mutations in FANCD1/BRCA2, in infants
less than 5 years of age [5, 6]. The second observation concerns the fact that the
types of malignancies prevalent in FA neither mirror the spectrum of malignan-
cies that is seen in the general population, nor do they mimic the kind of malig-
nancies that are seen in the other caretaker syndromes (cf. table 1). AML and
squamous cell carcinomas of the oropharynx and the genital area dominate in
FA. These types of malignancies are rarely if at all encountered in other care-
taker gene syndromes (cf. table 1). This suggests that susceptibility to certain
types of neoplasms primarily reflects the type of caretaker gene that is defective
in a given patient. Both the cellular and the clinical phenotype of FA indicate a
close connection between the FA proteins, DNA repair and tumorigenesis. This
article summarizes some basic information about the obvious FA–cancer con-
nection and thus focuses on the role of cancer in FA as well as on what is cur-
rently known on disruptions of FA genes in cancer.

Neoplasia in Fanconi Anemia

Myelodysplastic Syndrome (MDS)

MDS is a frequent hematological disorder in FA. In MDS, bone marrow
precursor cells fail to differentiate properly such that not enough mature

Cancer in Fanconi Anemia and Fanconi Anemia Genes in Cancer 61

erythrocytes, leukocytes or thrombocytes arrive in the periphery. MDS occurs
in around 7% of all patients with FA [4] amounting to around 50% of all
hemato-proliferative changes seen in FA [7]. The presence of MDS affects the
prognosis in FA: the estimated 5-year survival with MDS amounts to 0.09,
versus 0.92 without MDS [8]. If at all, FA-related MDS progresses to leukemia
more slowly compared to MDS in non-FA patients. FA patients may suffer
from MDS for many years without developing overt leukemia. Among
69 cases with MDS described by Alter et al. [8], only 10 developed leukemia.
MDS in FA obviously follows a different natural course than MDS in non-FA
patients [8].

Leukemia
The most common malignancy arising in FA is leukemia. The overall
cumulative incidence of leukemia is 37%, the median age of onset is 14 years
[4]. The types of leukemia prevalent in FA differ substantially from what is seen
in the general population: Acute myeloid leukemia (AML) dominates in FA
(94% of all leukemias), whereas acute lymphocytic leukemia (ALL) amounts to
only 6%. The reverse holds for non-FA patients (84% of leukemias in the gen-
eral population are ALL) [4]. Hematological malignancies other than AML or
ALL are relatively rare in FA. Chronic myelo-monocytic leukemia (CMMOL)
or Burkitt lymphoma have been described as single cases [7].
It is important to point out that AML may be the first and only manifes-
tation of FA in otherwise normal patients [9]. All defined FAB subclasses [10]
are represented in FA-associated AML [4], albeit with different proportions
compared to the general population. There are more FA patients in the M4 and
M6 and less in the M1 and M2 subclasses of AML compared to non-FA patients
[4]. Several recent studies show that certain clonal aberrations can affect the
risk of developing AML [8, 11–13]. The most frequently reported acquired
clonal aberrations in FA patients are trisomies of chromosome 1q and
monosomies of chromosome 7. In addition, partial trisomies and tetrasomies of
chromosome 3q are proven risk factors for the development of MDS and AML
[14] (see Chapter by Neitzel et al.).

Solid Tumors
The risk for solid tumors in FA patients exceeds that of the general popula-
tion by nearly the 50-fold [15]. In addition, cancer in FA strikes very early. The
median age is 16 years, whereas it is 68 years in the non-FA population [4]. FA
patients may have combinations of different malignancies [4, 7]. In 25% of FA
cases with cancer, the malignancy was present prior to diagnosis of FA [4], an
extreme example being a 49-year-old patient with bilateral carcinoma of the
breast who developed MDS and AML following chemotherapy [16].

Neveling/Kalb/Schindler 62
 Squamous Cell Carcinoma
Five percent of all FA patients summarized by Alter [4] had solid tumors,
the most typical tumors being squamous cell carcinomas of the head and neck
(HNSCC) in both males and females, and of the genital area in females.
HNSCC in FA include mostly tumors of the tongue, gingiva, upper esophagus,
larynx and oropharynx [15]. The risk for HNSCC in FA patients is more than
700 times greater than in the general population [15]. In non-FA patients,
HNSCC normally occurs in men older than 40 years who smoke and drink [17].
In contrast, FA patients with HNSCC are usually young, non-smokers and non-
drinkers. Females are more often affected than males, especially among those
patients in whom the diagnosis of FA was made after the diagnosis of cancer,
but this may reflect the overall survival advantage of female FA patients [15].
The mucous epithelia are common routes for viral infections, especially
for HPV (human papilloma virus) and HSV (herpes simplex virus). Immuno-
suppression associated with bone marrow failure might render patients vulner-
able to the integration of the viral DNA [7]. While HPV infections are proven
triggers for the development of cancer of the cervix uteri, their role in non-
anogenital cancers is still unclear [18]. The expression profile of HPV-associated
HNSCC differs from that in HPV-negative HNSCC: TP53 is mutated in HPV-
negative tumors, but remains functional in HPV-associated HNSCC because of
rapid p53 degradation by the viral protein E6 [18]. The presence of HPV in
tumors may affect the prognosis, as patients with HPV-associated HNSCC have
a lower risk for death than HPV-negative HNSCC patients [18]. Concerning FA,
it has been found that 54% of all patients with genital squamous cell carcino-
mas of cervix, vulva and anus had HPV-associated condylomas before occur-
rence of the tumor [7]. These findings were seemingly confirmed by a second
study showing that 84% of FA-associated squamous cell carcinomas (SCC) had
detectable HPV-DNA, while only 36% of non-FA SCC showed viral infection
[19]. However, other investigators were unable to demonstrate such high rates
of HPV contamination. For example, four cell lines freshly established from FA
HNSCC tumors had no detectable HPV DNA. No differences concerning loss
of heterozygosity pattern, TP53 mutations and TP53 polymorphisms were
found between these FA cell lines and non-FA HNSCC cell lines leading the
authors to conclude that FA-associated HNSCC are not fundamentally different
from sporadic HNSCC, except for their sensitivity to crosslinking agents [20].
In addition, chromosomal changes in oral squamous cell carcinomas of two FA
patients were very similar to what was seen in non-FA oral tumors [21].
Squamous cell carcinomas are aggressive tumors that initially present as minor
mucosal lesions, chronic inflammations or red or white cell spots. Diagnosis is
often delayed leading to poor therapeutic success and no better than 50% 2-year
survival rates [17]. A sensitive brush method for obtaining cells from tongue

Cancer in Fanconi Anemia and Fanconi Anemia Genes in Cancer 63

and oral cavity may facilitate monitoring of early cell changes and thereby con-
tribute to the prevention of this devastating type of malignancy.
Due to immune suppression and/or conditioning measures, the risk for
developing solid tumors in FA patients increases after hematopoetic stem cell
transplantation (HSCT). Conditioning but also post-transplant complications
(e.g. graft versus host disease and infections) sharply increase this risk for
malignancies with poor survival. The cumulative incidence of head and neck
carcinoma in FA patients is 20% and 53% by 10 and 15 years after transplanta-
tion, respectively [82]. The median age of FA patients who developed cancer
following HSCT was 21 years [4].

Liver Tumors
Liver tumors are frequent in FA and in the great majority of cases arise in
the context of androgen therapy for aplastic anemia. They mostly are adenomas
and rarely hepatocellular carcinomas. The cumulative incidence for these
tumors is 46% by age 50, and the median age of onset is 13 years [4].
Development of liver tumors due to androgen therapy is frequent in FA but not
specific for FA patients. The risk for hepatic carcinomas may depend on the
specific type of androgen used. Oxymetholone and methyltestosterone seem to
be associated more frequently with the development of hepatocellular carcino-
mas (HCC), whereas patients treated with danazol develop adenomas but may
be less affected by hepatic carcinomas [22]. Adenomas usually regress upon
discontinuation of therapy. In rare cases, both HCCs and adenomas can be
found in the same patient.

Other Tumors
Other tumors occasionally found in FA patients include brain and renal
tumors, breast carcinoma, basal cell carcinoma, neuroblastoma, desmoid
tumors, gonadoblastoma, melanoma, neurilemmona and osteogenic sarcoma
[7]. A major factor thought to promote tumor development in FA is oxidative
stress [23–26]. It should also be emphasized that biallelic mutations in two of
the downstream FA genes (FANCD1/BRCA2 and FANCN/PALB2) confer an
exceptionally high risk for early childhood leukemia and a variety of soild
tumors [6, 43, 84].

Fanconi Anemia Genes in Cancer

Disruption of the FA/BRCA pathway by a defect in one of the underlying

genes results in a defective response to certain types of DNA damage. Defects
in the FA/BRCA pathway may therefore contribute to non-FA tumorigenesis.

Neveling/Kalb/Schindler 64
So far, this has been especially reported for the FANCF gene which was found
to be epigenetically inactivated in a number of tumors in non-FA patients
(reviewed in [27]). In order to find evidence for the involvement of any of the
other FA genes in malignancies, a variety of tumors have been examined for
sequence variations in FA genes in non-FA patients. Studies searching for
germline changes including polymorphisms of FA genes as possible predispos-
ing factors of tumorigenesis in non-FA patients are summarized in table 2,
whereas the special case of FANCF will be discussed separately (see below).
Table 2 provides a survey of germline mutations in FA genes that were
found in tumors of non-FA patients. Listed are the respective FA genes, the type
of tumors investigated, the family history and the kind of mutations that were
detected. The question whether a given mutation is likely to be causal for
tumorigenesis is answered by ‘yes’ or ‘no’ depending on the respective authors’
interpretation. The conclusion suggested by the data presented in table 2 is that
there is no convincing evidence for a causal association between germline
changes in FA genes and propensity to tumorigenesis in non-FA patients.
In contrast to table 2, table 3 summarizes studies looking for purely
somatic changes in FA genes in tumor patients without FA germline mutations.
Again, studies involving FANCF are not included but are discussed below. Table 3
shows the respective FA genes analyzed, the types of tumors investigated, the
putative family history, and the kinds of mutations detected. Answers to the
question whether a mutation is likely to be causal for tumorigenesis are indi-
cated according to the respective authors’ interpretation. It should be pointed
out that in table 3 there are more ‘yes’ answers than ‘no’ answers to this ques-
tion. However, most of the ‘yes’ answers had to be marked with a question mark
which underlines the high degree of uncertainty that characterizes the interpre-
tation of most of these studies. Final conclusions must await additional studies,
but what is already clear at this point is that somatic mutations in FA genes,
with the possible exception of pancreatic carcinoma, appear to be surprisingly
infrequent events in various types of human malignancies.
As listed in tables 2 and 3, a number of malignancies in non-FA patients,
most notably AML, pancreatic cancer, carcinoma of the breast and ovary, and
squamous cell carcinoma, have been repeatedly investigated for possible asso-
ciations with FA germline and/or somatic mutations. The following is a brief
discussion of the most important findings of these studies (see also [28]).

AML
AML is the most common hematological malignancy in FA, and FA genes
were investigated in several cases of familial and sporadic non-FA AML.
Occasional mutations and polymorphisms were described in FANCA, FANCC,
FANCD1/BRCA2 and FANCG. Since more than 50% of FA patients belong to

Cancer in Fanconi Anemia and Fanconi Anemia Genes in Cancer 65

 Table 2. Germline sequence variations of FA genes in non-FA tumor patients

Gene Cancer typea Sporadic/ Mutation type Likely to Reference

familial found be causal

FANCA AML sporadic deletions yes [30]

breast cancer familial missense/polymorphism no [58]
sporadic polymorphism no [60]
both polymorphism no [59]
ovarian cancer both polymorphism yes? [59]
pancreatic cancer familial missense no [38]
FANCC AML/ALL (childhood) sporadic polymorphism yes? [32]
T-ALL familial frameshift yes [31]
breast cancer familial polymorphisms no [58]
pancreatic cancer familial polymorphisms no [37]
sporadic missense yes? [35]
frameshift, missense yes [36]
FANCD1/ AML (childhood) sporadic no IVS7 splice site mutation no [33]
BRCA2 esophageal SCC familial missense no? [61]
ovarian cancer both frameshift, stop yes? [50]
FANCD2 breast cancer familial polymorphisms no [58]
familial polymorphisms, silent variants no [56]
sporadic polymorphism yes [60]
FANCE breast cancer familial missense/polymorphisms no [58]
FANCF breast cancer familial polymorphism no [58]
FANCG AML (childhood) sporadic aberrant splicing, missense, no [34]
deletion
breast cancer familial – no [58]
pancreatic cancer familial polymorphisms no [37]
sporadic stop yes? [35]
FANCJ breast cancer familial missense yes? [53]
sporadic missense ? [53]
familial missense, polymorphisms no [55]
familial missense, frameshift, silent no [56]
familial polymorphism yes [54]
familial polymorphism no [57]
familial frameshift/stop yes [83]
FANCL breast cancer sporadic polymorphism no [60]
FANCN breast cancer familial frameshift/stop yes [84]
breast cancer familial missense/frameshift yes [85]
breast cancer familial frameshift yes [85]

a
ALL: Acute lymphocytic leukemia; AML: acute myeloid leukemia; SCC: squamous cell carcinoma.

Neveling/Kalb/Schindler 66
 Table 3. Somatic changes of FA genes in non-FA tumor patients

Gene Cancer typea Sporadic/ Mutation type Likely to Reference

familial be causal

FANCA AML (adult) sporadic not found yes? [29]

FANCC pancreatic cancer sporadic frameshift yes? [35]
FANCD1/BRCA2 breast cancer ? frameshift yes? [48]
sporadic frameshift, missense, yes? [49]
silent variants
esophageal SCC ? frameshift, missense, no? [61]
silent variants
oral SCC sporadic genomic alteration yes? [62]
ovarian cancer both aberrant splicing, yes? [50]
frameshift, stop
FANCD2 oral SCC sporadic genomic alteration yes? [62]
FANCG oral SCC sporadic genomic alteration yes? [62]

a
AML: Acute myeloid leukemia; SCC: squamous cell carcinoma.

complementation group FA-A, in many studies the FANCA gene has been the
primary target of investigation. For example, a leukemic cell line of an adult
non-FA AML patient was found to exhibit hypersensitivity towards mitomycin
C (MMC) and diepoxibutane (DEB) combined with a decrease of FANCA,
FANCG and FANCD2-L expression which could be corrected by transduction
with a FANCA containing vector. However, the authors of this study failed to
find any evidence for mutations in the FANCA gene. They therefore proposed
the idea of unidentified factors affecting the cellular localization and binding
activity of FANCA, leading to cytogenetic instability and clonal progression
[29]. In another study four cases of heterozygous germline FANCA deletions
were reported among 101 sporadic AML patients. In all four cases, there was no
evidence for inactivation of the second allele either by mutation or promoter
hypermethylation. Since the frequency of the detected deletions was 35-fold
higher than the expected frequency for germline FANCA deletions, the authors
concluded that deletions leading to reduced expression of FANCA may be
involved in a subset of cases of sporadic AML [30]. The possibility of a con-
nection between tumorigenesis and FA genes was also considered for FANCC:
Rischewski and coworkers described a heterozygous FANCC frameshift muta-
tion in two siblings with T-ALL [31]. No other mutations were found in these
patients, leading the authors to the conclusion that heterozygous carriers of
FANC mutations might have an increased risk for developing AML.

Cancer in Fanconi Anemia and Fanconi Anemia Genes in Cancer 67

Subsequently, Barber and colleagues found 12 FANCC sequence variants in
genomic DNA from sporadic AML/ALL cases [32]. However, there was no
evidence for known pathogenic FANCC mutations, but one of the polymor-
phisms (S26F) was observed four times more frequently in AML patients than
in random cord blood samples suggesting a possible contribution to the risk for
childhood AML. Other researchers found no convincing evidence for frequent
occurrence of FA gene mutations in AML. For example, FANCD1/BRCA2
patients who develop AML frequently show certain exon 7 splice mutations. In
contrast, such ‘signature’ mutations were not present in DNA from non-FA spo-
radic childhood AML patients [33]. In a subsequent study, 107 children with
sporadic AML were investigated for germline sequence variants in FANCG.
Different kinds of mutations were found in seven of these children. Two patients
had biallelic FANCG variants, suggesting that they might have been undiag-
nosed FA patients. The overall message from this study was that FA carrier sta-
tus does not predispose to sporadic AML [34]. Despite occasional positive
findings, it must be emphasized that in the great majority of non-FA patients
there is no convincing evidence for a clear association between defective FA
genes and occurrence of AML. However, one should be aware of the possibility
of undiagnosed FA cases among sporadic AML patients.

Pancreatic Cancer
FANC gene mutations in pancreatic cancer have been described for
FANCA, FANCC and FANCG. Again, in some cases the authors of these studies
consider a connection between detected mutations and tumorigenesis, but other
studies fail to find evidence for such an association. In particular, FANCC muta-
tions were found repeatedly in sporadic cases of pancreatic cancer, and inher-
ited as well as somatic mutations were reported in patients with early onset
pancreatic cancer. The mutations were hemizygous due to loss of one allele in
the tumor. Additionally, a probably inherited nonsense mutation was found in
FANCG. The authors concluded that their findings may point towards a general
involvement of FA genes in cancer [35]. Mutations in FANCC and FANCG in
sporadic pancreatic cancer were also described in another study, in which
germline truncating and missense FANCC mutations were reported as well as a
single somatic missense mutation in FANCG. The germline FANCC mutations
were associated with loss of heterozygosity (LOH) in the tumor. Again, the
authors suggested that inherited mutations in FANCC may predispose to pan-
creatic cancer [36]. However, in critically reviewing these and other data,
Rogers et al. [37] concluded that germline and somatic changes in FANCC and
FANCG, if at all, may have comparatively low penetrance for the pancreatic
cancer phenotype. The same was suggested regarding FANCA: although several
exonic FANCA variants, including disease-associated variants, were identified

Neveling/Kalb/Schindler 68
in lymphocyte DNA from familial pancreatic cancer patients, these variants
were either found in controls at the same frequency as in familial pancreatic
cancer, or there were no convincing segregation data for these variants. Rogers
et al. therefore maintain that germline FANCA mutations do not contribute to
familial pancreatic cancer susceptibility [38]. In reviewing these studies we
find that there is some degree of suggestive but no definitive evidence for
a causal association between defective FA genes and pancreatic cancer.

Breast and Ovarian Cancer

BRCA2 (breast cancer susceptibility gene 2) is one of the most important
caretaker genes associated with breast and ovarian cancer. BRCA2 was identi-
fied as the second gene mutated in high risk breast cancer families in 1994 [39,
40]. In 2002, biallelic mutations in BRCA2 were described to result in an FA-
like phenotype [41]. Therefore, monoallelic inactivation of this caretaker gene
causes breast and ovarian cancer predisposition, while biallelic mutations cause
a severe form of FA [42, 43]. The role of heterozygous germline mutations in
BRCA2 as a predisposing factor for familial breast and ovarian cancer has often
been described (for review see [44]) and germline mutations were also found to
be responsible for some kinds of apparently sporadic forms of breast and ovar-
ian cancer [45–47]. Somatic BRCA2 mutations are rare. A first case of a
somatic truncating mutation in a sporadic breast cancer involved a 1-bp dele-
tion resulting in premature truncation. This change was detected in a primary
ductual breast carcinoma with demonstrated LOH of BRCA2-flanking markers
[48]. Interestingly, somatic BRCA2 mutations were also described in rare cases
of sporadic male breast cancer involving frameshift, missense and silent muta-
tions in tumor tissues, all associated with allelic loss at the BRCA2 locus [49].
Regarding ovarian cancer (sporadic as well as familial), germline as well as
somatic BRCA2 mutations were occasionally found, all being frameshift muta-
tions or base substitutions resulting in premature termination and all associated
with LOH of flanking markers in the tumor [50]. Taken together, it is firmly
established that germline mutations of BRCA2 are associated with breast and
ovarian cancer. These alterations are accompanied by allelic loss in the tumor
tissue, suggesting a tumor suppressor function, whereas somatic mutations in
BRCA2 seem to be rare events in breast and ovarian cancer.
FANCJ/BRIP1/BACH1 is one of the recently identified FA genes (BRIP1:
BRCA1 interacting protein/BACH1: BRCA1 associated C-terminal helicase)
[51, 52]. Because of its direct interaction with BRCA1 at the protein level,
FANCJ is an obvious candidate for a breast cancer susceptibility gene. Several
studies investigated this hypothesis but arrived at different conclusions. Cantor
and colleagues described heterozygous germline missense mutations in two breast
cancer patients (P47A and M299I), one being familial and the other sporadic

Cancer in Fanconi Anemia and Fanconi Anemia Genes in Cancer 69

[53]. Co-segregation of the familial mutation with breast cancer could not be
tested because DNA from the family members was not available. However, both
mutations were not found in 200 controls. The authors concluded that mutant
BACH1/BRIP1 participates in breast cancer development. In a subsequent
study, genomic DNA from familial breast cancer cases was investigated for sin-
gle nucleotide polymorphisms (SNP) in BACH1/BRIP1. The polymorphism
P919S (rs4986764) was thought to be associated with increased breast cancer
risk [54]. In contrast, different heterozygous BACH1/BRIP1 germline alter-
ations (polymorphisms and missense mutations) were observed in a screen of
BRCA1/2-negative breast cancer families, and none of these sequence alter-
ations seemed to be associated with a detectable risk for breast cancer [55].
A subsequent study from Australia reported different heterozygous germline
BRIP1/FANCJ variants in a similar screen of BRCA1/2-negative breast cancer
families [56]. Investigation of relatives of the index patients provided no evi-
dence for co-segregation of the sequence variants with the tumors, contradict-
ing any causal association between BRIP1 alterations and familial breast
cancer. A comparable Scandinavian study found six germline alterations in
FANCJ, none of them segregating with the cancer phenotype. Notably, the
P919S polymorphism, previously reported to be associated with breast cancer,
showed no connection to increased breast cancer risk in the Finnish study [57].
However, the two largest studies to date of BRCA1/BRCA2 negative familial
breast cancer implicate both FANCJ/BRIP1 and the newly discovered FANCN/
PALB2 gene as low penetrance breast cancer susceptibility genes [83–85].
In addition to FANCD1/BRCA2 and FANCJ/BRIP1, other FANC genes
have also been investigated for being putative breast and ovarian cancer sus-
ceptibility genes. For example, germline DNA from 88 BRCA1/2-negative
breast cancer families was examined for mutations in FANCA, FANCC,
FANCD2, FANCE, FANCF and FANCG. Altogether, 69 sequence variants were
found, among them 25 exonic variants (including 14 polymorphisms), but no
frameshift and no nonsense mutations. Only two conservative missense vari-
ants were observed, one in FANCA and one in FANCE. These data suggest that
heterozygous FA gene mutations in genes other than FANCD1/BRCA2 do not
confer a high risk for breast cancer [58]. In a following study, a novel duplication-
polymorphism in the promoter region of FANCA was reported, and germline
DNA from breast cancer patients, ovarian cancer patients and controls was
investigated for this polymorphism [59]. The authors found a non-significant
decrease of the mutant allele compared to the common allele in breast cancer
patients, and a significant decrease in ovarian cancer patients. Accordingly,
this particular polymorphism may not influence the risk for breast cancer but
might rather convey a protective function regarding ovarian cancer. Concerning
FANCD2, a recent study reported lack of expression in 10–20% of sporadic

Neveling/Kalb/Schindler 70
and BRCA1-related breast cancer, but it was not stated whether this was due
to somatic mutations, defective processing, or epigenetic silencing [87].
Otherwise, there are contradictory data concerning genetic variation of
FANCD2 and breast cancer risk [56, 60]. The overall conclusion from these
studies is that a number of germline mutations have been identified in FA
genes of non-FA patients with early onset breast cancer. However, a likely role
as low-penetrance susceptibility genes has been demonstrated only for alter-
ations in the BRCA-associated genes FANCJ and FANCN.

Squamous Cell Carcinoma

Esophageal squamous cell carcinomas (ESCC) have been shown to have a
high rate of allelic loss involving chromosome 13, including the locus of the
BRCA2 gene. These observations prompted the screening of DNA from ESCC
patients for mutations in BRCA2. Altogether, three patients with germline muta-
tions and two patients with somatic sequence variants were found. These
included missense or silent variants. The authors suggested that BRCA2 muta-
tions do occur in ESCC, but are infrequent and of unknown consequence [61].
In another study, Sparano and colleagues searched for candidate genes respon-
sible for oral squamous cell carcinoma (OSCC) by array comparative genomic
hybridization (aCGH) [62]. They assessed 512 genes commonly altered in can-
cer. Among cancer related genes that were found to be changed in more than
25% of OSCC, the authors identified several FA genes, including FANCD1,
FANCD2 and FANCG. This might be an important clue. Given the high fre-
quency of OSCC in FA it would come as no surprise if FA genes would play a
role in the genesis of OSCC, but additional studies are needed to confirm these
preliminary observations.

FANCF: A Genuine Tumor Gene?

Whereas the data presented in tables 2 and 3 overwhelmingly contradict

the hypothesis of a non-random, causal association between altered FA genes
and the occurrence of neoplasia in non-FA patients, an important exception is
the FANCF gene. FANCF is the smallest of the presently known 12 FA genes.
It consists of only a single exon, and the coding region comprises only
1124 nucleotides. The 42 kDa protein has homology to the prokaryotic RNA
binding protein ROM, but no function similar to ROM. Known mutations in
FANCF are deletions and nonsense mutations [63]. The only function of
FANCF seems to be its role as a flexible adapter protein of the FA core complex
[64]. Contrary to its small size and to its presumably minor function in the FA
nuclear core complex, FANCF has been found silenced by hypermethylation of

Cancer in Fanconi Anemia and Fanconi Anemia Genes in Cancer 71

its promoter CpG island in a variety of human neoplasias. Table 4 contains a list
of different kinds of tumors in which FANCF hypermethylation has been
detected.
The prevailing concept is that inactivation of a caretaker gene such as
FANCF leads to genomic instability and therefore might be an early step in
tumorigenesis [65]. This concept has received strong support by the recent
observation of frequent epigenetic silencing of another caretaker gene, WRN, in
human tumors [66]. As summarized in table 4, epigenetic silencing of FANCF
has been investigated in a large number of different tumors, with positive
results in many of them. However, it is not clear at present whether disruption of
the FA/BRCA pathway by FANCF methylation is a causative factor in cancero-
genesis or whether FANCF methylation occurs because of its location adjacent
to a known hotspot region for hypermethylation at 11p15 [63, 67]. Silencing
effects due to physical proximity to hotspot regions for hypermethylation have
been described for other genes. For example, altered expression of the
CDKN1C gene by promoter hypermethylation has been reported for the
Beckwith-Wiedemann syndrome and for several other human cancers [68].
Therefore, the contribution of FANCF silencing to tumorigenesis might be a
coincidental rather than primary event [86].

General Conclusions

Individuals with biallelic mutations in any of the 12 known FA genes have

a strongly elevated risk for neoplastic disease at comparatively young age. The
types of neoplasias that predominate in FA patients include AML and squa-
mous cell carcinomas of the oropharynx and the genital area. Given the high
degree of genetic instability and elevated mutation rates seen in FA patients,
there is little doubt about a causative relationship between FA-specific defects
of genome maintenance and the emergence of malignant cells. If the immune
system was not relatively intact in most FA patients until the terminal stages of
bone marrow failure we would most likely encounter even higher frequencies
of malignant growth in these patients, and at even younger ages. In addition to
their increased sensitivity towards DNA crosslinking agents, FA cells are
exceedingly sensitive to reactive oxygen species which is unique for any of the
human caretaker gene syndromes and may explain part of their susceptibility
to malignant transformation. Ambient air and thus possibly harmful concentra-
tion of oxygen is present at the sites most frequently affected by squamous cell
carcinomas in FA patients (oral cavity and genital area), but other factors like
virus infections may also contribute to tumor development at these particular
sites.

Neveling/Kalb/Schindler 72
 Table 4. Silencing of FANCF in human malignancies

Cancer typea FANCF methylation Proportion (%) Reference

AML • one cell line (CHRF-288) 0 [71]

• 0/36 sporadic AML patients
Astrocytoma 0/1 gastrocytoma cell line 0 [72]
Sporadic childhood leukemia 0/81 diagnostic bone marrow aspitates 0 [73]
Biliary cancer 0/4 biliary cancer cell lines 0 [72]
Bladder cancer 0/2 bladder cancer cell lines 0 [74]
• 1/23 cell lines 4 [86]
• 1/41 bladder cancer tissues 2
Breast cancer 13/75 primary breast tumors 17 [75]
0/1 breast cancer cell line 0 [74]
1/15 breast cancer cell lines 6 [72]
(decreased expression, not tested
for methylation)
Cervical cancer • 27/91 primary tumors 30 [76]
• 3/9 cell lines 33
Colon cancer 0/3 colon cancer cell lines 0 [74]
HNSCC 0/8 HNSCC cell lines 0 [72]
13/89 primary HNSCC 15 [74]
0/10 HNSCC cell lines 0 [77]
NSCLC • 22/154 primary NSCLC 14 [74]
• 0/20 NSCLC cell lines
LCLC 0/1 LCLC cell line 0 [72]
SCLC 0/10 SCLC cell lines 0 [74]
Osteosarcoma 0/1 osteosarcoma cell line 0 [74]
Ovarian cancer • 2/25 ovarian tumor lines 8 [65]
• 4/19 primary ovarian tumors 21
6/25 granulosa cell tumors 24 [78]
0/106 stage III/IV epithelial ovarian tumors 0 [79]
• 1/7 ovarian cancer cell lines 14 [80]
• 5/18 primary ovarian tumors 28
Prostate cancer 0/1 prostate cancer cell line 0 [74]
0/6 prostate cancer cell lines 0 [72]
Testicular germ cell tumor 4/70 germ cell tumors 6 [81]
(non-seminoma)

a
AML: Acute myeloid leukemia; HNSCC: head and neck squamous cell carcinoma; LCLC: large-cell lung
carcinoma; NSCLC: non-small cell lung cancer; SCLC: small-cell lung cancer.

Cancer in Fanconi Anemia and Fanconi Anemia Genes in Cancer 73

 In contrast to the strongly elevated cancer risk in FA patients, there is little
evidence so far that germline mutations and/or polymorphisms of FA genes
play a major role in tumorigenesis of non-FA tumor patients. This is in good
agreement with the observations that heterozygous parents and siblings of FA
patients are not at increased risk for malignancies, indicating that FA genes,
with the possible exception of FANCD1/BRCA2, are unlikely to function as
classical tumor suppressor genes [69].
With respect to purely somatic mutations, the situation is less clear, but
again the evidence in favor of a major contribution of somatic FA mutations to
tumorigenesis in non-FA patients is tenuous at best. The only exception is the
FANCF gene which has been found to be epigenetically silenced in a variety of
tumors and tumor cell lines. However, FANCF maps to a known hotspot of
hypermethylation in the human genome such that the possibility of a
secondary rather than primary effect of FANCF silencing must be seriously
considered.
As pointed out by van der Heijden and colleagues, there might well be
selection for FA protein function in tumorigenesis in order to assure a certain
DNA repair capacity of malignant cells [70]. These authors speculate that ‘a
limited level of genetic instability could be beneficial to tumorigenesis, but
excessive DNA repair defect would not’. In reviewing the evidence for genetic
changes in FA genes in non-FA tumors, we tend to arrive at a similar conclu-
sion. The relative paucity of either germline or somatic mutations in FA genes
in non-FA tumor patients, and the possibility of a ‘bystander’ rather than pri-
mary effect of FANCF silencing in a variety of human neoplasias strongly sug-
gest that loss of FA gene function may be detrimental rather than beneficial for
most tumor cells. It seems that a family of genes like the FA family that partic-
ipates in diverse functions of DNA maintenance, including DNA replication
and DNA repair, is essential for cells with proliferative capacity. Loss of FA
gene function impairs proliferation of hematopoietic stem cells and might also
jeopardize proliferation of cancer stem cells.

References

1 Park Y, Gerson SL: DNA repair defects in stem cell function and aging. Annu Rev Med
2005;56:495–508.
2 Beckman RA, Loeb LA: Genetic instability in cancer: theory and experiment. Semin Cancer Biol
2005;15:423–435.
3 Thompson LH, Hinz JM, Yamada NA, Jones NJ: How Fanconi anemia proteins promote the four
Rs: replication, recombination, repair, and recovery. Environ Mol Mutagen 2005;45:128–142.
4 Alter BP: Cancer in Fanconi anemia, 1927–2001. Cancer 2003;97:425–440.
5 Meyer S, Fergusson WD, Oostra AB, Medhurst AL, Waisfisz Q, et al: A cross-linker-sensitive
myeloid leukemia cell line from a 2-year-old boy with severe Fanconi anemia and biallelic
FANCD1/BRCA2 mutations. Genes Chromosomes Cancer 2005;42:404–415.

Neveling/Kalb/Schindler 74
 6 Reid S, Renwick A, Seal S, Baskcomb L, Barfoot R, et al: Biallelic BRCA2 mutations are associ-
ated with multiple malignancies in childhood including familial Wilms tumour. J Med Genet
2005;42:147–151.
7 Kutler DI, Singh B, Satagopan J, Batish SD, Berwick M, et al: A 20-year perspective on the
International Fanconi Anemia Registry (IFAR). Blood 2003;101:1249–1256.
8 Alter BP, Caruso JP, Drachtman RA, Uchida T, Velagaleti GV, Elghetany MT: Fanconi anemia:
myelodysplasia as a predictor of outcome. Cancer Genet Cytogenet 2000;117:125–131.
9 Auerbach AD, Weiner MA, Warburton D, Yeboa K, Lu L, Broxmeyer HE: Acute myeloid leukemia
as the first hematologic manifestation of Fanconi anemia. Am J Hematol 1982;12:289–300.
10 Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton DA, et al: Proposals for the classification
of the myelodysplastic syndromes. Br J Haematol 1982;51:189–199.
11 Fundia A, Gorla N, Larripa I: Spontaneous chromosome aberrations in Fanconi’s anemia patients
are located at fragile sites and acute myeloid leukemia breakpoints. Hereditas 1994;120:47–50.
12 Ferti A, Panani A, Dervenoulas J, Raptis SA: Cytogenetic findings in a Fanconi anemia patient
with AML. Cancer Genet Cytogenet 1996;90:182–183.
13 Oliveira NI, Ribeiro EM, Raimondi SC, Bittencourt MA, Pasquini R, Cavalli IJ: Two different
karyotypes with 1q abnormalities in a patient with Fanconi anemia. Leuk Res 2002;26:
1047–1049.
14 Tonnies H, Huber S, Kuhl JS, Gerlach A, Ebell W, Neitzel H: Clonal chromosomal aberrations in
bone marrow cells of Fanconi anemia patients: gains of the chromosomal segment 3q26q29 as an
adverse risk factor. Blood 2003;101:3872–3874.
15 Alter BP, Joenje H, Oostra AB, Pals G: Fanconi anemia: adult head and neck cancer and
hematopoietic mosaicism. Arch Otolaryngol Head Neck Surg 2005;131:635–639.
16 Huck K, Hanenberg H, Gudowius S, Fenk R, Kalb R, et al: Delayed diagnosis and complications
of Fanconi anaemia at advanced age – a paradigm. Br J Haematol 2006;133:188–197.
17 Kutler DI, Auerbach AD, Satagopan J, Giampietro PF, Batish SD, et al: High incidence of head and
neck squamous cell carcinoma in patients with Fanconi anemia. Arch Otolaryngol Head Neck
Surg 2003;129:106–112.
18 Gillison ML, Lowy DR: A causal role for human papillomavirus in head and neck cancer. Lancet
2004;363:1488–1489.
19 Kutler DI, Wreesmann VB, Goberdhan A, Ben-Porat L, Satagopan J, et al: Human papillomavirus
DNA and p53 polymorphisms in squamous cell carcinomas from Fanconi anemia patients. J Natl
Cancer Inst 2003;95:1718–1721.
20 van Zeeburg HJ, Snijders PJ, Pals G, Hermsen MA, Rooimans MA, et al: Generation and molecu-
lar characterization of head and neck squamous cell lines of Fanconi anemia patients. Cancer Res
2005;65:1271–1276.
21 Hermsen MA, Xie Y, Rooimans MA, Meijer GA, Baak JP, et al: Cytogenetic characteristics of oral
squamous cell carcinomas in Fanconi anemia. Fam Cancer 2001;1:39–43.
22 Velazquez I, Alter BP: Androgens and liver tumors: Fanconi’s anemia and non-Fanconi’s condi-
tions. Am J Hematol 2004;77:257–267.
23 Pagano G, Youssoufian H: Fanconi anaemia proteins: major roles in cell protection against oxida-
tive damage. Bioessays 2003;25:589–595.
24 Pagano G, Degan P, d’Ischia M, Kelly FJ, Nobili B, et al: Oxidative stress as a multiple effector in
Fanconi anaemia clinical phenotype. Eur J Haematol 2005;75:93–100.
25 Joenje H, Arwert F, Eriksson AW, de Koning H, Oostra AB: Oxygen-dependence of chromosomal
aberrations in Fanconi’s anaemia. Nature 1981;290:142–143.
26 Schindler D, Hoehn H: Fanconi anemia mutation causes cellular susceptibility to ambient oxygen.
Am J Hum Genet 1988;43:429–435.
27 Taniguchi T, D’Andrea AD: The molecular pathogenesis of Fanconi anemia: recent progress.
Blood 2006;107:4223–4233.
28 Lyakhovich A, Surralles J: Disruption of the Fanconi anemia/BRCA pathway in sporadic cancer.
Cancer Lett 2006;232:99–106.
29 Lensch MW, Tischkowitz M, Christianson TA, Reifsteck CA, Speckhart SA, et al: Acquired
FANCA dysfunction and cytogenetic instability in adult acute myelogenous leukemia. Blood
2003;102:7–16.

Cancer in Fanconi Anemia and Fanconi Anemia Genes in Cancer 75

30 Tischkowitz MD, Morgan NV, Grimwade D, Eddy C, Ball S, et al: Deletion and reduced expres-
sion of the Fanconi anemia FANCA gene in sporadic acute myeloid leukemia. Leukemia
2004;18:420–425.
31 Rischewski JR, Clausen H, Leber V, Niemeyer C, Ritter J, et al: A heterozygous frameshift muta-
tion in the Fanconi anemia C gene in familial T-ALL and secondary malignancy. Klin Padiatr
2000;212:174–176.
32 Barber LM, McGrath HE, Meyer S, Will AM, Birch JM, et al: Constitutional sequence variation in
the Fanconi anaemia group C (FANCC) gene in childhood acute myeloid leukaemia. Br
J Haematol 2003;121:57–62.
33 Barber LM, Barlow RA, Meyer S, White DJ, Will AM, et al: Inherited FANCD1/BRCA2 exon 7
splice mutations associated with acute myeloid leukaemia in Fanconi anaemia D1 are not found in
sporadic childhood leukaemia. Br J Haematol 2005;130:796–797.
34 Meyer S, Barber LM, White DJ, Will AM, Birch JM, et al: Spectrum and significance of variants
and mutations in the Fanconi anaemia group G gene in children with sporadic acute myeloid
leukaemia. Br J Haematol 2006;133:284–292.
35 van der Heijden MS, Yeo CJ, Hruban RH, Kern SE: Fanconi anemia gene mutations in young-
onset pancreatic cancer. Cancer Res 2003;63:2585–2588.
36 Couch FJ, Johnson MR, Rabe K, Boardman L, McWilliams R, et al: Germ line Fanconi anemia
complementation group C mutations and pancreatic cancer. Cancer Res 2005;65:383–386.
37 Rogers CD, van der Heijden MS, Brune K, Yeo CJ, Hruban RH, et al: The genetics of FANCC and
FANCG in familial pancreatic cancer. Cancer Biol Ther 2004;3:167–169.
38 Rogers CD, Couch FJ, Brune K, Martin ST, Philips J, et al: Genetics of the FANCA gene in familial
pancreatic cancer. J Med Genet 2004;41:e126.
39 Wooster R, Neuhausen SL, Mangion J, Quirk Y, Ford D, et al: Localization of a breast cancer sus-
ceptibility gene, BRCA2, to chromosome 13q12-13. Science 1994;265:2088–2090.
40 Wooster R, Bignell G, Lancaster J, Swift S, Seal S, et al: Identification of the breast cancer sus-
ceptibility gene BRCA2. Nature 1995;378:789–792.
41 Howlett NG, Taniguchi T, Olson S, Cox B, Waisfisz Q, et al: Biallelic inactivation of BRCA2 in
Fanconi anemia. Science 2002;297:606–609.
42 Offit K, Levran O, Mullaney B, Mah K, Nafa K, et al: Shared genetic susceptibility to breast can-
cer, brain tumors, and Fanconi anemia. J Natl Cancer Inst 2003;95:1548–1551.
43 Wagner JE, Tolar J, Levran O, Scholl T, Deffenbaugh A, et al: Germline mutations in BRCA2:
shared genetic susceptibility to breast cancer, early onset leukemia, and Fanconi anemia. Blood
2004;103:3226–3229.
44 Wooster R, Weber BL: Breast and ovarian cancer. N Engl J Med 2003;348:2339–2347.
45 Lancaster JM, Wooster R, Mangion J, Phelan CM, Cochran C, et al: BRCA2 mutations in primary
breast and ovarian cancers. Nat Genet 1996;13:238–240.
46 Miki Y, Katagiri T, Kasumi F, Yoshimoto T, Nakamura Y: Mutation analysis in the BRCA2 gene in
primary breast cancers. Nat Genet 1996;13:245–247.
47 Teng DH, Bogden R, Mitchell J, Baumgard M, Bell R, et al: Low incidence of BRCA2 mutations
in breast carcinoma and other cancers. Nat Genet 1996;13:241–244.
48 Weber BH, Brohm M, Stec I, Backe J, Caffier H: A somatic truncating mutation in BRCA2 in a
sporadic breast tumor. Am J Hum Genet 1996;59:962–964.
49 Kwiatkowska E, Teresiak M, Breborowicz D, Mackiewicz A: Somatic mutations in the BRCA2
gene and high frequency of allelic loss of BRCA2 in sporadic male breast cancer. Int J Cancer
2002;98:943–945.
50 Hilton JL, Geisler JP, Rathe JA, Hattermann-Zogg MA, DeYoung B, Buller RE: Inactivation of
BRCA1 and BRCA2 in ovarian cancer. J Natl Cancer Inst 2002;94:1396–1406.
51 Levran O, Attwooll C, Henry RT, Milton KL, Neveling K, et al: The BRCA1-interacting helicase
BRIP1 is deficient in Fanconi anemia. Nat Genet 2005;37:931–933.
52 Levitus M, Waisfisz Q, Godthelp BC, de Vries Y, Hussain S, et al: The DNA helicase BRIP1 is
defective in Fanconi anemia complementation group J. Nat Genet 2005;37:934–935.
53 Cantor SB, Bell DW, Ganesan S, Kass EM, Drapkin R, et al: BACH1, a novel helicase-like
protein, interacts directly with BRCA1 and contributes to its DNA repair function. Cell 2001;105:
149–160.

Neveling/Kalb/Schindler 76
54 Sigurdson AJ, Hauptmann M, Chatterjee N, Alexander BH, Doody MM, et al: Kin-cohort esti-
mates for familial breast cancer risk in relation to variants in DNA base excision repair, BRCA1
interacting and growth factor genes. BMC Cancer 2004;4:9.
55 Luo L, Lei H, Du Q, von Wachenfeldt A, Kockum I, et al: No mutations in the BACH1 gene in
BRCA1 and BRCA2 negative breast-cancer families linked to 17q22. Int J Cancer 2002;98:638–639.
56 Lewis AG, Flanagan J, Marsh A, Pupo GM, Mann G, et al: Mutation analysis of FANCD2,
BRIP1/BACH1, LMO4 and SFN in familial breast cancer. Breast Cancer Res 2005;7:R1005–R1016.
57 Vahteristo P, Yliannala K, Tamminen A, Eerola H, Blomqvist C, Nevanlinna H: BACH1 Ser919Pro
variant and breast cancer risk. BMC Cancer 2006;6:19.
58 Seal S, Barfoot R, Jayatilake H, Smith P, Renwick A, et al: Evaluation of Fanconi Anemia genes in
familial breast cancer predisposition. Cancer Res 2003;63:8596–8599.
59 Thompson E, Dragovic RL, Stephenson SA, Eccles DM, Campbell IG, Dobrovic A: A novel
duplication polymorphism in the FANCA promoter and its association with breast and ovarian can-
cer. BMC Cancer 2005;5:43.
60 Barroso E, Milne RL, Fernandez LP, Zamora P, Arias JI, et al: FANCD2 associated with sporadic
breast cancer risk. Carcinogenesis 2006;27:1930–1937.
61 Hu N, Li G, Li WJ, Wang C, Goldstein AM, et al: Infrequent mutation in the BRCA2 gene in
esophageal squamous cell carcinoma. Clin Cancer Res 2002;8:1121–1126.
62 Sparano A, Quesnelle KM, Kumar MS, Wang Y, Sylvester AJ, et al: Genome-wide profiling of
oral squamous cell carcinoma by array-based comparative genomic hybridization. Laryngoscope
2006;116:735–741.
63 de Winter JP, Rooimans MA, van Der Weel L, van Berkel CG, Alon N, et al: The Fanconi anaemia
gene FANCF encodes a novel protein with homology to ROM. Nat Genet 2000;24:15–16.
64 Leveille F, Blom E, Medhurst AL, Bier P, Laghmani el H, et al: The Fanconi anemia gene product
FANCF is a flexible adaptor protein. J Biol Chem 2004;279:39421–39430.
65 Taniguchi T, Tischkowitz M, Ameziane N, Hodgson SV, Mathew CG, et al: Disruption of the Fanconi
anemia-BRCA pathway in cisplatin-sensitive ovarian tumors. Nat Med 2003;9: 568–574.
66 Agrelo R, Cheng WH, Setien F, Ropero S, Espada J, et al: Epigenetic inactivation of the premature
aging Werner syndrome gene in human cancer. Proc Natl Acad Sci USA 2006;103: 8822–8827.
67 de Bustros A, Nelkin BD, Silverman A, Ehrlich G, Poiesz B, Baylin SB: The short arm of chro-
mosome 11 is a ‘hot spot’ for hypermethylation in human neoplasia. Proc Natl Acad Sci USA
1988;85:5693–5697.
68 Hoffmann MJ, Florl AR, Seifert HH, Schulz WA: Multiple mechanisms downregulate CDKN1C
in human bladder cancer. Int J Cancer 2005;114:406–413.
69 Tischkowitz M, Dokal I: Fanconi anaemia and leukaemia - clinical and molecular aspects. Br
J Haematol 2004;126:176–191.
70 Van der Heijden MS, Brody JR, Elghalbzouri-Maghrani E, Zdzienicka MZ, Kern SE: Does
tumorigenesis select for or against mutations of the DNA repair-associated genes BRCA2 and
MRE11?: considerations from somatic mutations in microsatellite unstable (MSI) gastrointestinal
cancers. BMC Genet 2006;7:3.
71 Tischkowitz M, Ameziane N, Waisfisz Q, De Winter JP, Harris R, et al: Bi-allelic silencing of the
Fanconi anaemia gene FANCF in acute myeloid leukaemia. Br J Haematol 2003;123:469–471.
72 Van Der Heijden MS, Brody JR, Kern SE: Functional screen of the Fanconi anemia pathway in
cancer cells by Fancd2 immunoblot. Cancer Biol Ther 2004;3:534–537.
73 Meyer S, White DJ, Will AM, Eden T, Sim A, et al: No evidence of significant silencing of Fanconi
genes FANCF and FANCB or Nijmegen breakage syndrome gene NBS1 by DNA hyper-methylation
in sporadic childhood leukaemia. Br J Haematol 2006;134:61–63.
74 Marsit CJ, Liu M, Nelson HH, Posner M, Suzuki M, Kelsey KT: Inactivation of the Fanconi
anemia/BRCA pathway in lung and oral cancers: implications for treatment and survival.
Oncogene 2004;23:1000–1004.
75 Olopade OI, Wei M: FANCF methylation contributes to chemoselectivity in ovarian cancer.
Cancer Cell 2003;3:417–420.
76 Narayan G, Arias-Pulido H, Nandula SV, Basso K, Sugirtharaj DD, et al: Promoter hypermethyla-
tion of FANCF: disruption of Fanconi Anemia-BRCA pathway in cervical cancer. Cancer Res
2004;64:2994–2997.

Cancer in Fanconi Anemia and Fanconi Anemia Genes in Cancer 77

77 Snyder ER, Ricker JL, Chen Z, Waes CV: Variation in cisplatinum sensitivity is not associated
with Fanconi Anemia/BRCA pathway inactivation in head and neck squamous cell carcinoma cell
lines. Cancer Lett 2006;245:75–80.
78 Dhillon VS, Shahid M, Husain SA: CpG methylation of the FHIT, FANCF, cyclin-D2, BRCA2 and
RUNX3 genes in granulosa cell tumors (GCTs) of ovarian origin. Mol Cancer 2004;3:33.
79 Teodoridis JM, Hall J, Marsh S, Kannall HD, Smyth C, et al: CpG island methylation of DNA
damage response genes in advanced ovarian cancer. Cancer Res 2005;65:8961–8967.
80 Wang Z, Li M, Lu S, Zhang Y, Wang H: Promoter hypermethylation of FANCF plays an important
role in the occurrence of ovarian cancer through disrupting Fanconi Anemia-BRCA pathway.
Cancer Biol Ther 2006;5:256–260.
81 Koul S, McKiernan JM, Narayan G, Houldsworth J, Bacik J, et al: Role of promoter hypermethyl-
ation in cisplatin treatment response of male germ cell tumors. Mol Cancer 2004;3:16.
82 Guardiola P, Socié G, Li X, Ribaus P, Devergie A, et al: Acute graft-versus-host disease in patients
with Fanconi anemia or aquired aplastic anemia undergoing bone marrow transplantation from
HLA-identical sibling donors: risk factors and influence on outcome. Blood 2004;103:73–77.
83 Seal S, Thompson D, Renwick D, Elliott A, Kelly P, et al: Truncating mutations in the Fanconi
anemia J gene BRIP1 are low-penetrance breast cancer susceptibility alleles. Nat Genet 2006;38:
1239–1241.
84 Rahman N, Seil S, Thompson D, Kelly P, Renwick A, et al: PALB2, which encodes a BRCA2-
interacting protein, is a breast cancer susceptibility gene. Nat Genet 2007;39:165–167.
85 Erkko H, Xia B, Nikkila J, Schleutker J, Syrjakoski K, et al: A recurrent mutation in PALB2 in
Finnish cancer families. Nature 2007;446:316–319.
86 Neveling K, Kalb R, Florl AR, Herterich S, Friedl R, et al: Disruption of the FA/BRCA pathway in
bladder cancer. Cytogenet Genome Res 2007, in press.
87 Van der Groep P, Hoelzel M, Buerger H, Joenje H, de Winter JP, et al: Loss of expression of
FANCD2 protein in sporadic and hereditary breast cancer. Breast Cancer Res Treat 2007; Epub
ahead of print.

Kornelia Neveling
Department of Human Genetics, University of Würzburg, Biocenter
B107, Am Hubland, 97074 Würzburg (Germany)
Tel. ⫹49 931 888 4089, Fax ⫹49 931 888 4069
E-Mail [email protected]

Neveling/Kalb/Schindler 78
 Schindler D, Hoehn H (eds): Fanconi Anemia. A Paradigmatic Disease for the Understanding of
Cancer and Aging. Monogr Hum Genet. Basel, Karger, 2007, vol 15, pp 79–94

Clonal Chromosomal Aberrations in Bone

Marrow Cells of Fanconi Anemia Patients:
Results and Implications
H. Neitzela, J.-S. Kühl b, A. Gerlacha, W. Ebell b, H. Tönniesa
a
Institute of Human Genetics and the bDepartment of General Pediatrics, Bone
Marrow Transplant Unit, Charité – Universitätsmedizin Berlin, Berlin, Germany

Abstract
Fanconi anemia patients have a high risk for bone marrow failure, aplastic anemia,
myelodysplastic syndrome, and acute myeloid leukemia. Many FA patients acquire chromo-
somal aberrations in their bone marrow (BM) cells. The significance and predictive value of
these somatic aberrations for hematopoietic function and malignant progress are not fully
understood. Therefore, we initiated in cooperation with the ‘Deutsche Fanconi-Anämie-Hilfe
e.V.’ a prospective cytogenetic follow-up in BM cells of FA patients utilizing systematically
molecular cytogenetic techniques. The most frequent clonal aberrations are gains of material
of the long arm of chromosome 3, loss of most of the long arm of chromosome 7 or loss of
one entire copy of chromosome 7, and gains of the long arm of chromosome 1. The available
data suggest that these clonal chromosome aberrations in bone marrow cells of FA patients
represent an important step in the initiation of MDS and AML. Our data of patients with 3q
aberrations revealed that gains of 3q are an adverse risk factor. The overall survival in the 3q
group was extremely poor compared to FA patients without such aberrations. None of the FA
patients with 3q gains survived without undergoing HSCT. Because of the high MDS/AML
risk and the significantly higher mortality in the group of FA patients with 3q aberrations, we
recommend a systematic evaluation of all FA patients by molecular cytogenetics in order to
detect aberrations, including subtle aberrations, as early as possible. Such clonal aberrations
are a very strong clinical warning sign. All available evidence suggests that the finding of
any chromosomal abnormality in bone marrow cells, especially abnormalities involving
chromosomes 3 and 7, warrant very close clinical follow-up, as they may signal the develop-
ment of MDS or AML.
Copyright © 2007 S. Karger AG, Basel

Fanconi anemia patients have a high risk for bone marrow failure, aplastic
anemia (AA), myelodysplastic syndrome (MDS), acute myeloid leukemia
(AML), and, later in life, for epithelial malignancies. The most life-threatening
early event in most complementation groups is bone marrow failure (patients
with bi-allelic BRCA2 mutations, who seem to have early onset solid tumors,
may be exceptions). The hematologic complications of FA have been exten-
sively reviewed [1]. Patients with FA develop bone marrow (BM) failure, typi-
cally during the first decade of life. The actuarial risk of developing BM failure
is 90% by 40 years of age. The actuarial risk of developing hematologic neo-
plasms is 33% by 40 years of age [2]. The median age of patients who develop
cancer is 14 years for acute myelogenous leukemia (AML). Many FA patients
acquire chromosomal aberrations in their bone marrow cells. The significance
and the predictive value of such clonal alterations with respect to hematopoietic
function and malignant progress are not fully understood but are important to
define. This is the aim of our studies.

Cytogenetic Studies in Bone Marrow Cells of

Fanconi Anemia Patients

The first cytogenetic studies of bone marrow cells in FA were published

more than 30 years ago, most of them as single case reports (table 1). A review
of 1991 [3] summarized these early reports and updated them with data from
the International Fanconi Anemia Registry (IFAR). Out of those patients who
had a cytogenetic analysis 26 of 32 FA patients with AML had clonal aberra-
tions detectable by conventional cytogenetics and all patients with MDS
(n ⫽ 5) and AA (n ⫽ 13) presented with aberrations. The most frequent aberra-
tions detected were duplications of 1q, complete monosomy 7, deletions of 7q,
and deletions of 5q [3].
In a later study of MDS and AML cases in FA patients, a high incidence of
monosomy 7, deletions of 7q, and rearrangements involving 1q24–q34 were
reported [4] (table 1). The cytogenetic findings of further 20 patients with
Fanconi anemia were added in 1996 [5]. Out of these, seven patients had a mono-
somy 7 or a deletion in 7q. Alter et al. reported that approximately half of the
patients with clonal aberrations had chromosome 7 abnormalities, while a quar-
ter had chromosome 1 abnormalities [6]. Data of 16 FA patients with clonal aber-
rations and a cytogenetic follow up were presented in 2000 [7]. Out of these, two
had a monosomy 7, one a partial monosomy 7q, and four additional material of
1q. All studies performed until 2000 demonstrated that among the chromosomal
aberrations observed in bone marrow cells of FA patients the following stand out:
monosomy 7, deletions of 7q, gains of 1q, monosomy 5 or deletions of 5q.
Since the first publications of clonal aberrations in BM cells of FA patients
there is a dispute about the relevance of clonal aberrations in BM cells of FA

Neitzel/Kühl/Gerlach/Ebell/Tönnies 80
 Table 1. Overview of clonal chromosomal aberrations in bone marrow cells of Fanconi anemia patients
reported in the literature

Hematological statusa Clonal aberration Reference

Non-malignant 46,XX,Dq⫹ [17]

Non-malignant 47,XX,⫹21 [12]
AML-M4 45,XY,t(5;18)(q?;q?) [18]
Preleukemia 46,XX,dup(3)(q12q27)/add(12)(q24) [12]
AML 46,XX,add(1)(p36)dup(3)(q12q27),add(7)(p22),add(12)(q24) [12]
AML der(14)t(1;14)(q12 or q21;q23),4q⫺,add(6)(q27),7q⫺,add(13)(q34),20q⫺ [19]
Acute leukemia 46,XY,der(6)t(1;6)(q32;p25)/46,idem,der(11)t(11;?)(q25;?) [20]
Acute leukemia der(6)t(1;6)(q22 or 23;pter)/idem,5q⫺ [19]
Erythroleukemia 46,XY,dup(1)(p32p34),trp(3)(q12q27),del(7)(q11) [11]
Preleukemia 46,XX,dup(2)(q24q36)/47,idem,⫹?4p [11]
AML 46,XX,⫺7,⫺8,⫹mar1,⫹mar2/complex ⫺7,⫹mar/complex mar [21]
Erythroleukemia add(1)(p36),t(13;15) [22]
AML 45,XY,⫺7 [23]
Preleukemia 46,XX,t(10;?)(q26;?),del(12)(p11),⫺17,der(17)t(1;17)(q21;p13) [24]
AML 46,XX,7q⫺,9q⫺ [25]
Non-malignant 46,XX,i(7)(q10) [26]
Preleukemia-AML 46,XY,dup(1)(q24q44),del(6)(p15),del(7)(q11) [27]
Non-malignant 46,XY,trp(1)(q12q32) [28]
MDS 46,XY,der(6)t(1;6)(q12;p25)/46,idem,del(5)(q21q23)/46,t(1;2)(q12;q37),del(5) [29]
AML 47,XY,⫹20,⫺22,dup(1)(q22q34),del(22)(q11) [30]
Preleukemia der(2)t(2;3)(q35?;p21?),der(6)t(1;6)(q22;p25) [31]
MDS 47,XY,⫹8 [32]
AML 46,XX,der(1)t(1;?)(p36;?) [33]
MDS 46,X,⫺Y,⫹der(Y;1)(q12;q21) [34]
AML-M4 del(7)(p11),6p⫹,1q⫹ [3]
AML-M6 18q⫹,dup(12)(q),5q⫹,⫺16 [3]
NS del(7)(q) [3]
NS ⫺7 [3]
NS 2q⫹,6p⫹ [3]
MDS 2q⫹,⫹mar [3]
MDS t(1;5)(p34;q31) [3]
AA ⫺13,t(1q;13q),del(7)(q32) [3]
AA dup(1)(q24q32),t(17;?)(p12;?) [3]
AML 45,XY,⫺7 [35]
AML 47,XY,trp(1)(q32q44),⫹mar [35]
AML 46,XY,add(1)(q34),del(7)(q13) [35]
Non-malignant 46,XX,⫺5,⫹8 [35]
Non-malignant 46,XX,⫹5,⫺21 [35]
MDS 46,XY/47,XY,⫹21 [4]
AML 46,XX,⫺21,t(3;11)(q27;q21)/46,XX,t(3;11;21)(q27;q21;p11) [4]
MDS 45,XY,⫺7 [4]

Clonal Chromosomal Aberrations in Bone Marrow Cells of Fanconi Anemia Patients 81

 Table 1. (continued)

Hematological statusa Clonal aberration Reference

BMF 46,XY,del(7)(q22q36) [4]

MDS 46,XY,del(7)(q22q36)/46,XY,idem,del(6)(p23) [4]
BMF 46,XX,t(1;?)(p36;?)/46,XX [4]
MDS 46,XX,t(1;?)(p36;?)/46,XX,t(1;6)(q21;p21) [4]
NS add(6)(p22),t(1;17)(q21;p11) [4]
NS add(1)(p36),t(1;6)(q21;p21) [4]
NS del(6)(p21q25),dup(12)(q21q24) [4]
NS del(13)(q14q32) [4]
MDS 46,XX,add(1)(p34),del(5)(q31q35),inv(9)(p11q11)x2,add(16)(q24) [36]
NS 46,XY,der(2)t(1;2)(q12;q32) [36]
AML 47,XY,der(6)t(1;6)(q23;p22),⫹mar [37]
MDS der(1)dup(1)(q12 or q21q24) [6]
AML der(18)t(1;18)(q12;p11),del(12)(p12.1) [6]
MDS del(3)(q22q24) [6]
MDS dup(1)(q23q44) [6]
AML 47,XY,der(6)t(1;6)(q23;p22),⫹mar [38]
AML 46,XX,trp(1)(q23q42)/46,XX,idem,⫹13,⫺20 [39]
NS 46,XX,del(13)(q12q14)/46,XX [7]
MDS 46,XX,add(21)(q22)/46,idem,del(13)(q21q22) [7]
MDS 46,XX,add(1)(p11),add(2)(q33),add(6)(p11)/46,XX [7]
MDS 46,XY,der(11)t(1;11)(q23;q23)/46,idem,del(6)(p21)/46,XY [7]
AML 46,XX,t(q?;q?) [7]
AML 46,XX,add(14)(p11.2)/46,XX [7]
MDS 46,X,der(X)t(X;3)(p22.2;q13),⫹3 [7]
MDS 46,XX,del(7)(q31.2)/48,idem,add(1)(p36.1)/46,XX [7]
MDS 47,XY,⫹i(1)(q10)/46,XY [7]
MDS 47,XY,⫹i(1)(q10)/46,XY [7]
normal 49,XXX,⫹8,⫹21 [7]
AML 45,XY,⫺7/46,XY [7]
MDS 46,XX,dup(1)(q21q42)/46,XX,del(7)(q31)/ [7]
46,XX,del(11)(q21q25)/46,XX,der(20)t(1;20)(q10q13.3)
AML 46,XY,del(1)(q32) [40]

a
AML: Acute myeloid leukemia; BMF: bone marrow failure; NS: not specified; MDS: myelodysplastic syndrome.

patients. Auerbach and Allen mentioned that ‘considerable evidence shows that
the presence of chromosomally abnormal clones in the BM heightens the risk for
development of leukemia’ [3]. This is in agreement with Butturini et al. [8] who
demonstrated that the detection of a clonal cytogenetic abnormality correlates
with decreased survival of FA patients. In 1993, a provocative question was

Neitzel/Kühl/Gerlach/Ebell/Tönnies 82
formulated [9]: ‘Clonal chromosomal abnormalities in Fanconi’s anemia: what
do they really mean?’ Alter et al. [9] presented three of 11 FA patients with
clonal abnormalities after adequate cytogenetic analysis. The fact that some of
the clonal aberrations displayed transient appearance led those authors to reason
that ‘changing cytogenetic patterns may not be related to leukemic evolution in
patients with a DNA repair defect’. In contrast, in 2000 it was demonstrated that
the five-year probability of survival after detection of a clone is 0.40 compared
to 0.94 without a clone [7]. Notwithstanding these observations, those authors
stated that the morphological changes in the direction of MDS may be more
important than clonal aberrations with respect to the prediction of an adverse
outcome. Along these lines, Maarek et al. emphasized: ‘The significance of
chromosomal changes without apparent evolution to overt AML or MDS will
only be understood when more FA patients are cytogenetically studied and fol-
lowed over a sufficient period of time’ [5]. However, up to now long term follow-up
data are sparse and there is urgent need to collect such data.
All cytogenetic studies in the 70s to 90s have in common that the chromo-
somal aberrations could not be sufficiently defined by conventional cytogenet-
ics alone in a considerable number of patients. Karyotype descriptions denoted
as ‘⫹’, ‘add’, or ‘mar’ are common, indicating that additional material of
unknown origin was detected either as translocation to another chromosome
(⫹; add) or as marker chromosomes (⫹mar) of unidentified origin (table 1).
The results from these early cytogenetic studies in bone marrow cells of
FA patients illustrate the problem of all studies which were based solely on con-
ventional cytogenetics. There is a limited number of analyzable BM meta-
phases, especially in patients with progressive bone marrow failure, and there is
a limited resolution of GTG-banded chromosomes of BM metaphases resulting
in the inability to identify and assign chromosomal changes correctly, espe-
cially when these are subtle and/or mosaic. However, the correct assignment of
the karyotypic changes in BM cells is a prerequisite for any predication with
respect to hematopoietic function and progression to malignancy.

A Prospective Study of Bone Marrow Cells of FA Patients with

Systematic Application of Molecular Cytogenetic Techniques

The ongoing debate about the relevance of clonal aberrations in patients

with Fanconi anemia prompted us to initiate a systematic, prospective cytoge-
netic follow-up in BM cells of FA patients in Germany starting in 1996. The
study was supported by the German Patient Organisation ‘Deutsche Fanconi-
Anämie-Hilfe e.V.’. In our study, we utilized for the first time comparative
genomic hybridization (CGH) and fluorescence in situ hybridization (FISH) in

Clonal Chromosomal Aberrations in Bone Marrow Cells of Fanconi Anemia Patients 83

 Table 2. Characteristics and outcome of FA patients with or without trisomies or tetrasomies of chromoso-
mal segment 3q26–3q29 as published 2003 by Tonnies et al. [10]. FANCA indicates Fanconi anemia complemen-
tation group A; FANCC, Fanconi anemia complementation group C; FANCG, Fanconi anemia complementation
group G; MDS, myelodysplastic syndrome; AML, acute myeloid leukemia; and HSCT, hematopoietic stem cell
transplantation

Total With chromosome Without chromosome pd

3q gain 3q gain

Patientsa,b,c 53 18 35
Age
Median (months) 141 149 125
Range (months) 34–463 95–463 34–442
Gender
Male 28 11 17
Female 25 7 18
Complementation group
FANCA 28 10 18
FANCC 4 4 0
FANCG 8 3 5
unknown 13 1 12
genotypic reversiona 3 1 2
Outcome
MDSb 9 9 0 ⬍0.001
AMLc 5 4 1
MDS⫹AML 14 13 1 ⬍0.001
Alive 44 11 33 0.005
HSCT 20 12 8

a
Spontaneous reversion of the cellular FA phenotype in lymphocytes by back mutation or gene conversion in
compound heterozygotes.
b
MDS was defined according to the FAB classification; 8/9 MDS patients presented 3–26% blasts in the marrow.
c
AML was defined by more than 30% blasts in the marrow.
d
Fisher’s exact test (2-sided).

a systematic fashion. The investigation of all BM samples by CGH allows the

comprehensive analysis of the entire genome in just one experiment. Our
approach provides information not only about the chromosomal assignment but
also about the size of the chromosomal imbalances. All CGH results were vali-
dated by FISH with specific probes hybridized onto BM metaphases. The
results of this very first molecular cytogenetic study of clonal aberrations were
published in 2003 [10] yielding an unexpected finding: a high prevalence of
gains for chromosome 3q (table 2).

Neitzel/Kühl/Gerlach/Ebell/Tönnies 84
 wcp3 wcp10
3 10 3/12
a b c

Fig. 1. Conventional and molecular cytogenetic analysis in bone marrow cells of an FA

patient with a gain of 3q. (a) GTG-banding of chromosomes 3 and 10; the conventional
analysis revealed additional material on the short arm of one chromosome 10. (b) CGH indi-
cated a gain of material of 3q. (c) The result of the CGH was confirmed with whole chromo-
some paints for chromosome 3 (green) and chromosome 10 (red) indicating that additional
material of chromosome 3 was translocated to 10p.

Out of 53 FA patients, 28 had a normal karyotype in BM cells after con-

ventional cytogenetics and CGH, while 25 patients had clonal aberrations. Out
of these, 18 (72%) revealed partial trisomies or tetrasomies of the long arm of
chromosome 3 indicating an extremely high prevalence of 3q aberrations in our
FA cohort. The 3q aberrations of almost all FA patients were mosaics with sub-
tle aberrations due to unbalanced translocations of distal 3q to various other
chromosomes. Prior to our study, only very few other FA patients with addi-
tional material of chromosome 3q in BM cells have been described in the liter-
ature, and all had trisomies or tetrasomies for almost the entire long arm from
3q13 to 3qter permitting the identification by conventional cytogenetics [6,
11–13]. Considering the prevalence of 3q gains in our FA patient cohort com-
pared to the very low occurrence in the published cases, we assumed that other
studies might have failed to detect the more subtle partial trisomies and tetra-
somies of 3q. An example of such a subtle chromosomal change is shown in
figure 1 demonstrating that a translocation of additional material to the short
arm of chromosome 10 can be recognized by conventional cytogenetics.
However, the identification of the additional material is extremely difficult by
conventional cytogenetics alone. Using CGH the additional material could be
unambiguously assigned to 3q. Subsequent fluorescence in situ hybridization
using whole chromosome paints (wcp) for chromosomes 3 and 10 confirmed
this interpretation.
Evidently, the interpretation of chromosomal aberrations in bone marrow
cells of FA patients may be misleading when recognition is based solely on con-
ventional cytogenetics. Such a case is illustrated in figure 2. Conventional cyto-
genetic analysis of an FA patient with progressive bone marrow failure yielded
poor banding resolution and only few metaphases were available for analysis
which showed a deletion of 6q and additional material at 2q. Thus, based on

Clonal Chromosomal Aberrations in Bone Marrow Cells of Fanconi Anemia Patients 85

 a 2 3 6 c wcp3 wcp6

wcp3 wcp6
d
b 2/11 3/9 6/12

Fig. 2. Conventional and molecular cytogenetic analysis in bone marrow cells of

another FA patient with a gain of 3q. (a) GTG-banding of chromosomes 2, 3 and 6; from
conventional cytogenetics the interpretation of the karyotype is t(2;6)(q37;q23) indicating a
balanced translocation of 6q material to 2q. (b) CGH showed a more complex situation
revealing a deletion of 2q37, additional material of 3q25–qter and a deletion of 6q23–qter.
(c) FISH for chromosome 3 and 6 displayed that in fact 6q is translocated to 2q but only in
some of the cells while others had two normal chromosomes. Additional 3q material is
translocated to the long arm of chromosome 6 leading to a partial trisomy of 3q.

conventional cytogenetics, the interpretation of the karyotype would be t(2;6)

(q37;q23) indicating a balanced translocation of 6q material to 2q. However, the
application of CGH showed a more complex situation revealing a deletion
of 2q37, additional material of 3q25–qter, and a deletion of 6q23–qter.
Subsequent FISH analysis using whole chromosome paints of chromosomes 3
and 6 revealed that in fact 6q was translocated to 2q, but only in some of the cells
while others had two normal chromosomes 2. Furthermore and most impor-
tantly, additional 3q material was translocated to the q-arm of chromosome
6 resulting in partial trisomy 3q. The combined data of the CGH and FISH pro-
cedures imply that the first step of the aberration was a translocation of 6q to 2q
that was accompanied by a deletion of 2qter, and an unbalanced translocation of
3q to 6q with a gain of 3q. As a second event, the translocation chromosome
t(2;6) was lost and the normal chromosome 2 was duplicated resulting in a uni-
parental disomy (UPD) for chromosome 2.
In this context, we should mention that another case of UPD was observed in
bone marrow cells of another patient who had an unbalanced translocation
der(19)t(3;19)(q23;p13.3) (fig. 3). During subsequent clonal evolution, the deriv-
ative of chromosome 19 was duplicated and the normal chromosome 19 was lost.
There is at least one further report in the literature describing a duplication of an

Neitzel/Kühl/Gerlach/Ebell/Tönnies 86
 46,XX
90 46,XX,dup(1q)
46,XX,⫹3q
80
46,XX,⫹3q,⫹3q
Percentage of aberrant cells

70 46,XX,⫹3q,⫹3q,delXp

60

50

40

30

20

10

0
T0 ⫹7 m ⫹10m ⫹15m ⫹46m ⫹50m 19 3/10
(n ⫽56) (n ⫽ 50) (n ⫽51) (n⫽33) (n ⫽50) (n⫽41)
a Interval of the cytogenetic analyses b c

Fig. 3. Increase of a clonal 3q gain over time as determined by conventional cytoge-

netics. (a) The conventional cytogenetic analyses revealed first an outgrowth of a clone with
a trisomy of 3q. After 7 months a tetrasomic clone appeared due to a duplication of the aber-
rant chromosome 19 which carries 3q material on the short arm and the loss of the normal
chromosome 19 as depicted in GTG-banding (b) of the derivative chromosomes 19. (c) CGH
indicates a gain of 3q. In this patient, the tetrasomic clone replaced the trisomic clone over
time indicating either a higher proliferative advantage or an increased survival of bone mar-
row cells with the aberration.

aberrant derivative chromosome [47,XX,add(13)(q34), ⫹add(13) (q34)] [5].

This suggests that the formation of UPD might not be such a rare event in BM
cells of FA patients.
In 16 out of 18 FA patients with 3q gains, serial bone marrow analyses
were performed, amounting to an average of 5.6 analyses per patient. By con-
ventional cytogenetics, we observed in all of these patients a considerable
expansion over time of the clone with additional 3q material (figs. 3 and 4).
Furthermore, three patients developed a tetrasomic 3q clone, presumably
derived from the trisomic clone by a subsequent duplication event or by UPD.
In this patient group, the tetrasomic clones gradually replaced the trisomic
clones. Thus, our data strongly suggest that gains of 3q confer either a prolifer-
ative advantage or an increased survival to bone marrow cells. This notion is
supported by the fact that there were no transient 3q gains as described for other
clonal aberrations in FA patients.
Out of 18 patients eight had an additional monosomy 7. In two of these
patients the clone with the 3q gain and the monosomy 7 was already present at

Clonal Chromosomal Aberrations in Bone Marrow Cells of Fanconi Anemia Patients 87

 100

Percentage of aberrant cells 90

80

70

60 46,XY
45,X
50
45,X,⫹3q
40

30

20

10

0
T0 ⫹12m ⫹36m ⫹45m ⫹50m ⫹52m
(n⫽ 56) (n ⫽ 20) (n ⫽40) (n⫽8) (n⫽8) (n⫽55)
Time scale of cytogenetic analyses

Fig. 4. Another example for the outgrowth of cells with a clonal 3q aberration. The
first aberrant clone was detected 3 years after the first cytogenetic analysis of bone marrow
cells in this FA patient. At this time, more than 70% of the cells revealed the karyotype 45,X
indicating a loss of the Y chromosome. Only 9 months later, all cells carried additional mate-
rial of 3q. The analyses 5 and 7 months later indicated that the clone persisted.

the time of the first BM analysis while in the other 6 patients the monosomy 7
had developed in the 3q aberrant clone as a secondary event. Even though the
number of patients affected with both, a gain of 3q and monosomy 7, is fairly
small, our data imply that gains of 3q might increase the risk for the subsequent
development of monosomy 7.
In order to determine the percentage of aberrant cells in non-dividing BM
and peripheral blood mononuclear cells (PBMC), we performed interphase FISH
analyses using a YAC from 3q27–28. Our results demonstrate that FA patients
with 3q aberrations in up to 70–80% of their BM and PBMC cells have more
than two signals as compared to a maximum of 2–3% in normal control sub-
jects. Since there was a normal karyotype in their T- and B-lymphocytes, our
results suggest that almost all circulating granulocytes must carry the 3q aber-
ration in these patients. In three of the FA patients the 3q aberration was
detected only by screening PBMC by interphase FISH prior to conventional
cytogenetic analysis of BM cells. This proves the high sensitivity and speci-
ficity of interphase FISH as a screening tool for aberrant clones in PBMC (for
details regarding interphase FISH screening of PBMC see the chapter by
Tönnies et al.).

Neitzel/Kühl/Gerlach/Ebell/Tönnies 88
 1.0

0.8
Cum risk of MDS/AML

0.6

0.4

0.2

0
0 100 200 300 400 500
Age (months)

Fig. 5. Risk of developing MDS or AML in Fanconi anemia with (—) or without ( )
chromosome 3 aberration (Log Rank: p ⬍ 0.001) as published 2003 by Tonnies et al. [10].
(—) n ⫽ 18, MDS/AML 13, risk 0.90 ⫾ 0.09. ( ) n ⫽ 35, MDS/AML 1, risk 0.08 ⫾ 0.08.

Clonal Chromosomal Aberrations in Bone Marrow

Cells and Clinical Course

In order to determine the clinical relevance of our findings, we compared

cytogenetic data, morphologic features of bone marrows, and the clinical course
of 18 FA patients carrying chromosome 3 aberrations to 35 FA patients without
any evidence of clonal aberrations involving 3q. Both groups did not differ sig-
nificantly with respect to age, gender, complementation group or genotypic
reversion. There was a slight preponderance of males and of individuals belong-
ing to complementation group FA-C in the cohort with 3q aberrations (table 1).
Despite the fact that more hematopoietic stem cell transplants (HSCT) had been
performed in this cohort there was a significant survival advantage of patients
without abnormalities of chromosome 3q. Even more pronounced was the
increased risk of patients with gains of 3q material with respect to the develop-
ment of morphologic MDS and AML (table 2; fig. 5). Thus, our data of 18
patients with 3q aberrations revealed that gains of 3q are strongly associated
with a poor prognosis and represent an adverse risk factor in FA. An actual
update of our data is presented in table 3. Another five patients have been iden-
tified with 3q gains. The overall survival in the 3q group is extremely poor

Clonal Chromosomal Aberrations in Bone Marrow Cells of Fanconi Anemia Patients 89

 Table 3. Update of the data of table 2 in 2006

With chromosome 3q gain Without chromosome 3q gain

Patients 23 35
Male 16 17
Female 7 18
Overall survival 8/23 (35%) 27/35 (77%)
HSCT 17/23 (74%) 10/35 (29%)
Survival after HSCT 7/17 (41%) 7/10 (70%)
Survival without HSCT 0/5 (0%) 20/25 (80%)

Table 4. Correlation of clonal chromosomal aberrations and hematological status

(MDS/AML) as reported by Tonnies et al. [10] and by Hirsch et al. [14]

Karyotype Mean age (year) MDS/AML

(%) (%)

Patients (n ⫽ 53) [10] Normal 66 10.4 3

Aberrant 34 12.4 72
Patients (n ⫽ 99) [14] Normal 56 9.2 6
Aberrant 44 14.2 66

compared to FA patients without such aberrations. None of the FA patients with

3q gains (n ⫽ 5) survived without undergoing HSCT. Furthermore, there is a
clear prevalence of males in the cohort carrying 3q aberrations.
Our data of the high prevalence of 3q gains in FA patients were confirmed
at the 15th Annual Fanconi Anemia Research Scientific Symposium by the group
of Hirsch et al. [14] from the University of Minnesota School of Medicine
(table 4). The data of Hirsch et al. [14] were based on the cytogenetic investiga-
tion of 99 FA patients. Fifty-six percent of the patients had normal chromosome
findings, 44% had one or more clonal abnormalities. The most frequently seen
clonal aberrations were gains of material of the long arm of chromosome 3, loss
of most of the long arm of chromosome 7 or loss of one entire copy of chromo-
some 7, and gain of material of the long arm of chromosome 1. Like in our data
set there was a significant correlation between the degree of hematopathology
and the presence of clonal abnormalities. Only 6% of patients with normal
cytogenetics had MDS or AML, while two-thirds of patients with clonal aberra-
tions had apparent MDS or AML.

Neitzel/Kühl/Gerlach/Ebell/Tönnies 90
 At present, the data of Hirsch et al. [14] and our investigations are the only
studies utilizing molecular cytogenetic techniques for the investigation of
altogether more than 150 FA patients. Both groups report a high incidence of
3q gains which clearly stands out as the most frequent clonal aberration. The
second frequent clonal aberration in FA is monosomy 7 (partial 7q or complete)
followed by 1q gains.
The high frequency of monosomy 7 and the occasional occurrence of
monosomy 5 have led to the suggestion that AML in FA is similar to secondary
AML which is induced by alkylating agents in non-FA patients [15]. Various
alkylating agents are carcinogenic and augment the incidence of cancer in the
general population, and the incidence of AML is dramatically increased in can-
cer patients secondary to chemotherapy with alkylating agents. These patients
often develop nonrandom clonal chromosomal changes in the preleukemic
phase, especially monosomy 7, deletion of 7q, and monosomy 5. There is little
doubt that the presence of such chromosomally abnormal BM clones heightens
the risk for the development of leukemia in non-FA patients with secondary
AML. It therefore appears unlikely that these clonal aberrations should be pre-
dictors for the leukemic process in secondary AML of non-FA patients, but only
innocent bystanders in the malignant progress in patients with DNA repair defi-
ciencies. Thus, the statement made by Alter et al. in 1993 that ‘changing cyto-
genetic patterns may not be related to leukemic evolution in patients with a
DNA repair defect’ may no longer be valid [9].
While monosomy 5 and monosomy 7 are aberrations found in non-FA
patients with secondary myelodysplastic syndrome (MDS), gains of 3q have
not been described in secondary MDS. In AML of adults, 3q abnormalities are
seen in only 2% of patients and they are thought to be extremely rare in primary
childhood AML of non-FA patients. Common chromosomal aberrations in
childhood AML of non-FA patients include a number of balanced transloca-
tions like t(8;21), t(9;11), inv(16), t(11;19). These aberrations have not been
described in FA patients, indicating that AML in FA patients may indeed be
similar to secondary AML. With regard to the obvious rarity of 3q aberrations
in primary childhood AML a recent publication is of special interest. Haltrich
et al. [16] investigated 28 childhood AML cases with FISH and found aberrations
of chromosome 3 in nine out of the 28 patients. Four patients had a trisomy 3,
one patient a tetrasomy of 3q26.3–q28, one patient loss at 3q13–q21, and three
patients had structural rearrangements involving the 3q26 and/or 3q21 site. All
pediatric AML patients with 3q aberrations had a poor outcome [16].
The data of Haltrich et al. [16] thus suggest that it might turn out that 3q
aberrations are risk factors for the development of primary AML in both FA and
non-FA patients. It is noteworthy that the 3q chromosomal region is also
involved in multiple solid tumors indicating that its tumor promoting potential

Clonal Chromosomal Aberrations in Bone Marrow Cells of Fanconi Anemia Patients 91

may not be cell line or tissue specific but may be generally involved in tumor
initiation and/or progression. Four genes known to be involved in MDS and/or
AML have been identified in the critical region 3q that are shared by our FA
patients: myelodysplasia-myeloid leukemia factor 1 (MLF1), myelodysplasia
syndrome-associated sequence 1 (MDS1), murine myeloid leukemia-associated
gene EVI1, and Epstein-Barr associated protein EAP. Obviously, further data
on the role of these candidate genes located at 3q26–q29 are needed to elucidate
the pathomechanisms of MDS and AML in FA- and non-FA patients.

Conclusions

Because of a high MDS/AML risk and a significantly higher mortality in

the group of FA patients with 3q aberrations, we recommend a systematic eval-
uation of all FA patients by molecular cytogenetics in order to detect aberra-
tions, including subtle aberrations, as early as possible. The available data
suggest that specific clonal chromosome aberrations in bone marrow cells of
FA patients represent an important step in the initiation of MDS and AML.
Consequently, these aberrations seem to be of high prognostic value and might
thus serve as important criteria for the initiation of therapeutic measures such as
HSCT. At the moment, they are a very strong clinical warning sign. All avail-
able evidence suggests that the finding of any chromosomal abnormality in
bone marrow cells, especially abnormalities involving chromosomes 3 and 7,
warrant very close clinical follow-up, as they may signal the development of
MDS or AML.

Acknowledgement

Supported by grants from the Deutsche Fanconi Anämie Hilfe e.V. and the Charité
Research Fund (No.2000-627).

References

1 Alter BP, Greene MH, Velazquez I, Rosenberg PS: Cancer in Fanconi anemia. Blood 2003;101:
2072.
2 Kutler DI, Singh B, Satagopan J, Batish SD, Berwick M, et al: A 20-year perspective on the
International Fanconi Anemia Registry (IFAR). Blood 2003;101:1249–1256.
3 Auerbach AD, Allen RG: Leukemia and preleukemia in Fanconi anemia patients. A review of the
literature and report of the International Fanconi Anemia Registry. Cancer Genet Cytogenet 1991;51:
1–12.
4 Butturini A, Gale RP: Long-term bone marrow culture in persons with Fanconi anemia and bone
marrow failure. Blood 1994;83:336–339.

Neitzel/Kühl/Gerlach/Ebell/Tönnies 92
 5 Maarek O, Jonveaux P, Le Coniat M, Derre J, Berger R: Fanconi anemia and bone marrow clonal
chromosome abnormalities. Leukemia 1996;10:1700–1704.
6 Alter BP: Fanconi’s anemia and malignancies. Am J Hematol 1996;53:99–110.
7 Alter BP, Caruso JP, Drachtman RA, Uchida T, Velagaleti GV, Elghetany MT: Fanconi anemia:
myelodysplasia as a predictor of outcome. Cancer Genet Cytogenet 2000;117:125–131.
8 Butturini A, Gale RP, Verlander PC, Adler-Brecher B, Gillio AP, Auerbach AD: Hematologic
abnormalities in Fanconi anemia: an International Fanconi Anemia Registry study. Blood 1994;84:
1650–1655.
9 Alter BP, Sealise A, McCombs J, Najfeld V: Clonal chromosomal abnormalities in Fanconi’s
anaemia: what do they really mean? Br J Haematol 1993;85:627–630.
10 Tonnies H, Huber S, Kuhl JS, Gerlach A, Ebell W, Neitzel H: Clonal chromosomal aberrations in
bone marrow cells of Fanconi anemia patients: gains of the chromosomal segment 3q26q29 as an
adverse risk factor. Blood 2003;101:3872–3874.
11 Berger R, Bernheim A, Le Coniat M, Vecchione D, Schaison G: Chromosomal studies of leukemic
and preleukemic Fanconi’s anemia patients: examples of acquired ‘chromosomal amplification’.
Hum Genet 1980;56:59–62.
12 Berger R, Bussel A, Schenmetzler C: Somatic segregation and Fanconi anemia. Clin Genet 1977;11:
409–412.
13 Schaison G, Leverger G, Yildiz C, Berger R, Bernheim A, Gluckman E: Fanconi’s anemia.
Incidence of its development into leukemia. Presse Med 1983;12:1269–1274.
14 Hirsch B, Dolan M, Brandt K, MacMillan ML, Wagner J, et al: Recurring clonal chromosomal
abnormalities in Fanconi Anemia: characterization and association with bone marrow pathology.
15th Annual Fanconi Anemia Research Scientific Symposium, 2004.
15 Bagby GC, Alter BP: Fanconi Anemia. Semin Hematol 2006;43:147–156.
16 Haltrich I, Kost-Alimova M, Kovacs G, Klein G, Fekete G, Imreh S: Multipoint interphase FISH
analysis of chromosome 3 abnormalities in 28 childhood AML patients. Eur J Haematol 2006;76:
124–133.
17 Lisker R, Cobo de Gutierrez A: Cytogenetic studies in Fanconi’s anemia. Description of a case
with bone marrow clonal evolution. Clin Genet 1974;5:72–76.
18 Bourgeois CA, Hill FG: Fanconi anemia leading to acute myelomonocytic leukemia: cytogenetic
studies. Cancer 1977;39:1163–1167.
19 Tanzer J, Frocrain C, Desmarest MC: Anomalies du chromosome 1 dans 3 cas de leucémie com-
pliquant une aplasie Fanconi (FA). Nouv Rev Fr Hématol 1980;22(suppl):XCIII.
20 Obeid DA, Hill FG, Harnden D, Mann JR, Wood BS: Fanconi anemia. Oxymetholone hepatic
tumors, and chromosome aberrations associated with leukemic transition. Cancer 1980;46:
1401–1404.
21 Auerbach AD, Weiner MA, Warburton D, Yeboa K, Lu L, Broxmeyer HE: Acute myeloid leukemia
as the first hematologic manifestation of Fanconi anemia. Am J Hematol 1982;12:289–300.
22 Rotzak R, Kaplinsky N, Chaki R, Blei F, Berkowitz M, et al: Giant marker chromosome in
Fanconi’s anemia transforming into erythroleukemia in an adult. Acta Haematol 1982;67:
214–216.
23 Stivrins TJ, Davis RB, Sanger W, Fritz J, Purtilo DT: Transformation of Fanconi’s anemia to
acute nonlymphocytic leukemia associated with emergence of monosomy 7. Blood 1984;64:
173–176.
24 Nowell P, Bergman G, Besa E, Wilmoth D, Emanuel B: Progressive preleukemia with a chromo-
somally abnormal clone in a kindred with the Estren-Dameshek variant of Fanconi’s anemia.
Blood 1984;64:1135–1138.
25 Woods WG, Nesbit ME, Buckley J, Lampkin BC, McCreadie S, et al: Correlation of chromosome
abnormalities with patient characteristics, histologic subtype, and induction success in children
with acute nonlymphocytic leukemia. J Clin Oncol 1985;3:3–11.
26 Huret JL, Benz E, Guilhot F, Brizard A, Tanzer J: Fluctuation of a clone 46,XX,i(7q) in bone mar-
row in a Fanconi anaemia. Hum Genet 1986;74:98–100.
27 Barton JC, Parmley RT, Carroll AJ, Huang ST, Goodnough LT, et al: Preleukemia in Fanconi’s
anemia: hematopoietic cell multinuclearity, membrane duplication, and dysgranulogenesis.
J Submicrosc Cytol 1987;19:355–364.

Clonal Chromosomal Aberrations in Bone Marrow Cells of Fanconi Anemia Patients 93

28 Schindler D, Kubbies M, Hoehn H, Schinzel A, Rabinovitch PS: Confirmation of Fanconi’s ane-
mia and detection of a chromosomal aberration (1q12–32 triplication) via BrdU/Hoechst flow
cytometry. Am J Pediatr Hematol Oncol 1987;9:172–177.
29 Huret JL, Tanzer J, Guilhot F, Frocrain-Herchkovitch C, Savage JR: Karyotype evolution in the
bone marrow of a patient with Fanconi anemia: breakpoints in clonal anomalies of this disease.
Cytogenet Cell Genet 1988;48:224–227.
30 Berger R, Bussel A, Schenmetzler C: Cytogenetic studies in four cases of Fanconi’s anemia. Nouv
Rev Fr Hematol 1975;15:539–550.
31 Baron F, Sybert VP, Andrews RG: Cutaneous and extracutaneous neutrophilic infiltrates (Sweet
syndrome) in three patients with Fanconi anemia. J Pediatr 1989;115:726–729.
32 Standen GR, Hughes IA, Geddes AD, Jones BM, Wardrop CA: Myelodysplastic syndrome with
trisomy 8 in an adolescent with Fanconi anaemia and selective IgA deficiency. Am J Hematol
1989;31:280–283.
33 Bessho F, Mizutani S, Hayashi Y, Moriwaki K, Yokota S, Inaba T: Chronic myelomonocytic
leukemia with chromosomal changes involving 1p36 and hepatocellular carcinoma in a case of
Fanconi’s anemia. Eur J Haematol 1989;42:492–495.
34 Thompson PW, Standen GR, Geddes AD: Transient t(Y;1)(q12;q21) in a patient with Fanconi ane-
mia and myelodysplastic syndrome. Cancer Genet Cytogenet 1991;52:201–202.
35 Berger R, Le Coniat M, Schaison G: Chromosome abnormalities in bone marrow of Fanconi ane-
mia patients. Cancer Genet Cytogenet 1993;65:47–50.
36 Mozziconacci MJ, Arnoulet C, Stoppa AM, Mattei MG, Simonetti J, et al: Anomalies cytogéné-
tiques clonales dans l’anémie de Fanconi. Nouv Rev Fr Hématol 1995;37:52–53.
37 Ferti A, Panani A, Dervenoulas J, Raptis SA: Cytogenetic findings in a Fanconi anemia patient
with AML. Cancer Genet Cytogenet 1996;90:182–183.
38 Auerbach AD, Verlander PC: Disorders of DNA replication and repair. Curr Opin Pediatr 1997;9:
600–616.
39 Verbeek W, Haase D, Schoch C, Hiddemann W, Wormann B: Induction of a hematological and
cytogenetic remission in a patient with a myelodysplastic syndrome secondary to Fanconi’s ane-
mia employing the S-HAM regimen. Ann Hematol 1997;74:275–277.
40 Oliveira NI, Ribeiro EM, Raimondi SC, Bittencourt MA, Pasquini R, Cavalli IJ: Two different
karyotypes with 1q abnormalities in a patient with Fanconi anemia. Leuk Res 2002;26:
1047–1049.

Heidemarie Neitzel
Institute of Human Genetics, Charité, Medical University of Berlin
Augustenburger Platz 1
D–13353 Berlin (Germany)
Tel. ⫹49 30 450566411, Fax ⫹49 30 450566933, E-Mail [email protected]

Neitzel/Kühl/Gerlach/Ebell/Tönnies 94
 Schindler D, Hoehn H (eds): Fanconi Anemia. A Paradigmatic Disease for the Understanding of
Cancer and Aging. Monogr Hum Genet. Basel, Karger, 2007, vol 15, pp 95–109

Interphase FISH-Assay for the Detection of

MDS- and AML-Associated Chromosomal
Imbalances in Native Bone Marrow and
Peripheral Blood Cells
H. Tönnies, S. Huber, E. Volarikova, A. Gerlach, H. Neitzel
Institute of Human Genetics, Charité, Universitätsmedizin Berlin, Berlin, Germany

Abstract
Bone marrow (BM) failure in Fanconi anemia (FA) patients followed by myelodysplas-
tic syndrome or acute myeloid leukemia is frequently associated with the occurrence of spe-
cific clonal chromosomal imbalances in BM cells. Our previous data of patients with 3q
aberrations revealed that gains of 3q are strongly associated with a poor prognosis and repre-
sent an adverse risk factor in FA. In routine cytogenetic diagnostics, these clonal aberrations
can be detected after BM aspiration, culturing of bone marrow cells, and karyotyping of
GTG-banded metaphase chromosomes. However, some unbalanced mosaic aberrations in
bone marrow cells are too subtle to be detected by conventional cytogenetics alone. In order
to develop a sensitive and fast detection protocol for routine use, we studied FA cells by inter-
phase fluorescence in situ hybridization (I-FISH) using a YAC panel. We established a sensi-
tive interphase FISH assay for the early detection of prognostically poor clonal chromosome
aberrations in bone marrow and peripheral blood interphase cells of FA patients as a useful
complement to metaphase chromosome analysis of bone marrow cells. The analysis on the
single cell level allows the early and sensitive detection and the monitoring of chromosomal
imbalances as prerequisite for a timely intervention and treatment of affected patients.
Copyright © 2007 S. Karger AG, Basel

Common hematological findings in Fanconi anemia (FA) patients are

thrombocytopenia, aplastic anemia (AA), and myelodysplastic syndrome
(MDS), as well as an increased rate of malignancies, including transformation
into acute myeloid leukemia (AML) [1, 2]. Furthermore, FA patients have an
increased risk for the development of solid tumors [3]. Only relatively few stud-
ies and mostly single case reports investigating cytogenetic aberrations in FA
bone marrow cells are available to date [4–10]. Many FA patients with MDS
and AML develop clonal chromosome aberrations in their bone marrow cells,
and clonal cytogenetic abnormalities are often detected in close temporal proxi-
mity with or at the time of diagnosis of MDS or AML [5, 7, 11, 12]. Even
though clonal chromosome changes may persist in FA patients for many years
without conversion to full-blown MDS or AML, these alterations clearly are of
concern since chromosomal imbalances or chromosome rearrangements are
frequent manifestations of the conversion of normal to malignant cell growth.
More than 30 years ago, Traute Schroeder discovered the spontaneous chromo-
some instability in FA cells and postulated that the underlying defect in DNA
repair might itself initiate and promote, together with extrinsic factors, chromo-
somal and genetic instability and thus contribute to the emergence of malig-
nancy in FA patients [13, 14].
Recently, we could show that the most frequent chromosome imbalances
detected in a group of 53 FA patients with clonal bone marrow aberrations were
partial trisomies and tetrasomies for the long arm of chromosome 3 [12]. Our
follow-up data of eighteen patients with 3q aberrations, including eight patients
with additional monosomy 7, revealed that these aberrations were strongly
associated with a high risk for the development of MDS and AML.
This previous study suggests that the emergence of 3q aberrations is an
important warning sign that should have immediate consequences in the med-
ical care of FA patients. It is therefore extremely important to identify the rele-
vant aberrations with high diagnostic sensitivity. However, the detection of
these chromosome imbalances by classical metaphase cytogenetics alone may
be hampered by insufficient numbers of metaphases, and by the typically low
quality of morphology and banding resolution of bone marrow chromosome
preparations. Therefore, a fast and sensitive interphase fluorescence in situ
hybridization (I-FISH) assay operating at the single cell level would be highly
desirable. In the present study, we established such an interphase FISH assay
for the systematic detection of adverse clonal aberrations in mononuclear cells
from uncultured bone marrow and peripheral blood cells.

Materials and Methods

Conventional Cytogenetics
The diagnosis of Fanconi anemia was confirmed in all patients by a standard chromo-
some breakage test after mitomycin C treatment. Cell culture and high-resolution chromo-
some preparations from peripheral blood lymphocytes and synchronized bone marrow
metaphases for the detection of numerical and structural chromosome aberrations employed
standard techniques and GTG-banding [15]. Computer-assisted karyotyping of GTG-
banded chromosomes was performed with the Ikaros system (MetaSystems, Altlussheim,
Germany).

Tönnies/Huber/Volarikova/Gerlach/Neitzel 96
 Direct Cell Preparations
For interphase cell preparation, native bone marrow (BMD: bone marrow direct prepar-
ation) and peripheral blood (PBD: peripheral blood direct preparation) specimens (0.5 ml)
were incubated without prior culturing in hypotonic potassium chloride solution (0.4%;
10 min at 37⬚C) and fixed in ice-cold methanol/acetic acid (v/v, 3:1) three times. After drop-
ping on slides, cells were aged overnight at room temperature. For long time storage, slides
were kept at –20⬚C. Peripheral blood and bone marrow smears were prepared by standard
protocols.

Molecular Cytogenetics
In addition to conventional cytogenetics where up to 50 bone marrow metaphases were
analyzed, we utilized comparative genomic hybridization (CGH) and fluorescence in situ
hybridization (FISH) with whole chromosome painting probes for the detection and charac-
terization of chromosomal abnormalities.

Comparative Genomic Hybridization (CGH)

CGH was performed as described previously [16] with slight modifications. In brief,
test and control DNAs were differently labeled by nick translation with SpectrumGreen®-
dUTP and SpectrumOrange®-dUTP (VYSIS) respectively. 200 ng of labeled test DNA,
200 ng reference DNA, and 12.5 ␮g Cot-1 DNA were co-precipitated, denatured, and
hybridized to denatured metaphase spreads derived from a clinically normal male. After
incubation at 37⬚C for 72 h, standard post-hybridization washes were performed. Metaphase
images were evaluated using an epifluorescence microscope (Axioskop, ZEISS, Germany)
fitted with a cooled CCD-camera (Hamamatsu) and appropriate single band pass filter sets.
Image analysis and karyotyping were performed using the ISIS analysis system (Metasystems,
Germany). Positive CGH results were validated using the patients’ bone marrow metaphase
spreads for metaphase FISH with commercial whole chromosome painting probes (Oncor;
VYSIS) according to the manufacturers’ instructions.

Interphase FISH with YAC Clones

YAC clones for the critical regions of chromosome 3 and 7 (3q: 909d10; 7q: 942g09)
and a control region on the short arm of chromosome 3 and 7 (3p: 808b10; 7p: 956e01) were
selected from the MCN Reference Center at the Max-Planck Institute for Molecular
Genetics, Berlin, Germany (for detailed YAC-data see http://www.mpimg-berlin-
dahlem.mpg.de/⬃cytogen/probedat.htm). YAC-DNA was amplified (1st PCR) and labeled
(2nd PCR) by degenerate oligonucleotide-primed polymerase chain reaction (DOP-PCR) as
described by Telenius et al. [17] with minor modifications.
Interphase FISH with the labeled probe (e.g. chromosome 3: 808b10, red; 909d10,
green) was performed according to standard protocols. In brief, after treatment with RNase
solution (100 ␮g/ml) at 37⬚C for 1 h and denaturation for 5 min at 73⬚C in 70%
formamide/2⫻ SSC, slides were dehydrated in cold ethanol series. YAC-probes were dena-
tured at 73⬚C for 5 min and placed on denatured cells. After overnight hybridization at 37⬚C,
slides were washed in 0.4⫻ SSC/0.3% Igepal at 73⬚C for 2 min and rinsed in 2⫻ SSC/0.1%
Igepal. Slides were counterstained with 4⬘,6-diamidino-2-phenylindole dihydrochloride
(DAPI) and mounted with antifading solution (Vectashield, Vector Laboratories, Inc.
Burlingame, CA). Image capturing and analysis were performed as described for CGH. At
least 500 nuclei were analyzed by manual inspection of each hybridized aliquot.

FA-Specific Interphase FISH-Assay 97

 Determination of Positive Cut-off Values
The cut-off values for positive I-FISH results (e.g. three FISH signals for 3q) were
determined by statistical evaluation of I-FISH results of mononuclear cells (BMD and PBD)
from normal control subjects: the mean ⫾3 SD of false-positive nuclei was taken as the cut-
off level for diagnosing a duplication or deletion of chromosomal material.

Results

Sensitivity of I-FISH
To circumvent selective outgrowth of clonal aberrations in vitro and to
gain insights into the in vivo situation of the patient, we decided to analyze
uncultured interphase cells. As a first technical result, comparative analyses of
bone marrow or peripheral blood smears versus directly prepared cells showed
that hybridization quality and success rate were superior when cell suspension
slides (BMD and PBD) rather than smears were used (data not shown).
Therefore, we determined the positive cut-off levels of our YAC-clone probe
sets on bone marrow and peripheral blood direct preparations by hybridizing
70 control samples (56 PBD; 14 BMD). An average of 500 nuclei for each
hybridization were scored, which means that positive cut-off values were deter-
mined based on the evaluation of 70,000 interphase nuclei. The mean percent-
age of 3q positive cells (presence of three green signals; GGG) in BMD
preparations of control subjects was 1% (mean ⫹3 SD; SD ⫽ 0.25), whereas
using PBD the cut-off level was 2.9% (SD ⫽ 0.69). The cut-off values for par-
tial tetrasomy of chromosome 3q, which in FA specimens often results from
clonal progression of a partial trisomy 3q clone, were 0.98 (SD ⫽ 0.27) for
BMD and 0.26 (SD ⫽ 0.08) for PBD. The monosomy 7 cut-off values were
3.59 (SD ⫽ 0.99) in BMD and 0.46 (SD ⫽ 0.14) in PBD using the chromo-
some 7 probe set.

Comparison between Conventional Cytogenetic Data and I-FISH

The detection of chromosomal aberrations by classical cytogenetics after
GTG-banding of bone marrow metaphase chromosomes was performed by rou-
tinely analyzing up to 50 metaphase spreads. In order to determine the sensitiv-
ity of the FISH assay, we performed I-FISH experiments with the probe sets on
BMD and PBD of 3q-positive patients sampled at the same time as preparations
for conventional cytogenetics (see fig. 1).
In most analyses of samples taken at comparable time points, the number
of aberrant cells (partial trisomy 3q) was highest in metaphase preparations of
cultured bone marrow cells (BM-CC). In contrast, the number of aberrations
in non-cycling interphase cells, describing the in vivo situation, were lower in
BMD, and lowest in PBD. However, in all cases the number of aberrant cells in

Tönnies/Huber/Volarikova/Gerlach/Neitzel 98
 100 BM-CC
90 BMD
Aberrant cells (%)

80 PBD
70
60
50
40
30
20
10
0
Patient A Patient B Patient C Patient C Patient C
(0 month) (⫹5 months) (⫹8 months)

Fig. 1. Percentage of affected cells with partial trisomy 3q obtained by I-FISH with the
chromosome 3 YAC probe on cultivated bone marrow cells (BM-CC), bone marrow direct
preparations (BMD) and peripheral blood direct preparations (PBD) of three patients (A–C).
For patient C, the follow-up period extended over eight months.

direct preparations was higher than the positive cut-off values for the chromo-
some 3 probe set on BMD (1%) and PBD (2.9%).

Comparison between Concomitant BMD and PBD I-FISH

To further evaluate the feasibility of analyzing easily accessible peripheral
blood specimens (PBD) instead of BMD samples during the follow-up of FA
patients, we compared I-FISH data from analyses performed on ten paired sets
of BMD and PBD that had been collected from FA patients positive for partial
trisomy of the chromosome 3q critical region (fig. 2). In the BMD preparations,
percentages of partially trisomic cells ranged from 31.5 to 86.1%, whereas in
PBD preparations 22.4 to 78.2% of the cells were found to be trisomic. In eight
of these ten cases, the numbers of affected cells were lower in PBD cells com-
pared to BMD preparations. Interestingly, in two cases (cases 3 and 10) the
number of partially trisomic cells was higher in PBD compared to parallel
BMD preparations. The mean difference between both types of specimens was
11.5%; we therefore conclude that the detection sensitivity is somewhat lower
in PBD compared to BMD cells. However, an important point is that all positive
cases were successfully detected using PBD preparations.

Monitoring Clonal Evolution by Chromosome 3 I-FISH

In the beginning of our I-FISH studies, the significance of partial tri- or
tetrasomies of chromosome 3 was unknown. Therefore, we monitored patients
frequently in PBD in between their yearly bone marrow aspirations in order to

FA-Specific Interphase FISH-Assay 99

 100
BMD
90 PBD
80
Aberrant cells (%)

70
60
50
40
30
20
10
0
1 2 3 4 5 6 7 8 9 10
No. of analysis with matched BMD and PBD samples

Fig. 2. Percentage of cells affected by chromosome 3q imbalance in native bone mar-

row (BMD) and peripheral blood nuclei (PBD) screened by I-FISH. Bone marrow and blood
of the ten paired samples were taken at the same time. I-FISH was performed and 500 nuclei
per hybridization were analyzed.

ascertain whether aberrations of 3q show transient fluctuations. Figure 3

illustrates the results of such long term monitoring of a single FA patient. Via
repeated I-FISH examinations, information is gathered about quantitative (out-
growth of the clone) and qualitative (clonal evolution; evolution of a clone from
partial trisomy to partial tetrasomy) changes over the time (fig. 3). Starting with
approximately 7% tri- and tetrasomic cells in PBD at the time of the first inves-
tigation (month zero), the number of normal disomic cells decreased dramatic-
ally to less than 20% after 10 months of follow-up, whereas the tetrasomic
clone gradually replaced the trisomic clone. BMD analyses seven and ten
months after the first investigation also showed the clonal progression of the
tetrasomic clone.
There is a good correlation between the I-FISH data and CGH analyses
performed during the monitoring period. The imbalance of chromosome 3q was
clearly detectable by CGH using PBD-DNA, starting at month 4 of the obser-
vation period; however at month zero the number of affected cells (around
14%) was too low to be reliably detected by CGH alone.

Early Detection of Secondary Monosomy 7 in Peripheral Blood Cells

One of the goals of the I-FISH assay is the early and sensitive detection
of adverse chromosomal imbalances in FA cells by prospective monitoring of
FA patients. In a second example of long-term follow-up of an FA patient

Tönnies/Huber/Volarikova/Gerlach/Neitzel 100
 100
GGRR [33]
90 GGGRR [33 ⫹ 3q]
80 GGGGRR [33 ⫹3q3q]
Others
70
Cells (%)

60
50
40
30
20
10
0
th

s
th

th

th

th

th
on

on

on

on

on

on
m

m
0

10

10
⫹

⫹
⫹

⫹
Chromosome 3-PBD Chromosome 3-BMD

0 month ⫹4 months ⫹7 months

3 3 3

Fig. 3. Long term monitoring of the adverse imbalance of chromosome 3q in PBD

(left) and BMD (right) cells together with chromosome 3 ratio profiles of CGH analyses
(bottom) illustrating the outgrowth of two different chromosome 3q clones. GGRR[33], two
green (G) I-FISH signals for the critical region and two red (R) I-FISH signals for the control
region of chromosome 3; GGGRR[33⫹3q], three green I-FISH signals for the critical region
(aberrant) and two I-FISH signals for the control region of chromosome 3; GGGGRR
[33⫹3q3q], four green I-FISH signals for the critical region (aberrant) and two I-FISH
signals for the control region of chromosome 3. Bottom: CGH ratio profile of chromosome 3
after analysis of 10–12 single chromosomes at different time points.

with chromosome 3 aberration, we observed remarkable fluctuations of the

percentages of aberrant cells (partial trisomy 3q) over an 18-months interval
in PBD cells (fig. 4), even though the aberrant clone did at no time disappear.
At month 16 of monitoring, neither conventional cytogenetic analysis nor
I-FISH of bone marrow yielded any evidence for monosomy 7 (data not
shown). However, only two months later a complete monosomy 7 had
emerged in the cell line marked by the 3q aberrations. This clone had been
detected in PBD cells only two months after a routine BM standard cytoge-
netic analysis and BMD-I-FISH. The conventional cytogenetic analysis of bone
marrow cells at month 18, immediately after its detection in PBD, confirmed

FA-Specific Interphase FISH-Assay 101

 100
GGRR [33]
90 GGGRR [33⫹ 3q]
80 Others
Aberrant cells (%)

70
60
50
40
30
20
10
0
on D

m BD
m B

s)

s)

s)

s)

s)
m BD

m BD

m BD

m BD
th
(0 3 P

th

th

th

th

th
5 P

11 P

13 P

16 P

18 P
on

on

on

on

on
#

(⫹ #3

(⫹ #3

(⫹ #3

(⫹ #3

(⫹ #3
100 GGRR [77]
GR [7]
90 Others
80

70
Aberrant cells (%)

60

50

40

30

20

10

0
on D

m BD
m B

s)

s)

s)

s)

s)
m BD

m BD

m BD

m BD
th
(0 7 P

th

th

th

th

th
5 P

11 7 P

13 7 P

16 P

18 P
on

on

on

on

on
#

(⫹ #7

(⫹ #7

(⫹ #7
#

#
(⫹

(⫹

Fig. 4. Early detection of a secondary monosomy 7 in peripheral blood cells of a

patient with clonal chromosome 3q imbalance during long term follow-up (18 months).
GGRR[33], two green I-FISH signals for the critical region (G) and two red I-FISH signals
for the control region (R) of chromosome 3; GGGRR[33⫹3q], three green I-FISH signals
for the critical region (aberrant) and two red I-FISH signals for the control region of chro-
mosome 3. GGRR[77], two green I-FISH signals for the critical region of 7q and two red
I-FISH signals for the control region of chromosome 7; GR[7], one signal each for chromo-
some 7, indicating monosomy 7.

Tönnies/Huber/Volarikova/Gerlach/Neitzel 102
the monosomy 7 in cultured metaphase spreads (45% affected cells) and
BMD interphase cells (68% affected cells). This long term follow-up clearly
illustrates that I-FISH on PBD cells is a powerful tool for the early and sensi-
tive detection of adverse clonal evolution in uncultured peripheral blood cells
of FA patients.

Discussion

Common hematological findings in FA patients are thrombocytopenia,

aplastic anemia, and MDS, as well as an increased rate of malignancies, most
prominently AML [2] and squamous cell carcinomas [1]. Many FA patients
with MDS and AML develop clonal chromosome aberrations in their bone mar-
row cells [4–10, 12]. Most clonal aberrations are unbalanced while balanced
chromosome translocations such as t(8;21) and t(15;17) that are commonly
associated with M2, M3, and M5 AML of childhood leukemia of non-FA
patients have not been described in FA patients [5, 10]. The significance and the
predictive value of most clonal changes found in FA bone marrow cells are far
from being understood. However, in the last years, evidence has accumulated
that there is a significant correlation between the presence of specific clonal
abnormalities, the disease status of the patients (e.g. MDS), and an adverse
outcome [11, 12].
In our cohort of 53 FA patients published in 2003 [12], more than 50% of
patients had acquired clonal chromosome aberrations. The most frequent aber-
rations were partial trisomies and tetrasomies for the long arm of chromosome 3,
duplications involving 1q, and monosomy 7. Moreover, our data of 18 patients
with 3q aberrations and eight patients with monosomy 7 revealed that these
specific aberrations are strongly associated with poor prognosis and thus repre-
sent adverse risk factors for FA patients.
Based on these observations, we decided to establish and test a fast and
sensitive I-FISH assay. Such an assay could play an important role in the
timely detection of suspicious clones in FA patients. Moreover, follow-up and
prognosis might benefit from such a test. To establish the I-FISH assay,
we performed the following analyses: (1) determination of sensitivity by
testing the positive cut-off values for the different YAC clones on BMD and
PBD, (2) comparison of conventional cytogenetic data of clonal aberrations
with the BMD and PBD results at the same time points, (3) direct comparison
of I-FISH results obtained with bone marrow versus peripheral blood
specimens, (4) collection of quantitative data on clonal outgrowth and evolu-
tion, and (5) early sensitive detection of a secondary adverse chromosomal
imbalance.

FA-Specific Interphase FISH-Assay 103

 Determination of Positive Cut-off Values
For the definition of positive cut-off values of the YAC probes, we ana-
lyzed 500 hybridized nuclei each of 14 bone marrow and 56 peripheral blood
direct preparations of control subjects. Cut-off levels (mean ⫹3 SD) for the
chromosome 3 YAC probe for the detection of a partial trisomy 3q were 1% for
BMD and 2.9% for PBD, and 0.98% (BMD) and 0.26% (PBD) for a partial
tetrasomy. The monosomy 7 positive cut-off values were 3.5% for BMD and
0.46% for PBD. These values are comparable to cut-off values reported by
other groups using interphase FISH for the determination of chromosomal
imbalances [18–20].

Comparison between Conventional Cytogenetic and

I-FISH Data in 3q-Positive Patients
Traditionally, conventional cytogenetic analysis of bone marrow metaphase
spreads is used for the detection of chromosome aberrations in routine diagnos-
tics. However, karyotype analysis reflects genetic alterations only in the divid-
ing cell population (metaphase spreads). Thus, in previous FA BM studies,
many aberrations probably remained undetected by conventional cytogenetics
due to the low average resolution of the bone marrow chromosome preparations
[12]. To the best of our knowledge, there are only anecdotal reports of FA
patients with MDS or AML showing additional material of chromosome 3q in
BM cells [11, 21]. Considering the significant overrepresentation of 3q trisomies
and tetrasomies in our series of FA patients, we assume that other studies might
have failed to detect the more subtle aberrations of 3q because only conven-
tional cytogenetics rather than CGH or I-FISH were applied.
The precise characterization of chromosome aberrations is important for
the distinction between primary, pathogenetically relevant changes and sec-
ondary aberrations reflecting clonal evolution [22]. In addition to CGH, we rou-
tinely examined up to 50 metaphase spreads for the detection of chromosome
aberrations in bone marrow specimens (data to be published elsewhere). To test
the sensitivity of the I-FISH assay for the detection of unbalanced clones in
uncultured cells we performed comparative FISH experiments with probe sets
applied in parallel to bone marrow and peripheral blood cells (BMD, PBD), and
concomitant with conventional cytogenetics. In these experiments, the highest
numbers of aberrant cells with partial trisomy 3q were found in cultured bone
marrow metaphase preparations (BM-CC), whereas the figures for aberrant
interphase cells were lower in BMD preparations and still lower in PBD speci-
mens. Similar differences have also been described by other authors [23–25]. It
is generally assumed that cells with chromosome abnormalities in disorders
like MDS and AML often acquire a higher capacity for division in culture
than normal cells. Therefore karyotype analysis does not provide an accurate

Tönnies/Huber/Volarikova/Gerlach/Neitzel 104
estimate of the proportion of genetically altered cells in native bone marrow
preparations (in vivo situation) [25].
The number of affected cells in our test samples was clearly higher than the
positive cut-off values for the chromosome 3 probe set for BMD (1%) and PBD
(2.9%). We therefore conclude that in the case of FA patients who show aneu-
ploidy by conventional cytogenetics of BM cells, the application of I-FISH on
uncultured BMD and PBD cells leads to the accurate confirmation of the chro-
mosomal alterations.

Comparison between BMD and PBD I-FISH in 3q-Positive Patients

We compared I-FISH data from ten additional paired sets of BMD and
PBD collected from FA patients with a partial trisomy for the critical region of
chromosome 3q26q29. The percentages of aberrant cells varied among FA
patients, but in a given case there was good agreement between the number of
positive aberrant BMD and PBD cells. In the BMD cells, the percentages of
abnormal cells ranged from 31 to 86%, whereas in PBD preparations 22 to 78%
affected cells were detected. As expected, in most cases the number of affected
cells was lower in PBD compared to BMD preparations. The mean difference
between both analyses was 11%; indicating that the detection rate is slightly
lower when PBD cells are analyzed. Nevertheless, all positive cases were
detected successfully using PBD preparations. Significantly, using selective
enrichment of granulocytes from peripheral blood and subsequent I-FISH, we
could demonstrate that the chromosome aberrations are exclusively confined to
the granulocytes which represent approximately 80% of the peripheral
mononuclear cells (data not shown) while they were not detected in either B or
T lymphocytes. This may explain the smaller number of affected cells in PBD
in comparison to BMD.
Due to the small amount of blood needed and the fast results after I-FISH,
this technique is ideally suited for screening during the intervening time
between two bone marrow aspirations. As we have demonstrated, I-FISH is a
reliable and rapid method for the detection and monitoring of premalignant
clones in easily accessible cells of FA patients. Thus, a close follow-up is easily
achieved, especially if blood counts are fluctuating which could be a sign for the
appearance of cells with clonal aberrations. Despite these attractive attributes of
the I-FISH method, it can only complement but not replace regular bone mar-
row based cytogenetic analysis which obviously has the highest sensitivity for
the detection of aberrant clones.

Follow-up of Clonal Succession and Clonal Evolution by I-FISH

In the beginning of our I-FISH studies, the relevance of the partial tri- or
tetrasomy of chromosome 3q was not known. We therefore followed all patients

FA-Specific Interphase FISH-Assay 105

closely and examined their peripheral blood cells between the yearly bone mar-
row aspirations. With the help of I-FISH, quantitative (e.g. outgrowth of the
clone) and qualitative (e.g. clonal evolution from partial trisomy to partial
tetrasomy) information can be gathered. The example presented in this report
(fig. 3) demonstrates that even small numbers of aberrant cells with initially
only 7% tri- and tetrasomic cells can be detected by I-FISH. The dramatic
increase of the tetrasomic clone could be followed over time. Parallel analyses
of BMD and PBD cells during this follow-up (at months 7 and 10) clearly
showed that the number of cells with partial tetrasomy 3q is always higher in
BM than in PB. None of our FA patients with 3q aberrations experienced the
disappearance of the aberrant clone over time. The behavior of cell clones car-
rying 3q aberrations obviously follows a different pattern than is seen with
other types of clonal changes in FA patients [10, 21, 26, 27].

Early Detection of a Secondary Adverse Cell Clone in PBD Cells

During the follow-up of a second FA patient with chromosome 3 aber-
rations, we observed substantial fluctuations of the proportion of aberrant cells
in PBD over a period of 19 months. However, the aberrant clone did not disap-
pear over the entire observation period. After 18 months, a second adverse
aberration was detected by I-FISH in PBD cells: the cells with the 3q aberration
acquired in addition a monosomy 7. Cells with both aberrations must have
expanded rapidly, since a cytogenetic analysis and I-FISH of bone marrow two
months prior to the positive PBD findings did not reveal any such cells (data
not shown). Conventional cytogenetic analysis at month 18 on BM metaphase
spreads confirmed the monosomic clone in 45% of all metaphase spreads.
I-FISH was positive in 68% of BMD interphase cells.
Monosomy 7 has previously been reported in FA patients and is associated
with poor prognosis [8, 12, 19, 28, 29]. Interestingly, Thurston et al. [19]
described a patient, in whom they found two out of 30 bone marrow metaphase
spreads with monosomy 7. Parallel I-FISH with an alphoid centromeric probe
for chromosome 7 indicated that 40% of cells had a monosomy 7. Rescoring of
two earlier samples by I-FISH demonstrated that the first sample, taken
19 months before, had 4% nuclei with monosomy 7 and the second sample,
taken 12 months before, contained 21.5% of cells with a monosomy 7, whereas
the cytogenetic analyses at this date showed no monosomy 7. These results indi-
cate a slow evolution towards monosomy 7 in the patient’s bone marrow,
whereas in our patient, monosomy 7 could not be detected in bone marrow cells
by I-FISH. These data demonstrate that the dynamics of clonal outgrowth can
vary considerably among FA patients.
Thurston and coworkers concluded that the presence of a monosomy 7
clone in bone marrow aspirates can remain undetected by standard cytogenetic

Tönnies/Huber/Volarikova/Gerlach/Neitzel 106
analysis. Kearns et al. [20] examined bone marrow cells from 96 unselected
patients with bone marrow failure syndromes to assess the frequency of unde-
tected aneuploidy for chromosomes 7 and 8 by FISH. They identified
27 patients with cytogenetically undetected monosomy 7 or trisomy 8. Accord-
ing to these authors, undetected aneuploidy exists in bone marrow cells of a
significant percentage of patients with bone marrow failure syndromes. As we
have demonstrated in this report, the problem can be overcome by continued
monitoring of FA patients using I-FISH on PBD and, if available BMD cells.
This approach will assure the early detection of adverse clones with mono-
somy 7 or gain of 3q allowing earlier intervention, such as bone marrow trans-
plantation, to prevent the onset of leukemia in these patients. We therefore
have adopted I-FISH analysis with specific YAC clones for all FA patients as a
standard practice. Using this strategy, 3q aberrations were detected by screen-
ing PBD cells in three other FA patients (data not shown). This observation
emphasizes the high sensitivity and specificity of I-FISH as a screening assay
for the detection of aberrant clones in native peripheral blood mononuclear
cells.

Concluding Remarks

We have shown that continued monitoring of FA patients by I-FISH on

both BMD and PBD detects abnormal clones with monosomy 7 or additional
chromosome 3q material quickly and with high sensitivity. Whereas conven-
tional cytogenetics of bone marrow metaphases permits the evaluation of the
entire karyotype, only targeted aberrations are investigated by the I-FISH
assay. Nevertheless, systematic screening for the most adverse clonal aberra-
tions in Fanconi anemia patients can be recommended, and I-FISH has been
shown to be very useful for this purpose. In addition to monitoring by I-FISH
we recommend the application, in regular intervals, of conventional cytogenet-
ics and CGH in bone marrow cells of FA patients, since only conventional
cytogenetics is able to detect clonal aberrations affecting chromosomes other
than 3q or 7.

Acknowledgements

We are indebted to all Fanconi anemia patients, their parents, and their clinicians who
supported this study. YAC clones were provided by the Max-Planck Institute of Molecular
Genetics, Berlin, Germany. Supported by grants from the Charité Universitätsmedizin Berlin
and the Deutsche Fanconi-Anämie-Hilfe e.V. (www.fanconi.de).

FA-Specific Interphase FISH-Assay 107

 References

1 Rosenberg PS, Greene MH, Alter BP: Cancer incidence in persons with Fanconi anemia. Blood
2003;101:822–826.
2 Tischkowitz M, Dokal I: Fanconi anaemia and leukaemia – clinical and molecular aspects. Br
J Haematol 2004;126:176–191.
3 Alter BP, Greene MH, Velazquez I, Rosenberg PS: Cancer in Fanconi anemia. Blood 2003;
101:2072.
4 Berger R, Bernheim A, Le Coniat M, Vecchione D, Schaison G: Chromosomal studies of leukemic
and preleukemic Fanconi’s anemia patients: examples of acquired ‘chromosomal amplification’.
Hum Genet 1980;56:59–62.
5 Auerbach AD, Allen RG: Leukemia and preleukemia in Fanconi anemia patients: a review of the
literature and report of the International Fanconi Anemia Registry. Cancer Genet Cytogenet
1991;51:1–12.
6 Berger R, Le Coniat M, Schaison G: Chromosome abnormalities in bone marrow of Fanconi ane-
mia patients. Cancer Genet Cytogenet 1993;65:47–50.
7 Butturini A, Gale RP, Verlander PC, Adler-Brecher B, Gillio AP, Auerbach AD: Hematologic
abnormalities in Fanconi anemia: an International Fanconi Anemia Registry study. Blood
1994;84:1650–1655.
8 Berger R, Jonveaux P: Clonal chromosome abnormalities in Fanconi anemia. Hematol Cell Ther
1996;38:291–296.
9 Alter BP: Fanconi’s anemia and malignancies. Am J Hematol 1996;53:99–110.
10 Maarek O, Jonveaux P, Le Coniat M, Derre J, Berger R: Fanconi anemia and bone marrow clonal
chromosome abnormalities. Leukemia 1996;10:1700–1704.
11 Alter BP, Caruso JP, Drachtman RA, Uchida T, Velagaleti GV, Elghetany MT: Fanconi anemia:
myelodysplasia as a predictor of outcome. Cancer Genet Cytogenet 2000;117:125–131.
12 Tonnies H, Huber S, Kühl JS, Gerlach A, Ebell W, Neitzel H: Clonal chromosomal aberrations in
bone marrow cells of Fanconi anemia patients: gains of the chromosomal segment 3q26q29 as an
adverse risk factor. Blood 2003;101:3872–3874.
13 Schroeder TM, Kurth R: Spontaneous chromosomal breakage and high incidence of leukemia in
inherited disease. Blood 1971;37:96–112.
14 Schroeder TM: Relationship between chromosomal instability and leukemia. Hematol
Bluttransfus 1974;14:94–96.
15 Barch MJ, Knutsen T, Spurbeck JL: The AGT Cytogenetics Laboratory Manual. Philadelphia,
Lippincott-Raven, 1997.
16 Tonnies H, Stumm M, Wegner RD, Chudoba I, Kalscheuer V, Neitzel H: Comparative genomic
hybridization based strategy for the analysis of different chromosome imbalances detected in con-
ventional cytogenetic diagnostics. Cytogenet Cell Genet 2001;93:188–194.
17 Telenius H, Carter NP, Nordenskjold M, Ponder BA, Yunnacliffe A: Degenerate oligonucleotide-
primed PCR: general amplification of target DNA by a single degenerate primer. Genomics
1992;13:718–725.
18 Flactif M, Lai JL, Preudhomme C, Fenaux P: Fluorescence in situ hybridization improves the
detection of monosomy 7 in myelodysplastic syndromes. Leukemia 1994;8:1012–1018.
19 Thurston VC, Ceperich TM, Vance GH, Heerema NA: Detection of monosomy 7 in bone marrow
by fluorescence in situ hybridization. A study of Fanconi anemia patients and review of the litera-
ture. Cancer Genet Cytogenet 1999;109:154–160.
20 Kearns WG, Sutton JF, Maciejewski JP, Young NS, Liu JM: Genomic instability in bone marrow
failure syndromes. Am J Hematol 2004;76:220–224.
21 Berger R, Bussel A, Schenmetzler C: Somatic segregation and Fanconi anemia. Clin Genet
1977;11:409–412.
22 Tonnies H: Modern molecular cytogenetic techniques in genetic diagnostics. Trends Mol Med
2002;8:246–250.

Tönnies/Huber/Volarikova/Gerlach/Neitzel 108
23 Wilkens L, Komminoth P, Nasarek A, von Wasielewski R, Werner M: Rapid detection of kary-
otype changes in interphase bone marrow cells by oligonucleotide primed in situ hybridization
(PRINS). J Pathol 1997;181:368–373.
24 Velagaleti GV, Tharapel SA, Tharapel AT: Validation of primed in situ labeling (PRINS) for inter-
phase analysis: comparative studies with conventional fluorescence in situ hybridization and chro-
mosome analyses. Cancer Genet Cytogenet 1999;108:100–106.
25 Yan J, Zhang XX, Fetni R, Drouin R: Trisomy 8 and monosomy 7 detected in bone marrow using
primed in situ labeling, fluorescence in situ hybridization, and conventional cytogenetic analyses.
A study of 54 cases with hematological disorders. Cancer Genet Cytogenet 2001;125:30–40.
26 Huret JL, Benz E, Guilhot F, Brizard A, Tanzer J: Fluctuation of a clone 46,XX,i(7q) in bone mar-
row in a Fanconi anaemia. Hum Genet 1986;74:98–100.
27 Ferti A, Panani A, Dervenoulas J, Raptis SA: Cytogenetic findings in a Fanconi anemia patient
with AML. Cancer Genet Cytogenet 1996;90:182–183.
28 Auerbach AD, Weiner MA, Warburton D, Yeboa K, Lu L, Broxmeyer HE: Acute myeloid leukemia
as the first hematologic manifestation of Fanconi anemia. Am J Hematol 1982;12:289–300.
29 Stivrins TJ, Davis RB, Sanger W, Fritz J, Purtilo DT: Transformation of Fanconi’s anemia to acute
nonlymphocytic leukemia associated with emergence of monosomy 7. Blood 1984;64:173–176.

Dr. H. Tönnies
Institute of Human Genetics, Charité, Universitätsmedizin Berlin
Augustenburger Platz 1
D–13353 Berlin (Germany)
Tel. ⫹49 30 450566807, Fax ⫹49 30 450566933, E-Mail [email protected]

FA-Specific Interphase FISH-Assay 109

 Schindler D, Hoehn H (eds): Fanconi Anemia. A Paradigmatic Disease for the Understanding of
Cancer and Aging. Monogr Hum Genet. Basel, Karger, 2007, vol 15, pp 110–130

Applications of Cell Cycle Testing in

Fanconi Anemia
D. Schindlera, R. Friedla, I. Gavvovidisa, R. Kalba, K. Nevelinga, Y. Linkab,
H. Hanenbergb, M. Kubbiesc, H. Hoehna
a
Department of Human Genetics, University of Würzburg, Würzburg, bDepartment of
Pediatric Hematology, Oncology and Clinical Immunology, Heinrich Heine University
School of Medicine, Düsseldorf, cRoche Diagnostics, Department of Cell Analytics,
Pharmaceutical Research Oncology, Penzberg, Germany

Abstract
Chromosome breakage analysis following exposure of cultured cells to DNA-
crosslinking agents has long been considered the ‘gold standard’ for the confirmation or
exclusion of Fanconi anemia (FA). Cells containing DNA damage are arrested and accumu-
late, with a 4c DNA content, near the S/G2-phase border of the cell cycle. As manifestation
of their impaired DNA damage response, FA cells typically express elevated G2-phase cell
fractions which can easily be measured by flow cytometry. In our experience with more than
3,000 such analyses, at most 1 in 10 blood samples submitted for the exclusion or confirma-
tion of FA yields a positive result. Compared to traditional chromosome breakage analysis,
cell cycle testing is less demanding and offers the advantage of speed and low cost. We pre-
fer flow cytometric cell cycle testing for the initial screening of patients, in whom unex-
plained thrombocytopenia, macrocytic aplastic anemia or other clinical findings require the
exclusion of FA. In addition to its application in diagnostic screening, we here show that cell
cycle analysis has become a valuable tool for the determination of clastogen sensitivity (such
as required within the context of complementation studies), for the precise definition of cell
cycle checkpoints, and for the quantitative determination of compartment-specific cell cycle
delay or cell cycle arrest. Cell cycle analysis not only provides a reliable test system for the
initial confirmation or exclusion of FA, but also serves as a highly informative tool for the
comprehensive characterization of the FA cellular phenotype.
Copyright © 2007 S. Karger AG, Basel

Chromosome breakage analysis using diepoxybutane (DEB) or mitomycin

C (MMC) as clastogenic agents has long been considered the ‘gold standard’
for the laboratory confirmation of Fanconi anemia (FA) [1, 2]. Other methods
have been regarded as less reliable, and laboratory strategies for the confirma-
tion or exclusion of FA have largely followed longstanding experience with lit-
tle need for alternatives.
Early studies such as summarized in the 1989 monograph by Schroeder-
Kurth and colleagues [3] have established that FA cells are sensitive to a variety
of DNA-crosslinking agents. In addition to DEB and MMC, diagnostic proto-
cols have variously employed nitrogen mustard, cisplatinum, psoralen-UVA, and
others. As pointed out by Hans Joenje on several occasions, any differences in
the chromosome breakage response towards the various DNA-crosslinking
agents are likely to reflect differences in drug stability and laboratory procedures
rather than intrinsic, biological differences. From a theoretical point of view,
there is little reason to assume genuine differences in the mode of action of any
of these agents. Use of a certain drug by a given laboratory primarily reflects
empirical preferences and practical considerations. For instance, safety concerns
for the laboratory personnel would argue against the use of volatile agents.
Likewise, the necessity for metabolic activation of a substance may imply uncer-
tainty as to its efficacy, and the use of UV requires special equipment.
The observation by Traute Schroeder and co-workers [4] of increased
spontaneous chromosome breakage in patient cells turned the attention of diag-
nostic efforts to chromosome techniques. With the advent of complementation
assays, analysis of cell survival employing dose-response curves has been
added as a standard procedure with both diagnostic and research applications
[5]. Virtually all of the recent FA gene discovery reports include such cell sur-
vival assays. Most recently, FANCD2 immunoblotting has been advocated as a
convenient tool for the rapid recognition of FA patients belonging to subtypes
upstream of and including FA-D2 [6, 7]. The subject of the present chapter, cell
cycle testing using flow cytometry, was introduced into the FA field after chro-
mosome labeling techniques had suggested a cell cycle disturbance in FA [8].
G2 arrest, both spontaneous and enhanced by exposure to DNA-crosslinking
agents, was recognized as a hallmark cell cycle lesion of FA cells that could be
exploited for diagnostic purposes [9–12]. In 1995, it was shown that chromo-
some breakage studies and flow cytometric cell cycle analysis yield equivalent
results, thus validating the diagnostic use of cell cycle analysis [13]. In recent
years, flow cytometric cell cycle analysis has gained additional prominence as a
convenient read-out system for complementation studies [14, 15].
Recent insights into the FA/BRCA pathway have provided a new rationale
for the use of cell cycle analysis. DNA-interstrand crosslinks (ICLs) represent
obstacles to the progression of DNA synthesis at replication forks. Facing an
ICL, the DNA replication machinery is stalled. The barrier can be overcome by
a variety of mechanisms, including translesion synthesis and homologous
recombination, which both involve members of the FA pathway. Since FA cells

Applications of Cell Cycle Testing in Fanconi Anemia 111

are defective in the removal of ICLs, stalled replication forks activate cell cycle
checkpoint controls. These signaling cascades ensure that cell cycle progression
occurs only after successful bypass or repair of a given lesion. Whereas chro-
mosome breakage studies reflect the unsuccessful or faulty repair of DNA
lesions, cell cycle studies indicate whether or not checkpoint controls and DNA
repair function normally. Another advantage of cell cycle studies derives from
the fact that the number of cells that are investigated exceeds those examined in
chromosome studies by orders of magnitude.

MMC-Treated FA Cells Accumulate in G2 Phase in

a Dosage-Dependent Manner

The DNA histograms in the left hand panel of figure 1 (shaded grey) repre-
sent 48-h cell cycle distributions of FA fibroblasts exposed to increasing concen-
trations of MMC as indicated on the right. In the range from 10 to 100 nM MMC,
G2-phase accumulations increase from 21.9 to 72.8%, reflecting the innate
MMC sensitivity of FA cells. Via retroviral complementation, the fibroblast
strain shown in figure 1 had been typed as FA-G, since it was complemented by
stable transfer of FANCG cDNA in a bicistronic construct with GFP. Gene trans-
fer was achieved in 61.8–68.4% of cells. The series of cell cycle distributions
shown in the right hand panel of figure 1 represent the isogenic, but comple-
mented counterparts of the distributions shown on the left. Non-complemented
cells remain GFP-negative (shaded grey) whereas complemented cells are GFP-
positive (shaded green). The GFP marker permits the unambiguous discrimin-
ation between complemented and non-complemented cells via electronic gating
on green-positive and exclusion of green-negative cells. The MMC response of
complemented cells (green panels) is far less pronounced than that of FANCG-
defective cells (grey panels). In the concentration range between 10 and 100 nM
MMC, the respective G2-phase accumulations of complemented cells increase
only from 12.0 to 34.4%. Since the cell cycle distributions shown in figure 1 are
all derived from one and the same fibroblast culture of a single FA-G patient, the
striking differences in the response to MMC between the left and right hand pan-
els solely and uniquely reflect the functional competence of the FANCG gene in
protecting cells from the adverse effects of MMC.
Figure 2 illustrates dose-response curves of the G2-phase fractions of non-
complemented vs. complemented FA-G cells over a broader range of MMC
concentrations. In the range of 33–100 nM, differences of the G2-phase frac-
tions between non-complemented and complemented cells from these other-
wise isogenic cultures amount to 30% of the total cell count. This interval
indicates a diagnostic window allowing for optimal discrimination between

Schindler/Friedl/Gavvovidis/Kalb/Neveling/Linka/Hanenberg/Kubbies/Hoehn 112
 GFP
Negative Positive MMC (nM)
100
G1

10
G2

S
0
100

33
Relative number of cells (%)

0
100

66

0
100

100

0
1 512 1,024 512 1,024
Fluorescence intensity
(Channel number)

Fig. 1. MMC-induced G2-phase arrest of FA fibroblasts. Subcultures of fibroblasts

derived from an FA-G patient were transduced with a retroviral vector containing wtFANCG
cDNA in a bicistronic construct with GFP, and were exposed to different concentrations of
MMC for 48 h. Cells without (left) and with (right) effective gene transfer were gated by flow
cytometry according to the absence (left) or presence (right) of green fluorescence. At each
MMC concentration, Hoechst-33342-stained cells that have retained the FA genotype (left)
display much higher G2-phase peaks than cells, whose genotype was corrected by transduc-
tion with wtFANCG (right).

MMC-sensitive and MMC-resistant fibroblasts. Similar curves were obtained

for other fibroblast strains and other cultured cell types such as primary blood
lymphocytes and EBV-transformed lymphoblast cell lines. Whereas concentra-
tions between 150 and 300 nM MMC (corresponding to 50–100 ng/ml) are rou-
tinely used for chromosome breakage studies, we here show that a diagnostically
valid discrimination between MMC-sensitive and MMC-resistant cells can be

Applications of Cell Cycle Testing in Fanconi Anemia 113

 80
GFP negative
GFP positive

60
G2 phase (%)

40

20

0
1 10 100 1,000
MMC concentration (nM)

Fig. 2. Dose-response curves of G2-phase accumulations. Non-complemented (black

curve) and complemented (green curve) FA-G fibroblasts from one and the same cell culture
show differential G2-phase accumulations as a function of MMC concentration. Evaluation
was with cell cycle distributions of the same type as in figure 1. Cells corrected by transfec-
tion with wtFANCG were gated by their green fluorescence due to GFP co-expression from a
bicistronic vector.

achieved with much lower concentrations, if G2-phase arrest rather than chro-
mosome breakage is measured. This proves the higher sensitivity of the cell
cycle assay compared to chromosome breakage analysis.

ICL-Induced Cell Cycle Arrest Occurs after Completion of

S Phase but prior to G2-M Transition

As shown above for fibroblasts, following exposure to MMC, EBV-

transformed lymphoblast cell lines respond in the same way with an increase
of their G2-phase cell fractions. In the experiments summarized in figure 3, we
used a concentration of 45 nM and 48-h exposures for the discrimination
between FA and non-FA lymphoblast cultures. The top row panels comparing
untreated and treated control cultures show that there is only a very minor
MMC-induced G2-phase increase in a non-FA (CON) lymphoblast cell line. In
contrast, lymphoblast cultures of FA subtypes A, D2 and D1 (second, third and
fourth row) show higher baseline (untreated) G2-phase fractions and respond
to MMC with a more pronounced G2-phase increase, reflecting their innate
MMC sensitivity.

Schindler/Friedl/Gavvovidis/Kalb/Neveling/Linka/Hanenberg/Kubbies/Hoehn 114
 The horizontal axes of the cell cycle distributions displayed in figures 1
and 3 indicate DNA content as measured by the fluorescence intensity of a dye
that binds stoichiometrically to DNA. Thus, in each of these panels the distrib-
utions of cells within the G0/G1, S and G2/M compartments of the cell cycle
are visualized from left to right. Cells in G0/G1 have a 2c DNA content,
whereas cells in G2/M, having completed semiconservative DNA replication,
have a 4c DNA content. G2- and M-phase cells cannot be discriminated on the
basis of DNA content. In order to answer the question whether the cell cycle
arrest in response to MMC treatments affects both the G2- and the M-phase
compartments of the cell cycle, we included phospho-histone H3 (H3P)-staining
as a cellular protein marker, whose expression is limited to the M phase of
the cell cycle [16]. The simultaneous measurement of H3P and DNA content
permits the discrimination between G2- and M-phase cells. As shown by the
framed DNA histogram sections of figure 3, H3P-positive (mitotic) cells repre-
sent only a small proportion of the respective G2/M peaks. Following MMC
treatments, the proportion of H3P-positive cells changes inversely to the
increase of the G2/M peak. The higher the G2/M peak, the smaller the H3P-
positive M fraction. In a series of eight different non-FA control cell lines, we
observed a G2-phase increase of 5.2 ⫾ 1.6% (mean ⫾ 1 SD) and a H3P
decrease of 0.06 ⫾ 0.14%. In contrast, the G2-phase increase for five different
FA cell lines was 15.6 ⫾ 6.2% (mean ⫾ 1 SD) and the H3P decrease
0.49 ⫾ 0.23%. Thus, an approximately 3-fold increase of the G2/M peak
resulted in an approximately 8-fold decrease of the H3P-positive cell popula-
tion. These observations suggest that the MMC-induced accumulation of cells
with a 4c DNA content must occur prior to entry into the M phase.
This interpretation was strengthened by the experiment shown in the bot-
tom panels of figure 3: a non-FA control cell line was exposed to nocadazol, an
agent that binds to tubulin and thereby prevents microtubule protofilament
polymerization and spindle fiber formation. Nocadazol treatment resulted in a
G2/M-peak increase of 23.5%, which is very similar to FA cells exposed to
MMC. Strikingly however, the proportion of H3P-positive cells in the culture
exposed to nocadazol showed an increase of 15.8%, accounting for 2/3 of the
total G2/M-peak increase. In contrast to MMC, nocadazol clearly induces accu-
mulations of cells that have completed G2 and entered the M phase of the cell
cycle.
This series of experiments proves that the MMC-induced accumulation of
cells occurs prior to the M phase, but after completion or near completion of
DNA replication. Otherwise the accumulated cells would not display a 4c
DNA content. These observations localize the putative crosslink-sensitive
checkpoint into the early portion of the G2 phase, presumably close to the
S/G2 border.

Applications of Cell Cycle Testing in Fanconi Anemia 115

 ⫺MMC ⫹MMC
100

80
G1
60 H3P
1.2% 1.1% CON
40

20 G2

0
80

60
Relative number of cells (%)

1.1% 0.5% FA-A

40

20

80

60
0.45% 0.15% FA-D2
40

20

0
80

60
1.4% 0.6% FA-D1
40

20

⫺Nocadazol ⫹Nocadazol
100

80

60 0.7%
16.5% CON
40

20

0
0 512 0 512 1,024
Fluorescence intensity
(channel number)

Fig. 3. Flowcytometric differentiation between G2- and M-phase cells reveals an

inverse relationship between MMC-induced G2-phase arrest and H3P expression. FA lym-
phoblasts respond to exposure to 45 nM MMC for 48 h with increases of their G2-phase frac-
tions to high levels (FA-A, row 2, 16.3–42.1%; FA-D2, row 3, 25.2–36.9%; FA-D1, row 4,
16.0–31.2%), shown in DNA histograms recorded after Hoechst-33342 staining. Simul-
taneously, H3P-positive (mitotic) cells (scatter plots boxed above the G2-phase cells) become
substantially reduced as indicated. Normal control lymphoblasts show only modest G2-phase

Schindler/Friedl/Gavvovidis/Kalb/Neveling/Linka/Hanenberg/Kubbies/Hoehn 116
 Bivariate Flow Cytometry Enhances the Resolution of
Cell Cycle Analysis in FA

Following the pioneering work of Peter Rabinovitch at the University of

Washington in Seattle [17], we initiated studies with Hoechst-33258 fluores-
cence quenching following incorporation of the base analog 2-bromo-5⬘-
deoxyuridine (BrdU) into DNA, a phenomenon first observed in cytogenetics.
Added to cell cultures, BrdU is incorporated into newly synthesized DNA in
place of thymidine. As a consequence, Hoechst-33258 fluorescence is sup-
pressed in a DNA replication-dependent manner. In combination with a sec-
ond DNA-specific dye, e.g., ethidium bromide (EB) emitting at a different
wavelength, which does not show Hoechst-dependent fluorescence quenching,
it is possible to resolve the replicative heterogeneity within a given cell culture.
Figure 4 shows an example of the BrdU-Hoechst/EB double-staining tech-
nique as applied to a 96-h peripheral blood lymphocyte culture. In the presence
of PHA, these cells start cycling between 30 and 35 h after initiation of the cul-
ture. The first replicating cells are represented by a signal track (S) trailing
upward and to the left from the non-cycling cell cluster (G0/G1). As DNA con-
tent increases during replication, fluorescence due to EB staining also
increases on the y-axis, but the intensity of Hoechst-33258 fluorescence on the
x-axis decreases as result of Hoechst-fluorescence quenching by incorporation
of BrdU into newly synthesized DNA. The first cells reaching the G2/M-phase
compartment emerge at the 40-h time point. At 45 h, the first cycle is complete
and, following mitosis, a cluster of cells emerges with half the Hoechst and EB
fluorescence. These are cells that have reached the G1 phase of the second
cycle (G1⬘). At 50 h, a subset of cells has entered the S phase of the second
cycle (S⬘), whose signal track turns into a mirror image of that of the first
cycle. This reflects the smaller increment of Hoechst quenching with BrdU
substitution of second DNA strands as opposed to the substitution of first
strands. After the 55-h time point, the second cycle is complete. Still cycling
cells undergo division and enter the G1 phase of the third cycle (G1⬙), which
appears to the left of the G1⬘ cluster. This cell fraction increases over time and
at 60 h, some of these cells enter a new round of replication and their corres-
ponding S and G2/M phase signals (S⬙ and G2⬙) run parallel to those of the

increase (row 1, 8.3–14.1%) and corresponding H3P decrease. In contrast, exposure of

normal control lymphoblasts to 1.5 ␮M nocadazol for 24 h increases the G2/M phase (row 5,
8.3–31.8%) similar to FA lymphoblasts exposed to MMC, but most of this increase is due to
H3P-expressing cells.

Applications of Cell Cycle Testing in Fanconi Anemia 117

 35 h 40h 45h
G2 G2

S S

G0/G1 1st cycle

G1⬘
Ethidium bromide fluorescence

50 h 55h 60h
G2⬘

S⬘ S⬙
S⬘
G1⬙ G1⬙
2nd cycle
G1⬘

65 h 80h 96h
4th cycle
G2⬙ G2⬙
S⬙ S⬘⬙

G1⬘⬙ G1⬘⬙

3rd cycle

BRDU/HOECHST fluorescence

Fig. 4. Sequence of cell cycle distributions reveals entry and progression of PHA-
stimulated lymphocytes throughout four consecutive cell cycles. Bivariate scatter plots of
single cells, reflecting their Hoechst-33258- (x-axis) and EB- (y-axis) fluorescence intensi-
ties. Aliquots of BrdU-substituted PHA-stimulated lymphocytes from a healthy donor were
harvested at intervals from 35 to 96 h as indicated, stained, and flow measurements were
recorded to reveal static images of the dynamic progression through four consecutive cell
cycles (cycle one, labeled G0/G1, S, G2; cycle two, G1⬘, S⬘, G2⬘; cycle three, G1⬙, S⬙, G2⬙;
cycle four, G1⵮, S⵮). Modified from [28].

second cycle. The third cell cycle is almost complete after 65 h. At 80 h, cells
have visibly progressed into the G1 phase of a fourth cycle (G1⵮) and transi-
tion through this cycle is almost complete at 96 h. As the percentage increment
of BrdU substitution of DNA decreases from cycle to cycle, the additional
Hoechst-quenching effect becomes smaller with each cycle, which limits the
resolution of the BrdU-Hoechst technique to about four rounds of replication.
In addition to showing the distribution of cells throughout up to four cell
cycles within a single cell culture, the BrdU-Hoechst technique permits the

Schindler/Friedl/Gavvovidis/Kalb/Neveling/Linka/Hanenberg/Kubbies/Hoehn 118
 FA CON
G2⬘
S G2 G2⬘ S G2 Min. duration
100 Arrest Arrest
80 G2⬘
60
G2⬘
Cells (%)

40
G2
20 S G2
S

10
20 40 60 80 20 40 60 80
Time after stimulation (h) Time after stimulation (h)

Fig. 5. Kinetic display of successive cell cycle compartment transitions. The curves
represent fits of semi-logarithmic plots of the cell fractions in PHA-stimulated lymphocyte
cultures remaining in any given compartment as a function of time after culture initiation (␣-
plots according to Smith and Martin, modified as proposed by Rabinovitch [17]). The left
hand graph represents a PHA-stimulated lymphocyte culture derived from an FA patient, the
right hand graph that of a non-FA control. The distances of intercepts of exit curves with the
x-axis (representing minimum S, G2 and G2⬘ durations) are indicated by headline bars,
the corresponding arrest fractions are represented as distances of intercepts of the same
curves with the y-axis. The respective G2-phase compartments are shaded.

quantitative assessment of the fraction of cells that fails to respond to the mito-
gen (G0/G1 cell fraction). Unlike any other technique, the resolution of suc-
cessive cell cycles permits the characterization of disturbances of cell cycle
progression due to endogenous or exogenous influences. How this technique
can be applied to the study of the FA cellular phenotype is illustrated in the fol-
lowing section.

The Cell Cycle Disturbance in FA Includes Delay

and Arrest in S and G2

More than 20 years ago, Kubbies et al. [9] first described the cell kinetic
behavior of FA cells as revealed by the BrdU-Hoechst technique. Because of the
unique insights provided by this technique, we briefly summarize the essence of
these types of experiments in figure 5. During the course of a 96-h study of
PHA-stimulated lymphocytes, 12 harvests at 6-h intervals were performed, the
first 24 h after initiation of the culture and the last at 90 h. Each harvest was
recorded in the way shown in the previous section. Data on the fractions of cells

Applications of Cell Cycle Testing in Fanconi Anemia 119

in the various cell cycle compartments were extracted and the percentages of
cells remaining in each compartment were then plotted as function of time after
stimulation. The resulting curves were fitted to the transition probability model
of cell kinetics by Smith and Martin, modified to include a variable proportion
of non-cycling cells in addition to the growth fraction with probabilistic rates of
transition from phase to phase [17]. There are two advantages for FA studies of
using the BrdU-Hoechst data in this manner: first, minimum compartment tran-
sit times can easily be derived by computing the distances between the x-axis
intercepts of pairs of successive curves, each denoting the time course of cell
exit from a given compartment. Second, the fraction of cells incapable of exit
from a given compartment can be measured by comparing the plateau phases of
two successive exit curves. If all cells that enter also leave a given compart-
ment, then the extrapolated plateau levels of the two curves delimiting that
compartment must coincide. If, in contrast, a proportion of cells are perma-
nently halted after entry into such a compartment, then the two delimiting
curves will plateau at different levels, the difference being the fraction of cells
arrested within the respective compartment [9].
Single-time point, univariate flow cytometric measurements assign the dis-
turbance of FA cell cycle progression to the G2 phase. The low resolution of
such univariate and single-time measurements, however, does not permit to
decipher FA cell cycle disturbances other than global G2 accumulations, nor do
univariate measurements answer the question whether G2-phase accumulations
result from increased G2-phase transit times (delay or prolongation), or from
complete blockage of cells in G2 (arrest). Bivariate measurements and sequen-
tial analyses of untreated, PHA-stimulated lymphocyte cultures from an FA
patient (fig. 5, left) reveal S- and G2-phase durations in the first cycle of 10.4
and 5.6 h (indicated by bars above the x-axis) compared to 9.4 and 3.4 h (like-
wise indicated by bars) in a normal control culture (fig. 5, right). The arrest of
cells in the S and G2 phases of the first cycle, indicated on the y-axes, amounts
to 4.2 and 16.2% in the FA patient (fig. 5, left) whereas the corresponding num-
bers are 1.2 and 1.5% in the non-FA control (fig. 5, right). Thus, the cell cycle
disturbance in FA is more precisely described as a combination of S- and G2-
phase delay and S- and G2-phase arrest with a clear predominance, however, of
delay and arrest in the G2-phase compartment [9]. The relatively minor contri-
bution of S-phase delay and S-phase arrest to the cell cycle disturbance of FA
cell cultures is consistent with observations from univariate cell cycle assays
showing preferential accumulations of cells with a 4c DNA content. The analy-
sis of exit kinetic curves such as depicted in figure 5 also demonstrates that the
cell cycle disturbance in FA is not confined to just a single cell cycle, but rather
involves successive cell cycles consistent with the endogenous nature of the
genetic defect underlying FA.

Schindler/Friedl/Gavvovidis/Kalb/Neveling/Linka/Hanenberg/Kubbies/Hoehn 120
 FA CON
1,024 1,024
G2⬘ G2⬙
G2 G2⬘
G2

512 512
EB fluorescence

EB fluorescence
0 0
0 512 1,024 0 512 1,024

G2⬘
G2 G2⬙

G2⬘

G2

1,024 1,024

0 0
1,024 1,024
a BRDU/HOECHST fluorescence b BRDU/HOECHST fluorescence

Fig. 6. Bivariate Hoechst-33258- and EB-fluorescence flow cytograms of PHA-stimulated

72-h primary lymphocyte cultures. The left hand panel (a) represents a patient with FA, the
right (b) a normal control donor. Conspicuous are high G2-phase accumulations in successive
cell cycles (G2, 28.0%; G2⬘, 12.7% and G2⬙, 1.5%) of the cytograms from the FA patient as
opposed to only minor G2-phase cell fractions (G2, 4.4%; G2⬘, 6.1% and G2⬙, 3.9%) of the
normal control. The differences are apparent both in signal-density (upper panel) and signal-
height graphs (lower panel). The examples were selected for similar growth fractions to
emphasize the different G2 phase accumulations of cells. Modified from [18].

The FA Cell Cycle Disturbance is Consistently Present and can be

Exploited for Diagnostic Purposes

Figure 6 shows typical images of bivariate cell cycle distributions assayed

by Hoechst-33258- and EB-fluorescence of 72-h PHA-stimulated FA (a) and
non-FA (b) control cultures. Global G2-phase arrest observed in univariate cell
cycle distributions such as shown in figures 1 and 3 corresponds in figure 6 to
G2 arrest in each successive cell cycle (compare G2 and G2⬘ in a vs. b). This is
evident in signal-density representations (fig. 6, upper panels), where densities
reflect the cell numbers, but is also apparent in signal-height graphs, where

Applications of Cell Cycle Testing in Fanconi Anemia 121

peak heights represent the respective cell numbers (fig. 6, lower panels). In
the present example, the sum of G2 phases in the FA culture (fig. 6a) amounts
to 42.2% of all cells as compared to 14.4% in the non-FA control (fig. 6b).
Moreover, cell cycle progression encompasses four cycles in the control, but
only three in the FA culture. On the other hand, 27.5% of all cells of the control
and 20.7% of all cells in the FA culture did not respond to the mitogen within
the 72-h culture period and thus remain in the G0/G1-phase compartment. This
suggests that the mitogen response of the patient’s PHA-responsive lymphocyte
population is not systematically impaired. Vice versa, 72.5% of the control cells
and 79.3% of the FA culture have entered the cell cycle. This cycling population
is referred to as ‘growth fraction’ (GF). Whether or not G2 accumulations
indicate the diagnosis of FA does not only depend on absolute G2-phase peak
heights, but rather on G2-phase peak heights relative to the GF that may vary
over a wide range among tested cultures. In order to assess this relationship
numerically, we employ a ratio comprising the ‘sum of all G2 compartments
divided by the growth fraction’ (sG2/GF). In the non-FA culture (fig. 6b), this
ratio is close to 0.2. We would consider this ratio tentatively normal. The same
sum of G2 compartments relative to a much smaller growth fraction, such as a
GF of only 30% in a poorly growing culture, would result in an sG2/GF of
about 0.5 similar to that illustrated by the FA-positive example of figure 6a.
The type of bivariate 72-h cell cycle distributions depicted in figure 6 is
what our laboratory routinely uses for screening of blood samples in order to
confirm or exclude the clinical suspicion of FA [18]. The sG2/GF parameter
derived from the computer-assisted evaluation of such bivariate cytograms is
subsequently plotted against the percentage of non-cycling (G0/G1) cells of a
given culture. Figure 7 shows the results of altogether 394 FA and 630 non-FA
72-h peripheral blood lymphocyte cultures assessed in this way. It is important
to note that these measurements did not include prior exposures to MMC. What
we measure with obviously high sensitivity is the spontaneous, intrinsic cell
cycle disturbance of FA cells. We believe that this spontaneous cell cycle dis-
turbance, like spontaneous chromosome breakage, is a primary reflection of
unrepaired spontaneous DNA damage and of the innate sensitivity of FA cell
cultures to ambient air culture conditions with unphysiologically high oxygen
concentrations. Lowering oxygen concentrations during cell culture to near
physiologic conditions (5% v/v) all but eliminates the spontaneous cell cycle
disturbance [19].
In the diagnostic series shown in figure 7, the majority of non-FA cultures
has low sG2/GF ratios and their symbols (green circles) cluster within the left
hand side, whereas the majority of FA cultures has high sG2/GF ratios and their
symbols (red diamonds) come to lie to the right. With very few exceptions, there
is a clear separation of the FA and non-FA genotypes with the discriminatory

Schindler/Friedl/Gavvovidis/Kalb/Neveling/Linka/Hanenberg/Kubbies/Hoehn 122
 100
⫹4mo
80

⫺1 mo
G0/G1 (%)

60

⫹5 mo
40

20 ⫺39mo

⫹11 mo
0
0 0.2 0.4 0.6 0.8 1.0
sG2 vs GF

Fig. 7. Scattergram summarizing the results of flow cytometric cell cycle analyses of
blood samples submitted to the Würzburg laboratory for the confirmation or exclusion of
FA. 72-h, BrdU-substituted cultures of PHA-stimulated lymphocytes were subjected to
Hoechst-33258- and EB-fluorescence intensity recording. Cell-cycle-resolved cytograms
were electronically deconvoluted for quantitative assessment of each cell cycle compartment
and the growth and resting cell fractions. The ratio ‘sum of all G2 phases vs. growth fraction’
(sG2 vs. GF) was calculated and plotted against the G0/G1 fraction, with the latter indicating
the degree of lymphocyte activation. The panel represents 394 FA (red diamonds) and 630
non-FA (green circles) cases. Means of either group ⫾ 1 SD are denoted by the large dia-
mond and circle symbols and are given in the text. Blue squares denote successive cell cycle
assays of an FA patient prior to and after successful HSCT.

threshold empirically set at an sG2/GF of 0.285. The few FA symbols that are
scattered in the immediate vicinity below and above this threshold represent
either proven or likely mosaic cases that consist of a mixture of MMC-sensitive
and MMC-resistant (⫽reverted) cells, or they comprise cases with what we
consider ‘mild’ mutations, i.e. retention of partial protein function. With respect
to the sG2/GF parameter, the mean ⫾ 1 SD amounts to 0.174 ⫾ 0.035 for the
non-FA samples, whereas the mean ⫾ 1 SD of the FA samples is 0.456 ⫾ 0.098
(p ⬍ 0.001). Again, these cultures were not treated with MMC such that these
results would be comparable to the analysis of spontaneous chromosome break-
age rates, which, in contrast, are known to vary greatly and which are generally
judged to be of limited diagnostic value. Bivariate flow cytometry of cultures
exposed to MMC provides even better separation with the FA and non-FA clus-
ters drifting further apart (data not shown). Parallel exposure of cultures to
MMC is always included in diagnostic studies, and such additional MMC test-
ing is particularly helpful in cases of mild FA mutations or in mosaic cases (see
contribution by Hoehn et al.).

Applications of Cell Cycle Testing in Fanconi Anemia 123

 The FA and non-FA clusters depicted in figure 7 include cultures with a
wide range of their non-cycling (G0/G1) cell fraction, reflecting considerable
variability in their mitogen response. Empirically, cell samples with high pro-
portions of non-cycling cells but normal sG2/GF parameters are overrepre-
sented among patients with acquired non-FA types of aplastic anemia. However,
as far as the comparison between our non-FA and FA cohorts is concerned,
there is no significant difference of the mitogen response of lymphocytes of the
non-FA (G0/G1 mean ⫾ 1 SD, 25.9 ⫾ 11.7%) vs. those of the FA cohort
(G0/G1 mean ⫾ 1 SD, 32.2 ⫾ 18.3%).
How the FA genotype influences the outcome of the sG2/GF parameter is
shown by the example of successful hematopoietic stem cell transplantation
(HSCT) depicted by blue squares in the scattergram of figure 7. At 39 months
prior to HSCT, cell cycle analysis clearly assigned this patient to the FA group
of samples. This was confirmed at 1 month prior to HSCT, when the patient’s
sG2/GF had somewhat decreased but still remained within the cluster of FA
cases. Following engraftment, and at 4, 5, and 11 months after HSCT, sG2/GF
readouts of this patient are exclusively found among the non-FA cases, with
continuously decreasing non-cycling cell fractions. This complete shift to nor-
mal values indicates establishment of complete lymphocyte chimerism.
Immunosuppressive therapy most likely accounts for the rather high non-
cycling cell fractions of the sample taken early after transplantation.

Cell Cycle Analysis as Starting Point for the Comprehensive

Characterization of FA Patients

Figure 8 summarizes the diagnostic strategy currently adopted by the col-

laborating laboratories in Würzburg and Düsseldorf. Three steps follow each
other: the first step involves the laboratory confirmation or exclusion of the
clinical suspicion of FA via cell cycle analysis of 72-h cultures of peripheral
blood lymphocytes. In cases of a normal flow cytometric result but persisting
suspicion of FA on clinical grounds, MMC sensitivity testing of fibroblast cul-
tures is performed in order to exclude revertant mosaicism [20]. If FA is con-
firmed by either of these studies, the second step involves the assignment of the
patient to one of the known complementation groups. This may be performed
by means of CD3/CD28/Il-2-stimulated lymphocytes set apart from the initial

Fig. 8. Current strategy for characterization of FA patients including flow cytometry-

based cell cycle assays. The three-step procedure as currently performed in collaboration
between the Universities of Würzburg and Düsseldorf comprises laboratory confirmation of
the diagnosis, complementation group assignment and mutation analysis.

Schindler/Friedl/Gavvovidis/Kalb/Neveling/Linka/Hanenberg/Kubbies/Hoehn 124
 Clinically suspected diagnosis of FA

Blood sample DNA

Isolated lymphocytes EBV-transformed DNA

cell line (LCL) RNA

Stimulated cultures ⫹/⫺ MMC

Cell cycle analysis

(bivariate flow cytometry)

Absent G2 phase arrest:

Present G2 phase arrest:
negative diagnosis
positive diagnosis
or complete mosaicism

Fibroblast cultures
DNA
in case of continued clinical
RNA
suspicion

Absent G2 phase arrest: Present G2 phase arrest:

no FA FA mosaic

Retroviral complementation group analysis

(upstream FA genes and FANCD2)
with stimulated lymphocytes and confirmation with
LCL, or with fibroblasts in case of mosaicism

Unsuccessful: Successful:
FANCD2 immunoblot mutation analysis

Immunoblots for Immunoblots for

upstream (candidate) FA genes: downstream FA genes:
FAAP250 = FANCM FANCJ
FAAP100 FANCD1
FAAP16 FANCN

Negative for all:

new FA gene searches

Applications of Cell Cycle Testing in Fanconi Anemia 125

 100
EG A B C
80 G1

60
Relative number of cells (%)
G2
40

20
S
0
E F G L
80

60

40

20

0
0 512 0 512 0 512 0 512 1,024
Fluorescence intensity (channel number)

Fig. 9. Example of subgroup typing using retroviral transduction with FA cDNAs.

EG ⫽ vector containing only the enhanced green-fluorescent protein cDNA insert;
A, B, C, E, F, G, L ⫽ FA cDNA of the respective FA complementation groups in bicistronic
constructs with EG. In this example, successful complementation was achieved with the vec-
tor containing FANCB cDNA as evidenced by the decrease of the G2 cell fraction (arrow).

blood sample, by means of an EBV-transformed lymphoblast line, and/or by

means of cultured fibroblasts in case of proven mosaicism. As illustrated in
figure 9, complementation analysis includes transfer of cDNA of the FA genes
upstream of FANCD2 by retroviral vectors and examination of MMC sensitivity
of transduced cell cultures via univariate flow cytometry. The example shown in
figure 9 involves the separate transfer of FANC genes A, B, C, E, F, G and L,
with FANCB being the only cDNA that is capable of reducing the MMC-
induced elevation of the G2 phase cell cycle fraction. This result proves that the
tested cells belong to complementation group FA-B.
As outlined in figure 8, the complementation groups downstream of
FA-D2 (FA-J, FA-D1 and FA-N) and FA-D2 itself are assayed by means of
immunoblotting. Successful assignment of a patient to one of the known FA
complementation groups will be followed by mutation analysis in the corre-
sponding gene using DNA isolated from the initial blood sample and/or DNA
and RNA from a lymphoblast line or a fibroblast strain. If an FA patient cannot
be assigned to a known complementation group, FANCD2 immunoblotting will
provide the information whether an unknown up- or downstream FA gene may
be defective. This may then result in testing candidate genes or initiate other
strategies to search for new FA genes.

Schindler/Friedl/Gavvovidis/Kalb/Neveling/Linka/Hanenberg/Kubbies/Hoehn 126
 Concluding Remarks

Cell kinetic studies such as originally performed by Kubbies et al. [9] have
shown that the cell cycle disturbance of FA cells consists of both, delay and
arrest within the S and G2 phases of the cell cycle. Flow cytometric measure-
ments uniformly reveal a 4c DNA content of the accumulated cells. Since it has
become clear that defective FA proteins cause accumulation of ICLs and
replication-associated DNA-double strand breaks [21, 22], transient or permanent
delay of cell cycle progression appears as the logical consequence for cells that
have acquired such lesions. There are some seemingly contradictory findings as
to where precisely within the cell cycle the accumulation of damaged cells takes
place. When FA cells were treated with DNA-crosslinking agents after comple-
tion of DNA replication, i.e. during the G2 phase of the cell cycle, Akkari et al.
[23] observed neither G2/M arrest nor increased chromosome breakage. The
authors therefore concluded that crosslink-induced accumulations of FA cells
take place in the late S phase, even though these cells have a 4c DNA content,
which formally classifies them as having entered the G2 phase of the cell cycle.
Since the early cell kinetic studies by Kubbies et al. [9] clearly showed delay
and arrest during both, the S and G2 phases of the cell cycle, and since cell
cycle studies uniformly prove the 4c DNA content of cells that accumulate in
response to crosslinking agents, is comes down to a more or less semantic argu-
ment whether to call these accumulations late S or early G2. There is complete
agreement, however, that the accumulations take place prior to entry into the M
phase of the cell cycle, and the data summarized in figure 3 provide proof for
this conclusion. Our demonstration that M-phase cells accumulate in response
to a spindle damaging, but not in response to a DNA-crosslinking agent clearly
separates the M- and S/G2-phase checkpoints in FA cells. These observations
leave little doubt that the S/G2-phase checkpoint functions normally in FA
cells, as previously demonstrated by the careful studies of Heinrich et al. [14]
and Freie et al. [24]. As pointed out by these authors, spontaneous or crosslink-
induced accumulations of FA cells in the G2 phase of the cell cycle do not
reflect an abnormal cell cycle response per se, but rather represent a completely
normal cellular response to unresolved DNA damage. Normal function of the
S/G2 checkpoint is critical for the prevention of apoptosis or the erroneous re-
replication of DNA [25]. Most importantly, S/G2 checkpoint-mediated arrest of
damaged cells prevents their entry into mitosis, sets the stage for homology-
directed DNA repair, and thereby minimizes the undesired perpetuation of
genetic lesions.
In terms of diagnostic applications, we have outlined our approach and
summarized our experience with cell cycle analysis as an alternative to traditional

Applications of Cell Cycle Testing in Fanconi Anemia 127

chromosome breakage analysis for the laboratory confirmation or exclusion of
FA. To the best of our knowledge, there were no false positives among our 394
cases that were classified as FA via cell cycle testing. However, we have iden-
tified a number of patients in whom the initial cell cycle screen gave a close to
normal or completely normal result. Some of these false-negative patients
were retested in fibroblasts because of persisting clinical suspicion of FA. With
only few exceptions, the fibroblast cultures proved MMC-sensitive, classify-
ing these patients as peripheral blood mosaics. As illustrated in the contribu-
tion by Hoehn et al., long-term monitoring of proven or presumptive mosaic
patients via flow cytometry represents another highly informative application
of cell cycle testing. This also holds for the use of cell cycle analysis in
complementation assays. As illustrated in figure 9, failure or success of a
given cDNA to complement the genetic defect in unclassified FA cells is easily
recognized by the persistence or reduction of the G2-phase cell fraction.
We wish to emphasize that retroviral complementation assays are not trivial,
but with appropriate controls, and in experienced hands, the results of such
assays are highly reliable [15, 26, 27]. In our protocol (cf. fig. 8), subtyping of
newly diagnosed FA patients is regularly performed prior to mutation analysis.
There are, of course, many different protocols of how to optimize the
diagnostic process in a rare and genetically heterogeneous disease such as FA.
Most laboratories still rely on chromosome breakage analysis as the first diag-
nostic step, but FANCD2 immunoblotting has also been proposed as a primary
diagnostic tool which however is limited to subtypes upstream of and includ-
ing FA-D2 [5–7]. If mutation analysis is requested or important for clinical
reasons, there again are different strategies. The Amsterdam laboratory, for
example, starts with direct sequencing of the most frequent FA genes [2],
whereas other laboratories, including our own, first attempt to identify the
affected gene via retroviral complementation assays [15, 27]. As we have out-
lined in the flowchart of figure 8, it is important to secure DNA and/or cells at
each level of the diagnostic procedure. In our opinion, cell cycle testing offers
the opportunity for a fairly liberal, low cost screening of patients, in whom the
diagnosis of FA may be considered [18]. Since the clinical manifestations of
FA are so highly variable and frequently non-specific, the vast majority of the
submitted blood samples will test negative. This probably is one of the
strongest arguments in favor of cell cycle testing, which is less laborious and
costly than conventional chromosome breakage studies. However, as we have
shown in this chapter, there are many additional applications of cell cycle
analysis, which go beyond mere diagnostic use and which have made major
contributions to our understanding of the genetic and cell biological basis of
the disease.

Schindler/Friedl/Gavvovidis/Kalb/Neveling/Linka/Hanenberg/Kubbies/Hoehn 128
 Acknowledgements

We are grateful to Birgit Gottwald and Gitta Emmert, Würzburg, for dedicated cell cul-
ture work. We recognize and appreciate the long-standing and invaluable support by Peter S.
Rabinovitch, M.D., Ph.D., Seattle, with mathematical approaches to cell cycle models and
software development.

References

1 Auerbach AD: A test for Fanconi’s anemia. Blood 1988;72:366–367.

2 Joenje H, Oostra AB, Zwaan CM, Pals G: Fanconi anemia; in: Humana Press Series on Hereditary
Cancer Syndromes. 2006, in press.
3 Schroeder-Kurth TM, Auerbach AD, Obe G (eds): Fanconi Anemia, Clinical, Cytogenetic, and
Experimental Aspects. Berlin Heidelberg, Springer-Verlag, 1989, pp 1–378.
4 Schroeder TM, Anschütz F, Knopp A: Spontaneous chromosome aberrations in familial pan-
myelopathy. Humangenetik 1964;1:194–196.
5 Joenje H, Lo ten Foe JR, Oostra AB, van Berkel CG, Rooimans MA, et al: Classification of
Fanconi anemia patients by complementation analysis: evidence for a fifth genetic subtype. Blood
1995;86:2156–2160.
6 Shimamura A, Montes de Oca R, Svenson JL, Haining N, Moreau LA, et al: A novel diagnostic
screen for defects in the Fanconi anemia pathway. Blood 2002;100:4649–4654.
7 Soulier J, Leblanc T, Larghero J, Dastot H, Shimamura A, et al: Detection of somatic mosaicism
and classification of Fanconi anemia patients by analysis of the FA/BRCA pathway. Blood
2005;105:1329–1336.
8 Dutrillaux B, Aurias A, Dutrillaux AM, Buriot D, Prieur M: The cell cycle of lymphocytes in
Fanconi anemia. Hum Genet 1982;62:327–332.
9 Kubbies M, Schindler D, Hoehn H, Schinzel A, Rabinovitch PS: Endogenous blockage and delay
of the chromosome cycle despite normal recruitment and growth phase explain poor proliferation
and frequent endomitosis in Fanconi anemia cells. Am J Hum Genet 1985;37:1022–1030.
10 Kaiser TN, Lojewski A, Dougherty C, Juergens L, Sahar E, Latt SA: Flow cytometric characteri-
zation of the response of Fanconi’s anemia cells to mitomycin C treatment. Cytometry 1982;2:
291–297.
11 Miglierina R, Le Coniat M, Berger R: A simple diagnostic test for Fanconi anemia by flow cytom-
etry. Anal Cell Pathol 1991;3:111–118.
12 Fabio T, Crescenzio N, Saracco P, Leone L, Ponzio G, Ramenghi U: Cell cycle analysis in the diag-
nosis of Fanconi’s anemia. Haematologica 2000;85:431–432.
13 Seyschab H, Friedl R, Sun Y, Schindler D, Hoehn H, et al: Comparative evaluation of diepoxy-
butane sensitivity and cell cycle blockage in the diagnosis of Fanconi anemia. Blood 1995;85:
2233–2237.
14 Heinrich MC, Hoatlin ME, Zigler AJ, Silvey KV, Bakke AC, et al: DNA cross-linker-induced
G2/M arrest in group C Fanconi anemia lymphoblasts reflects normal checkpoint function. Blood
1998;91:275–287.
15 Chandra S, Levran O, Jurickova I, Maas C, Kapur R, et al: A rapid method for retrovirus-mediated
identification of complementation groups in Fanconi anemia patients. Mol Ther 2005;12:976–984.
16 Hendzel MJ, Wei Y, Mancini MA, van Hooser A, Ranalli T, et al: Mitosis-specific phosphorylation of
histone H3 initiates primarily within pericentromeric heterochromatin during G2 and spreads in an
ordered fashion coincident with mitotic chromosome condensation. Chromosoma 1997;106: 348–360.
17 Rabinovitch PS: Regulation of human fibroblast growth rate by both noncycling cell fraction and
transition probability is shown by growth in 5-bromodeoxyuridine followed by Hoechst 33258
flow cytometry. Proc Natl Acad Sci USA 1983;80:2951–2955.

Applications of Cell Cycle Testing in Fanconi Anemia 129

18 Schindler D, Hoehn H: Flow cytometric testing for syndromes with chromosomal instability,
aplastic anemia and related hematological disorders; in: Wegner R-D (ed): Diagnostic
Cytogenetics – Springer Lab Manual. Heidelberg, Springer, 1999, pp 269–281.
19 Schindler D, Hoehn H: Fanconi anemia mutation causes cellular susceptibility to ambient oxygen.
Am J Hum Genet 1988;43:429–435.
20 Joenje H, Arwert F, Kwee ML, Madan K, Hoehn H: Confounding factors in the diagnosis of
Fanconi anaemia. Am J Med Genet 1998;75:228–229.
21 Sala-Trepat M, Rouillard D, Escarceller M, Laquerbe A, Moustacchi E, Papadopoulo D: Arrest of
S-phase progression is impaired in Fanconi anemia cells. Exp Cell Res 2000;260:208–215.
22 Sobeck A, Stone S, Costanzo V, de Graaf B, Reuter T, et al: Fanconi anemia proteins are required
to prevent accumulation of replication associated DNA double-strand breaks. Mol Cell Biol
2006;26:425–437.
23 Akkari YM, Bateman RL, Reifsteck CA, D’Andrea AD, Olson SB, Grompe M: The 4N cell cycle
delay in Fanconi anemia reflects growth arrest in late S phase. Mol Genet Metab 2001;74:
403–412.
24 Freie BW, Ciccone SL, Li X, Plett PA, Orschell CMN, et al: A role for the Fanconi anemia C protein
in maintaining the DNA damage-induced G2 checkpoint. J Biol Chem 2004;279:50986–50993.
25 Zhu W, Dutta A: An ATR- and BRCA1-mediated Fanconi anemia pathway is required for activat-
ing the G2/M checkpoint and DNA damage repair upon rereplication. Mol Cell Biol 2006;26:
4601–4611.
26 Hanenberg H, Batish SD, Pollok KE, Vieten L, Verlander PC, et al: Phenotypic correction of pri-
mary Fanconi anemia T cells with retroviral vectors as a diagnostic tool. Exp Hematol
2002;30:410–420.
27 Casado JA, Callen E, Jacome A, Rio P, Castella M, et al: A comprehensive strategy for the subtyp-
ing of Fanconi Anemia patients: conclusions from the Spanish Fanconi Anemia research network.
J Med Genet 2006 [Epub ahead of print].
28 Poot M, Schmitt H, Seyschab H, Koehler J, Chen U, et al: Continuous bromodeoxyuridine label-
ing and bivariate ethidium bromide/Hoechst flow cytometry in cell kinetics. Cytometry
1989;10:222–226.

Detlev Schindler
Department of Human Genetics, University of Würzburg
Biozentrum, Am Hubland, D–97074 Würzburg (Germany)
Tel. ⫹49 931 888 4089, Fax ⫹49 931 888 4069
E-Mail [email protected]

Schindler/Friedl/Gavvovidis/Kalb/Neveling/Linka/Hanenberg/Kubbies/Hoehn 130
 Schindler D, Hoehn H (eds): Fanconi Anemia. A Paradigmatic Disease for the Understanding of
Cancer and Aging. Monogr Hum Genet. Basel, Karger, 2007, vol 15, pp 131–148

Prenatal Diagnosis of Fanconi Anemia:

Functional and Molecular Testing
A. Bechtolda, R. Kalba, K. Nevelinga, R. Friedla, B. Gottwalda, S. Herterichb,
M. Schmugge Linerc, C. Heilmannd, H. Hanenberge, D. Schindlera
a
Department of Human Genetics, University of Würzburg, bDepartment of Clinical
Chemistry and Pathobiochemistry, University of Würzburg, Würzburg, cDepartment of
Pediatrics, University of Zürich, Zürich, Switzerland; dDepartment of Pediatrics,
Rigshospitalet, Copenhagen, Denmark; eDepartment of Clinical Oncology and Hematology,
University Children’s Hospital, Heinrich-Heine-University, Düsseldorf, Germany

Abstract
There are two main approaches to the prenatal confirmation or exclusion of Fanconi
anemia: functional testing and molecular testing. Functional testing involves the determina-
tion of crosslink sensitivity either by chromosome breakage analysis or cell cycle testing.
Indications for functional testing include ultrasonographic findings of radial ray defects in
the absence of a family history of FA, but also testing of at risk pregnancies in families with
a prior affected child where for various reasons there is no information on complementation
group and disease causing mutations. Since laboratories offering functional prenatal testing
mostly use analysis of chromosome breakage, we here summarize our experience with flow-
cytometric testing of MMC-sensitivity in second trimester amniotic fluid cell cultures. We
show that among a series of 20 pregnancies at risk three amniotic fluid cell cultures were
highly sensitive to MMC as evidenced by their strong G2-phase elevations after exposure to
10 ng/ml of the drug. There were no false positives and no false negatives among our series
suggesting single parameter flowcytometry as a speedy and reliable alternative to conven-
tional chromosome breakage studies for the prenatal diagnosis of FA in situations where only
functional testing can be performed. Molecular testing of course is the method of choice but
requires prior knowledge of complementation group and mutations. Indirect genetic testing
is possible if at least the complementation group is known and DNAs from both parents and
an affected child are available. With the availability of retroviral vectors for rapid subtyping,
and owing to advances in high-throughput mutation analysis including MLPA, direct molec-
ular genetic testing is likely to replace functional testing for most but not all risk pregnancies
in the near future. We illustrate the practice of direct prenatal genetic testing with examples
from families belonging to complementation groups FA-A, FA-C, FA-G and FA-D2. Last but
not least we comment on the implications of preimplantation genetic testing (PGD) as a
high-tech but problematic procedure to preselect potential HLA-matched sibling donors.
Copyright © 2007 S. Karger AG, Basel
 Fanconi anemia (FA) is an inherited disease that is in principle amenable to
symptomatic and curative treatment. However, present day small families often
lack a matching sibling donor for lifesaving hematopoietic stem cell transplanta-
tion, and alternate donor transplants still carry a relatively high risk. Squamous
cell carcinomas remain a lifelong threat. Despite remarkable therapeutic progress
during recent years, Fanconi anemia still remains a serious and threatening dis-
ease. There are major impairments of the quality of life for the patients them-
selves, but also for their siblings and families [1]. Because of the severity of the
disease, the demanding care, and the limitations of therapeutic interventions par-
ents often feel overwhelmed and unable to care for a second or third affected
child. Given a 25% recurrence risk and the possibility of prenatal diagnosis, a
number of parents decide to use this option in their family planning. In contrast to
many other medical interventions, prenatal diagnosis is not a routine procedure
without impact on the physical and psychological well-being of patients [2, 3]. In
the case of Fanconi anemia use of prenatal diagnosis is a reliable but desperate
option in the face of our inability to provide a comprehensive cure for the disease.
The first prenatal diagnosis of FA was reported by Arleen Auerbach and her
coworkers in 1979 who treated amniotic fluid cell cultures from at risk pregnancies
with diepoxybutane (DEB) and studied chromosome breakage [4, 5]. Increased
breakage rates were observed in affected fetuses and also in heterozygotes, but
postnatal confirmation was obtained in only a single case [6]. Increased breakage
rates of chromosomes prepared from amniotic fluid cell cultures that had been
exposed to DEB were also reported in an affected fetus from South Africa [7]. In a
subsequent series, Auerbach et al. did chromosome breakage studies in 30 preg-
nancies at risk for FA and were able to confirm the correctness of their prenatal
findings by postnatal clinical and cytogenetic analyses [8]. This landmark study
demonstrated that the chromosome breakage test for the confirmation or exclusion
of FA could be used with amniotic fluid cell cultures much in the same way and
with similar high degrees of sensitivity and specificity as had been shown and put
into clinical practice for postnatally derived peripheral blood mononuclear cells.
The first prenatal diagnosis using fetal blood was reported by Shipley et al.
[9]. The umbilical cord blood mononuclear cells of the affected fetus showed
strongly elevated spontaneous and MMC-induced breakage rates. In 1985,
Dallapiccola et al. [10] reported the first prenatal diagnosis using cells from a
chorionic villus biopsy (CVS) with fetal trophoblast cells showing elevated
chromosome breakage rates following DEB treatments. The usefulness of CVS-
derived materials for the very early prenatal diagnosis of FA was quickly confirmed
by French and American investigators [11, 12]. By that time, ultrasonography
had become a standard monitoring procedure in prenatal care. A French group
reported a family with three affected children one of whom had been diagnosed
prenatally because of radial ray defects and elevated chromosome breakage rates

Bechtold/Kalb/Neveling/Friedl/Gottwald/Herterich/Schmugge Liner/Heilmann/ 132

Hanenberg/Schindler
[13]. This article is the first to mention that families without matching sibling
donors for their affected children might benefit from prenatal diagnosis and pre-
selection of healthy children as potential hematopoietic stem cell donors.
Prenatal identification of potential donors for umbilical cord blood derived stem
cell transplantation was subsequently advocated by Auerbach et al. [14] and suc-
cessfully applied, as an alternative to conventional bone marrow transplantation,
by Elaine Gluckman and coworkers [15]. Despite obvious ethical problems, a
number of children have since been conceived for the purpose of serving as stem
cell donors for their affected siblings, culminating since 2001 in preimplanta-
tion-based selection of matched sibling donors (see below).
Most recently, flowcytometric determination of MMC-sensitivity was
added to the spectrum of functional tests available for the prenatal confirmation
or exclusion of FA [16]. It was shown that assessment of G2 phase cell blockage
with and without prior exposure to mitomycin C (MMC) permits the distinction
between affected and unaffected fetuses. However, this holds only for conven-
tional single parameter flowcytometry of cultivated amniocytes, whereas
bivariate BrdU-Hoechst/Ethidium bromide staining that is reliable with short
term peripheral blood cultures proved unreliable with amniotic fluid cell cul-
tures, the likely reason being variable degrees of BrdU sensitivity of these cells.
In contrast, bivariate flowcytometry was shown to correctly predict affected
fetuses using umbilical cord mononuclear cells after short term (72 h) culture,
corresponding to what was known from the analysis of postnatally derived
blood cell cultures. Given its speed and simplicity, single parameter flowcytom-
etry was recommended by Bechtold et al. [16] as a screening procedure in preg-
nancies with low a priori risk of FA, e.g. ultrasonographic detection of radial
ray abnormalities in the absence of any family history of FA.
After the first FA gene (FANCC) had been cloned in 1992, the first molec-
ular prenatal diagnosis was reported in 1993 [17], when Murer-Orlando et al.
confirmed the increased chromosome breakage rates found in amniocytes from
an affected pregnancy by mutation analysis using DNA derived from chorionic
villus materials. The fetus was found to carry biallelic mutations in the FANCC
gene. The first molecular prenatal diagnosis of a twin pregnancy at risk for FA
was carried out by Kwee et al. [18] in a family with a previous affected child
identified with biallelic mutations in FANCC. The twins were shown to be het-
erozygous carriers of either the paternal or the maternal mutation and thus cor-
rectly predicted to be unaffected. The authors of these studies pointed out that
prior knowledge of the complementation group and, most importantly, knowl-
edge of the respective types of mutations assure a rapid and reliable diagnosis
of FA during the 10th to 14th week of pregnancy, whereas functional tests
(e.g. chromosome breakage studies using chorionic villus cells or amniocytes)
were considered more time consuming and less reliable.

Prenatal Diagnosis of Fanconi Anemia: Functional and Molecular Testing 133

 Categories of Prenatal Diagnosis

Depending on the family situation, there are two main categories of fami-
lies requesting prenatal confirmation or exclusion of FA. The first category
comprises pregnancies in which radial ray defects are observed as incidental
findings on routine ultrasonographic examination. Even though such radial ray
anomalies can be associated with a myriad of syndromes [19], this type of
anomaly may of course be an important clinical sign of FA. Increased nuchal
translucency as an ultrasonographic sign of an affected fetus has also been
reported [20], as has been abnormal fetal motor behavior [21]. In these situa-
tions, there is no prior evidence for FA in the family, and the only practical way
for confirmation or exclusion of FA is to carry out functional testing of fetal
cells. Functional testing is also required in families with a prior affected child
who had deceased or families who lack information on complementation group
and mutation. To arrive at a timely molecular diagnosis may also be difficult in
pregnancies where parents are not available for testing or in families where the
disease causing genes are large and rather time consuming to type (e.g. FANCA
or FANCD2). Subtyping and mutation analysis during the course of an ongoing
pregnancy has so far only been reported in two families assigned to subtype
FA-C, with FANCC being a relatively small gene where carrier testing and finding
the disease causing mutations is relatively straightforward [22]. The second cat-
egory, and of course optimal situation for the prenatal exclusion or confirma-
tion of FA comprises families with prior affected children in whom the affected
gene and both disease causing mutations are known. In such families, fetal
DNA can be obtained via chorionic villus biopsy early in the course of preg-
nancy, and the results of the molecular analysis, including testing for maternal
cell contamination, will be available within days [18, 23]. With the availability
of rapid subtyping using retroviral vectors [24] and owing to technical advances
of mutation analysis, including high throughput methods and MLPA, prenatal
molecular genetic testing is expected to replace much of the present functional
testing and thus be available for the majority of FA families in the near future.

Category 1: Functional Testing

Functional testing is based on the increased sensitivity of FA cells of all

subtypes to DNA bifunctional alkylating agents such as diepoxybutane (DEB),
mitomycin C (MMC), cisplatinum (CP) or nitrogen mustard (NM). Increased
sensitivity to these agents can be assessed in different ways with identical
results. The classical test assay is chromosome breakage [23, 25], but flowcyto-
metric determination of the G2 phase cell fractions or simple cell growth assays

Bechtold/Kalb/Neveling/Friedl/Gottwald/Herterich/Schmugge Liner/Heilmann/ 134

Hanenberg/Schindler
in the presence and absence of any of the damage inducing agents can also be
employed [26–28]. In addition to carefully controlled cell culture conditions,
monitoring of drug activities and testing of replicate cultures, an important
requirement for the validity of functional testing is the concomitant analysis of
both positive and negative control cells.
A potential drawback of chromosome breakage testing with amniotic fluid
cell cultures is their heterogeneous growth pattern reflecting a variety of cell
types of different fetal origins [29]. Under standard cell culture conditions that
utilize incubators with a mixture of 5% CO2 and ambient air, cells from affected
fetuses may show poor growth since FA cells are known to be sensitive to ambi-
ent oxygen concentrations [30, 31]. For these reasons the yield of good quality
metaphase spreads may be low or even insufficient for a timely and reliable
diagnosis. The problem of variable to poor growth of amniotic fluid cell cul-
tures is less critical and can largely be circumvented with the flowcytometric
assay which determines the DNA content distributions of interphase cells that
represent the vast majority of cells in amniotic fluid cell culture harvests.
Additional advantages of the flowcytometric measurements are their relative
speed, statistical accuracy and partial automation such that for some years our
laboratory has given preference to this procedure. Whereas chromosome break-
age testing is employed by most laboratories performing prenatal diagnosis of
FA, flowcytometric testing is less well known such that our personal experience
with this procedure will be briefly summarized below.
As shown in figure 1, ten-day amniotic fluid cell cultures established from
a second trimester pregnancy were stained with the fluorescent dye DAPI and
their DNA-content profiles were obtained using a flowcytometer of conven-
tional design. Ultrasonographic examination had revealed moderate fetal
growth retardation and bilateral radial ray abnormalities. Even without expo-
sure to MMC, the fetal cells from the pregnancy at risk showed an elevated
G2-phase fraction which became even more pronounced after treatment with
10 ng/ml MMC, indicating sensitivity to the crosslinking agent. Control cells
from a normal pregnancy had much smaller G2-phase fractions and did not
respond to MMC (negative control), whereas control cells from a pregnancy
with proven FA showed a distinct response to MMC (positive control).
Figure 2 depicts the quantitative evaluation of MMC sensitivity in a series
of 20 pregnancies belonging to category 1. For each of the pregnancies tested,
the respective G2-phase fractions without prior MMC treatments are repre-
sented by light squares, and the G2-phase fractions after exposure to 10 ng/ml
MMC are denoted by dark squares. For better visibility, the untreated and
treated G2-phase fractions of each of the tested pregnancies have been con-
nected by solid lines. Cases 5, 11, and 20 proved highly sensitive to MMC as
evidenced by the large difference between the respective G2-phase fractions of

Prenatal Diagnosis of Fanconi Anemia: Functional and Molecular Testing 135

 100 a b

MMC MMC

50 G2
G1 G2

S
0
Relative number of cells (%)

100 c d

MMC MMC

50
G2
G2

0
100 e f

MMC MMC

50 G2

G2

0
1 512 1,024 512 1,024
DNA content

Fig. 1. Flowcytometric determination of MMC sensitivity of amniotic fluid cell cul-

tures from a fetus at risk for FA (a, b) together with negative (c, d) and positive (e, f) control
cultures. Cells were grown without (left hand panels;
MMC) or with 10 ng/ml mitomycin C
(right hand panels, MMC) and analyzed after DAPI staining.

their untreated as opposed to their treated amniotic fluid cell cultures. These
three cases were confirmed as affected either subsequent to termination of
pregnancy or after birth. The data summarized in figure 2 illustrate two further
points: (1) The G2-phase cell fractions of untreated cultures vary considerably,
presumably due to different growth kinetics and variable amounts of G1-phase
tetraploid cells that occur rather frequently in second trimester amniotic fluid
cell cultures. (2) There is an impressive range of the degrees of MMC-response
among FA-negative cell cultures, with cases number 8 and 10 showing no
response at all, but cases 1 and 18 showing some minor degrees of MMC-
sensitivity which, however, is far below of what is observed in affected cases. In
the series of prenatal MMC sensitivity testing summarized in figure 2 there
were no false positive and no false negative cases, suggesting that a reliable

Bechtold/Kalb/Neveling/Friedl/Gottwald/Herterich/Schmugge Liner/Heilmann/ 136

Hanenberg/Schindler
 0 ng/ml MMC 10 ng/ml MMC

FA
19
17
15
Family number

13
11 FA
9
7
5 FA
3
1
0 10 20 30 40 50 60 70
Cells in G2 phase (%)

Fig. 2. Summary of flowcytometric MMC-sensitivity testing in 20 pregnancies at risk.

Light squares: G2-phase fractions without exposure to MMC; dark squares: G2-phase frac-
tions of parallel cultures after exposure to 10 ng/ml MMC. Cases 5, 11 and 20 were identified
as FA due to strongly increased MMC sensitivity.

confirmation or exclusion of FA can be achieved with amniotic fluid cell cul-

tures analyzed by conventional single parameter flowcytometry.

Category 2: Molecular Testing

As pointed out before, molecular testing is the method of choice for the
prenatal diagnosis of FA, provided that the affected gene and both mutations are
known. If only the complementation group is known, molecular testing can also
be applied as long as DNA from both parents and an affected child is available.
With the rapidly increasing catalogue of flanking and intragenic polymorphic
markers, indirect genotyping has turned into a practical and accurate tool in sit-
uations in which the affected gene is known but the disease causing mutations
remain to be detected. This is especially true for large and polymorphic genes
such as FANCA [32].

Indirect Genotyping
Two children of family S. suffered from FA and their complementation
group was determined as FA-A via retroviral gene transfer. The causal muta-
tions in FANCA, however, were still unknown at the time when a third preg-
nancy was under way. Indirect genotyping with microsatellite markers flanking

Prenatal Diagnosis of Fanconi Anemia: Functional and Molecular Testing 137

 D16S413 137 137 137 143 Exon 16: G501S   

D16S3121 79 77 77 83
FANCA

 Exon 22: P643A 



D16S3407 198 202 198 198
D16S303 121 121 119 121 Exon 40: T1328A 

D16S413 137 137 137 137 137 143 Exon 16: G501S     

4cM D16S3121 79 77 79 77 79 83
FANCA




 Exon 22: P643A






1cM
D16S3407 198 198 198 198 198 198
D16S303 121 119 121 119 121 121 Exon 40: T1328A

Fig. 3. Prenatal diagnosis via indirect genotyping in family S. with two affected girls.
Left panel: genotyping using microsatellite markers flanking FANCA. Right panel: genotyp-
ing using intragenic single nucleotide polymorphisms. Fetus at risk marked by arrow.

the FANCA gene was carried out using DNA of both parents, the affected chil-
dren, and fetal DNA obtained via chorionic villus biopsy. Using the markers
D16S413, D16S3121, D16S3407, and D16S303 haplotypes were constructed
encompassing the FANCA locus on chromosome 16. As shown in the left panel
of figure 3, the fetus was found to carry the defective paternal but the wildtype
maternal allele, identifying the fetus as heterozygous carrier not being affected
by FA.
The prediction of an unaffected fetus via microsatellite haplotyping was
confirmed by the analysis of intragenic single nucleotide polymorphisms of
FANCA (fig. 3, right panel). Informative polymorphisms were G501S in exon 16,
P643A in exon 22, and T1328A in exon 40. The parents were heterozygous for
one respectively two of these variants. The fetus was shown to carry the wild-
type maternal and the mutated paternal allele. Because of the concordance of
the results of indirect genotyping with both flanking and intragenic markers, the
fetus was confirmed as an unaffected carrier, and there were no signs of FA
after birth.

Direct Genotyping
Prenatal Diagnosis Involving FANCA: The first child of family V. had
been diagnosed with FA because of typical congenital malformations, aplastic
anemia and elevated chromosome breakage. The child had already died by the

Bechtold/Kalb/Neveling/Friedl/Gottwald/Herterich/Schmugge Liner/Heilmann/ 138

Hanenberg/Schindler
 FANCG Exon 13 1649delC

1 2 3 4

Fig. 4. Core pedigree of family A. Pregnancies 2, 3, and 4 underwent prenatal molecu-

lar testing as described in the text.

time of the second pregnancy, a dizygotic twin pregnancy, for which the parents
requested prenatal diagnosis. Using parental blood and frozen materials from
the deceased child, molecular analysis led to the diagnosis of compound het-
erozygosity for the following mutations in FANCA: a paternal 2-bp deletion at
position 1165 in exon 13 leading to a frameshift, and a maternal 18-bp deletion
within intron 40 affecting splicing (4010delG18), resulting in skipping of
exon 40. In the DNA prepared from chorionic villus biopsies of both twins
there was no evidence for either of the parental mutations indicating that the
twins were homozygous for the FANCA wildtype alleles. In order to exclude the
risk of double puncture of the same chorion in dizygotic twins seven polymor-
phic DNA markers were analyzed. The twins were found to differ in six out of
seven markers such that an erroneous puncture could be excluded. In summary,
family V. could be told that neither of their unborn twins was carrier of a
parental FANCA mutation, which excluded their being affected by FA. At birth,
the fraternal twins were healthy and free of signs of FA.
Prenatal Diagnosis Involving FANCG: As shown in the family pedigree
(fig. 4), the first child of family A. had been diagnosed with FA. During the sec-
ond pregnancy amniotic fluid cells were analyzed in an outside laboratory and
chromosome breakage rates were reported as suggestive of FA, even though the
number and quality of the available metaphases was reported as too low for a
reliable diagnosis. Complementation analysis with retroviral vectors assigned
the amniotic fluid cells to complementation group FA-G, and mutation analysis
detected a homozygous 1-bp deletion in exon 13 (1649delC). Homozygosity
was expected because of consanguinity. The mutation leads to a frameshift and
an unstable protein as described in the literature [33]. After induced abortion the
diagnosis was confirmed via flow cytometric measurement of lymphocytes from

Prenatal Diagnosis of Fanconi Anemia: Functional and Molecular Testing 139

fetal heart blood: the response of mononuclear blood cells to PHA was reduced,
there was increased cellular decay from the G0/G1 cell cycle fraction, and cell
cycle blockage in the G2 phase was prominent. Even without MMC challenge,
the ratio of sum of G2 phases vs. growth fraction was elevated to 0.405 which
corresponds to the level seen in proven FA patients.
During the following pregnancy genomic DNA from chorionic villus cells
was sequenced and the 1649delC mutation in FANCG exon 13 was detected
next to the wildtype allele, showing that the fetus was heterozygous and not at
risk of developing FA. Postnatally lymphocytes were analyzed via flow
cytometry. The results of these functional studies confirmed the conclusion of
the prenatal genotyping, as cell cycle distributions of neither untreated nor
MMC-treated fetal cells revealed evidence for G2 phase arrest.
The fourth pregnancy in this family ended with intrauterine fetal death.
DNA derived from the stillbirth was sequenced and the exon 13 mutation
(1649delC) was detected in a homozygous state indicating that the fetus would
have been affected by FA. Proving that the stillborn child would have suffered
from FA relieved the grief of the parents over the pregnancy loss.
Prenatal Diagnosis Involving FANCC: Family I. has a child with FA
assigned to complementation group FA-C. The parents requested prenatal diag-
nosis during their second pregnancy. cDNA from bone marrow cells of the first
child was isolated and subjected to exon scanning sequencing of the entire
FANCC gene. A homozygous frameshift mutation leading to premature termi-
nation of translation at seven codons downstream was found in exon 4
(454_455delAA; V152fsX159). The consanguineous parents were confirmed
as heterozygous mutation carriers. There was no doubt about the pathogenicity
of the mutation since it had been observed previously in another proven FA
patient. DNA analysis from fetal chorionic villus materials revealed homozy-
gosity for the mutation in exon 4, indicating an affected fetus. The results of the
prenatal testing in family I. are illustrated in figure 5.
Prenatal Diagnosis Involving FANCD2: Family G. already had two chil-
dren suffering from FA. The complementation group was determined via retro-
viral gene transfer and shown to be FA-D2. However, by the time of the third
pregnancy the underlying mutations had not yet been detected. Functional test-
ing using single parameter flow cytometry of cultured amniocytes showed a
normal response to MMC, suggesting an unaffected fetus. In order to confirm
the result of functional testing, an effort was made to detect the causal muta-
tions in FANCD2 via cDNA analysis of all family members. In case of the
FANCD2 gene, mutation analysis is complicated by the presence of pseudo-
genes, such that special primers had to be constructed [34]. The affected chil-
dren showed skipping of FANCD2 exon 22 resulting from the homozygous
mutation c.1948-16CT. This change could be shown to weaken the splice

Bechtold/Kalb/Neveling/Friedl/Gottwald/Herterich/Schmugge Liner/Heilmann/ 140

Hanenberg/Schindler
 FANCC, Exon 4, 454_455delAA

Fetus
homozygous
affected

Mother
heterozygous

Father
heterozygous

Fig. 5. Partial sequence of FANCC exon 4 of parents and fetus of family I. Both parents
are heterozygous carriers of an identical 2-bp deletion within exon 4. The fetus was found to
be homozygous for the defective parental allele and thus will be affected.

acceptor recognition of exon 21 by disruption of the pyrimidine-rich tract

preceding the splice acceptor in intron 21 [34]. The consanguineous parents
were heterozygous carriers, and the mutation was subsequently found in other
FA families [34]. Like the parents the fetus in question turned out to be a
heterozygous carrier, as evidenced by the simultaneous presence of exon 22
skipping and retention (fig. 6).
Since this was the first prenatal diagnosis in a FANCD2 family that was
carried out in our laboratory and given the slight background of aberrant exon
22 skipping that was observed in the affected sister (fig. 6) and even in healthy
control individuals, it was decided to confirm the prediction of an unaffected,
heterozygous carrier fetus by indirect genotyping. The affected children proved
homozygous for three polymorphisms (IVS5-14delTT, IVS1638AC und
IVS40103CT) whereas the paternal and the fetal cells showed heterozygos-
ity at these loci. Unfortunately, neither of the above polymorphic markers nor
six additional ones [1222AG (V374V), 1509CT (N503N), 2141AC
(P714L), 4098TG (L1366L), 4453GA and 4478AG] were informative
with regard to the maternal allele. They were homozygous in both mother and
fetus indicating a high homogeneity of the maternal alleles. Since the mother
clearly was not affected by FA, the observed constellation is not at variance
with the presumed heterozygosity of the fetus. It was concluded that the results
of prenatal molecular testing would be best explained by heterozygosity of the

Prenatal Diagnosis of Fanconi Anemia: Functional and Molecular Testing 141

 Exon 21 Exon 22
Exon 23

Fetus

Exon 21 Exon 23

Index patient

Fig. 6. Partial sequence of exons 21–23 of amniocytes (upper panel) and blood lym-
phoblasts (lower panel) showing homozygous skipping of exon 22 in the index patient of
family G. and the heterozygous nature of the mutation in amniocytes of the tested fetus. Note
slight background of exon 22 signals in the index patient.

FANCD2 mutation (c.1948-16CT, r.1948_2021del74), identifying the fetus as

not affected.
Another family (family S.) assigned to complementation group FA-D2
through an index patient was less fortunate. As shown in figure 7, testing of
amniocyte DNA revealed the presence of both parental mutations, identifying
the fetus as compound heterozygous and thereby affected. The index patient in
that family was compound heterozygous for a 4-bp deletion at position c3453
and an Alu insertion in intron 4 leading to skipping of exon 5 due to insertional
mutagenesis of the Alu element within the AT-rich target sequence of intron 4
[34]. Cultured amniocytes from the present pregnancy revealed exon 5 skipping
together with the deletion c3453_3456delCAAA. The mother was shown to be
a carrier of the Alu insertion in intron 4, whereas the father carried the 4-bp
deletion within a poly A stretch at position 3453. Given that ultrasonographic
examination revealed severe growth retardation, renal aplasia, and an unspeci-
fied heart defect, there was no doubt about the pathogenic nature of these

Bechtold/Kalb/Neveling/Friedl/Gottwald/Herterich/Schmugge Liner/Heilmann/ 142

Hanenberg/Schindler
 Intron 4 Intron 4
Alu Yb8

Maternal mutation
(IVS4-57ins298)

Exon 34

Paternal mutation
(c.3453delCAAA)

Fig. 7. Prenatal diagnosis in family S. Sequencing of FANCD2 intron 4 and exon 34

DNA derived from fetal cells revealed the presence of both the maternal (top) and the paternal
(bottom) defective allele, establishing the diagnosis of compound heterozygosity in the fetus.

mutations. In view of the result of the genetic testing the parents opted for ter-
mination of the pregnancy.

Comment on Molecular Testing in Pregnancies at Risk for FA

Even though prenatal molecular testing seems straightforward in families

of category 2 in whom the affected gene, or better yet both mutations are
known, our examples show that in practice the available molecular information
may be incomplete or, conversely, too novel to be interpreted with sufficient
confidence. This is the case with nearly all mutations that have not been previ-
ously associated with the disease phenotype, or where there is no possibility of
validation of the pathogenic nature of a given mutation via the examination of
an index patient. In the absence of comprehensive fetal pathology services in
many hospitals, every effort should be made to confirm a pathologic prenatal
result through functional testing of fetal cells. Although FA is a childhood
disease, we still witness second and third pregnancies before a correct clinical

Prenatal Diagnosis of Fanconi Anemia: Functional and Molecular Testing 143

diagnosis has been made in the affected child. A timely and correct diagno-
sis is therefore urgent, and in families wishing to have more children diagnos-
tic efforts should not be limited to the mere confirmation of the clinical
suspicion of FA by way of a chromosome breakage or cell cycle test. Ideally,
identification of the causal mutations should follow as soon as these tests have
yielded a positive result. Knowledge of the causal gene and of the causal muta-
tions not only provides valuable prognostic information (which will become
more useful with each tested child), but also enables parents to request the
optimal kind of prenatal diagnosis (i.e. direct genetic testing) in subsequent
pregnancies if they wish to make use of this option.

Preimplantation Genetic Diagnosis (PGD)

Currently hematopoietic stem cell transplantation (HSCT) represents the

only therapy for FA that is curative with respect to bone marrow function.
A complete cure of the disease is not possible but HSCT can reconstitute nor-
mal bone marrow function and prevent hematological malignancies. The risk
of solid tumors, however, persists in the patients since the transplanted cells
only replace those in the bone marrow. If HSCT is considered and no HLA-
matched sibling donor is available, PGD represents a novel option in many
countries. Successful HSCT of FA patients following PGD for selection of an
HLA-identical healthy donor has been reported from the US and Israel during
the last few years [35–37]. Whereas prenatal HLA testing for the purpose of
umbilical cord blood transplantation had been advocated and put into practice
since the early nineties [14, 38], the first successful PGD-based combined
HLA and FA testing was performed in 2001 by the Chicago group. Of 33 in
vitro fertilized eggs blastomere biopsy and molecular testing identified 5
embryos suitable for transfer of which a single embryo resulted in the birth of
a healthy child. His chord blood cells were harvested and used for HSCT of
his affected 6-year-old sister [35, 36]. Her clinical and hematological condi-
tion improved dramatically and remained stable after the procedure. The
worldwide second successful PGD involved the selection of an HLA compat-
ible donor free of FA as umbilical blood cell donor for his 3.5-year-old
affected brother [37].
One of the advantages of PGD as opposed to conventional prenatal diag-
nosis consists in the fact that couples are no longer confronted with the ethical
dilemma of induced abortion. Main disadvantages are health risks implicit in
the procedure, considerable financial and emotional burden for the families
involved, and the relatively low success rate of the procedure. In the above
mentioned successful cases 5 respectively 3 cycles of hormonal stimulation

Bechtold/Kalb/Neveling/Friedl/Gottwald/Herterich/Schmugge Liner/Heilmann/ 144

Hanenberg/Schindler
and egg harvesting were necessary to achieve a single surviving child. In addi-
tion to these low rates of success, there are many unsolved ethical questions.
For example, there is the questionable matter of parental motivation: parents
might want to conceive a child with the intention of raising it just like any
other of their children although the child was primarily conceived for the pur-
pose of stem cell donation [39]. Transplantation associated risks may also play
a role. If HSCT with umbilical cord blood cells fails (which is not infrequent
due to the relatively low amount of putative stem cells that can be obtained
from a single umbilical cord specimen), the donor child might be subjected to
repeated stem cell harvests during the first months of life which always carries
certain health risks. Some parents might even just want to obtain fetal stem
cells and abort the fetus after fulfilling its ‘purpose’ as stem cell donor.
Another major concern is the possible abuse of PGD for the purpose of sex
selection.
At present, the overall success rate of PGD-based selection of HLA-
matched sibling donors is still unknown, but certainly not as favorable as the
publicity surrounding the procedure suggests. Out of 20 families participating
in a recent meeting of the US American FA support groups 7 couples reported
PGD attempts (Ralf Dietrich, personal communication, 2006). A number of
these families have experienced repeated (3–5) unsuccessful PGD cycles. One
woman went through nine consecutive pregnancy losses without achieving a
liveborn child. Another family had attempted PGD 14 times without any suc-
cess. The families voiced concerns about their physical and psychological
strains, their disappointed hopes, their doubts and ambiguities, and last but not
least about the high financial burden of the procedure. To simultaneously
determine HLA type and to rule out FA in the minute amount of genetic mate-
rial that is available from the nucleus of a single blastomere is an unprece-
dented technical challenge whose obvious limitations one tends to forget. This
methodology clearly approaches the limits of what is biologically possible and
technically feasible. The optimistic reports of the very few cases where PGD
has been successful for the selection of suitable donors in FA families portray
PGD as a safe and reliable procedure that can be routinely applied. Nothing
could be further from the truth. In one of the US-families the first cycle
resulted in a pregnancy with twin girls who were assumed to be FA-negative
and HLA-identical to the affected child. However, molecular testing after
birth revealed that both girls were in fact affected by FA. In addition, only one
of the girls had the desired HLA characteristics. The family now faces the
problem of having three affected children, and needing two instead of just one
compatible HSCT donor. Due to the undisputable fact that in the case of FA
most attempts of PGD-based matched sibling donor selection have ended with
failures, there is increasing skepticism towards PGD in US-American families.

Prenatal Diagnosis of Fanconi Anemia: Functional and Molecular Testing 145

It is felt that the very few successful cases (at most 5 cases worldwide, of
which only two have been published to date) might not justify the very high
initial hopes of many families. An additional criticism voiced by the families
is that adequate information concerning the risks, strains and limitations of
the procedure have not been provided in many situations. Since allogeneic
HSCT is increasingly successful in FA, it is doubtful whether such a high-risk,
high-cost and low-success procedure like PGD will survive the test of time
and practicability.

Acknowledgements

The authors wish to thank the Fanconi Anemia families requesting prenatal diagnosis
for their patience and cooperation. Special thanks go to Ralf Dietrich of the German family
support group for sharing information on the experiences and attitudes of American FA fam-
ilies with PGD.

References

1 Hutson SP, Alter BP: Experiences of siblings of patients with Fanconi anemia. Pediatr Blood
Cancer 2006;48:72–79.
2 Schroeder TM: Ethical conflicts in genetic counselling during pregnancies (author’s transl.).
Monatsschr Kinderheilkd 1982;130:71–74.
3 Auerbach AD: Umbilical cord blood transplants for genetic disease: diagnostic and ethical issues
in fetal studies. Blood Cells 1994;20:303–309.
4 Auerbach AD, Warburton D, Bloom AD, Chaganti RS: Preliminary communication: prenatal
detection of the Fanconi Anemia gene by cytogenetic methods. Am J Hum Genet 1979;31:77–81.
5 Auerbach AD, Adler B, Chaganti RS: Prenatal and postnatal diagnosis and carrier detection of
Fanconi anemia by a cytogenetic method. Pediatrics 1981;67:128–135.
6 Voss R, Kohn G, Shaham M, Benzur Z, Arnon J, et al: Prenatal diagnosis of Fanconi anemia. Clin
Genet 1981;20:185–190.
7 Marx MP, Dawson B, Heyns AD: Prenatal diagnosis of Fanconi’s anemia. S Afr Med J 1982;62:
348.
8 Auerbach AD, Sagi M, Adler B: Fanconi anemia: prenatal diagnosis in 30 fetuses at risk.
Pediatrics 1985;76:794–800.
9 Shipley J, Rodeck CH, Garrett C, Galbraith J, Giannelli F: Mitomycin C induced chromosome
damage in fetal blood cultures and prenatal diagnosis of Fanconi’s anaemia. Prenat Diagn 1984;4:
217–221.
10 Dallapiccola B, Doria Lamba Carbone L, Ferranti G, Cristiani ML, Dagna Bricarelli F:
Monitoring of pregnancies at risk for Fanconi’s anemia by chorionic villi sampling. Acta Haematol
1985;73:157–158.
11 Boue J, Deluchat C, Nicolas H, Oury JF, Dumez Y: Prenatal diagnosis of genetic diseases using
chorionic villi. J Genet Hum 1986;34:221–233.
12 Auerbach AD, Min Z, Ghosh R, Pergament E, Verlinsky Y, Nicolas H, Boue J: Clastogen-induced
chromosomal breakage as a marker for first trimester prenatal diagnosis of Fanconi anemia. Hum
Genet 1986;73:86–88.
13 Bonnet-Gajdos M, Vasmant D, Baruchel S, Mamou JF, Boue J, Lasfargues G: Prenatal diagnosis
of Fanconi’s anemia. Presse Med 1985;14:1612.

Bechtold/Kalb/Neveling/Friedl/Gottwald/Herterich/Schmugge Liner/Heilmann/ 146

Hanenberg/Schindler
14 Auerbach AD, Liu Q, Ghosh R, Pollack MS, Douglas GW, Broxmeyer HE: Prenatal identification
of potential donors for umbilical cord blood transplantation for Fanconi anemia. Transfusion
1990;30:682–687.
15 Gluckman E, Devergie A, Bourdeau-Esperou H, Thierry D, Traineau R, et al: Transplantation of
umbilical cord blood in Fanconi’s anemia. Nouv Rev Fr Hematol 1990;32:423–425.
16 Bechtold A, Friedl R, Kalb R, Gottwald B, Neveling K, et al: Prenatal exclusion/confirmation of
Fanconi anemia via flow cytometry: a pilot study. Fetal Diagn Ther 2006;21:118–124.
17 Murer-Orlando M, Llerena JC Jr, Birjandi F, Gibson RA, Mathew CG: FACC gene mutations and
early prenatal diagnosis of Fanconi’s anaemia. Lancet 1993;342:686.
18 Kwee ML, Lo Ten Foe JR, Arwert F, Pals G, Madan K, et al: Early prenatal diagnosis of Fanconi
anaemia in a twin pregnancy, using DNA analysis. Prenat Diagn 1996;16:345–348.
19 Goldfarb CA, Wall L, Manske PR: Radial longitudinal deficiency: the incidence of associated
medical and musculoskeletal conditions. J Hand Surg 2006;7:1176–1182.
20 Tercanli S, Miny P, Siebert MS, Hosli I, Surbek DV, Holzgreve W: Fanconi anemia associated with
increased nuchal translucency detected by first-trimester ultrasound. Ultrasound Obstet Gynecol
2001;17:160–162.
21 de Vries JI, Laurini RN, Visser GH: Abnormal motor behaviour and developmental postmortem
findings in a fetus with Fanconi anaemia. Early Hum Dev 1994;36:137–142.
22 Merrill A, Rosenblum-Vos L, Driscoll DA, Daley K, Treat K: Prenatal diagnosis of Fanconi ane-
mia (Group C) subsequent to abnormal sonographic findings. Prenat Diagn 2005;25:20–22.
23 Auerbach AD: Fanconi anemia: genetic testing in Ashkenazi Jews. Genet Test 1997;1:27–33.
24 Chandra S, Levran O, Jurickova I, Maas C, Kapur R, et al: A rapid method for retrovirus-mediated
identification of complementation groups in Fanconi anemia patients. Mol Ther 2005;12:976–984.
25 Joenje H, Oostra AB, Zwaan CM, Pals G: Fanconi anemia; in Humana Press Series on Hereditary
Cancer Syndromes. Humana Press, 2006 in press.
26 Berger R, Le Coniat M, Gendron MC: Fanconi anemia. Chromosome breakage and cell cycle
studies. Cancer Genet Cytogenet 1993;1:13–16.
27 Seyschab H, Friedl R, Sun Y, Schindler D, Hoehn H, Hentze S, Schroeder-Kurth T: Comparative
evaluation of diepoxybutane sensitivity and cell cycle blockage in the diagnosis of Fanconi anemia.
Blood 1995;85:2233–2237.
28 Schindler D, Hoehn H: Flow cytometric testing for syndromes with chromosomal instability,
aplastic anemia and related hematological disorders; in Wegner R-D (ed): Diagnostic
Cytogenetics. Berlin/Heidelberg, Springer Verlag, 1999, pp 269–281.
29 Hoehn H: Amniotic fluid cell culture; in Milunsky A (ed): Genetic Disorder and the Fetus, ed 4.
Baltimore, The John Hopkins University Press, 1998, pp 128–149.
30 Joenje H, Arwert F, Eriksson AW, de Koning H, Oostra AB: Oxygen-dependence of chromosomal
aberrations in Fanconi’s anaemia. Nature 1981;290:142–143.
31 Schindler D, Hoehn H: Fanconi anemia mutation causes cellular susceptibility to ambient oxygen.
Am J Hum Genet 1988;43:429–435.
32 Levran O, Diotti R, Pujara K, Batish SD, Hanenberg H, Auerbach AD: Spectrum of sequence vari-
ations in the FANCA gene: an International Fanconi Anemia Registry (IFAR) study. Hum Mutat
2005;25:142–149.
33 Demuth I, Wlodarski M, Tipping AJ, Morgan NV, de Winter JP, et al: Spectrum of mutations in the
Fanconi anaemia group G gene, FANCG/XRCC9. Eur J Hum Genet 2000;8:861–868.
34 Kalb R, Neveling K, Hoehn H, Schneider H, Linka Y, et al: Hypomorphic mutations in the key
Fanconi anemia gene FANCD2 sustain a significant group of FA-D2 patients with severe pheno-
type. Am J Hum Genet 2007; in press.
35 Verlinsky Y, Rechitsky S, Schoolcraft W, Strom C, Kuliev A: Preimplantation diagnosis for
Fanconi anemia combined with HLA matching. JAMA 2001;285:3130–3133.
36 Grewal SS, Kahn JP, MacMillan ML, Ramsay NK, Wagner JE: Successful hematopoietic stem cell
transplantation for Fanconi anemia from an unaffected HLA-genotype-identical sibling selected
using preimplantation genetic diagnosis. Blood 2004;103:1147–1151.
37 Bielorai B, Hughes MR, Auerbach AD, Nagler A, Loewenthal R, Rechavi G, Toren A: Successful
umbilical cord blood transplantation for Fanconi anemia using preimplantation genetic diagnosis
for HLA-matched donor. Am J Hematol 2004;77:397–399.

Prenatal Diagnosis of Fanconi Anemia: Functional and Molecular Testing 147

38 Gluckman E, Broxmeyer HA, Auerbach AD, Friedman HS, Douglas GW, et al: Hematopoietic
reconstitution in a patient with Fanconi’s anemia by means of umbilical-cord blood from an HLA-
identical sibling. N Engl J Med 1989;321:1174–1178.
39 Alby N: The child conceived to give life. Bone Marrow Transplant 1992;9:95–96.

Dr. Astrid Bechtold

Department of Human Genetics, Biozentrum
Am Hubland, University of Würzburg
97074 Würzburg (Germany)
Tel. 49 931 8884068, Fax 49 931 8884434
E-Mail [email protected]

Bechtold/Kalb/Neveling/Friedl/Gottwald/Herterich/Schmugge Liner/Heilmann/ 148

Hanenberg/Schindler
 Schindler D, Hoehn H (eds): Fanconi Anemia. A Paradigmatic Disease for the Understanding of
Cancer and Aging. Monogr Hum Genet. Basel, Karger, 2007, vol 15, pp 149–172

Revertant Mosaicism in Fanconi Anemia:

Natural Gene Therapy at Work
H. Hoehna, R. Kalba, K. Nevelinga, R. Friedla, A. Bechtolda, S. Herterichb,
Y. Sunc, B. Gruhnd, H. Hanenberge, D. Schindlera
a
Department of Human Genetics, Biocenter, University of Würzburg, bDepartment of
Clinical Chemistry and Pathobiochemistry, University of Würzburg, Würzburg,
Germany; cKey Laboratory of Molecular and Medical Genetics, Nanjing Medical
University, Nanjing, P.R. China; dDepartment of Pediatrics, University of Jena, Jena,
e
Department of Pediatric Hematology and Oncology, Children’s Hospital, University
of Düsseldorf, Düsseldorf, Germany

Abstract
One out of four to five patients with Fanconi anemia experience a reversion or attenu-
ation of their constitutional mutations during their lifetime. If the reversion event takes place
in a bone marrow stem cell or in an early precursor cell of hematopoiesis, peripheral blood
cell counts may gradually recover, leading to improved quality of life. At the beginning of
this process, MMC testing will reveal a mixture of MMC sensitive and MMC resistant blood
lymphocytes, but after several years MMC sensitive cells (the original FA-cells) may be com-
pletely replaced by the progeny of the reverted progenitor cell such that the confirmation of
FA requires testing of patient fibroblasts. Molecular analysis reveals the presence of the dis-
ease causing biallelic mutations in fibroblast-derived DNA whereas MMC-resistant blood
cells show only a single defective allele, explaining their phenotypic reversion. The mech-
anisms leading to revertant mosaicism include intragenic mitotic recombination (crossing
over and gene conversion), back mutation, and compensatory second site mutations.
Evidence for each of these mechanisms has been obtained in mosaic FA-patients, but their
molecular details are not fully understood. Compound heterozygosity facilitates some of
these mechanisms, but reversions have also been observed in homozygous patients. Patients
belonging to subtypes FA-A, FA-C, FA-D1, FA-D2 and FA-L have developed revertant
mosaicism, with subtypes FA-A and FA-D2 being most frequently involved. Even though the
phenomenon of revertant mosaicism has been well documented in FA, there still are many
questions: we do not know whether the progeny of a single reverted blood stem or progenitor
cell would be able to sustain lifelong hematopoiesis, whether revertant mosaicism provides
protection against hematopoietic malignancy, or whether it would be possible to deliberately
increase the rate of somatic reversions in order to improve chances for ‘natural gene therapy’.
Prospective and long-term follow-up studies are needed to answer these questions.
Copyright © 2007 S. Karger AG, Basel
 The restoration to normal (or close to normal) function of a gene previ-
ously inactivated by a constitutional mutation is referred to as ‘reversion’. The
coexistence of reverted and wildtype cells within otherwise isogenic organisms
is referred to as ‘reverse’ or ‘revertant’ mosaicism. In contrast to Lo ten Foe
et al. [1] we prefer the term ‘revertant’ which has been defined as ‘a mutant
which has regained, partially or completely, the wildtype phenotype by either a
genetic or nongenetic mechanism of reversion’ [2]. As pointed out by Jonkman
et al. [3], the term ‘reverse’ simply implies a change in the opposite direction,
but does not necessarily denote complete or partial restoration of a defective
function. The phenomenon of revertant mosaicism has been observed in a num-
ber of inherited diseases, including a skin disease (epidermolysis bullosa), a
metabolic disease (tyrosinemia type I), a number of primary immunodeficiency
syndromes (Wiskott-Aldrich syndrome, X-linked severe combined immuno-
deficiency, ADA-deficiency), and in two chromosomal instability syndromes
(Bloom syndrome and Fanconi anemia) (see reviews by Youssoufian and
Pyeritz [4] and Hirschhorn [5]).

Revertant Mosaicism in FA: From Observations to Mechanisms

In the case of FA, revertant mosaicism has both clinical and theoretical
implications [1, 6]. These implications relate to the altered in vivo behavior of the
reverted cells and depend on the developmental stage, the blood cell lineage
affected by the reversion event, and the clinical situation of the mosaic patient. In
self-renewing tissues, the reversion event may convey a growth advantage to the
reverted cells such that their progeny may outgrow and, ultimately, replace the
defective cells. Such a positive outcome would suggest that gene therapy might be
successful in FA and other stem cell diseases. Because of these positive prospects,
revertant mosaicism has also been referred to as ‘natural gene therapy’ [3, 6–9].
As early as in 1983, the group of Hans Joenje in Amsterdam noted an unusual
response to bifunctional alkylating agents in a 22-year-old patient with Fanconi
anemia. Unexpectedly, only 40% of the patient’s peripheral blood mononuclear
cells proved highly sensitive to these agents as evidenced by strongly increased
chromosome breakage while 60% of the patient’s cells responded like cells
from healthy individuals [10]. The patient’s advanced age and the observation
of two distinctive peripheral blood cell populations with striking differences in
crosslink sensitivity leave little doubt that this patient represents one of the first
well-documented cases of revertant mosaicism in Fanconi anemia.
At the same time, other authors noted that among clinically diagnosed FA
patients between 1 to 3 out of 9 tested blood samples proved MMC- or DEB-
resistant [11, 12]. There was some concern that loss of activity during storage of

Hoehn/Kalb/Neveling/Friedl/Bechtold/Herterich/Sun/Gruhn/Hanenberg/Schindler 150
the DEB or MMC stock solutions could give rise to false-negative chromosome
breakage tests [13], but it was generally felt that a subgroup of patients with
clinical manifestations of FA might lack the FA-typical pattern of crosslink sen-
sitivity. The existence of such a subgroup was convincingly documented by
Arleen Auerbach and Traute Schroeder-Kurth in their first report on the
International Fanconi Anemia Registry (IFAR) [14]. Among a series of 162 FA
patients Auerbach and colleagues noted 7 patients who appeared to have two
populations of cells. 60–81% of DEB-treated cells from these patients dis-
played no evidence for chromosome breakage, while the remainder exhibited
high numbers of breaks and exchanges typical of FA cells. Since three of these
cases were from multiplex sibships, the authors came to the conclusion that the
‘phenomenon of two-cell populations does not seem to be related to genetic
heterogeneity in FA’ [14]. Clearly, what the authors had observed but not recog-
nized as such at that time was the phenomenon of revertant mosaicism.
Laboratories which routinely established lymphoid cell lines from FA patients
noted that around 30% of such EBV transformed cell lines that originate from B-
lymphocytes became crosslink resistant during the course of prolonged in vitro
culture. However, it was not until 1997 when a landmark study by the Amsterdam
group confirmed that between 20 and 30% of FA patients show evidence for a
mixture of MMC-sensitive and MMC-resistant peripheral blood lymphocytes [1].
Molecular analysis of these two types of cells provided unambiguous proof of the
concept of revertant mosaicism. In the study by Lo ten Foe and colleagues [1] the
expected biallelic mutations in a given FANC gene were present in DNA from the
MMC-sensitive subpopulation, whereas the MMC-resistant cells were found to
carry only a single defective allele. Since in recessive diseases a single intact copy
of a disease gene suffices for the maintenance of function, the study by Lo ten Foe
et al. [1] provided molecular proof of rescue of defective gene function via somatic
reversion of a constitutional mutation. In the same year, Jonkman and coworkers
published their landmark study on somatic reversion as molecular cause of mRNA
rescue in epidermolysis bullosa, suggesting that gene conversion might be the
mechanism by which the defective function of one of the collagen XVII alleles had
been restored in unaffected skin areas of their patient [15].

Detection of Somatic Reversion in FA

In FA patients, unexplained improvements of blood counts may herald the

emergence of revertant mosaicism. A mosaic patient may also be older than the
average FA patient, and the diagnosis of revertant mosaicism is frequently
established at a time when these patients reach adolescence or adulthood
without experiencing prior major medical problems [9, 10, 16–19]. However,

Revertant Mosaicism in Fanconi Anemia: Natural Gene Therapy at Work 151

long before an improving hematological situation suggests the possibility of
reversion, the diagnostic laboratory may have noted the co-existence of MMC-
sensitive and MMC-resistant cell populations in peripheral blood mononuclear
cell cultures as they are routinely used for the confirmation or exclusion of the
clinical suspicion of FA. Likewise, FA research laboratories that routinely
establish EBV-transformed lymphoid cell lines from peripheral blood cells may
have noted the emergence of DEB or MMC resistance in a considerable propor-
tion of these (B-lymphocyte derived) cell lines. Emergence of drug resistance
signals loss of the FA cellular phenotype in these cell lines and precludes their
use for complementation and other studies. Between 20 and 30% of lym-
phoblastoid cell lines derived from FA patients have been observed to undergo
the transition from MMC sensitivity to MMC resistance [6, 20].

Chromosome Breakage Test for the Detection of Reverted Cells

Landmark studies by the Auerbach and Schroeder-Kurth laboratories (sum-

marized in their common monograph on Fanconi anemia [21]) have defined cul-
ture conditions, clastogen types, clastogen concentrations, and numerous other
criteria for evaluation of chromosome breakage rates. In particular, the parame-
ter ‘number of breaks per aberrant cell’ was found to optimally define whether or
not cells are sensitive to a given clastogen [22]. These criteria are still valid today
and prove useful for the recognition of revertant mosaicism. In mosaic cases,
laboratories using chromosome breakage analysis will typically detect a bimodal
distribution of breakage rates whereas non-mosaic cases show unimodal break-
age distributions. Figure 1 depicts chromosome breakage distributions of 72-h
peripheral blood mononuclear cell cultures evaluated (from top to bottom) at 0,
50 and 100 ng/ml MMC. The panels of figure 1a and 1c illustrate examples of
unimodal distributions. Such distributions classify the respective cell sample as
either MMC-resistant (fig. 1a) or MMC-sensitive (fig. 1c). According to these
criteria, the donor of blood sample 1a would be classified as non-FA, and the
donor of blood sample 1c would be classified as FA-positive. Following expo-
sure to MMC, FA positive cases typically exhibit cells with ten or more breaks
per metaphase, reflecting the innate MMC-sensitivity of the FA genotype (panel
1c). MMC-treated cultures of exceptional individuals with a questionable diag-
nosis of FA may show intermediate types of MMC-sensitivity, such as shown in
panel 1b. This pattern is in essence very similar to the pattern observed in gen-
uine FA-mosaic cases (illustrated in panels 1d–f). However, the patient whose
chromosome breakage study is shown in figure 1b developed B-cell lymphoma
rather than AML, and he tolerated chemotherapy with alkylating agents rather
well such that a diagnosis other than FA appeared more likely [23].

Hoehn/Kalb/Neveling/Friedl/Bechtold/Herterich/Sun/Gruhn/Hanenberg/Schindler 152
100
0 ng 0ng 0ng 0ng 0ng 0ng
80
60
40
20
0

100
50ng 50ng 50ng 50 ng 50 ng 50 ng
80
60
40
20
0

100
100ng 100ng 100ng 100ng 100ng 100ng
80
60
40
20
0
0
1
2
3
4
5
6
7
8
ⱖ9
10

0
1
2
3
4
5
6
7
8
ⱖ9
10

0
1
2
3
4
5
6
7
8
ⱖ9
10
0
1
2
3
4
5
6
7
8
ⱖ9
10

0
1
2
3
4
5
6
7
8
ⱖ9
10

0
1
2
3
4
5
6
7
8
ⱖ9
10
a b c d e f

Fig. 1. Chromosomal breakage testing of 72-h peripheral blood mononuclear cell cul-
tures without (top row) and with (middle row, 50 ng/ml; bottom row, 100 ng/ml) mitomcycin
C. (a) MMC-resistant: non-FA; (b) intermediate type of MMC sensitivity: questionable diag-
nosis of FA; (c) MMC-sensitive: FA-positive; (d–f) intermediate types of MMC sensitivity:
different grades of FA-mosaicism.

The panels shown in figures 1d–f illustrate three examples of revertant

mosaicism in patients with clinical features of FA. The simultaneous existence
of cells with zero and with ten or more breaks per metaphase at two different
MMC concentrations indicates the presence of cell populations with striking
differences in MMC sensitivity. Even if (such as shown in panel 1f) only a sin-
gle metaphase without any breaks is identified among 50 cells studied after
exposure to the highest clastogen concentration, the patient could be a mosaic,
just like a single metaphase with more than 10 breaks/cell raises the suspicion
of FA. In the examples shown in figure 1d–f one would classify the sample in
figure 1d as a high-grade mosaic with predominance of MMC-resistant, i.e.
reverted cells. The sample shown in figure 1f would qualify as a low-grade
mosaic with a predominance of MMC-sensitive, i.e. non-reverted cells. The
exclusive presence of reverted cells in a peripheral blood sample obscures the
underlying disease genotype such that study of fibroblasts is required in order
to arrive at the correct diagnosis (see below). However, because of the remark-
able longevity of T-lymphocyte subsets, complete lack of MMC-sensitive cells
is only rarely observed.

Revertant Mosaicism in Fanconi Anemia: Natural Gene Therapy at Work 153

100 100
CON
FA
80 80
PAT

60 60

40 40

20 20

0 0
a 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 b 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

100 100

80 80

60 60

40 40

20 20

0 0
c 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 d 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

Fig. 2. Flowcytometric monitoring of 72-h peripheral blood mononuclear cell culture

samples obtained from possible and proven mosaic FA patients. Horizontal axis: sum of
G2/GF ratios as a measure of G2-phase arrest; vertical axis: G0/G1 cell fraction (%) as a
measure of mitogen response [24]. Green symbols denote proven non-FA samples; red sym-
bols denote proven FA-positive cases; blue square symbols denote repeat measurement over
time of four different (a–d) mosaic cases. Years of the respective measurements are as fol-
lows (from right to left): (a) 2000, 2001, 2005, 2004; (b) 2005, 2004, 1998, 1997; (c) 5/2004,
2001, 1999, 1998, 2002, 2003, 1/2004; (d) 2004, 4/2005, 10/2005.

Cell Cycle Testing for the Recognition and Monitoring of

Revertant Mosaicism

In order to see whether and how the two populations of MMC-sensitive

and MMC-resistant cells evolve over time, mosaic patients should be monitored
by repeated chromosome breakage or cell cycle studies. Examples of such
long-term monitoring by way of cell cycle analysis are shown in figure 2. In this
figure, the patient samples are marked by blue squares that are plotted against a
background of laboratory cohorts comprising non-FA (green dots) and FA
(red diamonds) samples. Using standardized procedures for bivariate cell cycle

Hoehn/Kalb/Neveling/Friedl/Bechtold/Herterich/Sun/Gruhn/Hanenberg/Schindler 154
analysis of 72-h peripheral blood mononuclear cell cultures [24] FA-negative
samples display a sum of G2/GF ratios below 0.3 while FA-positive samples
have a sum of G2/GF ratios higher than 0.3, reflecting their FA-typical
G2-phase arrest (see contribution by Schindler et al.). As shown in figure 2a,
mosaicism usually develops slowly over a time course of several years, as evi-
denced by the gradual shift of the sum of G2/GF parameter from above to below
the 0.3 threshold. In the example shown in figure 2b, the first measurement in
1997 suggested nearly complete mosaicism, but subsequent measurement in
2004 and 2005 provided evidence for a persisting mixture of FA-positive
and FA-negative cells. Such fluctuations over time are also documented in
figure 2c, where the measurement 1/2004 yielded a majority of FA-negative,
and the repeat measurement only 4 months later (5/2004) a majority of FA-
positive cells.
The example illustrated in figure 2d implies progression of low-grade
mosaicism to complete mosaicism within a time interval of little over a single
year, with evidence for complete replacement of the original FA cells by
reverted cells. Unless such a patient has been followed with repeat investiga-
tions over time, he or she can no longer be recognized as FA-positive by a one
time chromosome breakage or cell cycle study. If in such instances the labor-
atory study yields a normal result but the possibility of FA persists on clinical
grounds, examination of DEB- or MMC-sensitivity of patient fibroblast
cultures is the only way to arrive at a correct diagnosis [25].

Skin Fibroblast Testing in Cases Discrepant for

Clinical and Blood Findings

In order to resolve the discrepancy between a normal blood cell based

chromosome breakage or cell cycle test and the persisting clinical suspicion of
FA, a skin fibroblast culture needs to be established and tested. Because of the
known oxygen sensitivity of FA fibroblasts, such cultures should ideally be
propagated under hypoxic (i.e. 5% v/v oxygen) cell culture conditions [26, 27].
Figure 3 shows an example of MMC-sensitivity testing of a fibroblast culture
derived from a patient with a completely normal cell cycle study of his peri-
pheral blood cells. Daily cell counts reveal that patient fibroblasts grow nor-
mally in the absence of MMC, but fail to proliferate in the presence of the drug.
Control fibroblasts from a non-FA donor also show lower cell counts after
exposure to MMC, but this reduction is far less pronounced than in the FA pos-
itive sample. Using this simple test will assure the correct diagnosis despite the
presence of 100% reverted peripheral blood lymphocytes. Similar results can of
course be obtained by chromosome breakage testing or cell cycle studies of

Revertant Mosaicism in Fanconi Anemia: Natural Gene Therapy at Work 155

 70
CON
60

50

40

30

20
Number of cells

10

0
70
60 FA

50

40

30

20

10

0
1 2 3 4 5 6
Days after seeding

Fig. 3. Fibroblast growth assay without (black bars) and in the presence of 1␮g/ml
(light grey bars) or 3 ␮g/ml mitomcyin C (dark grey bars). Upper panel: control culture from
a non-FA individual; lower panel: fibroblast culture derived from an FA patient. Bars indicate
means and SD of daily triplicate cell counts.

patient derived fibroblast cultures, but simple cell counts such as shown in
figure 3 are the easiest way to confirm or exclude an FA cellular phenotype.

Molecular Confirmation of Revertant Mosaicism in

T- and B-Lymphocytes

Peripheral blood mononuclear cells (consisting mainly of subclasses of T-

and B-lymphocytes) can be easily obtained by venipuncture. These cells are the
most convenient and accessible source of cells for diagnostic and research pur-
poses. Since bone marrow failure of FA normally progresses in the order of
thrombopenia, erythropenia, leukopenia and lymphopenia, the lymphocytic cell
compartment is relatively well preserved until the final stages of pancytopenia.

Hoehn/Kalb/Neveling/Friedl/Bechtold/Herterich/Sun/Gruhn/Hanenberg/Schindler 156
In most of the cases of revertant mosaicism reported in the literature, molecular
confirmation of the somatic reversion event has been achieved by comparison
of the respective types of mutations found in lymphocyte and fibroblast DNA of
a given patient. This assumes that fibroblast-derived DNA reflects the constitu-
tional pattern of mutations, whereas lymphocyte-derived DNA may have been
altered by somatic mutations. So far there is no evidence that would contradict
the assumption of relative stability of the fibroblast as opposed to the clonally
expanding lymphocyte cell pool, notwithstanding the fact that in certain disease
situations there can be clonal expansion of skin cell populations [3]. Molecular
confirmation of revertant mosaicism has so far been documented mostly for the
FANCA and FANCC genes (cf. table 1) which prompts us to illustrate the mole-
cular confirmation of revertant mosaicism in two mosaic patients belonging to
complementation group FA-D2 [28].

Examples of Somatic Reversions Involving FANCD2

Figure 4 shows the result of DNA sequencing of part of exon 29 of
FANCD2 in a patient whose EBV transformed lymphoid cell line proved resis-
tant to MMC (patient FAD2-14 in table 1). One of the disease causing muta-
tions of the compound heterozygous patient was a substitution of two cytidine
by two thymidine residues within exon 29 (c.2775_2776CC⬎TT). Figure 4a
and 4b shows this constitutional alteration both by gDNA and cDNA sequenc-
ing of fibroblast-derived DNA and RNA. As illustrated in figure 4c and 4d, the
double base substitution present in fibroblasts had disappeared in DNA pre-
pared from the patient’s MMC-resistant lymphoblasts. The reemergence of
exon 29 wildtype sequence at position c.2776 explains the phenotypic reversion
of the patient’s lymphoid cells to MMC-resistance. Both back mutation and
gene conversion qualify as potential molecular mechanisms of reversion in this
patient (see below).
A second example of a somatic reversion in FANCD2 is illustrated in figure 5.
This example involves a splice mutation which is the most common type of
genetic alterations observed in FA-D2 patients [28]. The disease causing muta-
tion in this patient is a homozygous T⬎G base change within intron 21 (panel
5a). This alteration affects splicing and results in skipping of exon 22 as shown
in panel 5b. Genomic DNA derived from the patient’s MMC-resistant lymphoid
cell line (panel 5c) revealed the constitutional mutation within intron 21, but an
additional base change (1954G⬎A) within exon 22 (panel 5c). This obviously
somatic alteration within exon 22 resulted in the reappearance of exon 22
sequence at the cDNA level (panel 5d), indicating restoration of regular splic-
ing of the respective allele transcript. The newly emerged 1954G⬎A within
exon 22 therefore qualifies as a second site compensatory mutation that
explains the conversion to MMC resistance of the patient’s lymphoid cells.

Revertant Mosaicism in Fanconi Anemia: Natural Gene Therapy at Work 157

 Table 1. Molecular analysis of revertant mosaicism in FA patients reported until the end of year 2006, including cases from our own
laboratories. Except for cases with mitotic crossover the allele targeted by a reversion event is listed as allele 2.
Hoehn/Kalb/Neveling/Friedl/Bechtold/Herterich/Sun/Gruhn/Hanenberg/Schindler

Patient code FANC Allele 1 Allele 2 Reversiona Probable Revertant Cellular Ref.a
Genes (target) mechanisma cellsa phenotypea

URD A 3976C⬎T 856C⬎T 856C (WT) back mutation PBLs, LCL mild pancytopenia [8]
STT A intragenic 862G⬎T 862G (WT) back mutation PBLs, LCL mild pancytopenia [8]
deletion
MRB A IVS9-1G⬎T 971T⬎G 971T (WT) back mutation PBLs, LCL mild pancytopenia [8]
EUFA704 A 1615delG 1615delG 1637delA/ comp. mutation PBLs, LCL NR [40]
1641delT
FA67 A 3720-3724del 2546delC 2546C⬎T comp. mutation granulocytes, mild pancytopenia [18]
mononuclear
phagocytes,
LCL
FA98 A 3720-3724del 2546delC 2546C⬎T⫹ comp. mutation PBLs progressive [18]
3720-3724del pancytopenia
NR A 790C⬎T 2585delCT 2585CT gene conversion PBLs, LCL normal CBC [19]
PD839 A 2555⌬T 2670G⬎A 2927G⬎A comp. mutation PBLs, LCL normal CBC [9]
PD845 A 2555⌬T 2670G⬎A 2927G⬎A comp. mutation PBLs, LCL normal CBC [9]
IFAR557-2 A genomic 2815-2816 2815-2816ins back mutation PBLs, LCL mild [6]
deletion at ins19 19 (WT) stem cells pancytopenia
3⬘ end of gene progression/
clonal
evolution
EUFA393 A 3559insG 3559insG 3580insCG comp. mutation PBLs and LCL NR [40]
CTG or 3576
insGCTGC
158
 EUFA173 A Del exon 2852G⬎A 2852G (WT) back mutation PBLs and LCL, mild pancytopenia [8]
17-31 respectively
Revertant Mosaicism in Fanconi Anemia: Natural Gene Therapy at Work

EUFA806 C 67delG 1806insA heterozygote mitotic crossover PBLs, LCL normal CBC [1]
segregants
EUFA449 C L554P 67delG 67delG (WT) gene conversion PBLs, LCL mild pancytopenia [1]
RNT C 67delG IVS11- heterozygote mitotic crossover PBLs, LCL progressive [8]
2A⬎G segregants pancytopenia
EUFA506 C 1749T⬎G 1749T⬎G 1748C⬎T comp. mutation PBLs, LCL NR [40]
FA-AML1 D1 8415G⬎T 8732C⬎A 8731T⬎G comp. mutation Leukemic cells resistance to MMC [41]
FAD2-3 D2 1948-16T⬎G 1948-16T⬎G c.1954G⬎A comp. mutation PBL, LCL, resistance to MMC [28]
bone marrow
FAD2-14 D2 1948-6C⬎A 2775_2776 2776CC (WT) back mutation/ PBL, LCL resistance to MMC [28]
CC⬎TT conversion
FAD2-15 D2 1948-6C⬎A 2775_2776 (heterozygote (mitotic PBL resistance to MMC [28]
CC⬎TT segregants?) crossover?)
FAD2-26 D2 3803G⬎A 376A⬎G c.376A (WT) back mutation PBL, LCL resistance to MMC [28]
FAL-1 L 891C⬎G 483delATCAC c.472-2A⬎G comp. mutation PBL, LCL, BM resistance to MMC [28]
EUFA N exon 4 exon 4 stop in-frame fusion comp. mutation LCL resistance to MMC [34]
1341 (R) deletion codon of exons 3 and (Alu-mediated
mutation 5; elimination recombination)
of exon 4

a
BM ⫽ bone marrow cells; CBC ⫽ comprehensive blood counts; Comp. mutation ⫽ compensatory or second site mutation; LCL ⫽ Lymph-
oblastoid cell line; MMC ⫽ mitomycin C; NR ⫽ not reported; PBL ⫽ peripheral blood lymphocytes; Ref. ⫽ reference; WT ⫽ wild type. Tentative
interpretations are shown in parentheses.
159
 gDNA cDNA

Exon 29 Exon 29

T A A T T CC C A T G C T T T T T T N N G A G A G C T GG A C A T T G A A T T CCC A T GC T T T T T T CCGAGAGC T GGA C A T

170 180 190 200 60 70 80

c.2775_2776CC>TT c.2775_2776CC>TT

a b

Exon 29 Exon 29

T A A T T CC C A T G C T T T T T T C C G A G A G C T GG A C A T T G A A T TCCC A T GC T T T T T T CCGAGAGC T GGA C A T

170 180 190 200 370 380 390

c.2775_2776CC (wt) c.2775_2776CC (wt)

c d

Fig. 4. Partial sequence of exon 29 of FANCD2 at both gDNA and cDNA levels.
Panels a and b show the sequence obtained from fibroblasts (constitutional mutations).
Panels c and d depict the sequence obtained from lymphoid cells that had reverted to MMC-
resistance.

Molecular Confirmation of Somatic Reversion in

Hematopoietic Precursor Cells

Significant clinical improvement due to revertant mosaicism can only be

expected if (1) the reversion event has taken place in a hematopoietic stem or
early precursor cell, and if (2) the reversion event conveys a proliferative and
functional advantage to the progeny of the reverted cell. A number of recent
studies have successfully looked for somatic mutations in blood cell lineages
other than peripheral blood T- and B-lymphocytes [6, 9, 18]. These studies have
demonstrated multilineage somatic reversions of constitutional FA mutations
that are consistent with a reversion event during early blood cell differentiation,
or even at the stem cell level. A particularly instructive example is the observa-
tion by Mankad et al. [9] who detected, in lymphocyte DNA of twin sisters, an

Hoehn/Kalb/Neveling/Friedl/Bechtold/Herterich/Sun/Gruhn/Hanenberg/Schindler 160
 gDNA cDNA

Intron 21 Exon 22 Exon 21 Exon 23

C T T T T TGT TGGT T TGCT TCC TGA AGGAA TGGGTTGGGC AT A GA T CC AAAAGCCC T G T G A C T T T CC A T T T CC T G
240 250 260 270 150 160 170
c.1948-16T>G

a b
Intron 21 Exon 22 Exon 23 Exon 22
Exon 21
C T T T T TGT TGGT T TGCT TCC TGA AGGAA TGGGTTGGGCA T A GG A A A T GG A A A G T C A CC T T NCNG A A C A A C A C
230 240 250 260 180 190 200 210
c.1948-16T>G
c.1954G>A

c d

Fig. 5. Partial sequence of FANCD2 extending from intron 21 through exon 23 illus-
trating heterozygous restoration of exon 22 sequence in patient lymphocytes. The best
sequencing result showing reappearance of exon 22 sequence was obtained by use of reverse
primers, which explains the reverse sequence orientation in panel d.

identical compensatory missense mutation in exon 30 of FANCA that corrected

a constitutional missense mutation in exon 28, restoring nuclear localization
and function of the resulting protein. Both twin sisters had normal blood cell
counts since birth, suggesting that the somatic reversion event may have
occurred prenatally in a hematopoietic stem cell in one of the twins. The other
twin may have acquired the corrected stem cell progeny via common intrauter-
ine blood circulation, and the lifelong persistence of reverted cells supports the
idea of a hematopoietic stem cell reversion [9].

Examples of Multilineage Somatic Reversions in FA-A Patients

In order to emphasize that multilineage somatic reversion is necessary for a sus-
tained improvement of the hematological situation of mosaic patients, we
review two examples in which in addition to DNA derived from lymphoblasts
and fibroblasts, DNA from other cell lineages was analyzed as well. The first
patient (patient EUFA173 in table 1) has been previously described by Joenje

Revertant Mosaicism in Fanconi Anemia: Natural Gene Therapy at Work 161

et al. [29] and Gross et al. [8]. He manifested with severe pancytopenia at age
15. His maternal FANCA allele carries the base change 2852G⬎A and his
paternal allele shows a large genomic deletion involving exons 17–31 [29]. As
shown in figure 6a–c, the patient’s blood counts unexpectedly improved over a
time course of 3 to 8 years such that his hematological status was close to nor-
mal at age 22. The obvious improvement of more than a single cell lineage sug-
gested a reversion event at the level of a hematopoietic precursor if not stem
cell. Sequencing at age 26 revealed almost complete reversion to wildtype of
the 2852G⬎A mutation in granulocyte DNA and in DNA from a newly estab-
lished B-lymphoblast cell line, whereas a mixture of mutated and wildtype
sequence was observed in the T-lymphocyte derived sample (fig. 6d–f). The
persistence of a fraction of non-reverted cells in the T-lymphocyte sample may
reflect mitogenic activation of long-lasting, quiescent subsets of T cells that
were already present in vivo before the reversion took place [30]. Because of
the remarkable longevity of certain T cell types, one may encounter traces of
non-reverted DNA sequence in T-lymphocyte samples even decades after the
initial reversion event has taken place in a lymphocyte precursor cell. Since the
region corresponding to the 2852G⬎A alteration is deleted on the other allele
and thus cannot serve as template for a gene conversion event, back mutation is
the most likely explanation for the reversion to wildtype.
The second example of multilineage reversion involves a patient who
developed severe thrombocytopenia and anemia at age 6 years but experienced
unexpected improvements of his blood counts shortly thereafter (patient MRB
in table 1). Using retroviral complementation analysis [31] the patient was
assigned to complementation group FA-A and mutation analysis of fibroblast
DNA revealed a maternally transmitted base change (971T⬎G) and a splice
mutation (IVS9-1G⬎T) on the paternal allele. As illustrated in figure 7a–c,
introduction of an empty vector containing only EGFP failed to resolve the G2
phase accumulation of the patient’s cells, while transfection with the vector
containing the wildtype FANCA sequence resulted in reduction of the G2 phase
arrest, proving complementation. The patient’s hemoglobin and thrombocyte
counts started to improve at around three years after he had been diagnosed
with FA such that androgen treatment could be reduced to a low maintenance
level (cf. fig. 7g, h). Sequencing at that time revealed that the maternal mutation
had all but disappeared in whole blood cell DNA (fig. 7e) while it was of course
present in fibroblast DNA (fig. 7d). In order to prove multilineage reversion,
leukocytes were separated into monocytes, T and B lymphocytes, and individ-
ual white cell (CFU-GM) and red cell (BFU-E) colonies were isolated using a
standard bone marrow progenitor cell assay. A representative result of this study
is shown in figure 7f which proves complete reversion to wildtype of the mater-
nal 971T⬎G change in DNA derived from BFU-E colonies. A similar result

Hoehn/Kalb/Neveling/Friedl/Bechtold/Herterich/Sun/Gruhn/Hanenberg/Schindler 162
 6 200
15

Thrombocyte counts (⫻1,000/␮l)

Leukocyte counts (⫻1,000/␮l)
5
13 160
Hemoglobin (g/dl)

4
11
120
3
9

80
7 2

Hb WBC Platelets
5 1 40
19 8
19 0
92

19 4
19 6
98
00

88

19 0
92

19 4
19 6
98
00

88

19 0
92

19 4
96

20 8
00
8
9

9
9

9
9

9
a b c
19

19

20

19
19

19

20

19
19

19

19
Reversion of 2852G>A in:
Granulocyte T-lymphocyte B-lymphoblast, new

2848 G A A C A/G G C A G 2856 2848 G A A C A/G G C A G 2856 2848 G A A C A/G G C A G 2856

950 Glu Gln/Arg Gln 952 950 Glu Gln/Arg Gln 952 950 Glu Gln/Arg Gln 952
Exon 29 Exon 30 Exon 29 Exon 30 Exon 29 Exon 30

d e f

Fig. 6. Evidence for phenotypic (panels a–c) and genotypic (panels d–f ) reversion of
the constitutional mutation 2852G⬎A in patient EUFA173 (cf. table 1). The horizontal axis in
panels a–c denotes the years 1988–2000 during which time the recovery of blood counts took
place. Panels d–f illustrate FANCA sequence stretches containing the 2852G⬎A change.

was obtained for the other cell lineages tested, assigning the reversion event to
at least an early hematopoietic precursor, if not to a stem cell.

Which FA Genes are Frequent Targets of Somatic Reversions?

Table 1 provides a summary of somatic reversions reported in the literature

up to 2006, including cases from our own laboratories [28]. Of the total 23 rever-
sions that have been analyzed at the molecular level, twelve were found in
FANCA, four each in FANCC and FANCD2, and a single case each in FANCD1/
BRCA2, FANCL and FANCN. The preponderance of FANCA might be explained
by the fact that in most populations FANCA is the most frequently affected FA

Revertant Mosaicism in Fanconi Anemia: Natural Gene Therapy at Work 163

 EGFP FANCA FANCA 971T > G

G1 S G2

a b DNA content c
Fibroblast Blood BFU-E
967 A C C C T /G G A C T A C C C T /G G A C T A C C C T /G G A C T 975
323 Thr Lou/Arg Thr Thr Lou/Arg Thr Thr Lou/Arg Thr 325

d e f
18 240
Hemoglobin (g/dl)

16
Thrombocytes

180
(1,000/␮l))

14
120
12
10 60

8 0
0 1 2 3 4 5 0 1 2 3 4 5
g Time after diagnosis (y) h Time after diagnosis (y)

Fig. 7. Assignment of patient MRB to complementation group FA-A via retroviral com-
plementation (panels a–c). Panels d–f: evidence for multilineage reversion (e and f) of the con-
stitutional base substitution 971T⬎G present in fibroblast-derived DNA (d). Improvement of
blood counts was noted at 3 years after the patient had been diagnosed with FA (panels g and h).

gene. In addition, FANCA harbors a number of repetitive sequence elements that

may facilitate reversion, for example by polymerase slippage [32]. Taking into
account that Soulier et al. [33] reported additional 5 FA-D2 patients (not listed in
table 1 due to lack of molecular confirmation) with evidence for revertant
mosaicism, revertant mosaicism may also be a relatively frequent event in sub-
type FA-D2. Most recently, a somatic reversion event has been reported in the
lymphoid cell line derived from a patient belonging to complementation group
FA-N [34], adding this subtype to the FA genes that potentially exhibit somatic
reversions.

Hoehn/Kalb/Neveling/Friedl/Bechtold/Herterich/Sun/Gruhn/Hanenberg/Schindler 164
 Molecular Mechanisms of Somatic Reversion

As listed in table 1, there are four principal mechanisms that have

been implicated in revertant mosaicism in FA: two of these involve mitotic
recombination (mitotic crossover and gene conversion). The other two mecha-
nisms are back mutation and compensatory or second site mutation. Different
molecular types of reversion can give rise to compensatory mutations (see
below). The mechanisms leading to revertant mosaicism have first been explored
in a number of diseases other than FA, as described in the cited reviews [3–5].
A particularly instructive disease with respect to the molecular mechanisms of
revertant mosaicism is the skin disease epidermolysis bullosa where areas of
reverted skin can be easily recognized and probed as to their molecular constitu-
tion. The group of Professor Jonkman at the University of Groningen has pub-
lished a number of key articles on this topic, including a recent report in which
two patients are presented who each display multiple reversion events affecting
different body parts [35]. In these two patients, the mechanisms underlying
somatic reversion were studied side by side within the same individual. They
were found to include back mutation, gene conversion, and second site (com-
pensatory) mutations. Even though damaged skin and skin renewal in epider-
molysis bullosa represents a very special situation, these impressive ‘experiments
of nature’ suggest that, by analogy, multiple reversion events within a given bone
marrow and possibly other cellular compartments of FA patients might occur.
What we presently know about revertant mosaicism in FA may therefore repre-
sent only the literal tip of the iceberg. We strongly encourage the interested
reader to study the instructive examples of revertant mosaicism presented by the
Jonkman group, but limit ourselves to briefly review the molecular mechanisms
of reversion that have been described in FA patients. Figure 8 summarizes the
various proposed mechanisms in a simplified form.
Intragenic crossover is a conservative mechanism of mitotic recombin-
ation that was first shown in a mosaic FA-C patient by Lo ten Foe et al. [1]. By
comparing the haplotypes of polymorphic markers flanking the FANCC gene
on chromosome 9 between MMC-sensitive fibroblasts and MMC-resistant lym-
phoblasts, Lo ten Foe et al. provided evidence for haplotype switching in the
lymphoblast-derived DNA sample, suggesting a mitotic recombination event as
explanation for the phenotypic reversion to MMC-resistance. Indirect evidence
for intragenic recombination was also provided by Gross et al. [8] who demon-
strated the simultaneous presence, in their reverted lymphoid cell culture, of
wildtype alleles and alleles carrying either a single or both mutations which
would be the expected result after segregation of mitotic crossover recom-
binants. The simultaneous presence of these three allele types suggests an in vitro
rather than in vivo reversion event. Both patients were compound heterozygotes,

Revertant Mosaicism in Fanconi Anemia: Natural Gene Therapy at Work 165

 a b a b a b a b

c d c d c d d c

1 2 3 4

Fig. 8. Molecular mechanisms leading to somatic reversion in FA. Simplified scheme

depicting pre-replication chromosome pairs carrying either compound heterozygous (1, 3) or
homozygous (2) biallelic mutations (black boxes). (1) Either back mutation or gene conver-
sion reverts one of the two alleles to wildtype. There is no evidence for strand exchange, and
the original wildtype sequence is restored. (2) Compensatory second site mutation leading to
restoration of the open reading frame upstream or downstream of a constitutional mutation.
(3 and 4) Intragenic mitotic crossing over converts compound heterozygous biallelic muta-
tions to (3) functional heterozygosity with the resulting alleles (4) carrying either two or
none of the original mutations. Analysis of flanking markers (marked in red) provides proof
of strand switching between homologous chromosomes (3).

and the mitotic recombination events might have been facilitated by the mutu-
ally distant locations of the respective paternal and maternal mutations.
Gene conversion is another conservative mechanism of mitotic recombi-
nation without, however, mutual strand exchange. Proof of such locus-restricted
recombinational events requires the analysis of polymorphic markers flanking
the mutation which retain their parent-of-origin specific patterns. An intact
homologous copy of the mutated region is required in order to serve as template
for the conversion of the mutated to wildtype sequence. Gene conversion can
only function in compound heterozygous patients, and it does not function if
the homologous gene region on the other allele is deleted. Gene conversion as a
probable mechanism of revertant mosaicism was first suggested by Lo ten Foe
et al. [1]. These authors observed two mosaic siblings harboring segregant
alleles with only single mutations, but without evidence for haplotype switch-
ing. Alter et al. [19] also suggested gene conversion as mechanism for revertant
mosaicism in their compound heterozygous patient even though there was no
study of polymorphic markers within or flanking the presumptive site of gene

Hoehn/Kalb/Neveling/Friedl/Bechtold/Herterich/Sun/Gruhn/Hanenberg/Schindler 166
conversion. Given the relatively wide spacing of polymorphic markers, and
given the presumptive shortness of conversion tracts in mammals (less than
58 bp; [36]), gene conversion as a causal mechanism of revertant mosaicism is
still difficult to prove. An additional problem is the requirement for somatic
pairing of homologous chromosomes which may be mediated, at least in the
case of FANCA, by a large heterochromatin block on chromosome 16 [37], but
which otherwise has been rarely observed in mitotic cells.
Back mutation is another possible mechanism underlying revertant
mosaicism that, again, is difficult to prove since there is no change of polymor-
phic markers surrounding the back-mutated gene locus. Back mutation has
been invoked to explain mosaicism for blood cells with either high and low
SCE phenotypes in two patients carrying homozygous mutations in the Bloom
syndrome gene [38]. In compound heterozygous FA-A patients, back mutation
has been assumed as probable mechanism of reversion since (1) the homolo-
gous gene region opposite to the reverted mutation was deleted in some of these
patients, and (2) the reverted allele displayed the original wildtype rather than
any random sequence [8]. With two exceptions, all presumptive cases of back
mutation listed in table 1 were observed in FANCA, a gene with a high number
of small and large repetitive elements such that slipped-strand mispairing has
been assumed as major mechanism of mutagenesis in this gene [32]. The first
case of revertant mosaicism in an FA-A patient ascribed to back mutation was
reported by Gregory et al. [6] who observed the reversion to wildtype of a
maternally inherited 19-bp insertion within exon 29 of FANCA. The authors
argued that the insertion creates a tandem repeat sequence in close proximity to
a deletion hot spot consensus sequence with the possible consequence of for-
mation of a loop structure. If this loop is deleted prior to a next round of repli-
cation, the wildtype sequence would be restored. A very similar line of
arguments was followed by Gross et al. [8] who observed that the reversion in
three of their mosaic FA-A patients affected the region of exons 10–11 of
FANCA which is known as a highly mutable region due to an abundance of
repetitive elements and sequence motifs, including palindromic sequences.
Such structures are known to be involved in the breakage and rejoining of
DNA. Gross et al. [8] therefore assumed that back mutation (via mechanisms
such as slipped-strand mispairing and others) would be the most likely explan-
ation for the fact that one of the mutated single base pair alleles in all of their
FA-A patients had reverted precisely to wildtype. They further argued that
random base alterations would have yielded mostly non-conservative amino
acid exchanges such that the resulting protein may have been functionally
impaired. As suggested by experiments in mice [39] selection would favor cells
with reversions to the original nucleotide since only these would assure full
restoration of protein function. Back mutation combined with selection for the

Revertant Mosaicism in Fanconi Anemia: Natural Gene Therapy at Work 167

restored wildtype allele was therefore considered as most likely explanation for
the surprisingly uniform pattern of reversion observed in these patients [8].
Compensatory or second site mutations in cis as first observed by
Waisfisz et al. [40] turn out to be the most frequent type of mechanism under-
lying revertant mosaicism in FA (cf. table 1). Even though this type of reversion
may not result in complete restoration of protein function, it is the prevailing
mechanism that leads to at least partial functional rescue in the case of homozy-
gous mutations. The original report by Waisfisz et al. [40] documents such res-
cue in the case of homozygous microdeletions, microinsertions and missense
mutations in two FA-A siblings and an FA-C patient. A frameshift mutation
(1615delG) in FANCA was compensated by two downstream single-base pair
deletions, another frameshift mutation (3559insG) was compensated by a
downstream five base pair insertion, and a homozygous missense mutation in
FANCC (1749T⬎G) was altered by a base change in the preceding position. In
all these cases the predicted proteins were different from wildtype, but their
cDNAs were shown to complement the typical MMC sensitivity of mutant FA
cells, explaining the phenotypic reversion of the patient’s peripheral blood cells.
The landmark study by Waisfisz et al. [40] makes it clear that compensating
second site mutations can lead to partial or complete mRNA rescue without the
necessity for elimination of the constitutional mutation. Even though the result-
ant protein may differ from wildtype, its function is likely to be sufficient for
phenotypic reversion.
As already mentioned above, Mankad et al. [9] observed a most impressive
example of a compensatory second site mutation in the FANCA gene that must
have taken place early in embryogenesis since the FA twin siblings were both
found to carry the compensatory change.
Compensatory mutations may not only be found upstream or downstream
of the constitutional mutation, but may also affect the mutation itself [3]. Such
an example was documented by Hamanoue et al. [18] who found the constitu-
tional FANCA frameshift mutation 2546delC reverted to 2546C⬎T, resulting in
expression of a functional missense protein. Compensatory second site muta-
tions have also been observed in FA-D1 and FA-D2 patients (cf. table 1). In the
case of the FA-D1 patient, a two-year-old boy affected by AML, the compen-
satory change emerged in a patient derived leukemic cell line [41]. The child
had biallelic mutations in FANCD1/BRCA2. In the patient’s leukemic cells the
8732C⬎A nonsense mutation was changed into the missense alteration
8731T⬎G, leading to restoration of the open reading frame and explaining the
MMC-resistance of the leukemic cells [41]. An example of a compensatory sec-
ond site mutation in an FA-D2 patient has been illustrated in figure 5. In this
patient, a homozygous mutation in intron 21 of FANCD2 caused skipping of
exon 22 which was restored by a second site compensatory mutation [28]. Most

Hoehn/Kalb/Neveling/Friedl/Bechtold/Herterich/Sun/Gruhn/Hanenberg/Schindler 168
recently, the newly discovered FANCN gene has been added to the list of genes
that exhibit somatic reversions, at least in cultured lymphoid cells [34]. The
reversion in FANCN was shown to result from Alu-mediated recombination
leading to an in-frame fusion of exons 3 and 5, thereby eliminating one of the
disease causing mutations located in exon 4. Alu-mediated recombination can
therefore be added to the list of mechanisms instrumental in compensatory
mutations [34].

Clinical Consequences of Revertant Mosaicism

Improvement of the hematological situation of a given patient can be

expected in cases of multilineage reversions, meaning that the reversion event
has to take place in an early hematopoietic precursor cell, ideally in a
hematopoietic stem cell. We have documented two such cases in detail (cf.
figs. 6 and 7), and a number of other examples are listed in table 1. In fact, the
majority of mosaic patients summarized in table 1 showed improvement of
their hematological status, mostly characterized as mild or non-progressive
pancytopenia. However, with few exceptions [6, 9, 18] the hematological con-
sequences of revertant mosaicism have not been fully documented over time,
such that definite statements concerning the clinical benefits of revertant
mosaicism must await the results of prospective studies. The patients docu-
mented in figures 6 and 7 in whom multilineage reversion has been firmly
established, and the case reports by Hamanoue et al. [18] and Mankad et al. [9]
leave little doubt about a positive, long lasting effect on hematopoiesis if the
reversion event has taken place in an early progenitor or stem cell. An overall
beneficial effect of revertant mosaicism has also been documented by Soulier
et al. [33] who compared the hematological status of 8 mosaic with that of 45
non-mosaic patients. As shown in figure 9, blood counts were on average
higher in mosaic patients. P values comparing mosaic vs. non-mosaic patients
were 0.00017 for white blood counts, 0.0001 for neutrophils, 0.001 for hemo-
globin, and likewise 0.001 for platelets, whereas differences in MCV were not
significant.
Whereas lack of hematological improvement is not surprising in cases
where the reversion is limited to a precursor cell of the lymphoid cell lineage
(e.g. patient RNT in table 1; [8]), the patient reported by Gregory et al. [6]
demonstrates that despite proven multilineage reversion a clonal cytogenetic
change may arise in the remaining fraction of non-revertant bone marrow cells
such that the possibility of malignant cell growth cannot be entirely excluded
in mosaic patients. A similar observation was made by Manoharan [42]
whose patient developed clonal trisomy 1q despite longstanding and apparently

Revertant Mosaicism in Fanconi Anemia: Natural Gene Therapy at Work 169

 14
45FA
8mosaic
12

10

0
WBC ANC Hb Platelets

Fig. 9. Comparison of blood cell counts between mosaic and non-mosaic FA patients.
Data plotted from table 3 of Soulier et al. [33]. ANC ⫽ Absolute neutrophil count; Hb ⫽ hemo-
globin; WBC ⫽ white blood cell count; vertical axis in arbitrary units. See text for details.

multilineage revertant mosaicism [16]. In addition, our laboratory has observed

at least two patients in whom the fraction of FA-negative (i.e. reverted) periph-
eral blood mononuclear cells has not, as expected, increased but decreased over
time. It will be important to follow these cases closely in order to answer the
question whether these are random fluctuations due to clonal expansion and
clonal attenuation of lymphocytic cells, for example in response to infections,
or whether these changes may have clinical significance. At this point we sim-
ply do not know whether reverted cell clones may in fact be limited in the num-
ber of progeny and lifespan. Assuming that reversion has taken place in a
hematopoietic stem or early progenitor cell, we also do not know whether the
clonal expansion of the reverted cell in a milieu of non-corrected stromal cells
would in each instance be successful. Sensitivity to oxidative stress surely
would remain a major obstacle for the stromal cell population [43] which may
adversely impact the proliferative capacity of the reverted cell lineage. Finally,
it is clear that mosaic patients need special attention during preparation for
HSCT which still may be required if the reversion is limited to the lymphocytic
cell lineage. Conversely, if there is proven multilineage reversion, HSCT may
not be a prime therapeutic consideration. Clearly, much more has to be learned
about the unique phenomenon of revertant mosaicism, its molecular mecha-
nisms, and its clinical consequences.

Hoehn/Kalb/Neveling/Friedl/Bechtold/Herterich/Sun/Gruhn/Hanenberg/Schindler 170
 References

1 Lo Ten Foe JR, Kwee ML, Rooimans MA, Oostra AB, Veerman AJ, et al: Somatic mosaicism in
Fanconi anemia: molecular basis and clinical significance. Eur J Hum Genet 1997;5:137–148.
2 Rieger R, Michaelis A, Green MM: Glossary of Genetics and Cytogenetics: Classical and
Molecular, ed 4. New York, Springer, 1996.
3 Jonkman MF, Castellanos Nuijts M, van Essen AJ: Natural repair mechanisms in correcting path-
ogenic mutations in inherited skin disorders. Clin Exp Dermatol 2003;28:625–631.
4 Youssoufian H, Pyeritz RE: Mechanisms and consequences of somatic mosaicism in humans. Nat
Genet Rev 2002;3:748–758.
5 Hirschhorn R: In vivo reversion to normal of inherited mutations in humans. J Med Genet
2003;40:721–728.
6 Gregory JJ Jr, Wagner JE, Verlander PC, Levran O, Batish SD, et al: Somatic mosaicism in
Fanconi anemia: evidence of genotypic reversion in lymphohematopoietic stem cells. Proc Natl
Acad Sci USA 2001;98:2532–2537.
7 Youssoufian H: Natural gene therapy and the Darwinian legacy. Nat Genet 1996;13:255–256.
8 Gross M, Hanenberg H, Lobitz S, Friedl R, Herterich S, et al: Reverse mosaicism in Fanconi
anemia: natural gene therapy via molecular self-correction. Cytogenet Genome Res 2002;98:126–135.
9 Mankad A, Taniguchi T, Cox B, Akkari Y, Rathbun RK, et al: Natural gene therapy in monozygotic
twins with Fanconi anemia. Blood 2006;107:3084–3090.
10 Kwee ML, Poll EHA, Van de Kamp JJP, De Koning H, Eriksson AW, Joenje H: Unusual response
to bifunctional alkylating agents in a case of Fanconi anaemia. Hum Genet 1983;64:384–387.
11 Cohen MM, Simpson SJ, Honig GR, Maurer HS, Nicklas JW, Martin AO: The identification of
Fanconi anemia genotypes by clastogenic stress. Am J Hum Genet 1982;34:794–810.
12 Duckworth-Rysiecki G, Hulten M, Mann J, Taylor AM: Clinical and cytogenetic diversity in
Fanconi’s anemia. J Med Genet 1984;21:197–203.
13 Cervenka J, Hirsch BA: Cytogenetic differentiation of Fanconi anemia, ‘idiopathic’ aplastic ane-
mia, and Fanconi anemia heterozygotes. Am J Med Genet 1983;15:211–223.
14 Auerbach AD, Rogatko A, Schroeder-Kurth TM: International Fanconi Anemia Registry: First
Report; in Schroeder-Kurth TM, Auerbach AD, Obe G (eds): Fanconi Anemia, Clinical
Cytogenetic and Experimental Aspects. Berlin/Heidelberg, Springer-Verlag, 1989, pp 3–17.
15 Jonkman MF, Scheffer H, Stulp RP, Pas HH, Nijenhuis M, et al: Revertant mosaicism in epider-
molysis bullosa caused by mitotic gene conversion. Cell 1997;88:543–551.
16 Hopkins NC, Manoharan A, Lammi AT: Spontaneous improvement in Fanconi’s anemia.
Pathology 1991;23:254–255.
17 Kwee ML, van der Kleij JM, van Essen AJ, Begeer JH, Joenje H, et al: An atypical case of Fanconi
anemia in elderly sibs. Am J Med Genet 1997;68:362–366.
18 Hamanoue S, Yagasaki H, Tsuruta T, Oda T, Yabe H, et al: Myeloid lineage-selective growth of
revertant cells in Fanconi anaemia. Br J Haematol 2006;132:630–635.
19 Alter BP, Joenje H, Oostra AB, Pals G: Fanconi anemia: adult head and neck cancer and
hematopoietic mosaicism. Arch Otolaryngol Head Neck Surg 2005;131:635–639.
20 Joenje H, Oostra AB, Zwaan CM, Pals G: Fanconi anemia; in Humana Press Series on Hereditary
Cancer Syndromes. Totowa, Humana Press, 2006; in press.
21 Schroeder-Kurth TM, Auerbach AD, Obe G (eds): Fanconi Anemia, Clinical Cytogenetic and
Experimental Aspects. Berlin/Heidelberg, Springer-Verlag, 1989, pp 1–378.
22 Schroeder-Kurth TM, Zhu TH, Hong Y, Westphal I: Variation in cellular sensitivities among
Fanconi Anemia patients, non-Fanconi Anemia patients, their parents and siblings, and control
probands; in Schroeder-Kurth TM, Auerbach AD, Obe G (eds): Fanconi Anemia, Clinical
Cytogenetic and Experimental Aspects. Berlin/Heidelberg, Springer-Verlag, 1989, pp 106–136.
23 Bakhshi S, Joenje H, Schindler D, Oostra A, Mohamed AN, et al: A case report of a patient with
microcephaly, facial dysmorphism, mitomycin-C-sensitive lymphocytes, and susceptibility to
lymphoma. Cancer Genet Cytogenet 2006;164:168–171.
24 Schindler D, Hoehn H: Flowcytometric testing for syndromes with chromosomal instability, aplas-
tic anemia and related haematological disorders; in Wegner RD (ed): Diagnostic Cytogenetics.
Berlin/Heidelberg, Springer-Verlag, 1999, pp 269–281.

Revertant Mosaicism in Fanconi Anemia: Natural Gene Therapy at Work 171

25 Joenje H, Arwert F, Kwee ML, Madan K, Hoehn H: Confounding factors in the diagnosis of
Fanconi anemia. Am J Med Genet 1998;79:403–405.
26 Schindler D, Hoehn H: Fanconi anemia mutation causes cellular susceptibility to ambient oxygen.
Am J Hum Genet 1988;43:429–435.
27 Kalb R, Duerr M, Wagner M, Herterich S, Gross M, et al: Lack of sensitivity of primary Fanconi’s
anemia fibroblasts to UV and ionizing radiation. Radiat Res 2004;161:318–325.
28 Kalb R, Neveling K, Hoehn H, Schneider H, Linka Y, et al: Hypomorphic mutations in the key
Fanconi anemia gene FANCD2 sustain a significant group of FA-D2 patients with severe pheno-
type. Am J Hum Genet 2007; in press.
29 Joenje H, Levitus M, Waisfisz Q, D’Andrea A, Garcia-Higuera I, et al: Complementation analysis
in Fanconi anemia: assignment of the reference FA-H patient to group A. Am J Hum Genet
2000;67:759–762.
30 Globerson A, Effros RB: Ageing of lymphocytes and lymphocytes in the aged. Immunol Today
2000;10:515–521.
31 Hanenberg H, Batish SD, Pollok KE, Vieten L, Verlander PC, et al: Phenotypic correction of pri-
mary Fanconi anemia T cells with retroviral vectors as a diagnostic tool. Exp Hematol
2002;30:410–420.
32 Levran O, Diotti R, Pujara K, Batish SD, Hanenberg H, Auerbach AD: Spectrum of sequence vari-
ation in the FANCA gene: an International Fanconi Anemia Registry (IFAR) study. Hum Mutat
2005;25:142–149.
33 Soulier J, Leblanc T, Larghero J, Dastot H, Shimamura A, et al: Detection of somatic mosaicism
and classification of Fanconi anemia patients by analysis of the FA/BRCA pathway. Blood
2005;105:1329–1336.
34 Xia B, Dorsman JC, Ameziane N, de Vries Y, Rooimans MA, et al: Fanconi anemia is associated
with a defect in the BRCA2 partner PALB2. Nat Genet 2006;39:159–161.
35 Pasmooij AMG, Pas HH, Deviaene FCL, Nijenhuis M, Jonkman MF: Multiple correcting
COL17A1 mutations in patients with revertant mosaicism of Epidermolysis bullosa. Am J Hum
Genet 2005;77:727–740.
36 Elliott B, Richardson C, Winderbaum J, Nickoloff JA, Jasin M: Gene conversion tracts from
double-strand break repair in mammalian cells. Mol Cell Biol 1998;18:93–101.
37 Haaf T, Steinlein K, Schmid M: Preferential somatic pairing between homologous heterochro-
matic regions of human chromosomes. Am J Hum Genet 1986;38:319–329.
38 Ellis NA, Ciocci S, German J: Back mutation can produce phenotype reversion in Bloom syn-
drome somatic cells. Hum Genet 2001;108:167–173.
39 Battaile KP, Bateman RL, Mortimer D, Mulcahy J, Rathbun RK, et al: In vivo selection of wild-
type hematopoietic stem cells in a murine model of Fanconi anemia. Blood 1999;94:2151–2158.
40 Waisfisz Q, Morgan NV, Savino M, de Winter JP, van Berkel CG, et al: Spontaneous functional
correction of homozygous Fanconi anaemia alleles reveals novel mechanistic basis for reverse
mosaicism. Nat Genet 1999;22:379–383.
41 Ikeda H, Matsushita M, Waisfisz Q, Kinoshita A, Oostra AB, et al: Genetic reversion in an acute
myelogenous leukemia cell line from a Fanconi anemia patient with biallelic mutations in BRCA2.
Cancer Res 2003;63:2688–2694.
42 Manoharan A: Fanconi anaemia. Br J Haematol 2006;135:139.
43 Cohen-Haguenauer O, Peault B, Bauche C, Daniel MT, Casal I, et al: In vivo repopulation ability
of genetically corrected bone marrow cells from Fanconi anemia patients. Proc Natl Acad Sci
USA 2006;103:3240–3245.

H. Hoehn
Department of Human Genetics, Biocenter
Am Hubland, 97074 Würzburg (Germany)
Tel. ⫹49 931 888 4071, Fax ⫹49 931 888 4434, E-Mail [email protected]

Hoehn/Kalb/Neveling/Friedl/Bechtold/Herterich/Sun/Gruhn/Hanenberg/Schindler 172
 Schindler D, Hoehn H (eds): Fanconi Anemia. A Paradigmatic Disease for the Understanding of
Cancer and Aging. Monogr Hum Genet. Basel, Karger, 2007, vol 15, pp 173–182

Stem Cell Transplantation in Fanconi

Anemia – Recent Advances with
Alternative Donors
M. Eyrich, B. Winkler, P.G. Schlegel
Pediatric Stem Cell Transplant Program, Children’s Hospital,
University of Würzburg, Würzburg, Germany

Abstract
Bone marrow transplantation from an HLA-identical sibling donor is the treatment of
choice for Fanconi anemia (FA) patients with bone marrow failure. However, with today’s
small size families less than 25% of FA patients have a matching sibling donor. The remain-
ing patients can be treated with stem cell transplantations from alternative donors such as
matched unrelated or haploidentical donors. While results with conventional bone marrow
transplantation continually improved, outcome after alternative donor transplantation remained
poor due to high rates of transplant-related complications, graft rejection and graft-versus-host
disease (GvHD). Recently, advances with less myeloablative and more immunosuppressive
conditioning regimens have demonstrated promising results. Insights into the mechanisms of
immune reconstitution could show that at least in children immune recovery after haploiden-
tical stem cell transplantation is fast and results in complete restoration of normal immune
function within the first year after transplantation. Finally, innovations in the field of stem
cell collection and processing have led to cellular graft compositions which now provide reli-
able engraftment in almost all patients with concomitant low incidence of GvHD. This
review discusses these recent advances in alternative donor transplantations and concludes
that this treatment option should be an early recommendation for FA patients with bone mar-
row failure who lack an HLA-identical sibling donor.
Copyright © 2007 S. Karger AG, Basel

Although the name of the disease suggests a purely hematological dis-

order, Fanconi anemia (FA) is a systemic disease which is characterized by
physical abnormalities, progressive bone marrow failure and increased risk of
malignancies. A typical diagnostic feature, which might in part explain some of
these symptoms, is increased chromosome breakage at baseline, and even more
dramatic upon exposure to clastogenic agents such as mitomycin C. Complete
bone marrow failure is often not present at diagnosis. Mild to moderate throm-
bocytopenia or leukopenia may precede pancytopenia. However, if pancytope-
nia develops, 80% of patients would not survive more than two years without
therapies other than blood transfusions [1]. Another threat for FA patients is the
occurrence of acute leukemias which has been reported to be in the range of
10–30% [2]. Once overt leukemia has developed, patients suffer from a poor
prognosis since conventional chemotherapy is difficult to implement. Therefore,
it is important to treat bone marrow failure before complicating events such as
aplastic anemia or leukemia can develop. Up to now, the only curative option to
achieve this goal involves allogeneic bone marrow transplantation. However, it
has to be kept in mind that this procedure will not correct the genetic back-
ground, physical abnormalities and the increased risk of solid tumors in older
long-term survivors.
Early attempts to treat bone marrow failure in FA patients with allogeneic
bone marrow transplantation demonstrated the feasibility of this approach,
however with limited success [3]. High cyclophosphamide toxicity, severe
graft-versus-host disease (GvHD), and graft rejection urged researchers to look
for modified conditioning regimens and better GvHD prophylaxis. The use of
low dose cyclophosphamide, limited field irradiation, plus antithymocyte
globulin (ATG) for pretransplant conditioning and cyclosporine for GvHD-
prophylaxis resulted in a 66% two-year survival after transplantation of bone
marrow from a human leukocyte antigen (HLA-) identical sibling [4]. However,
results with transplantations from alternative donors (unrelated or mismatched
related) remained poor (29%) [4]. Taken together, these early reports about the
use of bone marrow transplantation in FA patients suggested that this approach
is feasible and may be extended to patients who lack an HLA-identical sibling
donor. In order to improve results of alternative donor transplantations, less
toxic conditioning regimens and novel techniques for tighter control of the bal-
ance between GvHD and graft rejection had to be developed.

Transplantation beyond HLA-Barriers

Until recently, transplantation of bone marrow from an HLA-identical sib-

ling was considered the standard treatment regimen for children requiring a
stem cell graft. Transplantation of complete bone marrow, however, has major
limitations: first, the limited availability of an HLA-identical sibling donor, and
second the potential risk of developing GvHD, which may attack skin, gut and
liver of the transplant recipient. From 1990 on peripheral blood stem cells were
increasingly used instead of bone marrow, unfortunately this did not result in a

Eyrich/Winkler/Schlegel 174
decline of the incidence of acute or chronic GvHD, probably due to the even
higher number of T cells in peripheral blood stem cell grafts. For years, it has
been believed that T cells in the graft are indispensable to allow for a stable
engraftment in an allogeneic environment. Only patients with an inherent lack
of T cells like children with severe combined immunodeficiency were able to
accept even T-cell depleted, haploidentical grafts without intensive pretrans-
plant conditioning [5]. However, this view changed with introduction of the
‘megadose concept’: high numbers of CD34⫹ stem cells without T-cell support
are able to suppress anti-donor cytotoxic T lymphocyte (CTL) responses and
ensure engraftment even after transplantation beyond HLA-barriers [6].
Innovations in the field of stem cell processing techniques made it possible to
isolate the required large numbers of highly purified hematopoetic progenitor
cells [7]. This approach was rapidly translated into clinical protocols, first in
high-risk adult leukemia patients [8], later also in a pediatric cohort with child-
hood leukemia [9]. The first case report of a successful haploidentical trans-
plantation in a girl with FA was published by Elhasid et al. in 2000 [10].
Interestingly, this report followed exactly the megadose concept with transplan-
tation of large numbers of highly purified hematopoietic progenitor cells and
only minimal T-cell content. Another study, published in the same year reported
29 FA patients transplanted with partially T-cell depleted bone marrow from
alternative donors after receiving cyclophosphamide (40 mg/kg), ATG and total
body irradiation (450–600 cGy) as conditioning regimen [11]. In this cohort,
the probability of survival at 1 year was only 34%. Although a final assessment
about the use of haploidentical stem cell transplantation based on these results
is not possible so far, it is tempting to speculate that the megadose concept will
result in superior engraftment rates with a concomitant low incidence of acute
and chronic GvHD. In fact, Lang et al. reported a series of 25 patients with non-
malignant diseases (excluding FA) where transplantation of large numbers of
positively enriched CD34⫹ hematopoietic stem cells from related or unrelated
donors resulted in a survival rate of 88% without acute GvHD greater than
grade II and limited chronic GvHD in only 8% of the patients [12]. Future stud-
ies have to show whether these encouraging results will also apply to FA
patients. If haploidentical stem cell transplantation in combination with the
megadose concept proves to be sufficiently safe and effective in FA then this
therapeutic option could be offered early on to all FA patients with deterioration
of bone marrow function. Either parents or siblings qualify as potential hap-
loidentical donors such that almost every patient with FA is likely to have at
least one haploidentical donor. However, successful outcome of stem cell trans-
plantation is not only based on selection of the best donor and appropriate stem
cell processing techniques but also on a pretransplant conditioning regimen
which has been optimized for the special needs of FA patients.

Stem Cell Transplantation in Fanconi Anemia 175

 Breaking Resistance of Host T Cells – The Way to New
Immunological Conditioning Regimens

The first conditioning regimens used for preparation of patients for bone
marrow transplantation contained mainly total body irradiation (TBI) plus one
or two alkylating agents. These regimens created ‘marrow space’ for the allo-
geneic graft by eradicating host hematopoiesis but also had a profound
immunosuppressive effect. Initial reports with these conditioning regimens in
FA patients showed an unacceptable high toxicity especially due to the use of
cyclophosphamide and its metabolites [3]. Another problem was the high rate
of graft rejection frequently seen in FA patients [13]. These data suggested that
existing protocols for pretransplant conditioning were too toxic albeit not
immunosuppressive enough for FA patients to allow for safe allogeneic engraft-
ment. In 1991 Fischer et al. demonstrated that addition of a monoclonal anti-
body against leukocyte function-associated antigen (LFA-1, CD11a) to the
conditioning regimen significantly reduced rates of graft rejection in patients
with nonmalignant diseases [14]. Furthermore, it could be shown that patients
with primary nonengraftment or graft rejection can be successfully retrans-
planted after a purely immunological reconditioning based on the anti-T-cell
antibody OKT3 and methylprednisolone [15]. In 1997, the first fludarabine-
based protocol with addition of low-dose cyclophosphamide and ATG was pub-
lished [16]. Fludarabine monophosphate is a purine analogue and has a much
more favorable toxicity profile for FA patients compared to alkylating agents or
irradiation. Most of the conditioning regimens published thereafter used low-
dose cyclophosphamide and ATG in combination with either limited thoraco-
abdominal irradiation [11] or fludarabine [17–19] or both [20–22]. All authors
reported a remarkable primary engraftment rate of almost 100% with sec-
ondary graft rejections in 4–25% of patients [17, 19, 21–23]. Most of these lat-
ter patients could be successfully retransplanted. The rates of acute and chronic
GvHD also decreased over the past decade. However, the incidence of GvHD
seems to be more related to donor selection and the degree of T-cell depletion
than to the use of TBI, cyclophosphamide, or fludarabine.
Taken together, over the last two decades considerable improvements in
pretransplant conditioning regimens were achieved. The application of irradia-
tion or alkylating agents which carry a major risk for secondary malignancies in
FA patients could be dramatically reduced to the benefit of less toxic and highly
T-cell suppressive regimens. These advances in patient conditioning not only
resulted in higher rates of engraftment but also paved the way for the use of
more sophisticated grafts like T-cell depleted stem cells from alternative
donors. Recently, a study has been initiated which completely avoids irradiation
and alkylating agents. Prepared by a conditioning regimen with fludarabine and

Eyrich/Winkler/Schlegel 176
therapeutic antibodies against CD45 and CD52, FA patients in this study will be
grafted with bone marrow from matched related or CD34⫹ peripheral blood
stem cells from haploidentical donors (Persis Amrolia, personal communica-
tion). This ongoing study surely forms the preliminary endpoint of a logical
development, which was fostered by experiences with immunological condi-
tioning regimens from the last 25 years.

Immune Reconstitution after Haploidentical Stem

Cell Transplantation

The high expectations after the first successful stem cell transplantations
from haploidentical donors were attenuated by a high rate of infectious compli-
cations, especially in adults [24]. This was due to the extremely low numbers of
T cells in the graft and a subsequently delayed immune reconstitution.
However, the patterns of immune recovery after transplantation of highly
enriched, CD34⫹ peripheral blood stem cells from haploidentical donors were
initially only poorly understood. Our group has prospectively analyzed the pat-
terns of immune reconstitution in children after haploidentical transplantations.
In a first study, twenty children transplanted with high doses of purified CD34⫹
hematopoietic stem cells were prospectively monitored for their immune recon-
stitution during the first post-transplant year [25]. T, B, and NK cells began to
reconstitute (as a median) on day ⫹72, ⫹68, and ⫹30 respectively (fig. 1).
During extended follow-up, their numbers and proliferative capacity upon
mitogen stimulation continually increased. Early reconstituting T cells were
predominantly of a memory (primed, activated) phenotype. Naive T cells
emerged after approximately 6 months post transplantation. All patients self-
maintained sufficient immunoglobulin levels after day ⫹200. This study
demonstrated for the first time that at least pediatric recipients of highly puri-
fied, haploidentical stem cells are able to reconstitute functioning T-, B- and
NK-cell compartments within the first post-transplant year.
In the subsequent follow-up study, we analyzed the factors governing
the normalization of the initially restricted T-cell receptor repertoire [26].
Normalization of the initially restricted repertoire is of prime importance to the
patient with regard to fighting off potentially life-threatening infections. We
determined the relative contributions of naive and memory T-cell subsets within
the CD4⫹ and CD8⫹ compartments to the evolution of overall TCR-repertoire
complexity following transplantation of CD34⫹ selected peripheral blood prog-
enitors. During the first post-transplant year, sorted CD4/45RA, CD4/45RO,
CD8/45RA and CD8/45RO subsets were analyzed at three monthly intervals for

Stem Cell Transplantation in Fanconi Anemia 177

 1,400 900
CD3⫹
1,200 CD8⫹ 800
CD4⫹
700
1,000
600
T-cells/␮l

B-cells/␮l
800 500
600 400
300
400
200
200
100

50 100 150 200 250 300 350 400 450 50 100 150 200 250 300 350 400 450
a Days post transplantation b Days post transplantation

500
450
400
350
NK-cells/␮l

300
250
200
150
100
50

50 100 150 200 250 300 350 400 450

c Days post transplantation

Fig. 1. Immune reconstitution after haploidentical stem cell transplantation in children

[25]. (a) Increase in peripheral T-cell numbers (CD3⫹, ) after haploidentical SCT (n ⫽ 20,
median values). T-cell counts ⬎100/␮l were present after a median of 72 days (range 14–123).
Cytotoxic, suppressor CD8⫹ T-cells ( ) recovered faster than CD4⫹ T-helper cells ( ), result-
ing in an inverted CD4/CD8 ratio. For CD4⫹ T-cells the median time to reach ⬎100 cells/␮l was
82 days (range 13–131), for CD8⫹ T-cells the median time to reach ⬎100 cells/␮l was 74 days
(range 15–123). (b) Increase in peripheral CD19⫹ B-cell numbers ( ) after haploidentical SCT
(n ⫽ 20, median values). The median time to reach ⬎100 CD19⫹ cells/␮l was 68 days (range
13–340). (c) Increase in peripheral CD16⫹/56⫹ NK-cell numbers ( ) after haploidentical SCT
(n ⫽ 20, median values). NK-cell values recovered more rapidly than T- and B-cell numbers and
reached a median of 305 CD16⫹/56⫹ NK-cells/␮l already on day 30 post-transplant.

their respective TCR-repertoire complexity by CDR3-size-spectratyping. Skewing

of the repertoire was observed only in early memory-type T cells. With the help
of sorting techniques, we found direct evidence that the emerging naive T cells
were responsible for correcting the initially skewed TCR repertoire.

Eyrich/Winkler/Schlegel 178
 Thymus-dependent T-cell regeneration (de novo generation of naive
T-cells) is a key pathway for immune reconstitution after stem cell transplanta-
tion. In a third study, we prospectively assessed T-cell dynamics and thymic
function in 164 pediatric patients between 1 and 124 months post-transplant by
measuring T-cell-receptor recombination excision circles (TRECs) as a mea-
sure for de novo thymic T-cell emigrants and spontaneous expression of Ki67 in
peripheral T-cell subsets (as a measure for proliferation of T cells in the periph-
eral pool) [27]. We analyzed the impact of recipient age, conditioning regimen,
type of donor and graft, stem cell dose, and graft-versus-host disease (GvHD)
on the onset as well as on the plateau of thymic output. Multivariate analysis
revealed that the onset of thymic recovery was inversely correlated only with
recipient age (p ⬍ 0.0002), whereas the plateau of thymic output was higher
in patients receiving increased (⬎107 CD34⫹/kg BW) stem cell numbers
(p ⬍ 0.0022). Interestingly, donor type, stem cell source and conditioning regi-
men influenced none of the analyzed parameters. Onset and plateau of thymic
activity were found to be independently regulated by different transplant-related
factors.
Taken together our data demonstrated that in children transplantation of
highly purified stem cells results in fast reconstitution of NK-, T-, and B-cell
compartments. These data are in contrast to results from adult patients after
T-cell depleted haploidentical transplantations, whose T-cell reconstitution can
take up to two years or more. They also correlate with the clinical observation
that transplanted children have a lower incidence of infectious complications
than adult patients. One of the major differences between adults and children
with regard to T-cell reconstitution seems to be the enhanced thymic recovery in
children [28]. Thus, transplantation of highly purified haploidentical stem cells
should especially be considered in pediatric patients, since immune recovery in
children is fast and no long-lasting immunodeficiency due to the transplant pro-
cedure has to be expected.

Future Directions

Bone marrow transplantation from an HLA-identical sibling donor is still

the treatment of choice for FA patients with signs of bone marrow failure.
However, recent advances in pretransplant conditioning regimens, stem cell
processing techniques, and peritransplant supportive care have made transplan-
tations from alternative donors an interesting option for FA patients lacking an
HLA-identical sibling donor. To date, no randomized data exist about the pref-
erential use of matched unrelated versus haploidentical donors, or cord blood
versus peripheral blood stem cells. Several aspects point towards a beneficial

Stem Cell Transplantation in Fanconi Anemia 179

role of haploidentical stem cell transplantations in FA patients: hematologic and
immune recovery are well documented and obviously fast, especially in chil-
dren. Haploidentical donors are readily available for a second donation when
graft rejection occurs or for immunotherapy in case of increasing mixed
chimerism, which should be treated promptly in FA patients [11]. A number of
innovations are currently under investigation, namely the use of conditioning
regimens, which completely avoid irradiation and alkylating agents, and novel
stem cell processing techniques such as CD3/CD19 depletion instead of CD34
selection which will further promote engraftment rates by cotransplantation of
large numbers of graft-facilitating cells. These improvements in the field of
alternative donor transplantation will hopefully ensure that each FA patient with
bone marrow failure can be offered a safe and effective allogeneic stem cell
transplantation before aplastic anemia, myelodysplasia or even leukemia develops.

Acknowledgements

This work was supported in part by IZKF Wuerzburg (Z-2/7) and by a grant from the
Childhood Cancer Parent’s Initiative Wuerzburg.

References

1 Bernard J, Mathé G, Najean Y: Contribution de l’étude clinique et physiopathologique de la mal-

adie de Fanconi. Rev Fr Etudes Clin Biol 1958;3:599–612.
2 Schaison G, Leverger G, Yildiz C, Berger R, Bernheim A, Gluckman E: Fanconi’s anemia.
Incidence of its development into leukemia. Presse Med 1983;12:1269–1274.
3 Gluckman E, Devergie A, Schaison G, Bussel A, Berger R, et al: Bone marrow transplantation in
Fanconi anaemia. Br J Haematol 1980;45:557–564.
4 Gluckman E, Auerbach AD, Horowitz MM, Sobocinski KA, Ash RC, et al: Bone marrow trans-
plantation for Fanconi anemia. Blood 1995;86:2856–2862.
5 Friedrich W, Goldmann SF, Ebell W, Blutters-Sawatzki R, Gaedicke G, et al: Severe combined
immunodeficiency: treatment by bone marrow transplantation in 15 infants using HLA-
haploidentical donors. Eur J Pediatr 1985;144:125–130.
6 Rachamim N, Gan J, Segall H, Krauthgamer R, Marcus H, et al: Tolerance induction by ‘mega-
dose’ hematopoietic transplants: donor-type human CD34 stem cells induce potent specific reduc-
tion of host anti-donor cytotoxic T lymphocyte precursors in mixed lymphocyte culture.
Transplantation 1998;65:1386–1393.
7 Schumm M, Lang P, Taylor G, Kuci S, Klingebiel T, et al: Isolation of highly purified autologous
and allogeneic peripheral CD34⫹ cells using the CliniMACS device. J Hematother 1999;8:
209–218.
8 Aversa F, Tabilio A, Terenzi A, Velardi A, Falzetti F, et al: Successful engraftment of T-cell-
depleted haploidentical ‘three-loci’ incompatible transplants in leukemia patients by addition of
recombinant human granulocyte colony-stimulating factor-mobilized peripheral blood progenitor
cells to bone marrow inoculum. Blood 1994;84:3948–3955.
9 Handgretinger R, Klingebiel T, Lang P, Schumm M, Neu S, et al: Megadose transplantation of
purified peripheral blood CD34(⫹) progenitor cells from HLA-mismatched parental donors in
children. Bone Marrow Transplant 2001;27:777–783.

Eyrich/Winkler/Schlegel 180
10 Elhasid R, Ben Arush MW, Katz T, Gan Y, Shechter Y, et al: Successful haploidentical bone mar-
row transplantation in Fanconi anemia. Bone Marrow Transplant 2000;26:1221–1223.
11 MacMillan ML, Auerbach AD, Davies SM, Defor TE, Gillio A, et al: Haematopoietic cell trans-
plantation in patients with Fanconi anaemia using alternate donors: results of a total body irradia-
tion dose escalation trial. Br J Haematol 2000;109:121–129.
12 Lang P, Klingebiel T, Bader P, Greil J, Schumm M, et al: Transplantation of highly purified peripheral-
blood CD34⫹ progenitor cells from related and unrelated donors in children with nonmalignant
diseases. Bone Marrow Transplant 2004;33:25–32.
13 Zwaan CM, Van Weel-Sipman MH, Fibbe WE, Oudshoorn M, Vossen JM: Unrelated donor bone
marrow transplantation in Fanconi anaemia: the Leiden experience. Bone Marrow Transplant
1998;21:447–453.
14 Fischer A, Friedrich W, Fasth A, Blanche S, Le Deist F, et al: Reduction of graft failure by a mon-
oclonal antibody (anti-LFA-1 CD11a) after HLA nonidentical bone marrow transplantation in
children with immunodeficiencies, osteopetrosis, and Fanconi’s anemia: a European Group for
Immunodeficiency/European Group for Bone Marrow Transplantation report. Blood 1991;77:
249–256.
15 Schlegel P, Eyrich M, Bader P, Handgretinger R, Lang P, et al: OKT-3 based reconditioning regi-
men for early graft failure in HLA-non-identical stem cell transplants. Br J Haematol 2000;111:
668–673.
16 Kapelushnik J, Or R, Slavin S, Nagler A: A fludarabine-based protocol for bone marrow trans-
plantation in Fanconi’s anemia. Bone Marrow Transplant 1997;20:1109–1110.
17 Motwani J, Lawson SE, Darbyshire PJ: Successful HSCT using nonradiotherapy-based condition-
ing regimens and alternative donors in patients with Fanconi anaemia – experience in a single UK
centre. Bone Marrow Transplant 2005;36:405–410.
18 Rossi G, Giorgiani G, Comoli P, Nobili B, Salvaneschi L, et al: Successful T-cell-depleted, related
haploidentical peripheral blood stem cell transplantation in a patient with Fanconi anaemia using a
fludarabine-based preparative regimen without radiation. Bone Marrow Transplant 2003;31:
437–440.
19 George B, Mathews V, Shaji RV, Srivastava V, Srivastava A, Chandy M: Fludarabine-based condi-
tioning for allogeneic stem cell transplantation for multiply transfused patients with Fanconi’s
anemia. Bone Marrow Transplant 2005;35:341–343.
20 Boulad F, Gillio A, Small TN, George D, Prasad V, et al: Stem cell transplantation for the
treatment of Fanconi anaemia using a fludarabine-based cytoreductive regimen and T-cell-
depleted related HLA-mismatched peripheral blood stem cell grafts. Br J Haematol 2000;111:
1153–1157.
21 Boyer MW, Gross TG, Loechelt B, Leemhuis T, Filipovich A, Harris RE: Low risk of graft-
versus-host disease with transplantation of CD34 selected peripheral blood progenitor cells from
alternative donors for Fanconi anemia. J Pediatr Hematol Oncol 2003;25:890–895.
22 Yabe H, Inoue H, Matsumoto M, Hamanoue S, Koike T, et al: Allogeneic haematopoietic cell
transplantation from alternative donors with a conditioning regimen of low-dose irradiation, flu-
darabine and cyclophosphamide in Fanconi anaemia. Br J Haematol 2006;134:208–212.
23 Torjemane L, Ladeb S, Ben Othman T, Abdelkefi A, Lakhal A, Ben Abdeladhim A: Bone marrow
transplantation from matched related donors for patients with Fanconi anemia using low-dose
busulfan and cyclophosphamide as conditioning. Pediatr Blood Cancer 2006;46:496–500.
24 Aversa F, Tabilio A, Velardi A, Cunningham I, Terenzi A, et al: Treatment of high-risk acute
leukemia with T-cell-depleted stem cells from related donors with one fully mismatched HLA
haplotype. N Engl J Med 1998;339:1186–1193.
25 Eyrich M, Lang P, Lal S, Bader P, Handgretinger R, et al: A prospective analysis of the pattern of
immune reconstitution following transplantation of HLA-disparate hematopoietic stem cells from
parental donors. Br J Haematol 2001;114:422–432.
26 Eyrich M, Croner T, Leiler C, Lang P, Bader P, et al: Distinct contributions of CD4⫹ and CD8⫹
naive and memory T-cell subsets to overall TCR-repertoire complexity following transplantation
of T-cell depleted CD34 selected hematopoietic progenitor cells from unrelated donors. Blood
2002;100:1915–1918.

Stem Cell Transplantation in Fanconi Anemia 181

27 Eyrich M, Wollny G, Tzaribaschev N, Dietz K, Brugger D, et al: Onset of thymic recovery and
plateau of thymic output are differentially regulated after stem cell transplantation in children.
Biol Blood Marrow Transplant 2005;11:194–205.
28 Storek J, Joseph A, Dawson MA, Douek DC, Storer B, Maloney DG: Factors influencing
T-lymphopoiesis after allogeneic hematopoietic cell transplantation. Transplantation 2002;73:
1154–1158.

Matthias Eyrich
University of Würzburg
Children’s Hospital, Pediatric Stem Cell Transplantation Program
Josef-Schneider-Strasse 2, D30
D–97080 Würzburg (Germany)
Tel. ⫹49 931 201 27640, Fax ⫹49 931 201 27649
E-Mail [email protected]

Eyrich/Winkler/Schlegel 182
 Schindler D, Hoehn H (eds): Fanconi Anemia. A Paradigmatic Disease for the Understanding of
Cancer and Aging. Monogr Hum Genet. Basel, Karger, 2007, vol 15, pp 183–199

Fanconi Anemia Genes in Vertebrates:

Evolutionary Conservation, Sex-Linkage,
and Embryonic Expression of FANCC and
FANCG in Avian Cells
I. Nandaa, A. Buwea, A. Wizenmanb, M. Takatac, T. Haafd,
M. Schartle, M. Schmida
a
Department of Human Genetics, University of Würzburg, Würzburg, bGSF- National
Research Center for Environment and Health, Institute for Stem Cell Research,
Neuherberg/Munich, Germany; cDepartment of Human Genetics, Research Institute
for Radiation Biology and Medicine, Hiroshima University, Hiroshima, Japan;
d
Department of Human Genetics, University of Mainz, Mainz, eDepartment of
Physiological Chemistry I, University of Würzburg, Würzburg, Germany

Abstract
Orthologs of the human Fanconi anemia (FANC) genes have been identified in several
vertebrate and invertebrate model organisms, indicating variously conserved functions of the
FANC protein complex. In particular, the analysis of chicken DT40 cells has made important
contributions to the functional characterization of FANC genes. Orthologs for most human
FANC genes have been found in the chicken genome which is considerably smaller than the
human genome. Even though a Fanconi anemia-like phenotype has not been described in
avian species, FANC-deficient chicken cells display the same sensitivity to DNA crosslinking
agents as mammalian cells. In addition to a brief review of FANC gene orthologs in verte-
brates and lower organisms, we here show that the chicken FANCC and FANCG genes are
highly expressed in multiple avian embryonic tissues, whereas in adult birds strong expres-
sion was only observed in gonads which underlines the putative function of these genes dur-
ing premeiotic DNA replication and meiotic recombination. We further show that the avian
homologs of two important members of the FANC gene family, FANCC and FANCG, are
located on the chicken Z sex chromosome. We tested the sensitivity of chicken embryonic
fibroblasts from males (ZZ) and females (ZW) towards mitomycin C and observed a gender
difference. This observation is consistent with the absence of Z chromosome dosage com-
pensation in birds and supports the essential role of FANCC and FANCG in the cellular
defense against DNA crosslinking agents.
Copyright © 2007 S. Karger AG, Basel
 Cells of all FA patients are highly susceptible to the induction of chromo-
some breaks with DNA crosslinking agents such as diepoxybutane (DEB) or
mitomycin C (MMC). This indicates that mutations in the different FANC genes
cause a similar DNA repair defect(s) [1–4]. Sensitivity to crosslinking agents is
a cytological ‘FA hallmark’, which is widely used for diagnostic testing of FA.
Somatic cell fusion and biochemical analyses have delineated 13 different FA
complementation groups (FA-A, B, C, D1, D2, E, F, G, I, J, L, M, and N). With
the exception of FANCI, the disease genes underlying these groups have already
been cloned [5–7].
Orthologs of the centrally important FANCD2 protein have been detected
in such diverged organisms as Arabidopsis, Caenorhabditis, Drosophila, fugu
and zebrafish [8–10], indicating likely conservation of FANCD2 function. In
this report we review recent findings on the FA pathway in different vertebrates.
Special attention is given to birds which phylogenetically are closest to mam-
mals. Since they can be easily manipulated by homologous recombination,
chicken DT40 cells are increasingly used in FA research. An interesting addi-
tional aspect of the avian cell model in FA research is the increased susceptibil-
ity of FA cells to oxidative stress because, for various reasons, birds appear to
be less or more resistant than mammals to the detrimental effects of reactive
oxygen species, in general [11, 12].

Conserved and Divergent Functions of FANC Genes

Double-strand break (DSB) intermediates formed by interstrand cross-

links in the stalled replication fork can be resolved through different mecha-
nisms such as homologous recombination (HR), non-homologous end joining
(NHEJ), and error-prone translesion synthesis (TLS). The finding that
FANCD1 (BRCA2) and the HR protein RAD51 directly interact provided the
first evidence connecting the FA pathway with HR-mediated DNA repair [13].
Subsequent studies found HR-defects in BRCA2-deficient cells [14] whereas
the in vitro interaction between BRCA2 and RAD51 stimulates RAD51-
mediated recombination [15]. In meiotic cells, antibodies against FANCD2
specifically label the synaptonemal complexes which are required for recombi-
nation between homologous chromosomes [16]. This supports the hypothesis
that many DNA repair proteins that are required for the detection and process-
ing of damaged DNA in somatic cells perform related functions during meiotic
recombination [17].
DSBs induced by crosslinking agents are preferentially repaired by HR.
Experimental evidence suggests that monoubiquitinated FANCD2 promotes
efficient HR repair, possibly by interacting with the BRCA2-RAD51 complex

Nanda/Buwe/Wizenman/Takata/Haaf/Schartl/Schmid 184
and other HR proteins like FANCJ [18, 19]. So far, there are only few studies
that directly test recombination activity, for example by introducing a plasmid
substrate in normal and FA cell extracts [20] or in murine cells with a targeted
mutation at the Fanca locus [21]. However, there are a number of studies that
have evaluated FANC protein function using cell-free assays with Xenopus egg
extracts [22], knockdown zebrafish embryos [10] and chicken DT40 knockout
cells [23]. Collectively, these results suggest a role(s) for FANC proteins in
DSB repair most likely through error-free HR. As such, FANC proteins appear
to be important for maintaining genomic stability [24]. This guardian role is
consistent with evolutionary conservation of at least part of the FA pathway.
Besides their crucial role in HR-mediated DSB repair, FANC proteins have
been functionally linked to oxygen metabolism, cell cycle regulation, hemato-
poiesis, and apoptosis. Human FA cells show increased susceptibility to oxida-
tive stress which suggests a role for FANC proteins in the protection against the
endogenous production of reactive oxygen species (ROS). In spite of a higher
metabolic rate, body temperature and blood sugar level, which all contribute to
increased DNA damage after oxidative stress, birds produce relatively fewer
ROS than mammals [11, 12]. This suggests that birds must be endowed with
preventive or protective mechanisms that may also include the FA family of
genes.

Evolutionary Conservation of FANC Genes

The phenotypes of Fanca, Fancc, and Fancg knockout mice and even dou-
ble knockouts are very similar, indicating involvement of different Fanc genes
in the same biochemical pathway(s). Although Fanc-deficient mouse fibrob-
lasts are also sensitive to crosslinking agents, unlike humans KO mice are not
affected by bone marrow failure or malignancies [25]. Interestingly, the differ-
ent KO mice all showed hypogonadism and impaired fertility [26], indicating a
functional role for FANC proteins during meiosis.
Fancd2-deficient mice and zebrafish embryos exhibit some malforma-
tions, i.e. microcephaly and microphthalmia that are also observed in FA patients.
Increased apoptosis in these embryos can be partially corrected by injection of
human FANCD2 protein [10]. The zebrafish orthologs of nine human FANC
genes have been cloned and mapped to syntenic regions shared by the zebrafish
and human genomes [27]. Evidently, the FANC gene network and most likely
the FA pathway already existed in a common ancestor of teleost fish and
tetrapods. Because of a genome duplication event in the ray fin fish lineage [28,
29] most zebrafish orthologs of human genes exist as two paralogous gene
copies. However, none of the zebrafish Fanc genes has such a second copy,

Fanconi Anemia Genes in Vertebrates 185

suggesting loss of the duplicated Fanc genes after the genome duplication
event. The high degree of evolutionary conservation of FANC genes in lower
vertebrates is also documented by the identification of FANCD2 and FANCL
orthologs in the chordate Ciona intestinalis [27].
Similar to the situation in humans, Xenopus FANCF protein appears to be
crucial for proper assembly of the FANC core complex [30]. Both FANCA and
FANCD2 along with FANCL prevent the accumulation of DNA breaks that
arise during unperturbed DNA replication [22]. Functional homologs of
FANCD2 and its activator FANCL have been found in Caenorhabditis and
Drosophila [8, 31]. Homozygous C. elegans mutants are sensitive to crosslink-
ing agents but viable [9, 32]. The fungus Ustilago maydis is endowed with a
FANCD1 (BRCA2) ortholog [33], and a homolog of the recently identified
FANCJ (Hef) gene is even present in archaebacteria where it is needed to
process blocked replication forks. FANCJ (Hef) homologs are indeed highly
conserved among single and multicellular organisms [34], whereas orthologs of
the sex-linked human FANCB gene are only found in mouse, chicken, and
zebrafish but not in worm and fly [35].

FANC Orthologs in Birds

During the course of evolution, the avian genome has been subject to
extensive structural reorganization. Modern bird genomes are compartmental-
ized into macro- and micro-chromosomes. Although the size of the chicken
genome amounts to only 40% of the human genome it is endowed with almost
the same gene content [36]. It is therefore not surprising that orthologs of
almost all human FANC genes were found in chicken.
The first well characterized FANC gene in chicken was FANCD1 (BRCA2)
[37, 38], which maps to GGA chromosome 1. Most other avian FANC genes
have been isolated from the highly recombinogenic chicken B cell line DT40
[39, 40]. The overall amino acid sequence similarity between human and
chicken FANC orthologs is not particularly high, amounting to 39% for FANCC,
40% for FANCD1 (BRCA1), 49% for FANCG, 54% for FANCJ (BRIP1) and
57% for FANCD2. Despite this overall low level of sequence conservation
between avian and mammalian FANC genes, functionally important domains,
including the PHD finger domain in FANCL, the helicase and degenerated
nuclease domains in FANCM, the helicase domain in FANCJ, the RAD51-binding
BRC repeats in FANCD1, and to some extent the tetratricopeptide motif in
FANCG are highly conserved in the chicken genome. This is consistent with the
view that these domains act as scaffolds to co-ordinate the functions of different
FANC proteins. One remarkable species difference is the absence of a specific

Nanda/Buwe/Wizenman/Takata/Haaf/Schartl/Schmid 186
 L
B

NC
NC

FA

M F
D1
FA

NC C
FA FAN
NC
FA

1 2 3 4 5 6 7

D2
A
NC

NC
FA

FA
8 9 10 11 12 13 14

J
NC
FA
15 16 17 18 19 20 21

E
NC
FA
22 23 24 25 26 27 28

C G
NC NC
29 30 31 32 33 34 35

FA FA
36 37 38 Z Z

Fig. 1. DAPI-stained male chicken karyotype indicating the chromosomal localiza-

tions of human FANC gene orthologs. The prelimary assignments of FANCE to GGA26 and
FANCJ to GGA19 are indicated in green. The newly mapped FANCC and FANCG in the
Z chromosome are designated in red.

phosphorylation site in the chicken FANCJ protein. In human cells this site is
critical for interaction of FANCJ with BRCA1 [39].
The availability of the complete chicken genome sequence [36] facilitates
the in silico identification of avian orthologs of human disease genes. By BLAST
searches of the chicken genome (http://www.ncbi.nlm.nih.gov/mapview/;
http://www.ensembl.org/Gallus_gallus) with human FANC gene sequences, we
have mapped orthologs of the human FANC genes to chicken chromosomes
(fig. 1). With exception of FANCE and FANCJ which are represented by several
contigs in the current version of the chicken genome, the chicken and human
FANC orthologs lie in syntenic regions.
In order to confirm the in silico data, we localized FANCC and FANCG by
fluorescence in-situ hybridization (FISH) on chicken metaphase spreads.
Z-linkage of these genes is not unexpected because hemizygous (ZW) chicken
DT40 cells are known to contain only single copies of FANCC and FANCG
[23, 41]. Indeed, we found specific FISH signals for both genes (fig. 2a and b)
on the short arm of the chicken Z sex chromosome with FANCC close to the

Fanconi Anemia Genes in Vertebrates 187

 ZZ ZW ZZ ZW
23.1

FANCG
6.6

p 4.4
Z W Z W
a Ch FANCG FANCC

q 2.3
2.0

Z Z Z Z c GGA Z
b Ch FANCC

d PstI

Fig. 2. Physical mapping of chicken FANCG and FANCC to the Z sex chromosome. (a)
FANCG hybridization signals (green FITC fluorescence) on the Z short arm. The ZW cut-
outs are counterstained with both DAPI (left) and propidium iodide (right). The bright PI flu-
orescence of terminal heterochromatin allows the distinction between the long and short arm
of the metacentric Z. (b) FANCC signals appear close to the centromere of the Z short arm.
(c) Ideogram of the chicken Z chromosome illustrating the location of both genes. (d)
Southern hybridization of FANCG cDNA to PstI-digested male and female chicken genomic
DNA. Note the increased hybridization intensity of bands in males compared to females.
Size markers in kb are indicated on the left.

centromere and FANCG at a telomeric position of Zp (fig. 2c). Z-linkage of

these genes was also indicated from Southern hybridization of FANCG cDNA
to male and female chicken genomic DNA (fig. 2d). The respective signals
showed an approximately twofold intensity in males compared to females.
Z-linkage, deduced from both the FISH and the Southern blot data is consistent
with the position of FANCC and FANCG on human chromosome 9 which
exhibits a high degree of synteny conservation with the chicken Z chromosome
[42]. In contrast, the FANCB gene is X-linked in humans but autosomal in
chicken, mapping to GGA1. Avian and mammalian sex chromosomes evolved
independently from different ancestral autosome pairs [43, 44]. The localization
of two FANC genes on the chicken Z which other than the mammalian X may
not be subject to dosage compensation, is consistent with a higher gene dosage
in male (ZZ) than in female (ZW) birds.

Nanda/Buwe/Wizenman/Takata/Haaf/Schartl/Schmid 188
 600 Male
50 Wild Female
FANCG– 500
40
Total aberrations

Total aberrations
400
30
300
20
200

10 100

0 0
Control 20ng/ml 40ng/ml Control 150ng/ml 250ng/ml 350ng/ml
a MMC concentrations b MMC concentrations

Fig. 3. (a) Frequency of aberrations in normal DT40 (blue bars) and fancgDT40
knockout (purple bars) cells after treatment with the indicated MMC concentrations (for
details, see [23]). Note the increased level of MMC-induced DNA damage in fancg deficient
cells. (b) Number of chromosome aberrations in diploid chicken embryonic fibroblasts of
both sexes (male: blue bar and female: red bar) exposed to the indicated MMC concentra-
tions. Note the significantly higher chromosome aberrations in female. 200 metaphases were
screened for each experiment.

Lessons Learnt from Chicken DT40 Cells

One powerful approach to investigate FANC protein function is reverse

genetics. The chicken B lymphocyte line DT40 has been widely used in differ-
ent research areas such as receptor signaling, cell cycle regulation, gene con-
version and apoptosis, because it can be much more easily manipulated than
mammalian model systems. DT40 cells exhibit an exceptionally high integra-
tion rate of gene targeting constructs at homologous loci [45]. This high gene
targeting efficiency facilitates the generation of null mutants through deletion
of specific genomic regions. An added advantage compared to murine ES cells
is that DT40 cells maintain a stable karyotype and cellular phenotype over pro-
longed culture periods. In contrast to the mouse, DT40 knockouts for many
genes essential for recombinational types of DNA repair, including the Rad6
group of genes [46] are viable [47]. Another important but also critical aspect of
these cells is their lack of functional p53 which enables DT40 cells to prolifer-
ate rapidly, with very short doubling times.
The disruption of different FANC genes in DT40 cells renders them
sensitive to cross-linking agents and causes a high incidence of chromosome
aberrations after MMC treatment (fig. 3a). Thus, the cellular phenotype (MMC
sensitivity) of fancDT40 cells is comparable to that of human FA cells, which

Fanconi Anemia Genes in Vertebrates 189

makes them an excellent model for molecular analyses of the FA pathway. The
DNA repair defect in fancDT40 cells has been examined with an elegant HR
assay that measures repair of a genomic DSB induced at an I-SceI restriction
site, which then facilitates the expression of the reporter gene. So far five fanc
genes have been disrupted in chicken DT40 cells and all knockouts displayed
decreased HR activity [23, 41, 48, 49]. This is consistent with the homogeneity
of the HR defect in human FA. Although fancd2DT40 cells show a 40-fold
decrease in their HR activity compared to wild-type (which is much more pro-
nounced than in human FA cells), the basic cellular defect is likely to be similar
to that in human FA fibroblasts [21]. In mammals, cells deficient for bonafide
HR repair enzymes (e.g. XRCC3) exhibit much more severe DNA repair
defects than FANC mutants [49]. One plausible explanation is that human FANC
genes are not directly involved in HR but rather modulate HR efficiency.
Matushita et al. [50] generated a chicken fancd2-ubiquitin fusion gene
replacing the natural monoubiquitination site (D2KR-Ub). These cells reverse
cisplatin hypersensitivity, and the fusion protein localized to chromatin in
FANCD2-deficient DT40 cells. The observation that this fusion protein could
not complement FANCC, FANCG or FANCL knockouts suggests that apart from
FANCD2 monoubiquitination the core complex FANC proteins may have addi-
tional functions. Recent research using DT40 cells has contributed to the detec-
tion of a new down stream member of the FA pathway. FANCJ (BRIP1), which
is a DEAH helicase that interacts with the BRCT domain of BRCA1, has an
important function in BRCA1-dependent DNA repair and checkpoint control.
BRIP1-deficient DT40 cells display the same sensitivity to DNA-crosslinking
agents as fanccDT40 [39], but their cellular phenotype is clearly different from
BRCA1 knockouts. Their crosslink hypersensitivity is corrected by expression
of human BRIP1 lacking the C-terminal BRCT-interacting domain. BRIP1-
deficient DT40 cells therefore demonstrate that FANCJ (BRIP1) has a novel
function in the FA pathway which is independent of BRCA1.
Another very interesting property of FANC-deficient DT40 cells is their
increased rate of spontaneous sister chromatid exchanges (SCE) which is not
typical for human FA mutant cells. Because several vertebrate cell lines with
TLS repair defects including RAD18-deficient DT40 show increased SCE
levels [51], it is assumed that TLS (possibly including the BLM helicase) may
be impaired in fancd2DT40 cells. To elucidate the mechanism underlying ele-
vated SCE levels, Hirano et al. [41] deleted FANCC in DT40 cells lacking the
Rad51 paralog XRCC3 and TLS factor Rad18. The spontaneous SCE rate was
clearly decreased in xrcc3DT40 cells, but elevated in fanccDT40 cells. In
xrcc3/fancc double mutants the SCE rate was similar to that of xrcc3-deficient
cells. Thus, spontaneous SCE in fanccDT40 cells require XRCC3. The higher
SCE rate in fancc/rad18 double knockout DT40 cells compared to single

Nanda/Buwe/Wizenman/Takata/Haaf/Schartl/Schmid 190
mutants supports a role for FA proteins in facilitating HR but not global TLS
during crosslink repair. Taken together, FANC-deficient DT40 cells have clearly
improved our understanding of the molecular mechanisms in the FA pathway.

Sex-Specific Sensitivity to DNA Crosslinking Agents in Birds

FA cells are characterized by increased sensitivity to DNA crosslinking

agents such as MMC, DEB and others, but the molecular mechanism(s) under-
lying the various types of induced chromosome aberrations in FA in response to
crosslinking agents are not fully understood. MMC appears to exert its main
biochemical effects in the cytoplasm. Several lines of evidences suggest that
chromosome damage in FA cells is associated with ROS generated by these
crosslinking agents [52]. FA cells exhibit higher MMC sensitivity when grown
at 20% ambient oxygen than at 5% oxygen [53, 54], indicating that ROS gener-
ated by MMC are toxic. Vice versa, expression of the anti-oxidant protein
thioredoxin can decrease the MMC or DEB sensitivity of FA cells [55].
Additional evidence for a link between ROS and chromosome hypersensitivity
comes from the interaction of FANCC with cytoplasmic enzymes that are asso-
ciated with the production of ROS, i.e. NADPH cytochrome-P450 reductase
[56], glutathione transferase [57], and FANCG with CYP2E1 [58]. Apart from
their nuclear location, both FANCC and FANCG also appear to be present in
the cytoplasm [59]. Collectively, these results suggest that FA cells exhibit
increased susceptibility to oxidative stress induced by DNA crosslinking agents
and that both FANCC and FANCG somehow interfere with the generation of
such oxidative stress.
We have shown that both FANCC and FANCG are Z-linked in birds. Since
there is no evidence for chromosome wide inactivation of the additional Z chro-
mosome in the homogametic avian (ZZ) sex, males should express a higher
dosage for the two Z-linked FANC genes. We treated male and female chicken
cells with MMC to test whether FANCC and FANCG dosage influences the cel-
lular response to oxidative stress. Whereas FANCC- and FANCG-deficient
DT40 cells are known to be hypersensitive to MMC treatments [23, 41], it is
important to note that DT40 cells lack the TP53 tumor suppressor gene and are
trisomic for GGA2 which carries the c-myc oncogene [60]. Thus, MMC hyper-
sensitivity in DT40 cells may not fully reflect the situation in wild type avian
cells. In our experiments, exposure of male versus female chicken embryonic
fibroblasts to different MMC concentrations resulted in both chromosome and
chromatid breaks (fig. 4) reflecting unresolved DNA damage sustained during
the G1 and G2 phase of the cell cycle. As predicted, the frequency of chromatid
and chromosome aberrations was significantly higher in heterogametic (ZW)

Fanconi Anemia Genes in Vertebrates 191

 Fig. 4. Representative chicken metaphase spreads showing chromatid-type (G2 phase)
and chromosome-type (G1 phase) aberrations and occasional chromosome pulverizations
after exposure to the crosslinking agent MMC.

female cells than in homogametic (ZZ) male cells (fig. 3b). The increased sen-
sitivity of female cells is particularly evident when the cells are exposed to rel-
atively high MMC concentrations. At low MMC concentration the gender
difference was less pronounced, however the SCE frequency was markedly
increased in female compared to male chicken cells (data not shown).
As far as we can ascertain, the sex-specific cellular response to MMC
treatments appears to be unique for birds. Although the FANCB gene is
X-linked in mammals, this should not cause a sex-specific DNA damage response
because of X-inactivation in females. It is tempting to speculate that because of
their higher FANCC and FANCG gene dosage male birds may be better pro-
tected from oxidative damage through reducing ROS production. In most ratites
(flight-less birds) the Z and W sex chromosomes are still very similar in size
and gene content [61]. Thus, in contrast to modern birds female ratites most
likely will have two copies of FANCC and FANCG. It will therefore be interest-
ing to test whether sensitivity to MMC treatment and oxidative stress differs
between male and female cells in ratites as we here show for modern birds.

Function of Avian FANC Genes during Embryogenesis and

Gonadal Development

FA proteins are thought to interact in a common pathway for the repair of

stalled replication forks which are common events in proliferating cells.
Considering the high rate of cell proliferation in developing embryos, FANC
genes are likely to play a critical role during early embryogenesis. The congen-
ital defects in FA patients point towards impaired function(s) of FA genes

Nanda/Buwe/Wizenman/Takata/Haaf/Schartl/Schmid 192
 Fancc HH19 Fancc HH26 Fancg HH24

Fig. 5. Expression of FANCC and FANCG during chicken embryogenesis. Non-radioactive

whole mount in situ hybridization with anti-sense RNA was performed on chicken embryos of
HH stages 19, 24 and 26. Abbreviations used: Ba ⫽ Branchial arches, Di ⫽ Diencephalon,
Lb ⫽ Limb bud, Mes ⫽ Mesencephalon, Ot ⫽ Otic placode, Rh ⫽ Rhombencephalon
Sc ⫽ Spinal cord, So ⫽ Somites, Tel ⫽ Telencephalon, Wb ⫽ Wing bud.

during development. Mouse studies revealed that Fancc, Fancg and Fanca are
preferentially expressed in highly proliferating and embryonic tissues [62–64].
The importance of Fanc genes during early development was also demonstrated
in zebrafish embryos where Fancd2 prevented inappropriate apoptosis in neural
cells and other highly proliferating tissues [10].
In order to explore the activity of FANC genes during avian development,
we used whole mount RNA in situ hybridization on chicken embryos at differ-
ent Hamburger Hamilton (HH) stages to delineate the expression patterns of
chicken FANCC and FANCG genes. FANCC mRNA was first detected in
embryos at HH19, whereas FANCG expression started at HH24 (fig. 5). At
HH19 (fig. 5, left image) FANCC was highly expressed in the midline of the
entire dorsal neural tube (including brain and spinal cord), in the eye and otic
vesicle, at the tips of the branchial arches (white arrow), and in the mesenchyme
around the diencephalon. FANCC was also expressed in both limb buds at
HH19 and in the developing paws at later stages (fig. 5, middle image). In the
limb buds, FANCC mRNA was enriched in the zone of polarizing activity
(ZPA) and the apical epidermal ridge (AER). At HH26 and later stages FANCC
was still expressed in the dorsal spinal cord and in the somites (fig. 5, middle
image).
FANCG expression became detectable in HH24 embryos. Compared to
FANCC it was expressed in a broader region of the dorsal neural tube with
mRNA present in the entire dorsal midbrain and telencephalon (fig. 5, right
image). In the branchial arches FANCG was only weakly expressed at the poste-
rior edge of Ba1 and Ba3 (white arrows). In the limb buds FANCG mRNA cov-
ered the entire proximal-distal extent. In contrast to FANCC, FANCG was not

Fanconi Anemia Genes in Vertebrates 193

 lee M)
)
Ki y (M

)
lee )

ng )
Lu n (F
Sp y (F

Lu (M

st )
Br (M)

Te s (F
er )

Ki (F)

Sp n (
Liv (M

Br F)

s
(

is

y
e

e
ain

ain

ng
er

ar
dn

dn
Liv

Ov
5.1kb

2.0kb

Fig. 6. Northern blot with total RNAs from different adult male and female chicken
tissues hybridized with chicken FANCG cDNA. Note the prominent signals in the gonads.
The two different bands in testis tissue may be due to alternative splicing.

detectable in eye and otic vesicle. At HH30 and later embryonic stages both
FANCC and FANCG were expressed concomitantly in the urogenital region
(data not shown), possibly indicating a function of these genes during gonad
and kidney development. In contrast to our findings on chicken embryos,
FANCG expression was not detectable by Northern blot analysis in adult
somatic tissues. Only testis and ovary showed high FANCG mRNA levels on
Northern blots (fig. 6). RT-PCR revealed low FANCG transcript levels in adult
somatic tissues and high levels in gonads (data not shown).
The observed expression patterns support an important role for both
FANCC and FANCG during chicken development. Clearly, expression of these
genes is tightly regulated in a tissue- and developmental stage-specific manner.
The FANCC and FANCG expression patterns are partially overlapping, but not
identical. Abundant transcript levels in the developing eye, ear, spinal cord, and
brain suggest a function for central nervous system development. This is con-
sistent with the observation that in addition to other congenital malformations
many FA patients suffer from eye and ear defects. The expression of chicken
FANCC in limb and wing buds is consistent with the typical radial ray defects of
FA patients.

Nanda/Buwe/Wizenman/Takata/Haaf/Schartl/Schmid 194
 Strong expression of FANC genes in chicken and mouse gonads, the
impaired fertility of Fanc knockout mice, and the strongly reduced fertility of
FA patients [64, 65] argue in favor of a germ-cell specific function of FA pro-
teins. Interestingly, the function(s) of FANC genes and the FA pathway during
germ cell development and meiotic recombination appear to be highly con-
served throughout vertebrate evolution, whereas the effects of FANC gene defi-
ciency on somatic development appear to vary widely among different species.

Perspective

Orthologs for almost all mammalian FANC genes have been identified in
chicken. Because most FANC orthologs are not found in yeast, genetic analysis
in this convenient eukaryotic model organism cannot be performed. In this
light, the possibility to efficiently disrupt endogenous FANC genes in chicken
DT40 cells has been extremely useful for the investigation of individual FANC
genes, their specific functions, and the FA pathway.
In addition, birds have revealed interesting novel properties of FANC genes
that are not evident in other vertebrates. With the exception of FANCD1/ BRCA2,
monoallelic inactivation of FANC genes does not appear to cause serious health
problems in humans, whereas compound heterozygotes are phenotypically
affected. In birds we have the exceptional situation that the female (ZW) genome
contains only one copy of both FANCC and FANCG. Consequently, female
chicken cells exhibit a higher sensitivity to MMC treatments. Because male
birds have a higher level of the Z-linked FANCC and FANCG proteins, they are
likely to be more resistant to oxidative stress. This is consistent with observa-
tions that in many bird species males are longer-lived than females [66], and also
have higher survival rates [67]. Whether the survival advantage of male birds
results chiefly from their higher dosage of FANC caretaker genes or whether
additional protective factors are involved remains to be elucidated.
The high expression of FANC genes in avian gonads, and the fact that fertil-
ity is compromised in Fanc knockout mice and in FA patients argues for an essen-
tial function of the FANC family of genes in germ cell development, most likely
during recombination of homologous chromosomes during meiotic prophase I.
The DNA replication process in avian gonads must be particularly efficient to
ensure massive sperm production in male birds and ovary maturation in females.
High levels of FANC gene and protein expression may be instrumental in repair-
ing DSBs arising from DNA replication machinery slippage. However, FANC
proteins may not only be involved in DSB repair by promoting homologous
recombination, but may also contribute to the repair of base changes by transle-
sion synthesis [68]. The study of the Fanconi anemia family of genes in vertebrate

Fanconi Anemia Genes in Vertebrates 195

model systems, including fish, amphibian and avian cells, contributes greatly to
our understanding of the mechanisms of DNA maintenance without which the
remarkable longevity of our species would not be possible.

Acknowledgements

This work was supported by a grant from German Research Foundation (DFG; Schm
484/20-2).

References

1 Sasaki MS, Tonomura A: A high susceptibility of Fanconi’s anemia to chromosome breakage by

DNA cross-linking agents. Cancer Res 1973;33:1829–1836.
2 German J, Schonberg S, Caskie S, Warburton D, Falk C, Ray JH: A test for Fanconi’s anemia.
Blood 1987;69:1637–1641.
3 Auerbach AD: A test for Fanconi’s anemia. Blood 1988;72:366–367.
4 Grompe M, D’Andrea A: Fanconi anemia and DNA repair. Hum Mol Genet 2001;10:2253–2259.
5 Levitus M, Waisfisz Q, Godthelp BC, de Vries Y, Hussain S, et al: The DNA helicase BRIP1 is
defective in Fanconi anemia complementation group J. Nat Genet 2005;37:934–935.
6 Meetei AR, Medhurst AL, Ling C, Xue Y, Singh TR, et al: A human ortholog of archaeal DNA
repair protein Hef is defective in Fanconi anemia complementation group M. Nat Genet 2005;37:
958–963.
7 Reid S, Schindler D, Hanenberg H, Barker K, Hanks S, et al: Biallelic mutations in PALB2 cause
Fanconi anemia subtype FA-N and predispose to childhood cancer. Nat Genet 2007;39:162–164.
8 Timmers C, Taniguchi T, Hejna J, Reifsteck C, Lucas L, et al: Positional cloning of a novel Fanconi
anemia gene, FANCD2. Mol Cell 2001;7:241–248.
9 Dequen F, St-Laurent JF, Gagnon SN, Carreau M, Desnoyers S: The Caenorhabditis elegans
FancD2 ortholog is required for survival following DNA damage. Comp Biochem Physiol B
Biochem Mol Biol 2005;141:453–460.
10 Liu TX, Howlett NG, Deng M, Langenau DM, Hsu K, et al: Knockdown of zebrafish Fancd2
causes developmental abnormalities via p53-dependent apoptosis. Dev Cell 2003;5:903–914.
11 Herrero A, Barja G: H2O2 production of heart mitochondria and aging rate are slower in canaries
and parakeets than in mice: sites of free radical generation and mechanisms involved. Mech
Ageing Dev 1998;103:133–146.
12 Ogburn CE, Carlberg K, Ottinger MA, Holmes DJ, Martin GM, Austad SN: Exceptional cellular
resistance to oxidative damage in long-lived birds requires active gene expression. J Gerontol A
Biol Sci Med Sci 2001;56:B468–B474.
13 Wong AK, Pero R, Ormonde PA, Tavtigian SV, Bartel PL: RAD51 interacts with the evolutionar-
ily conserved BRC motifs in the human breast cancer susceptibility gene brca2. J Biol Chem
1997;272:31941–31944.
14 Tutt A, Bertwistle D, Valentine J, Gabriel A, Swift S, et al: Mutation in Brca2 stimulates error-
prone homology-directed repair of DNA double-strand breaks occurring between repeated
sequences. EMBO J 2001;20:4704–4716.
15 Yang H, Jeffrey PD, Miller J, Kinnucan E, Sun Y, et al: BRCA2 function in DNA binding and
recombination from a BRCA2-DSS1-ssDNA structure. Science 2002;297:1837–1848.
16 Zickler D, Kleckner N: Meiotic chromosomes: integrating structure and function. Annu Rev
Genet 1999;33:603–754.
17 Marcon E, Moens PB: The evolution of meiosis: recruitment and modification of somatic DNA-
repair proteins. Bioessays 2005;27:795–808.

Nanda/Buwe/Wizenman/Takata/Haaf/Schartl/Schmid 196
18 Mirchandani KD, D’Andrea AD: The Fanconi anemia/BRCA pathway: a coordinator of cross-link
repair. Exp Cell Res 2006;312:2647–2653.
19 Kitao H, Yamamoto K, Matsushita N, Ohzeki M, Ishiai M, Takata M: Functional interplay between
BRCA2/FancD1 and FancC in DNA repair. J Biol Chem 2006;281:21312–21320.
20 Thyagarajan B, Campbell C: Elevated homologous recombination activity in Fanconi anemia
fibroblasts. J Biol Chem 1997;72:23328–23333.
21 Nakanishi K, Yang YG, Pierce AJ, Taniguchi T, Digweed M, et al: Human Fanconi anemia
monoubiquitination pathway promotes homologous DNA repair. Proc Natl Acad Sci USA 2005;
102:1110–1115.
22 Sobeck A, Stone S, Costanzo V, de Graaf B, Reuter T, et al: Fanconi anemia proteins are required
to prevent accumulation of replication-associated DNA double-strand breaks. Mol Cell Biol
2006;26:425–437.
23 Yamamoto K, Ishiai M, Matsushita N, Arakawa H, Lamerdin JE, et al: Fanconi anemia FANCG
protein in mitigating radiation- and enzyme-induced DNA double-strand breaks by homologous
recombination in vertebrate cells. Mol Cell Biol 2003;23:5421–5430.
24 Levitus M, Joenje H, de Winter JP: The Fanconi anemia pathway of genomic maintenance. Cell
Oncol 2006;28:3–29.
25 Noll M, Battaile KP, Bateman R, Lax TP, Rathbun K, et al: Fanconi anemia group A and C double-
mutant mice: functional evidence for a multi-protein Fanconi anemia complex. Exp Hematol
2002;30:679–688.
26 Koomen M, Cheng NC, van de Vrugt HJ, Godthelp BC, van der Valk MA, et al: Reduced fertility
and hypersensitivity to mitomycin C characterize Fancg/Xrcc9 null mice. Hum Mol Genet 2002;11:
273–281.
27 Titus TA, Selvig DR, Qin B, Wilson C, Starks AM, et al: The Fanconi anemia gene network is con-
served from zebrafish to human. Gene 2006;371:211–223.
28 Wittbrodt J, Meyer A, Schartl M: More genes in fish? Bioessays 1998;20:511–515.
29 Taylor JS, Braasch I, Frickey T, Meyer A, Van de Peer Y: Genome duplication, a trait shared by
22000 species of ray-finned fish. Genome Res 2003;13:382–390.
30 Leveille F, Blom E, Medhurst AL, Bier P, Laghmani el H, et al: The Fanconi anemia gene product
FANCF is a flexible adaptor protein. J Biol Chem 2004;279:39421–39430.
31 Meetei AR, de Winter JP, Medhurst AL, Wallisch M, Waisfisz Q, et al: A novel ubiquitin ligase is
deficient in Fanconi anemia. Nat Genet 2003;35:165–170.
32 Collis SJ, Barber LJ, Ward JD, Martin JS, Boulton SJ: C. elegans FANCD2 responds to replication
stress and functions in interstrand cross-link repair. DNA Repair (Amst.) 2006;5:1398–1406.
33 Kojic M, Kostrub CF, Buchman AR, Holloman WK: BRCA2 homolog required for proficiency in
DNA repair, recombination, and genome stability in Ustilago maydis. Mol Cell 2002;10:683–691.
34 Mosedale G, Niedzwiedz W, Alpi A, Perrina F, Pereira-Leal JB, et al: The vertebrate Hef ortholog
is a component of the Fanconi anemia tumor-suppressor pathway. Nat Struct Mol Biol 2005;12:
763–771.
35 Meetei AR, Levitus M, Xue Y, Medhurst AL, Zwaan M, et al: X-linked inheritance of Fanconi ane-
mia complementation group B. Nat Genet 2004;36:1219–1224.
36 International chicken genome sequence consortium: Sequence and comparative analysis of the
chicken genome provide unique perspectives on vertebrate evolution. Nature 2004;432:695–716.
37 Takata M, Tachiiri S, Fujimori A, Thompson LH, Miki Y, et al: Conserved domains in the chicken
homologue of BRCA2. Oncogene 2002;21:1130–1134.
38 Warren M, Smith A, Partridge N, Masabanda J, Griffin D, Ashworth A: Structural analysis of the
chicken BRCA2 gene facilitates identification of functional domains and disease causing muta-
tions. Hum Mol Genet 2002;11:841–851.
39 Bridge WL, Vandenberg CJ, Franklin RJ, Hiom K: The BRIP1 helicase functions independently of
BRCA1 in the Fanconi anemia pathway for DNA crosslink repair. Nat Genet 2005;37:953–957.
40 Takata M, Kitao H, Ishia M: Fanconi anemia: genetic analysis of a human disease using chicken
system. Cytogenet Genome Res 117 (2007, in press).
41 Hirano S, Yamamoto K, Ishiai M, Yamazoe M, Seki M, et al: Functional relationships of FANCC
to homologous recombination, translesion synthesis, and BLM. EMBO J 2005;24:418–427.

Fanconi Anemia Genes in Vertebrates 197

42 Nanda I, Shan Z, Schartl M, Burt DW, Koehler M, et al: 300 million years of conserved synteny
between chicken Z and human chromosome 9. Nat Genet 1999;21:258–259.
43 Fridolfsson AK, Cheng H, Copeland NG, Jenkins NA, Liu HC, et al: Evolution of the avian sex
chromosomes from an ancestral pair of autosomes. Proc Natl Acad Sci USA 1998;95:8147–8152.
44 Marshall Graves JA, Shetty S: Sex from W to Z: evolution of vertebrate sex chromosomes and sex
determining genes. J Exp Zool 2001;290:449–462.
45 Buerstedde JM, Takeda S: Increased ratio of targeted to random integration after transfection of
chicken B cell lines. Cell 1991;67:179–188.
46 Nojima K, Hochegger H, Saberi A, Fukushima T, Kikuchi K, et al: Multiple repair pathways medi-
ate tolerance to chemotherapeutic cross-linking agents in vertebrate cells. Cancer Res 2005;65:
11704–11711.
47 Bemark M, Khamlichi AA, Davies SL, Neuberger MS: Disruption of mouse polymerase zeta
(Rev3) leads to embryonic lethality and impairs blastocyst development in vitro. Curr Biol 2000;
10:1213–1216.
48 Yamamoto K, Hirano S, Ishiai M, Morishima K, Kitao H, et al: Fanconi anemia protein FANCD2
promotes immunoglobulin gene conversion and DNA repair through a mechanism related to
homologous recombination. Mol Cell Biol 2005;25:34–43.
49 Niedzwiedz W, Mosedale G, Johnson M, Ong CY, Pace P, Patel KJ: The Fanconi anaemia gene
FANCC promotes homologous recombination and error-prone DNA repair. Mol Cell 2004;15:
607–620.
50 Matsushita N, Kitao H, Ishiai M, Nagashima N, Hirano S, et al: A FancD2-monoubiquitin fusion
reveals hidden functions of Fanconi anemia core complex in DNA repair. Mol Cell 2005;19:
841–847.
51 Yamashita YM, Okada T, Matsusaka T, Sonoda E, Zhao GY, et al: RAD18 and RAD54 coopera-
tively contribute to maintenance of genomic stability in vertebrate cells. EMBO J 2002;21:
5558–5566.
52 Ahmad SI, Hanaoka F, Kirk SH: Molecular biology of Fanconi anaemia – an old problem, a new
insight. Bioessays 2002;24:439–448.
53 Clarke AA, Philpott NJ, Gordon-Smith EC, Rutherford TR: The sensitivity of Fanconi anaemia
group C cells to apoptosis induced by mitomycin C is due to oxygen radical generation, not DNA
crosslinking. Br J Haematol 1997;96:240–247.
54 Liebetrau W, Runger TM, Mehling BE, Poot M, Hoehn H: Mutagenic activity of ambient oxygen
and mitomycin C in Fanconi’s anaemia cells. Mutagenesis 1997;12:69–77.
55 Ruppitsch W, Meisslitzer C, Hirsch-Kauffmann M, Schweiger M: Overexpression of thioredoxin
in Fanconi anemia fibroblasts prevents the cytotoxic and DNA damaging effect of mitomycin C
and diepoxybutane. FEBS Lett 1998;422:99–102.
56 Kruyt FA, Hoshino T, Liu JM, Joseph P, Jaiswal AK, Youssoufian H: Abnormal microsomal detox-
ification implicated in Fanconi anemia group C by interaction of the FAC protein with NADPH
cytochrome P450 reductase. Blood 1998;92:3050–3056.
57 Cumming RC, Lightfoot J, Beard K, Youssoufian H, O’Brien PJ, Buchwald M: Fanconi anemia
group C protein prevents apoptosis in hematopoietic cells through redox regulation of GSTP1. Nat
Med 2001;7:814–820.
58 Futaki M, Igarashi T, Watanabe S, Kajigaya S, Tatsuguchi A, et al: The FANCG Fanconi anemia
protein interacts with CYP2E1: possible role in protection against oxidative DNA damage.
Carcinogenesis 2002;23:67–72.
59 Thomashevski A, High AA, Drozd M, Hunt DF, Grant PA, Kupfer GM: The Fanconi anemia core
complex forms four complexes of different sizes in different subcellular compartments. J Biol
Chem 2004;279:26201–26209.
60 Sonoda E, Sasaki MS, Buerstedde JM, Bezzubova O, Shinohara A, et al: Rad51-deficient verte-
brate cells accumulate chromosomal breaks prior to cell death. EMBO J 1998;17:598–608.
61 Ogawa A, Murata K, Mizuno S: The location of Z- and W-linked marker genes and sequence on
the homomorphic sex chromosomes of the ostrich and the emu. Proc Natl Acad Sci USA 1998;95:
4415–4418.
62 Krasnoshtein F, Buchwald M: Developmental expression of the Fac gene correlates with congeni-
tal defects in Fanconi anemia patients. Hum Mol Genet 1996;5:85–93.

Nanda/Buwe/Wizenman/Takata/Haaf/Schartl/Schmid 198
63 Abu-Issa R, Eichele G, Youssoufian H: Expression of the Fanconi anemia group A gene (Fanca)
during mouse embryogenesis. Blood 1999;94:818–824.
64 Wong JC, Buchwald M: Disease model: Fanconi anemia. Trends Mol Med 2002;8:139–142.
65 Hölzel M, van Diest PJ, Bier P, Wallisch M, Hoatlin ME, et al: FANCD2 protein is expressed in
proliferating cells of human tissues that are cancer-prone in Fanconi anaemia. J Pathol 2003;201:
198–203.
66 Holmes DJ, Thomson SL, Wu J, Ottinger MA: Reproductive aging in female birds. Exp Gerontol
2003;38:751–756.
67 Payevsky VA, Vysotsky VG, Yefremov VD, Markovets MY, Shapoval AP: Sex-specific survival
rates in birds. Zh obshch Biol 1997;58:5–20.
68 Thompson LH, Hinz JM, Yamada NA, Jones NJ: How Fanconi anemia proteins promote the four
Rs: replication, recombination, repair, and recovery. Environ Mol Mutagen 2005;45:128–142.

Indrajit Nanda
Department of Human Genetics, University of Würzburg
Biozentrum, Am Hubland
D–97074 Würzburg (Germany)
Tel. ⫹49 931 888 4078, Fax ⫹49 931 888 4058
E-Mail [email protected]

Fanconi Anemia Genes in Vertebrates 199

 Schindler D, Hoehn H (eds): Fanconi Anemia. A Paradigmatic Disease for the Understanding of
Cancer and Aging. Monogr Hum Genet. Basel, Karger, 2007, vol 15, pp 200–210

Studying Homologous Recombination

in Fanconi Anemia
I. Demuth, M. Digweed
Institut für Humangenetik, Charité Universitätsmedizin Berlin,
Campus Virchow-Klinikum, Berlin, Germany

Abstract
Cells from patients suffering from the recessive syndrome Fanconi anemia (FA), are
characterized by increased sensitivity to agents that induce DNA interstrand cross-links
(ICLs). This hypersensitivity manifests as a dramatically elevated rate of chromosome
breaks in FA cells when compared to controls and led, more than thirty years ago, to the
suggestion that the repair of ICLs is disturbed in FA. Today, a DNA repair defect as the basis
of FA is widely accepted, however, the exact role of the 12 known genes is still elusive. The
past several years have brought growing evidence that FA cells are compromised in homol-
ogy dependent DNA repair processes. This review will summarize these studies with a
focus on integrated plasmid reporter assays which are used to investigate repair products
after induction of a single defined DNA double-strand break (DSB).
Copyright © 2007 S. Karger AG, Basel

Establishing a disturbance in repair of DNA damage as a possible

explanation for the basic defect in Fanconi anemia (FA) was the merit of two
papers: The discovery of an increased level of spontaneous chromosome aber-
rations in cells from FA patients published by Schroeder and colleagues in
1964 and, almost ten years later, Sasaki’s and Tonomura’s finding that the chro-
mosomes of FA patients were extremely and exclusively sensitive to agents
that produce interstrand cross-links (ICLs) in DNA [1, 2]. The authors of
the latter paper concluded: ‘. . ., although the mechanism is unidentified,
mammalian cells have an efficient repair mechanism to tolerate the DNA
cross-links, and the FA cells are genetically handicapped in functioning the
system’.
 The mechanisms of ICL processing within the mammalian cell are still
poorly understood but there is clear evidence that repair of these lesions utilizes
proteins from different DNA repair pathways [3]. Most importantly, it has been
shown that DNA double-strand breaks (DSBs) are generated as an intermediate
of ICL repair [4, 5] so that the repair of this particular DNA lesion plays a crit-
ical role in the cellular response to ICLs.
There are two principle mechanisms by which DSBs are processed
depending on the requirement for sequence homology. Non-homologous end-
joining (NHEJ), as the name suggests, rejoins free DNA ends with no, or only
limited, requirement for homology and is error prone (reviewed in [6]).
The repair mechanisms that involve sequence homology can be divided
into two types based on whether homologous associations arise from strand
exchange or strand annealing (fig. 1). The central protein of the pathway of
homology directed repair (HDR) is RAD51. This evolutionarily conserved pro-
tein, with orthologues in prokaryotes and archaea, initiates strand exchange
reactions between a broken sequence and its undamaged sister chromatid, or
homologue, to allow restoration of the damaged region. As many as five
RAD51 paralogues have been identified in mammals, all playing important
roles in HDR. In addition, the genes BRCA1 and BRCA2 showed reduced HDR
in experiments using integrated plasmid reporters (see below) indicating
involvement in this repair pathway also.
The second homology dependent pathway, single strand annealing (SSA),
uses homologies of the region flanking the DSB, e.g. repetitive sequences,
which are present at high frequency within the mammalian genome. DNA end
resection removes the sequence between the homologous elements and is fol-
lowed by annealing of 3⬘ overhangs and local DNA synthesis to restore an intact
DNA molecule.
During recent years, several important findings suggested a link between
FA and homology-driven repair mechanisms. Most of the known FA proteins
form the so called ‘FA core’ complex, which activates the FANCD2 protein
by adding an ubiquitin molecule in S-phase or in response to DNA damage.
Monoubiquitinated FANCD2-Ub then relocates to nuclear foci, where it
colocalizes with the HDR proteins RAD51 and BRCA1 [7]. The iden-
tification of BRCA2 as the gene mutated in FA patients of complementation
group D1 [8], the finding of direct interaction of BRCA2/FANCD1 with the
FA proteins FANCG and FANCD2 [9, 10] and the observation of attenuated
RAD51 foci formation in FA cells [11] provided additional evidence for a
connection between FA and HDR. These findings stimulated research on
homology-driven repair in FA and much of the recent work in this field took
advantage of integrated plasmid reporter assays originally introduced by
Maria Jasin.

Studying Homologous Recombination in Fanconi Anemia 201

 a b c a' d

Homology directed repair (HDR) Single strand annealing (SSA)

a b a' d a b a' d

a b c a' d
a b a' d

a b a' d

a d

a b c a' d

a b c a' d a d

Fig. 1. DNA double-strand break repair mechanisms involving sequence homology.

The graphic shows a highly schematic representation of two double-strand break repair path-
ways, homology directed repair (HDR) and single strand annealing (SSA). In HDR, a single
strand, originating at the double-strand break, invades a homologous double helix (here in
grey) and uses it as a template for DNA synthesis. In SSA, the DNA ends are processed back
to the repetitive sequences a and a⬘. Annealing of the single stranded, complementary DNAs
allows resynthesis of the missing DNA strands. Only in the HDR pathway is the original
c-sequence restored.

Integrated Plasmid Reporter Assays

The reporter plasmids typically used to evaluate DSB repair contain two dif-
ferently modified, non-functional reporter genes, one of these inactive because of
the insertion of a recognition site for the rare cutting I-SceI endonuclease and the
other consisting of an internal fragment of the gene. Repair of a DSB induced in
one copy of the gene, which exploits the second undamaged copy leads to the
restoration of its function. The most commonly used reporter gene is the green
fluorescent protein (GFP), but reporter plasmids with various drug resistance

Demuth/Digweed 202
genes have also been described. Expression of I-SceI in cells containing a copy of
the reporter introduces a single defined DSB. Subsequent analysis allows the effi-
ciency of repair of the DSB to be quantified. For example, when GFP is used as
the reporter, the HDR pathway restores the GFP function by recombination
between the two non-functional GFP gene copies. Thus the fraction of cells that
successfully employed HDR to repair the DSB can be directly assessed by count-
ing green fluorescent cells. In addition, amplification and analysis of the
sequence originally containing the I-SceI recognition site can allow further deter-
mination of the exact repair pathway used (e.g. HDR, SSA or NHEJ).
Analysis of chicken DT40 cells with targeted mutations in FA genes has
been shown to be a powerful strategy to investigate FA protein function and has
been used by many researchers to analyse DNA repair with plasmid reporters
(for references see table 1). Studies of DT40 cells with knocked-out FA genes,
fancg or fancd2, revealed a mild but clear reduction in HDR whereas DT40
cells mutated in fancj and fancm were not affected in this pathway (table 1).
Even the HDR analysis of the same fancc knock-out DT40 cell line yielded
conflicting results [12–14].
Mouse ES cells, homozygous for a Fanca null mutation or for ‘mild’ hypo-
morphic mutations in the Fancd1/Brca2 gene, displayed reduced HDR compara-
ble to that of HDR-compromised DT40 FA cells [15–17].
Interestingly, in addition to the HDR defect, we found Fanca⫺/⫺ mouse
cells, as well as FA patient derived cells from complementation groups A, G
and D2, to be deficient in a further DSB repair pathway, SSA, a finding also
reported for Brca2 mutated cells [17–19].
The relatively mild reduction in HDR capacity of human FA-A and FA-G
cells stands in strong contrast to the more than 100-fold decrease in HDR of
FANCD1/BRCA2-deficient CAPAN-1 tumour cells [15, 18]. In addition to
FA-A and FA-G cells, we also found reduced HDR in FA-D2 cells, however in
another report the same FA-D2 cell line (PD20) was not deficient in this path-
way [18, 19]. MCF7 cells, depleted for FANCJ by siRNA mediated knock-
down, showed a 7–10-fold reduction in HDR which is contradictory to the data
obtained from the analysis of DT40 fancj cells [14, 20].

RAD51 Foci Formation

In response to ionizing radiation (IR) or treatment with other DNA damag-

ing agents, RAD51 accumulates at the sites of DSBs, and this is visible, by
immunofluorescence staining, as discrete nuclear foci. Attenuated RAD51 foci
formation has been reported for a number of cell lines mutated in genes involved
in HDR [16, 21–24]. Consequently, cell lines with knock-out mutations in FA

Studying Homologous Recombination in Fanconi Anemia 203

 Table 1. Homology directed repair (HDR) in Fanconi anemia

FA gene HDR (plasmid Gene targetinga Immunoglobulin Sister chromatid Reference

disrupted reporter assay) gene conversiona exchangesa

Chicken DT40
Demuth/Digweed

fancg 9-fold reduced reduced n.d. normal (sp./MMC) [41]

fancd2 reduced reduced deficient sp. elevated; MMC, normal [42]
fancj (bach1/brip1) normal n.d. n.d. sp. elevated [14]
fanccb 3-fold reduced reduced n.d. MMC/Cisplatin, [12]
defective induction
fanccb normal n.d. n.d. sp. elevated [14]
fanccb 3-fold reduced n.d. deficient sp. elevated; MMC, [13]
no further induction
fancm (hef1) normal n.d. proficient sp. elevated; MMC, [13]
no further induction
Mouse
Fancd1 (Brca2) 5–6-fold reduced reduced n.d. [15]
Fancd1 (Brca2) reduced n.d. reduced (sp./MMC) [16]
Fanca reduced n.d. n.d. [18]
Fanca 3-fold reduced reduced n.d. [17]

Human
FANCA reduced n.d. [18]
FANCG reduced n.d. [18]
FANCD1 (BRCA2)c ⬎100-fold reduced n.d. [15]
FANCD2d reduced n.d. [18]
FANCD2d normal n.d. [19]
FANCJ (BACH1/BRIP1)e 7–10-fold reduced n.d. [20]

a
MMC: Mitomycin C; n.d.: not determined; sp.: spontaneous.
b
Identical cell line.
c
Tumour cell line.
204

d
Identical cell line.
e
siRNA mediated depletion.
 Table 2. RAD51 foci formation in Fanconi anemia cells

FA gene(s) disrupted RAD51 foci induction (treatment, RAD51 foci formation Reference
time between treatment and
analysis)a

Human
FANC A/B/C/D1/D2/E/F/G IR (12 Gy, 8 h) reduced in all groups [11]
FANC A/B/C/D1/D2/E/F/G IR (12 Gy, 8 h), reduced in FANCD1 cells [25]
MMC (1 h/2.4 ␮g/ml, 24 h) reduced in FANCD1 cells
FANC A/G/C IR (5 Gy, 8 h) slightly reduced [43]
MMC (1 h/0.1 ␮g/ml, time course) delayed
FANCA IR (8 Gy, 2 h) reduced [44]
FANCD2 IR (12 Gy) normal [19]
MMC (40 ng/ml) normal
FANC A/D1/F/I/J/L IR (12 Gy, 7 h/24 h) reduced in FANCD1 cells [26]
MMC (1 h/2.4 ␮g/ml, 7 h/24 h) reduced in FANCD1 cells
FANCJ (BACH1/BRIP1) MMC normal [20]
HU (1 mM, 18 h) normal
FANCD2 IR (2 and 15 Gy, 8 h) reduced [45]
Rodent
CHO-FANCG IR (8 Gy, 4 h) normal [46]
CHO-BRCA2 IR (12 Gy, 8 h) reduced [47]
MMC (1 h/2.4 ␮g/ml, 8 h) reduced
Brca2 IR (10 Gy, 5 h) reduced [16]
Fanca MMC (1 h/0.1 ␮g/ml, 9 h) reduced [17]
DT40
fancd2 IR (8 Gy) normal [40]
MMC (1 h/0.5 ␮g/ml, time course) normal
fancg IR (8 Gy) normal [41]
MMC (1 h/0.5 ␮g/ml, time course) normal

a
HU: Hydroxyurea; IR: ionizing radiation; MMC: mitomycin C.

genes and FA patient cell lines have been analyzed for the integrity of RAD51
foci formation in response to IR or mitomycin C (MMC) (table 2).
As shown in figure 2, we found IR induced RAD51 foci formation was
attenuated in FA cells of complementation groups FA-A, B, C, D1, D2, E, F and
G [11]. This has been confirmed by several studies (table 2). However, the
analysis of FA cells of the same complementation groups and cells of the newer
complementation groups, FA-I, FA-J and FA-L, revealed a specific defect in
RAD51 foci formation only in FA-D1 cells [25, 26]. Other reports in this con-
text on single cell lines are also somewhat contradictory (table 2).

Studying Homologous Recombination in Fanconi Anemia 205

 a
40
Control FAD1 FAD2 FAA, FAB, FAC
FAE, FAF, FAG

30
Foci-positive cells (%)

20

10

0
b Gy 0 12 0 12 0 12 0 12

Fig. 2. a RAD51 nuclear foci observed in irradiated control cells. Primary fibroblasts
were irradiated with 12 Gy and were fixed 8 h later, permeabilized and incubated with a pri-
mary rabbit antibody directed towards RAD51 followed by detection with a secondary Cy3-
conjugated goat anti-rabbit Ig. Cells were counterstained with DAPI and examined
microscopically using the appropriate filters. Representative digital images are shown.
b Quantification of RAD51 foci in control and FA fibroblasts. The levels of RAD51 foci-
positive cells are shown for control fibroblasts and fibroblasts from various FA complemen-
tation groups. Open columns represent unirradiated cells, filled columns are cells irradiated
with 12 Gy and processed for immunofluorescence 8 h later.

Other Measures of HDR

The frequency with which plasmid-carried, homologous sequences are

integrated into the mammalian genome, gene targeting efficiency, is a further
sensitive method to analyze HDR [27]. Cell lines with mutations in FA genes

Demuth/Digweed 206
tested for this feature were concordant in targeting efficiency and HDR of a
plasmid reporter (table 1).
The previously mentioned chicken DT40 cell line is derived from B lym-
phocytes and continuously undergoes immunoglobulin gene conversion in cul-
ture, a feature that can also be used to assess HDR at the sequence level. In
those cases analyzed, results were in agreement with data from plasmid reporter
assays, fancc and fancd2 cells being deficient, and fancm cells showing profi-
ciency, in immunoglobulin conversion (table 1).
Symmetrical exchanges between the two sister chromatids (sister chro-
matid exchanges, SCEs) can be visualized cytogenetically after DNA labeling
with 5-bromodeoxyuridine (BrdU) during the preceding S-phase. SCEs occur
spontaneously but their frequency can be enhanced by treatment with various
DNA damaging agents. The exact mechanism by which SCEs are formed is
not known, however, the interpretation of SCEs as the result of homologous
recombination processes is widely accepted. Indeed, cells lacking key homol-
ogous recombination genes exhibit significantly reduced spontaneous and
DNA-damage induced SCE levels [28]. Although spontaneous and induced
SCE levels have been investigated in FA cells for more than thirty years, the
reported studies are rather inconsistent (e.g. [29–31], see also table 1).
However, cells from Bloom syndrome patients, deficient in a protein, BLM,
which associates with the FA core complex [32], do show consistently exces-
sive SCEs [33].

Concluding Remarks

Recent results have implicated at least a subset of FA proteins in HDR

processes. FA proteins interact with HDR proteins in multiple direct and indi-
rect ways: FANCD2 and FANCG interact physically with BRCA2/FANCD1, as
shown in Yeast-Two-Hybrid analysis and in co-immunoprecipitations [9, 10].
Colocalizations in nuclear foci have been found in response to DNA damage for
FANCD2, RAD51 and BRCA1 as well as for FANCG, BRCA2 and RAD51 [7,
9]. FANCD1/BRCA2 has been consistently reported to be essential for RAD51
foci formation [11, 25, 34]. The reason for contradictory reports on disturbed
RAD51 foci formation in FA cells of complementation groups other than
FA-D1 is not clear (table 2) but may possibly be related to cell cycle effects, or
perhaps just reflect differences in foci formation kinetics.
The majority of studies on HDR based on the analysis of plasmid reporters
after induction of a single defined DSB and tracking its repair, show reduced
HDR, as summarized in table 1. In fact, for all analyzed FA genes, except
fancm, there is at least one study reporting a deficiency in HDR. In two cases,

Studying Homologous Recombination in Fanconi Anemia 207

the analysis of the same cell lines (PD20, mutated in FANCD2 and DT40,
mutated in fancc) yielded contradictory results in different laboratories.
The observed deficiencies in HDR are relatively mild, especially when
compared with the severe recombination defects seen in mutants of BRCA1,
BRCA2, RAD51 and the Rad51 paralogues, XRCC2/XRCC3 [35–39]. This is
in agreement with the viability of FA patients, and of mice mutated in FA genes,
and suggests that the FA genes, except for FANCD1/BRCA2, are likely to be
non-essential components of HDR.
In addition to recombination repair proteins, proteins from the excision
repair pathway, FA proteins and components involved in translesion synthesis
(TLS) have all been implicated in the repair of ICLs and there is growing evi-
dence that FA proteins promote TLS [40]. Similarly, SSA is yet another path-
way seemingly promoted by FA proteins [17, 18]. These recent discoveries
suggest that FA proteins may be involved in early processing of DNA lesions or
in stabilization of DNA repair intermediates before channeling them into an
appropriate repair pathway. Future research will aim to define these FA protein
functions more precisely by exploiting the powerful battery of experimental
approaches already available.

References

1 Schroeder TM, Anschütz F, Knopp A: Spontane Chromosomenaberrationen bei familiärer

Panmyelopathie. Humangenetik 1964;1:194–196.
2 Sasaki MS, Tonomura A: A high susceptibility of Fanconi’s anemia to chromosome breakage by
DNA cross-linking agents. Cancer Res 1973;33:1829–1836.
3 McHugh PJ, Spanswick VJ, Hartley JA: Repair of DNA interstrand crosslinks: molecular mech-
anisms and clinical relevance. Lancet Oncol 2001;2:483–490.
4 De Silva IU, McHugh PJ, Clingen PH, Hartley JA: Defining the roles of nucleotide excision repair
and recombination in the repair of DNA interstrand cross-links in mammalian cells. Mol Cell Biol
2000;20:7980–7990.
5 Rothfuss A, Grompe M: Evidence for a DNA double-strand break-dependent activation of the
Fanconi anemia/BRCA pathway during the repair of DNA-interstrand crosslinks. Am J Hum
Genet 2004;73:575.
6 Weterings E, van Gent DC: The mechanism of non-homologous end-joining: a synopsis of synap-
sis. DNA Repair 2004;3:1425–1435.
7 Garcia-Higuera I, Taniguchi T, Ganesan S, Meyn MS, Timmers C, et al: Interaction of the Fanconi
anemia proteins and BRCA1 in a common pathway. Mol Cell 2001;7:249–262.
8 Howlett NG, Taniguchi T, Olson S, Cox B, Waisfisz Q, et al: Biallelic inactivation of BRCA2 in
Fanconi anemia. Science 2002;297:606–609.
9 Hussain S, Witt E, Huber PA, Medhurst AL, Ashworth A, Mathew CG: Direct interaction of the
Fanconi anaemia protein FANCG with BRCA2/FANCD1. Hum Mol Genet 2003;12:2503–2510.
10 Hussain S, Wilson JB, Medhurst AL, Hejna J, Witt E, et al: Direct interaction of FANCD2 with
BRCA2 in DNA damage response pathways. Hum Mol Genet 2004;13:1241–1248.
11 Digweed M, Rothe S, Demuth I, Scholz R, Schindler D, et al: Attenuation of the formation of
DNA-repair foci containing RAD51 in Fanconi anaemia. Carcinogenesis 2002;23:1121–1126.
12 Niedzwiedz W, Mosedale G, Johnson M, Ong CY, Pace P, Patel KJ: The Fanconi anaemia gene FANCC
promotes homologous recombination and error-prone DNA repair. Mol Cell 2004;15: 607–620.

Demuth/Digweed 208
13 Mosedale G, Niedzwiedz W, Alpi A, Perrina F, Pereira-Leal JB, et al: The vertebrate Hef ortholog is a
component of the Fanconi anemia tumor-suppressor pathway. Nat Struct Mol Biol 2005;2:763–771.
14 Bridge WL, Vandenberg CJ, Franklin RJ, Hiom K: The BRIP1 helicase functions independently of
BRCA1 in the Fanconi anemia pathway for DNA crosslink repair. Nat Genet 2005;37:953–957.
15 Moynahan ME, Pierce AJ, Jasin M: BRCA2 is required for homology-directed repair of chromo-
somal breaks. Mol Cell 2001;7:263–272.
16 Tutt A, Bertwistle D, Valentine J, Gabriel A, Swift S, et al: Mutation in Brca2 stimulates error-
prone homology-directed repair of DNA double-strand breaks occurring between repeated
sequences. EMBO J 2001;20:4704–4716.
17 Yang YG, Herceg Z, Nakanishi K, Demuth I, Piccoli C, et al: The Fanconi anemia group A protein
modulates homologous repair of DNA double-strand breaks in mammalian cells. Carcinogenesis
2005;26:1731–1740.
18 Nakanishi K, Yang YG, Pierce AJ, Taniguchi T, Digweed M, et al: Human Fanconi anemia
monoubiquitination pathway promotes homologous DNA repair. Proc Natl Acad Sci USA 2005;
102:1110–1115.
19 Ohashi A, Zdzienicka MZ, Chen J, Couch FJ: Fanconi anemia complementation group D2
(FANCD2) functions independently of BRCA2- and RAD51-associated homologous recombin-
ation in response to DNA damage. J Biol Chem 2005;280:14877–14883.
20 Litman R, Peng M, Jin Z, Zhang F, Zhang J, et al: BACH1 is critical for homologous recombin-
ation and appears to be the Fanconi anemia gene product FANCJ. Cancer Cell 2005;8:255–265.
21 Bishop DK, Ear U, Bhattacharyya A, Calderone C, Beckett M, et al: Xrcc3 is required for assem-
bly of Rad51 complexes in vivo. J Biol Chem 1998;273:21482–21488.
22 Takata M, Sasaki MS, Sonoda E, Fukushima T, Morrison C, et al: The Rad51 paralog Rad51B pro-
motes homologous recombinational repair. Mol Cell Biol 2000;20:6476–6482.
23 O’Regan P, Wilson C, Townsend S, Thacker J: XRCC2 is a nuclear RAD51-like protein required
for damage-dependent RAD51 focus formation without the need for ATP binding. J Biol Chem
2001;276:22148–22153.
24 Bhattacharyya A, Ear US, Koller BH, Weichselbaum RR, Bishop DK: The breast cancer suscepti-
bility gene BRCA1 is required for subnuclear assembly of Rad51 and survival following treatment
with the DNA cross-linking agent cisplatin. J Biol Chem 2000;275:23899–23903.
25 Godthelp BC, Artwert F, Joenje H, Zdzienicka MZ: Impaired DNA damage-induced nuclear Rad51
foci formation uniquely characterizes Fanconi anemia group D1. Oncogene 2002;21:5002–5005.
26 Godthelp BC, Wiegant WW, Waisfisz Q, Medhurst AL, Arwert F, et al: Inducibility of nuclear
Rad51 foci after DNA damage distinguishes all Fanconi anemia complementation groups from
D1/BRCA2. Mutat Res 2006;594:39–48.
27 Yamazoe M, Sonoda E, Hochegger H, Takeda S: Reverse genetic studies of the DNA damage
response in the chicken B lymphocyte line DT40. DNA Repair 2004;3:1175–1185.
28 Sonoda E, Sasaki MS, Morrison C, Yamaguchi-Iwai Y, Takata M, Takeda S: Sister chromatid
exchanges are mediated by homologous recombination in vertebrate cells. Mol Cell Biol 1999;19:
5166–5169.
29 Sperling K, Wegner RD, Riehm H, Obe G: Frequency and distribution of sister-chromatid
exchanges in a case of Fanconi’s anemia. Humangenetik 1975;27:227–230.
30 Fujiwara Y, Kano Y, Yamamoto Y: DNA interstrand cross-linking, repair, and SCE mechanism in
human cells in special reference to Fanconi anemia. Basic Life Sci 1984;29(pt B):787–800.
31 Wilson JB, Johnson MA, Stuckert AP, Trueman KL, May S, et al: The Chinese hamster FANCG/
XRCC9 mutant NM3 fails to express the monoubiquitinated form of the FANCD2 protein, is
hypersensitive to a range of DNA damaging agents and exhibits a normal level of spontaneous sis-
ter chromatid exchange. Carcinogenesis 2001;22:1939–1946.
32 Meetei AR, Sechi S, Wallisch M, Yang D, Young MK, et al: A multiprotein nuclear complex con-
nects Fanconi anemia and Bloom syndrome. Mol Cell Biol 2003;23:3417–3426.
33 Chaganti RS, Schonberg S, German J: A manyfold increase in sister chromatid exchanges in
Bloom’s syndrome lymphocytes. Proc Natl Acad Sci USA 1974;71:4508–4512.
34 Yuan SS, Lee SY, Chen G, Song M, Tomlinson GE, Lee EY: BRCA2 is required for ionizing
radiation-induced assembly of Rad51 complex in vivo. Cancer Res 1999;59:3547–3551.

Studying Homologous Recombination in Fanconi Anemia 209

35 Westermark UK, Reyngold M, Olshen AB, Baer R, Jasin M, Moynahan ME: BARD1 participates
with BRCA1 in homology-directed repair of chromosome breaks. Mol Cell Biol 2003;23:
7926–7936.
36 Moynahan ME, Cui TY, Jasin M: Homology-directed DNA repair, mitomycin-C resistance, and
chromosome stability is restored with correction of a Brca1 mutation. Cancer Res 2001;61:
4842–4850.
37 Stark JM, Pierce AJ, Oh J, Pastink A, Jasin M: Genetic steps of mammalian homologous repair
with distinct mutagenic consequences. Mol Cell Biol 2004;24:9305–9316.
38 Johnson RD, Liu N, Jasin M: Mammalian XRCC2 promotes the repair of DNA double-strand
breaks by homologous recombination. Nature 1999;401:397–399.
39 Pierce AJ, Johnson RD, Thompson LH, Jasin M: XRCC3 promotes homology-directed repair of
DNA damage in mammalian cells. Genes Dev 1999;13:2633–2638.
40 Thompson LH, Hinz JM, Yamada NA, Jones NJ: How Fanconi anemia proteins promote the four
Rs: replication, recombination, repair, and recovery. Environ Mol Mutagen 2005;45:128–142.
41 Yamamoto K, Ishiai M, Matsushita N, Arakawa H, Lamerdin JE, et al: Fanconi anemia FANCG
protein in mitigating radiation- and enzyme-induced DNA double-strand breaks by homologous
recombination in vertebrate cells. Mol Cell Biol 2003;23:5421–5430.
42 Yamamoto K, Hirano S, Ishiai M, Morishima K, Kitao H, et al: Fanconi anemia protein FANCD2
promotes immunoglobulin gene conversion and DNA repair through a mechanism related to
homologous recombination. Mol Cell Biol 2005;25:34–43.
43 Pichierri P, Averbeck D, Rosselli F: DNA cross-link-dependent RAD50/MRE11/NBS1 subnuclear
assembly requires the Fanconi anemia C protein. Hum Mol Genet 2002;11:2531–2546.
44 Djuzenova C, Flentje M, Plowman PN: Radiation response in vitro of fibroblasts from a Fanconi
anemia patient with marked clinical radiosensitivity. Strahlenther Onkol 2004;180:789–797.
45 Wang X, Andreassen PR, D’Andrea AD: Functional interaction of monoubiquitinated FANCD2
and BRCA2/FANCD1 in chromatin. Mol Cell Biol 2004;24:5850–5862.
46 Tebbs RS, Hinz JM, Yamada NA, Wilson JB, Salazar EP, et al: New insights into the Fanconi ane-
mia pathway from an isogenic FancG hamster CHO mutant. DNA Repair 2005;4:11–22.
47 Kraakman-van der Zwet M, Overkamp WJ, van Lange RE, Essers J, van Duijn-Goedhart A, et al:
Brca2 (XRCC11) deficiency results in radioresistant DNA synthesis and a higher frequency of
spontaneous deletions. Mol Cell Biol 2002;22:669–679.

Ilja Demuth
Institut für Humangenetik, Charité Universitätsmedizin Berlin
Campus Virchow-Klinikum
Augustenburger Platz 1
13353 Berlin (Germany)
Tel. ⫹49 30 450 566306, Fax ⫹49 30 450 566904, E-Mail [email protected]

Demuth/Digweed 210
 Schindler D, Hoehn H (eds): Fanconi Anemia. A Paradigmatic Disease for the Understanding of
Cancer and Aging. Monogr Hum Genet. Basel, Karger, 2007, vol 15, pp 211–225

Functional Knock-Down of Human RAD51

for Testing the Fanconi Anemia-BRCA
Connection
P. Rioa,b, H. Hanenberga,c
a
Department of Pediatric Hematology, Oncology and Clinical Immunology, Children’s
Hospital, Heinrich-Heine-University, Düsseldorf, Germany; bHematopoiesis and Gene
Therapy Division, CIEMAT, Madrid, Spain; cDepartment of Pediatrics, Herman B
Wells Center for Pediatric Research, Indiana University School of Medicine,
Indianapolis, Ind., USA

Abstract
One of the proteins essential in homologous recombination is RAD51, a recombinase
involved in nucleoprotein filament formation. After DNA damage, RAD51 colocalizes in
nuclear foci with other proteins involved in DNA repair. This foci formation is regulated by
BRCA2/FANCD1. As Rad51 deficiency is lethal in mice and a human disorder due to
defects in RAD51 does not exist, the aim of the current study was to develop a system to
knock-down human RAD51 in an inducible form. For this purpose, we took advantage of the
regulatable shRNA delivery system developed by Wiznerovicz and Trono [1] based on the
cotransduction of target cells with two lentiviral vectors. The first vector carries expression
cassettes for the shRNA and for a marker gene. The second vector constitutively expresses
the transcription suppressing protein tTRKRAB whose activity can be regulated via doxycy-
cline (DOX). Here, we show the ability of lentiviral vectors expressing shRNAs designed
against human RAD51 to downregulate the protein expression in primary human fibroblasts
and in HeLa cells. This effect on the protein levels of RAD51 was controllable by DOX and
allowed to visualize RAD51 foci formation in cells via a gammaretroviral vector expressing
RAD51 fused to the enhanced green fluorescent protein (EGFP). This inducible system for
visualization of fluorescence-tagged essential cellular proteins might facilitate to study the
role of RAD51 and its interaction with members of the FA/BRCA pathway in response to
defined DNA damages.
Copyright © 2007 S. Karger AG, Basel

As described in the preceding chapter by Demuth and Digweed, there is

increasing evidence for the involvement of at least some of the FA-genes in
recombinational types of DNA repair. A central player in this type of repair is the
BRCA2 gene which recruits the RAD51 recombinase to the site of DNA damage.
Formal proof for a direct connection between the FA and BRCA pathways was
the identification of BRCA2 as the gene mutated in patients belonging to the clin-
ically most severe form of FA, the FA-D1 complementation group [2]. The recent
discoveries of the BRCA1 interacting protein, BRIP1, as FANCJ [3, 4] and of the
partner and localizer of BRCA2, PALB2, as FANCN [5] further strengthens the
notion that FA genes acting downstream of FANCD2 are involved in homology
directed types of DNA repair [6]. Although several interactions of upstream FA
proteins with BRCA1 and BRCA2 have also been described [7, 8], the exact role
of the FA pathway in the repair of DNA lesions induced by crosslinking agents
such as mitomycin C (MMC) still remains elusive.
One of the proteins essential for the execution of DNA repair via homolo-
gous recombination is RAD51. RAD51 is a recombinase involved in the forma-
tion of the nucleoprotein filaments on single-stranded (ss) DNA mediating
homologous pairing and strand exchange reactions between ssDNA and homolo-
gous double-stranded (ds) DNA [9, 6]. After DNA damage, RAD51 localizes in
repair foci in the nucleus of cells together with other proteins essential for recom-
binational DNA repair [10]. In addition to BRCA1 and BRCA2/FANCD1, this
also includes RPA which in close cooperation with RAD51 facilitates homolo-
gous pairing and DNA strand exchange [11]. The presence of RAD51 in these
foci depends on BRCA2/FANCD1. The BRC repeats in the exon 11 of BRCA2
directly interact with RAD51 and control the correct RAD51 oligomerization and
filament formation [12, 13]. At least in murine cells, BRCA2 can also interact
with RAD51 via a C-terminal motif [14]. In patients belonging to complementa-
tion group FA-D1, who have been shown to carry biallelic mutations in BRCA2,
nuclear RAD51 foci are almost completely absent and not inducible by irradia-
tion or DNA crosslinking agents [15]. In contrast to its counterpart RecA in
E. coli (31) and scRad51 in yeast [16, 17], the mammalian RAD51 protein
appears to be essential for cell survival, as targeted disruption of Rad51 induces
early embryonic lethality in mice [18]. In addition, there is neither a RAD51 gene
defect known in humans nor does a human cell line deficient in RAD51 exist.
Therefore, the aim of the current study was to develop a tool to conditionally
knock down RAD51 levels in FA and normal primary cells, thereby facilitating
functional studies on the role of RAD51 in the FA/BRCA pathway.
RNA interference (RNAi)-mediated gene silencing has emerged as a pow-
erful approach to regulate levels of endogenous proteins and to analyze gene
functions in many organisms [19]. In mammalian cells, RNAi-mediated gene
silencing can be obtained by the delivery of chemically synthesized short
(⬍30 nt) double-stranded siRNA molecules [20]. Alternatively, plasmid and
viral vector-based systems have been developed for stable expression of short

Rio/Hanenberg 212
hairpin RNAs (shRNAs) which then are processed to siRNA in the target cells
[21]. In some instances, where the knock-out/down of a target gene is lethal for
the cell, such as RAD51, alternative approaches for the controlled suppression
of protein levels are needed. In the present study, we took advantage of the gene
targeting system developed by Wiznerovicz and Trono [1] using lentiviral vec-
tors for the production of siRNAs in a drug inducible system. In the targeting
vector, the shRNA is expressed under the control of human H1 promoter
located in the 3⬘ SIN LTR. Also included in the 3⬘ LTR are tet operator (tetO)
sequences which directly bind the tetracycline repressor (tTR). Via a second
vector, the KRAB domain of human KOX1 fused to tTR is constitutively
expressed in the target cell. In the absence of doxycycline (DOX), this
tTRKRAB protein will bind to the tetO element and suppress any polymerase II
and III promoters within 3 kb in an orientation independent manner. In the pres-
ence of DOX, tTRKRAB is sequestered away from tetO thereby allowing gene
expression from both promoters, the H1 promoter driving the shRNA and the
internal EF1␣ promoter driving EGFP as a marker gene for the easy detection
of transduced cells and promotor activities. Using this targeting strategy, we
established an inducible RAD51 knock-down system that can be applied to the
study of homologous recombination in human primary cells and cell lines,
including cells derived from FA patients.

Materials and Methods

siRNA Transfection
Two different siRNAs against RAD51 were used (Invitrogen, Karlsruhe, Germany). A
fluorescent oligo labeled with FITC was used as a control (Invitrogen). HeLa cells were
transfected using lipofectamin 2000 following manufacturer’s instructions (Qiagen, Hilden,
Germany). Transfection efficiency was assessed by flow cytometry analysis of cells trans-
fected with the fluorescent oligo.

Vector Construction
shRNAs against RAD51 were ordered as phosphorylated oligos with ClaI and MluI
overhangs, subsequently annealed and then ligated into the LV-THM vector cut in the same
way using standard protocols.
S11EGRAIN was constructed by introducing the EGFP cDNA into the retroviral vector
S11RAIN (Velleuer and Hanenberg, unpublished) expressing human RAD51 linked via an
EMCV IRES site to the neomycin phosphotransferase (NEO) gene under the control of the
SFFV promotor. The presence of the NEO gene in the vector allowed us to select transduced
cells using G418 (0.5 mg/ml, GIBCO/Invitrogen).

Supernatant Production
Lentiviral Vectors: HEK293T cells were cultivated in Dulbecco’s Modified Eagle’s
Medium (DMEM) supplemented with 10% fetal calf serum, 2 mM L-glutamine, penicillin

Functional Knock-Down of RAD51 and Fanconi Anemia 213

G and streptomycin (all from GIBCO/Invitrogen). Subconfluent HEK293T cells were cotrans-
fected with 10 ␮g of the plasmid lentiviral vector, 10 ␮g of pCMV-DR8.91 and 10 ␮g of
pMD2G-VSV-G using Fugene 6 (Roche Diagnostics, Mannheim, Germany). After 24 h,
medium was changed and recombinant lentivirus containing supernatants were harvested 24 h
later, filtered to 0.45 ␮m and then used freshly or stored at ⫺80⬚C.
Retroviral Vectors: Ecotropic Phoenix (eFNX) cells were cultured under the same con-
ditions as 293T cells. After seeding, eFNX cells were transfected with 20 ␮g of S11EGRAIN
or S11EGIN using Fugene 6. Supernatants were harvested 48 h after transfection as
described. 3.5 ⫻ 104 PG13 cells were seeded on a plate in complete DMEM medium and
then transduced with supernatant obtained from eFNX cells. 48 h later, transduced PG13
cells were selected as bulk cultures with 0.5 mg/ml G418 (Gibco) and expanded.
Supernatants were obtained as described above.

Primary Fibroblast and HeLa Cell Transduction

Lentiviral Vectors: 3.5 ⫻ 104 primary human fibroblasts were seeded per well of gela-
tinized 6-well plates and the next day transduced by LV-THM vectors expressing different
shRNAs against human RAD51: shRNA5 5⬘-GCGCCAAAGAAGGAGCTAA-3⬘, shRNA6
5⬘-CCACCAGACCCAGCTCCTTTA-3⬘, shRNA7 5⬘-GCAGTGATGTCCTGGATAA-3⬘.
HeLa cells were plated in 6-well plates (3.5 ⫻ 104 cells/well). After 24 h, 1 ml of medium
containing LV-THM virus was added. The following day, supernatants were replaced by com-
plete DMEM. To study the ability of the regulatable system to control EGFP expression, cells
were cotransduced with 1:1 mixture of LV-THM and LV-tTRKRAB, respectively. Two days
after transduction, doxycycline (DOX, Ratiopharm, Ulm, Germany) was added in different
concentrations ranging from 2.5 ␮g/ml to 12.5 ␮g/ml. Five days after treatment, cells were
analyzed by flow cytometry for the expression of EGFP. For experiments using 5 ␮g/ml of
DOX, cells were analyzed every day for five days to determine the time period necessary for
maximal EGFP expression.
Retroviral Vector: HeLa cells were plated in 6-well plates (3.5 ⫻ 104 cells/well). After
24 h, 1 ml of medium containing S11EGRAIN vector was added. The following day the
supernatant was replaced by fresh DMEM supplemented with G418 for 7 days.

Flow Cytometry Analysis

HeLa cells were trypsinized, washed with PBS and then analyzed by flow cytometry
using a FACScan (Becton Dickinson, Heidelberg, Germany) for EGFP fluorescence using
standard filters. Cell quest software was used for analysis of the files containing
20,000–50,000 events.

Western Blot
Western blot analyses were performed using extracts of fibroblasts, collected by
centrifugation, washed twice in PBS, lysed in 1⫻ lysis buffer (50 nM Tris-HCl, 70 mM
2-mercaptoethanol, and 2% sodium dodecylsulfate (SDS), then boiled for 5 min, and sub-
jected to 4–12% Nupage (Invitrogen). After electrophoresis, proteins were transferred to a
nitrocellulose membrane using a submerged transfer apparatus (BioRad, Hercules, CA),
filled with 25 mM Tris Base, 200 mM glycine, and 20% methanol. After blocking with 5%
non-fat dried milk in TBS-T (50 mM Tris-HCl (pH 8.0), 150 mM NaCl, and 0.1% Tween 20)
the membrane was incubated with the primary antibodies (anti-RAD51, Calbiochem; anti-GFP
or anti-␤-actin, Abcam, Cambridge, MA) diluted in TBS-T. The detection was performed

Rio/Hanenberg 214
with the Western Breeze Immunodetection Kit (Invitrogen). The concentration of protein was
measured by the Bradford assay.

Immunofluorescence Studies of RAD51

For immunofluorescence studies, cells were fixed with 3.7% paraformaldehyde in PBS
for 15 min followed by permeabilization with 0.5% Triton X-100 in PBS for 5 min. After
30 min in blocking buffer (10% FBS, 0.1% NP-40 in PBS) cells were incubated with anti-
RAD51 anti-rabbit antibody (PC130) from Calbiochem and anti-GFP (Abcam) at 1/200 dilu-
tion. Cells were then washed three times in TBS and subsequently incubated with the
TexasRed-conjugated anti-rabbit polyclonal antibody (Jackson Immunoresearch
Laboratories, Cambridgeshire, CA). After 45 min, cells were washed three times with TBS
and the slides were mounted in Vectashield (Vector Laboratories Burlingame, CA) with 4,6-
diamidino-2-phenylindole (DAPI).

Results

Down-Regulation of RAD51 Protein Expression Using siRNA

HeLa cells were transfected with two different siRNAs designed against
hRAD51 that are commercially available. In parallel, a fluorescent oligonu-
cleotide labeled with FITC was utilized to verify by flow cytometry that 75%
of the HeLa cells were transfected with the siRNA (fig. 1a). On these cells,
western blot analysis was performed to analyze the kinetics of RAD51 knock-
down 24, 48 and 72 h after transfection. siRNA1 but not siRNA2 inhibited
RAD51 mRNA as evidenced by the absence of RAD51 protein (fig. 1b). When
both siRNAs were transfected simultaneously, a similar inhibition of RAD51
protein translation was achieved. The downregulation of RAD51 protein was
limited to the first 24 h interval after transfection, whereas there was no change
of RAD51 protein levels at 48 h and 72 h. When we analyzed the ability of
HeLa cells to form RAD51 foci 24 h after transfection, a marked depletion of
MMC-induced RAD51 foci was only observed in cells transfected with
siRNA1 (fig. 1c).

Stable Inhibition of RAD51 in Primary Fibroblast by Lentiviral shRNA

As the inhibition of RAD51 translation via siRNA1 was only transient,
lasting no longer than 24 h, we subsequently explored a lentivirus-based stable
shRNA system for RAD51 downregulation. Lentiviral vectors were constructed
using the LV-THM as backbone kindly provided by Didier Trono [1]. Different
shRNAs under the control of H1 promoter were cloned into the ClaI-MluI sites
(fig. 2a) and tested for their ability to transduce primary human fibroblasts
detectable by the simultaneous analysis of EGFP and knock-down of RAD51
protein.

Functional Knock-Down of RAD51 and Fanconi Anemia 215

 siRNA1,2
siRNA1

siRNA2
Control
512
Control 128
siRNA-FITC

24h
Events

Events

transfection
Time after
M1 M1 48h
0 0
100 101 102 103 104 100 101 102 103 104
72h
a EGFP EGFP
b Anti-RAD51

Control siRNA1 siRNA2

Fig. 1. Transfection of HeLa cells with two different siRNAs against RAD51. (a)
HeLa cells were cotransfected with a fluorescent oligo (FITC) to analyze the efficiency of
transfection. (b) Western blot of HeLa cells transfected with siRNA1, 2, or both. Western blot
analyses were performed on cells harvested 24, 48 and 72 h after transfection. (c) Ability of
HeLa cells transfected with RAD51 siRNAs to form RAD51 foci after induction of DNA
damage with MMC. RAD51 foci appeared red due to staining with a Tx-Red labeled
secondary antibody.

Lentivirus containing supernatant was produced in HEK293T cells with

VSV-G as envelope protein and used to transduce primary human fibroblasts.
Flow cytometric analysis revealed that more than 90% of the primary cells were
EGFP positive, indicating a high rate of transduction (data not shown). As
RAD51 protein appears in multiple discrete foci at sites of DNA damage, we
next analyzed the ability of fibroblasts transduced with lentiviral vectors
expressing different shRNA against RAD51 to form RAD51 foci after induc-
tion of DNA damage with MMC. Figures 2b and 2c demonstrate that cultures
with untransduced cells (control) and cells transduced with a control vector
(LV-THM) had similar percentages of cells with RAD51 foci after DNA dam-
age. In contrast, cells transduced with LV-THM expressing RAD51 shRNA5,
-6 and -7 displayed a distinctly reduced ability to form foci after DNA damage.
When quantified as the number of foci per EGFP positive cell (fig. 2b and 2c),
the effects are particularly pronounced for shRNA6 and -7, with the number of

Rio/Hanenberg 216
 LV-THM
MluI 5612
5894 bp
ClaI 5628 Cloning sites
for shRNAs
SD 744 SA 1907

env env cPPT WPRE tetO H1

a HIV LTR GAG* RRE EF1-␣ EGFP 3'LTR 3' LTR
LoxP

Number of cells with 5–10 foci Number of cells with ⬎10 foci
60 60
Number of cells with

Number of cells with

50 50
foci/200 cells

foci/200 cells
40 40
30 30
20 20
10 10
0 0
b Control LV-THM shRNA5 siRNA6 siRNA7 c Control LV-THM shRNA5 shRNA6 shRNA7

Fig. 2. (a) Map of the lentiviral vector LV-THM expressing shRNA. The shRNAs were
cloned between the MluI and ClaI sites and expressed under control of the human H1 pro-
moter. The EF1␣ promoter was used to express the enhanced green fluorescent protein
(EGFP) as marker gene. (b, c) Number of RAD51 foci per 200 cells was counted in primary
fibroblasts after transduction with LV-THM vectors expressing the shRNA5, -6 and -7
against RAD51. As control, nontransfected cells and cells transduced with LV-THM vector
were utilized. (b) Number of cells with 5–10 foci per cell. (c) Number of cells with more than
10 foci per cell. Number of cells with foci after MMC treatment (purple column) and with-
out MMC (yellow).

cells containing more than 10 foci per cell being strongly reduced compared to
control cells. These results suggest that although formation of RAD51 foci is
not completely abolished, shRNA6 and -7 expressing cells experience at least
partial inhibition of RAD51 mRNA processing.

Doxycyclin-Regulated Activity of the LV-THM Lentiviral Vector

As there are no human cell lines deficient for RAD51, obviously due to
the absolute requirement of the protein for long-term cellular survival [22], we
next modified our lentiviral shRNA delivery system, aiming at an inducible
interference with RAD51 expression via the DOX system. To this end, cells
were transduced in parallel with the LV-tTRKRAB vector kindly provided by
Didier Trono [1]. Expression of the tTRKRAB fusion protein inhibits the

Functional Knock-Down of RAD51 and Fanconi Anemia 217

MOI LV-THM: 10⫺1 1 3
512 512 512

Events

Events

Events
HeLa 67.42% 67.42% 99.7%

M1 M1 M1

0 0 0
100 101 102 103 104 100 101 102 103 104 100 101 102 103 104
EGFP EGFP EGFP

512 512 512

Events

Events

Events
HeLa transduced 0% 0.59% 3.34%
LV-tTRKRAB M1 M1 M1

0 0 0
100 101 102 103 104 100 101 102 103 104 100 101 102 103 104
EGFP EGFP EGFP

Fig. 3. Regulation of EGFP expression by LV-tTRKRAB vector. HeLa cells were

transduced with LV-THM vector using different MOIs. The same cells were transduced with
LV-tTRKRAB vector. The percentage of EGFP positive cells was determined by flow cytom-
etry analysis.

transcription of both promotors, H1 and EF1␣. Thus, without DOX in the

medium, cells are neither green nor do they express the transfected shRNAs.
In order to test the efficiency of this system, HeLa cells were infected with dif-
ferent multiplicities of infection (MOI) of the LV-THM vector while keeping
the LV-tTRKRAB vector at a constant MOI. As shown in figure 3, LV-
tTRKRAB can efficiently inhibit EGFP expression if the MOI for LV-THM is
1 or lower. At higher MOIs, the tTRKRAB mediated control of transcription
appeared to be less stringent.
As a next step, we analyzed the optimal conditions for DOX dependent
regulation of transcription in cells transduced with both the lentiviral shRNA
vector and the LV-tTRKRAB vector. As read-out, the recovery of EGFP expres-
sion was analyzed by flow cytometry using 5 different concentrations of DOX
ranging from 2.5 to 12.5 ␮g/ml (fig. 4). The results show that DOX concentra-
tions of 2.5 and 5 ␮g/ml are sufficient to reexpress EGFP in all transduced
cells. As concentrations of 7.5 ␮g/ml and higher markedly influenced
cell growth kinetics (data not shown), 5 ␮g/ml DOX was used in all further
experiments.

Rio/Hanenberg 218
 HELA LV-THM:LV-tTRKRAB (1:1)
HELA control 0 ␮g/ml DOX 2.5 ␮g/ml DOX
1,023 512 512 512

Events

Events
Events
SSC

5.4% 96.4%
M1 M1 M1

0 0 0 0
0 1,023 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104
FSC EGFP EGFP EGFP

5 ␮g/ml DOX 7.5 ␮g/ml DOX 10␮g/ml DOX 12.5␮g/ml DOX

512 512 512 512

Events
Events

Events
Events

97.5% 97.4% 96.25% 95.35%

M1 M1 M1 M1

0 0 0 0
100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104
EGFP EGFP EGFP EGFP
HELA LV-THM:LV-tTRKRAB (1:1)

Fig. 4. Cells cotransduced with LV-THM and LV-tTRKRAB vector were incubated
with different concentrations of doxycycline (DOX). HeLa cells were analyzed by flow
cytometry five days after incubation with 0–12.5 ␮g/ml DOX.

We also tested the time period after adding DOX to LV-tTRKRAB express-
ing cells that would be required to completely abrogate the negative effect of the
tTRKRAB protein on transcription from the H1 and EF1␣ promoters of the LV-
THM vector. To this end, the percentage of cells transduced with both LV-THM
and LV-tTRKRAB at a ratio of 1:1 was analyzed for the expression of EGFP in
the presence of 5 ␮g/ml DOX. As shown in figure 5, flow cytometric analysis at
24 h intervals revealed that the tTRKRAB mediated inhibition of transcription
starts to decrease at 24 h after incubation with DOX. It nevertheless took 72 h
until the maximum EGFP expression occurred as determined by the percentage
of EGFP positive cells and the mean fluorescence intensity (MFI) of cells in
region M1 (data not shown).

DOX-Controlled Regulation of RAD51 and EGFP Protein Levels

With the above established conditions for optimal application of the LV-
THM/LV-tTRKRAB system, we determined the time course of inhibition of

Functional Knock-Down of RAD51 and Fanconi Anemia 219

 HELA LV-THM: LV-tTRKRAB (1:1)
HELA control DAY 0 DAY 1
1,023 512 512 512

46.74%

Events

Events
Events
SSC

4.55% M1
M1
M1

0 0 0 0
0 1,023 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104
FSC EGFP EGFP EGFP

Day 2 Day 3 Day 4 Day 5

512 512 512 512

Events
Events

Events
Events

94.43% 95.84% 96.67%

M1 95.45% M1 M1
M1

0 0 0 0
100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104
EGFP EGFP EGFP EGFP

Fig. 5. Time for DOX to sequester tTRKRAB thereby inducing EGFP expression. Cells
cotransduced with LV-THM and LV-tTRKRAB were treated with 5 ␮g/ml of DOX and the
expression of EGFP was analyzed by flow cytometry between day 0 and day 5 after treatment.

RAD51 and of EGFP expression in HeLa cells cotransduced with LV-THM

shRNA-6 and LV-tTRKRAB at a ratio of 1:1. Western blot analyses revealed
that after three days of incubation with 5 ␮g/ml DOX, transduced HeLa cells
achieved the maximal upregulation of EGFP in concert with downregulated
RAD51 (fig. 6). The inverse parallel regulation of the two proteins by simple
addition of DOX to the culture medium underlines the effectiveness of this
lentiviral vector system for the induced expression of the shRNA cassette as
well as for visualization of cells in which the shRNA cassette is actively
expressed.

Expression of EGFP-Tagged RAD51

Controlled knockdown of endogenous RAD51 expression offers the
unique possibility to visualize RAD51 protein trafficking in live cells by over-
expression of a fluorescence tagged protein. For this purpose we constructed a
retroviral vector expressing a fusion protein between RAD51 and EGFP (fig. 7a).

Rio/Hanenberg 220
 shRNA6 RAD51
0 1 2 3 4

Anti-RAD51

Anti-GFP

Anti-␤-actin

Fig. 6. Regulation of RAD51 and EGFP expression in primary fibroblasts cotrans-

duced with the LV-THM/RAD51_shRNA6 and LV-tTRKRAB vectors. After transduction,
cells were incubated with DOX (5 ␮g/ml) for 4 days. Samples were analyzed daily by west-
ern blot for RAD51 and EGFP expression. ␤-actin was used as loading control.

S11EGRAIN
4938 bp

a PCMV LTR EGFP RAD51 IRES NEO SFFV LTR

Fig. 7. Cotransduction of HeLa cells with LV-THM/RAD51_shRNA6 and

S11EGRAIN. (a) The gammaretroviral vector S11EGRAIN expressing a fusion between
EGFP and human RAD51 genes under the control of the SFFV promoter linked via an EMCV
IRES site to the neomycin phosphotransferase (NEO) was used to transduce HeLa cells also
infected with LV-THM/RAD51_shRNA6. (b) The expression of RAD51 protein in cells was
visualized by the stainings with antibodies directed against EGFP and hRAD51. The expres-
sion of the EGFP-RAD51 fusion protein after antibody staining is depicted in green. RAD51
foci formation is shown by staining with an anti-RAD51 antibody and a TxRed labeled sec-
ondary antibody. The overlap of both pictures revealed the normal distribution of EGFP-
RAD51 fusion protein in cells. The DAPI staining of the nucleus is shown in blue.

Functional Knock-Down of RAD51 and Fanconi Anemia 221

In order to allow for selection of transduced cells, the vector constructs also
coexpressed the neomycin phosphotransferase gene (NEO) via an encephalo
myocarditis virus (EMCV) internal ribosomal entry site (IRES). To test the fea-
sibility of this approach, HeLa cells stably expressing the shRNA6 directed
against RAD51 were co-transduced with the S11EGRAIN vector and selected
with G418. Using foci formation after staining with GFP and RAD51 antibod-
ies as read-out (fig. 7b) demonstrates that exogenous EGFP-tagged RAD51,
much like wildtype protein, is assembled into nuclear foci after DNA
damage. These results suggest that the fusion protein retains its RAD51 spe-
cific function.

Discussion

The RAD51 recombinase is a key player for the execution of homologous

recombination. Downregulation of RAD51 sensitizes cells to DNA-damaging
agents [23], whereas increased RAD51 protein levels have been implicated in
uncontrolled recombination, genome instability and increased resistance of
tumor cells to radio- and chemotherapy [24]. In addition to its multiple interac-
tions with proteins such as p53, BLM, c-Abl, stat 5, BRCA1 and others, the
interaction between BRCA2 and RAD51 has revealed its crucial contribution to
homology directed DNA repair [24]. As biallelic mutations in BRCA2 cause a
severe form of FA with an early tumor phenotype, the exploration of the precise
role of FA-related proteins in homology directed DNA repair would benefit
greatly from an experimental system in which the activity of the RAD51 recom-
binase could be fine-tuned and controlled. Our present results suggest that we
are very close to this goal.
The elegant lentiviral shRNA system described by Trono’s group [1, 21]
has several advantages over siRNA transfection or other viral shRNA transfer
systems such as AAV2 or gammaretroviral vectors. Recombinant lentiviruses
especially when pseudotyped with the VSV-G glycoprotein are remarkably
efficient for the delivery of expression cassettes of genes or shRNAs to adher-
ent cells, usually resulting in transduction efficiencies in excess of 95% of cells.
We here show that this high transfer efficiency is also observed in primary cells.
Use in combination with the visualization of single transduced cells by simulta-
neous EGFP expression avoids artifact-prone selection or enrichment proce-
dures which are essential in case the suppressed gene is critical for cell survival.
Our experiments have established suitable conditions for creating a regu-
latable RAD51 knock-down in HeLa cells. Maximum targeting effects were
achieved after a 3-day exposure to 5 ␮g/ml DOX in cells simultaneously

Rio/Hanenberg 222
transduced at a 1:1 ratio with shRNA and tTRKRAB expressing vectors. We
also showed that knock-down using a single shRNA construct still allowed
RAD51 foci formation in primary human cells albeit at reduced levels. We
finally demonstrated that overexpression of a 5⬘ EGFP tagged human RAD51
protein in combination with knockdown of endogenous RAD51 protein led
to unimpaired foci formation as detected in transduced cells through the
EGFP tag.
Previous results showed that a vector expressing RAD51 fused to GFP
was mainly expressed in the cytoplasm of cells. Translocation into the cell
nucleus occurred only in response to DNA damage [25]. Although we still
have to formally prove that our EGFP-RAD51 fusion protein functions nor-
mally and that its in vivo expression is stable over time, the normal distribu-
tion of EGFP positive RAD51 foci after DNA damage that we observed in
cells with knock-down of the endogenous RAD51 suggests that the combina-
tion of the vectors we describe here could be an important approach for study-
ing the role of RAD51 in cells of FA patients and normal controls. The
possibility to induce local DNA damage employing a UV laser and confocal
microscopy offers an exciting tool to describe the kinetics of RAD51 mobi-
lization in living cells and to visualize the movement of FA proteins tagged
with other fluorescent dyes to the site of damaged DNA. This approach might
also be useful to study how homologous recombination (HR) takes place in
cells of FA patients belonging to complementation group FA-D1, and how dif-
ferent mutations in BRCA2 resulting in truncated proteins might still allow
residual HR activity.
There is clear need for further improvement of our retroviral vectors as
the shRNA against the RAD51 ORF will also influence the EGFP-tagged
overexpressed protein. This problem can be addressed by introducing silent
mutations in the RAD51 cDNA thereby changing the particular sequences that
are used for the shRNA targeting. In addition, the mutated and fluorescent
tagged protein could be introduced into cells with the same lentiviral vector as
the shRNA. With this design, it would be possible to utilize one construct to do
both, overexpression of a tagged exogenous and knock-down of the endoge-
nous protein. In addition, EGFP as a marker to visualize transduced cells
should be replaced in the shRNA vector either by selectable genes such as
NEO or by surface expressed molecules that enable selection of transduced
cells by MACS or FACS. With these modifications, a multicolor approach
using the same regulatable shRNA vector construct to target the endogenous
protein and to overexpress a fluorescence-tagged version of the same protein
appears to provide a fascinating starting point to visualize and better under-
stand the complex mechanisms in FA cells undergoing HR following exposure
to crosslinking agents.

Functional Knock-Down of RAD51 and Fanconi Anemia 223

 Acknowledgements

We would like to thank Didier Trono for providing his inducible lentiviral shRNA vec-
tor system. We are deeply indebted to Juan Bueren for his continuous support and helpful
discussions. This study was supported by the Fanconi Anemia Research Fund, (FARF), Inc.,
Portland, Oregon, U.S.A., and the Elterninitiative Kinderkrebsklinik e. V., Duesseldorf,
Germany.

References

1 Wiznerowicz M, Trono D: Conditional suppression of cellular genes: lentivirus vector-mediated

drug-inducible RNA interference. J Virol 2003;77:8957–8961.
2 Howlett NG, Taniguchi T, Olson S, Cox B, Waisfisz Q, et al: Biallelic inactivation of BRCA2 in
Fanconi anemia. Science 2002;297:606–609.
3 Levitus M, Waisfisz Q, Godthelp BC, de Vries Y, Hussain S, et al: The DNA helicase BRIP1 is
defective in Fanconi anemia complementation group J. Nat Genet 2005;37:934–935.
4 Levran O, Attwooll C, Henry RT, Milton KL, Neveling K, et al: The BRCA1-interacting helicase
BRIP1 is deficient in Fanconi anemia. Nat Genet 2005;37:931–933.
5 Reid S, Schindler D, Hanenberg H, Barker K, Hanks S, et al: Biallelic mutations in PALB2 cause
Fanconi anemia subtype FA-N and predispose to childhood cancer. Nat Genet 2007;39:162–164.
6 Litman R, Peng M, Jin Z, Zhang F, Zhang J, et al: BACH1 is critical for homologous recombina-
tion and appears to be the Fanconi anemia gene product FANCJ. Cancer Cell 2005;8:255–265.
7 Levitus M, Joenje H, de Winter JP: The Fanconi anemia pathway of genomic maintenance. Cell
Oncol 2006;28:3–29.
8 Mirchandani KD, D’Andrea AD: The Fanconi anemia/BRCA pathway: a coordinator of cross-link
repair. Exp Cell Res 2006;312:2647–2653.
9 Baumann P, Benson FE, West SC: Human Rad51 protein promotes ATP-dependent homologous
pairing and strand transfer reactions in vitro. Cell 1996;87:757–766.
10 Haaf T, Golub EI, Reddy G, Radding CM, Ward DC: Nuclear foci of mammalian Rad51 recombi-
nation protein in somatic cells after DNA damage and its localization in synaptonemal complexes.
Proc Natl Acad Sci USA 1995;92:2298–2302.
11 Golub EI, Gupta RC, Haaf T, Wold MS, Radding CM: Interaction of human rad51 recombination
protein with single-stranded DNA binding protein, RPA. Nucleic Acids Res 1998;26:5388–5393.
12 Galkin VE, Esashi F, Yu X, Yang S, West SC, Egelman EH: BRCA2 BRC motifs bind RAD51-
DNA filaments. Proc Natl Acad Sci USA 2005;102:8537–8542.
13 Pellegrini L, Yu DS, Lo T, Anand S, Lee M, et al: Insights into DNA recombination from the struc-
ture of a RAD51-BRCA2 complex. Nature 2002;420:287–293.
14 Sharan SK, Morimatsu M, Albrecht U, Lim DS, Regel E, et al: Embryonic lethality and radiation
hypersensitivity mediated by Rad51 in mice lacking BRCA2. Nature 1997;386:804–810.
15 Godthelp BC, Wiegant WW, Waisfisz Q, Medhurst AL, Arwert F, et al: Inducibility of nuclear
Rad51 foci after DNA damage distinguishes all Fanconi anemia complementation groups from
D1/BRCA2. Mutat Res 2006;594:39–48.
16 Muris DF, Vreeken K, Carr AM, Broughton BC, Lehmann AR, et al: Cloning the RAD51 homo-
logue of Schizosaccharomyces pombe. Nucleic Acids Res 1993;21:4586–4591.
17 Shinohara A, Ogawa H, Ogawa T: Rad51 protein involved in repair and recombination in S. cerevisiae
is a RecA-like protein. Cell 1992;69:457–470.
18 Tsuzuki T, Fujii Y, Sakumi K, Tominaga Y, Nakao K, et al: Targeted disruption of the Rad51 gene
leads to lethality in embryonic mice. Proc Natl Acad Sci USA 1996;93:6236–6240.
19 Sandy P, Ventura A, Jacks T: Mammalian RNAi: a practical guide. Biotechniques 2005;39:
215–224.

Rio/Hanenberg 224
20 Elbashir SM, Harborth J, Lendeckel W, Yalcin A, Weber K, Tuschl T: Duplexes of 21-nucleotide
RNAs mediate RNA interference in cultured mammalian cells. Nature 2001;411:494–498.
21 Wiznerowicz M, Szulc J, Trono D: Tuning silence: conditional systems for RNA interference. Nat
Methods 2006;3:682–688.
22 Raderschall E, Bazarov A, Cao J, Lurz R, Smith A, et al: Formation of higher-order nuclear Rad51
structures is functionally linked to p21 expression and protection from DNA damage-induced
apoptosis. J Cell Sci 2002;115:153–164.
23 Ohnishi T, Taki T, Hiraga S, Arita N, Morita T: In vitro and in vivo potentiation of radiosensitivity
of malignant gliomas by antisense inhibition of the RAD51 gene. Biochem Biophys Res Commun
1998;245:319–324.
24 Henning W, Stuerzbecher HW: Homologous recombination and cell cycle checkpoints: Rad51 in
tumour progression and therapy resistance. Toxicology 2003;193:91–109.
25 Kraakman-van der Zwet M, Overkamp WJ, van Lange RE, Essers J, van Duijn-Goedhart A, et al:
Brca2 (XRCC11) deficiency results in radioresistant DNA synthesis and a higher frequency of
spontaneous deletions. Mol Cell Biol 2002;22:669–679.

Helmut Hanenberg, M.D.

Department of Pediatric Hematology, Oncology and Clinical Immunology
Children’s Hospital, Heinrich-Heine-University, Moorenstr. 5
D–40225 Düsseldorf (Germany)
Tel. ⫹49 211 811 6103, Fax 49 211 811 9436, E-Mail [email protected]

Functional Knock-Down of RAD51 and Fanconi Anemia 225

 Author Index

Bechtold, A. 131, 149 Herterich, S. 39, 131, 149 Schlegel, P.G. 173
Buwe, A. 183 Hoehn, H. 23, 110, 149 Schmid, M. 183
Huber, S. 95 Schmugge Liner, M. 131
Demuth, I. 200 Schroeder-Kurth, T. 1
Dietrich, R. 9 Kalb, R. 39, 59, 110, 131, Sperling, K. 23
Digweed, M. 23, 200 149 Sun, Y. 149
Kubbies, M. 110
Ebell, W. 79 Kühl, J.-S. 79 Takata, M. 183
Eyrich, M. 173 Tönnies, H. 79, 95
Linka, Y. 110
Friedl, R. 110, 131, 149
Velleuer, E. 9
Nanda, I. 183
Gavvovidis, I. 110 Volarikova, E. 95
Neitzel, H. 79, 95
Gerlach, A. 79, 95 Neveling, K. 39, 59, 110,
Gottwald, B. 131 Winkler, B. 173
131, 149
Gruhn, B. 149 Wizenman, A. 183
Rio, P. 211
Haaf, T. 183
Hanenberg, H. 110, 131, Schartl, M. 183
149, 211, Schindler, D. 39, 59, 110,
Heilmann, C. 131 131, 149

226
 Subject Index

Acute myeloid leukemia (AML) 62, 65 Comparative genomic hybridization (CGH)

Aging 6, 34 84, 97
Alkylating agents Compensatory mutation 168
sensitivity to 40, 134 Complementation group 28, 43
Amniotic fluid 135 Compound heterozygotes 166
Androgen therapy 9 Conditioning regimens 176
Avian cells 183 Cytogenetic analysis 80

Back-mutation 34, 167 DNA crosslinking agents

Bone marrow exposure to 111
failure 33 sensitivity to 25, 111, 191
cytogenetic studies 80 DNA damage 60
clonal aberrations 81, 89 DNA double-strand-break (DSB) repair
transplantation 173 26, 201
BRCA2 69 Doxycyclin (DOX) 217
BrdU/Hoechst technique 117 DT40 cells 189, 203
Breast cancer 69
Exon skipping 140
Cancer risk 4, 25, 59
Caretaker genes 60 FA/BRCA pathway 23, 40, 64
Case reports 11 FA core complex 31, 41
Cell cycle FA genes
analysis 124 description 47, 71
disturbance 121 evolutionary conservation 44, 185
testing 110, 154 frequency 39
Cell lines 27, 151 identification 28, 40
Chicken 183 location 29, 43
Chromosome aberrations 23 mutations 39, 44, 139
clonal 79, 81, 89 orthologs in birds 186
Chromosome breakage test 152 size 43, 45
Chromosome instability 3, 25 somatic reversion 163
Clonal structure 39, 45
expansion 170 FA protein function 31, 41
evolution 87, 99, 105 FA registry 5, 39

227
FA research 23 Neoplasia 61
milestones in 26 Nocadazol 115
FANCF silencing 71, 73 Non-FA tumor patients 66
Fibroblasts 112, 155 AML 65, 91
Flow cytometry 111, 117, 135, 214 breast cancer 69
Fluorescence in situ hybridization (FISH) ovarian cancer 69
84 pancreatic cancer 68
Founder mutations 50 squamous cell carcinoma 71
Non-homologous end-joining (NHEJ) 201
Gene conversion 34, 166
Genetic heterogeneity 27, 53 Ovarian cancer 69
Genotype-phenotype correlation 52
Genotyping in PGD Pancreatic cancer 68
direct 138 Peripheral blood cells (PBC) 95
indirect 137 Phenotype variability 10
Germline mutations 65 Phosphohistone H3 (H3P) 115
Gonad 192 Preimplantation genetic diagnosis (PGD)
Graft-versus-host disease (GvHD) 174 33, 144
Green fluorescent protein (GFP) 112, 202 Prenatal diagnosis of FA 131
G2-phase accumulation 112 functional testing 134
molecular testing 137
H3P staining 115 Protein-protein interaction 31
Hematopoietic stem cell transplantation
(HSCT) 9, 33, 64 RAD51
HLA-identical sibling 174 foci formation 203, 211
Hoechst 33258 117 functional knock-down 211
Homologous recombination 200 immunofluorescence 215
Homology directed repair (HDR) 201, 204 Reactive oxygen species 185
Reverse mutations 5
Immune reconstitution 177 Revertant mosaicism 149
Interphase FISH (I-FISH) 95 cell cycle testing 154
sensitivity 98 chromosome breakage test 152
Interstrand crosslinks (ICL) 40, 111, 200 clinical consequences 169
Intragenic crossover 165 hematopoietic precursor cells 160
lymphocytes 156
Leukemia 62 molecular confirmation 156
Liver tumors 64 skin fibroblast testing 155
Loss of heterozygosity (LOH) 68
Sex chromosomes 183
Megadose concept 175 sG2/GF ratio 122
Microsatellite markers 137 shRNA delivery system 211
Mitomycin C (MMC) Solid tumors 62
sensitivity to 133, 149 Somatic reversion 34, 53
treatment 112 cell lines 151
Monosomy 7 80, 87, 100 FANCD2
Multiplex ligation-dependent probe molecular mechanisms 165
amplification (MLPA) 46 multilineage 161, 169
Myelodysplastic syndrome (MDS) 61 target FA genes 163

Subject Index 228

Single strand annealing (SSA) 201 Trisomy 3q 84, 99
Squamous cell carcinoma 63, 71 tTRKRAB 201
-head and neck (HNSCC) 63 Tumorigenesis 65
Stem cell transplantation 173 Tumors 64
Support groups 10, 34
Uniparental disomy (UDP) 86
Tetrasomy 3q 84, 99
Therapy 32 Vertebrates 183
Thymic function 179

Subject Index 229