Genetics of The Mouse

Jean Louis
Guénet · Fernando Benavides
Jean-Jacques Panthier · Xavier Montagutelli
Genetics of
the Mouse
Genetics of the Mouse
Jean Louis Guénet · Fernando Benavides
Jean-Jacques Panthier · Xavier Montagutelli
Genetics of the Mouse
13
Jean Louis Guénet Jean-Jacques Panthier
Institut Pasteur (Emeritus) Mouse Functional Genetics Unit
Paris Institut Pasteur
France Paris
France
Fernando Benavides and
Division of Basic Science Research
Department of Molecular Carcinogenesis Ecole Nationale Vétérinaire d’Alfort
The University of Texas MD Anderson Maisons-Alfort
Cancer Center France
Smithville, TX
USA Xavier Montagutelli
Mouse Functional Genetics Unit
Institut Pasteur
Paris
France
ISBN 978-3-662-44286-9 ISBN 978-3-662-44287-6 (eBook)

DOI 10.1007/978-3-662-44287-6
Library of Congress Control Number: 2014945779
Springer Heidelberg New York Dordrecht London
© Springer-Verlag Berlin Heidelberg 2015

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or
information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed. Exempted from this legal reservation are brief excerpts
in connection with reviews or scholarly analysis or material supplied specifically for the purpose of
being entered and executed on a computer system, for exclusive use by the purchaser of the work.
Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright
Law of the Publisher’s location, in its current version, and permission for use must always be obtained
from Springer. Permissions for use may be obtained through RightsLink at the Copyright Clearance
Center. Violations are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of
publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for
any errors or omissions that may be made. The publisher makes no warranty, express or implied, with
respect to the material contained herein.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)

This book is dedicated to
the memory of
Professor François Jacob
(June 1920–April 2013)
Foreword
The science of experimental biology rests on the analysis of causative factors, fol-
lowed by synthesis. Commonly, the analytic step involves determining the con-
sequences of a known perturbation. Classical experimental biology rested on
perturbations of the environment, or on surgical operations such as transplantation.
When the science of genetics reached molecular resolution in the twentieth cen-
tury, mutational perturbation became prominent. In organisms for which sophisti-
cated genetic methods have been developed, it is feasible, either through positional
cloning of the mutated gene or through directed mutagenesis, to make connections
between changes in phenotype and specific molecular changes. The laboratory
mouse is the first experimental mammalian species allowing these sophisticated
methods. Thus, The Genetics of the Mouse by Guénet, Benavides, Montagutelli,
and Panthier is more than a genetics textbook. It is also a talisman, containing
instructions by which the experimental mammalian biologist can analyze a pro-
cess of interest at molecular resolution. It is a twenty-first-century version of the
twelfth-century tome on the crafts of the medieval guilds authored by Theophilus:
On Divers Arts.
The chapters delve deeply into the biology of the mouse. They range from
detailed presentations of the natural history of the species, its handling in the labo-
ratory, and its classical genetics, to contemporary issues including the epigenet-
ics of parental imprinting and X-chromosome inactivation. Further, they provide a
detailed discussion of the strategies for creating and cloning constitutive and con-
ditional mutant alleles. Finally, they present a platform from which the analysis of
complex quantitative traits is currently addressed. When the in-depth details of a
subject exceed reasonable limits in length, the authors provide footnotes to more
extensive treatments. As experienced geneticists, the authors appreciate the impor-
tance of phenotyping, not letting it get lost in the details of analyzing and manipu-
lating the genotype. At the core of their presentation is the importance of inbred
strains and isogenicity for the identification of single causative factors.
The ultimate goal of many mammalian experimental biologists is to develop
an understanding of issues in human biology. The authors recognize the circum-
stances in which a particular mouse model fails to present the phenotype expected
vii
viii Foreword
from the cognate condition in the human, and they outline ways in which mice can
be made chimeric for human tissues. Because any one model gives at best only a
first approximation to the human case, a diverse set of models may provide further
approximations. The methods presented can lead to the development of a homol-
ogous series of mouse models in any of their distinct inbred backgrounds, or in
their genetically homogeneous F1 hybrids, or in other mammalian genera that can
be inbred.
Seen broadly, The Genetics of the Mouse connects the past, present, and future
in the experimental biology of mammals.
William F. Dove
Streisinger Professor of Experimental Biology, Emeritus
University of Wisconsin
Madison
Preface
This book is intended for several different categories of potential readers. First,
are students who have completed their university studies in biology or medical
sciences and wish to undertake a PhD project making use of mice but who have
no experience with this model organism. Reading this book will enable them to
acquire, rapidly and in a relatively condensed form, a background that will be
helpful for the critical reading of primary scientific publications and for the opti-
mal design of their projects. Genetics instructors will also find useful examples
to illustrate undergraduate biology courses. Molecular and developmental biolo-
gists whose research program is focused on a gene or gene family will also be
interested and will realize that the mouse is an exceptional model with which they
may be able to develop studies impossible or difficult to achieve with any other
mammalian species. For example, they may be able to produce a variety of point
mutations in the same genetic background or exactly the same point mutation in a
variety of different backgrounds, allowing exploration of the function of this gene
and its interplay within gene networks. This book will also be helpful to physi-
cians and pediatricians by allowing them to choose or design the best possible
model for their research related to a specific human pathology. This would be true
not only for the diseases resulting from point mutations in orthologous genes but
also, and more interestingly, for those mutations whose phenotypic expression is
influenced by the environment or the genetic background of the animal. Finally,
laboratory animal veterinarians and technicians, who are in charge of the breed-
ing and preservation of mouse models, will find useful explanations about their
increasing complexity.
This book covers all aspects of mouse genetics. The first four chapters describe
the origin of laboratory mice, the reproductive biology, the cytogenetics, and the
mapping of genes. The establishment of highly detailed genetic maps was a major
and fundamental contribution to mouse geneticists during the twentieth century
that ultimately led to the complete sequencing of the genome. This topic has been
presented in a relatively condensed form in this book, because we have consid-
ered that the excellent book published in 1995 by Lee M. Silver, which is freely
available on the site “Mouse Genome Informatics”, is still a major reference in
ix
x Preface
this matter. On the contrary, the transcriptome and the parental imprinting of the
genome are topics that have been the subject of intensive research over the last 10
years. For this reason they are presented in more detail along with the techniques
for the production of mutations, which is one of the most attractive features of the
mouse. Finally, quantitative genetics, a branch of genetics that is in expansion, is
presented in a didactic manner.
This book greatly benefited from the contributions of some of our colleagues
whom we would like to cordially thank. François Bonhomme, an old friend with
whom we have collaborated many times in the past, reviewed and commented on
Chap. 1. Marie-Geneviève Mattei read and amended Chaps. 3 and 6 and allowed
us to share her extensive knowledge of cytogenetics. Yann Herault also made inter-
esting suggestions about Chap. 3 and provided us with a schematic figure repre-
senting the best models of Down syndrome. Benoît Robert accepted the difficult
task of writing an original synthesis concerning the regulation of gene expres-
sion (Chap. 5). Edith Heard, Luisa Dandolo, and Deborah Bourc'his abundantly
corrected and commented on Chap. 6 dealing with X-inactivation and parental
genetic imprinting. Michel Cohen-Tannoudji corrected and completed our initial
versions of Chap. 8, and Tomoji Mashimo read the section of the same chapter
dealing with the production of targeted alterations using engineered nucleases and
provided a summary picture. Finally, Robert P. Erickson kindly read the whole
of our manuscript, making many insightful comments. The authors also wish to
thank Drs. Hesed M. Padilla-Nash and Thomas Ried from the Genetics Branch,
National Cancer Institute, National Institutes of Health, Bethesda for providing
a picture of a mouse spectral karyotyping, Dr. Dianne Creasy, Huntingdon Life
Sciences, East Millstone, for providing a picture of the seminiferous epithelium
with identification of the different cell types, and Ms Annie Orth for providing a
picture of a sample of her unique collection of wild mice. Finally, the authors are
greatly indebted to their colleague Dominique Simon, who helped in the prepara-
tion of many illustrations and to Mrs. Sarah Adai, MD Anderson Cancer Center,
who undertook to “translate” their awkward English into a more readable form.
Writing this book has kept us busy for nearly two years, but it was really an
enthralling experience. Whatever the chapter, we realized that the Genetics of the
Mouse has changed considerably over the last 20 years and, with an increasing
number of transnational collaborative projects, we can expect even more dramatic
changes in the years to come.
Contents
1 Origins of the Laboratory Mouse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.1 Phylogenetic Relationships of Laboratory Mice
with Other Mammals. . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.2 How the House Mouse Became a Domestic Species. . . 5
1.1.3 How the House Mouse Became a Model
for Geneticists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.1.4 The Community of Mouse Geneticists . . . . . . . . . . . . . 12
1.1.5 The Main Institutions Involved in Mouse Genetics. . . . 12
1.1.6 Books and Other Sources of Information
Concerning the Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.1.7 The Future of Mouse Genetics. . . . . . . . . . . . . . . . . . . . 14
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2 Basic Concepts of Reproductive Biology and Genetics. . . . . . . . . . . . 19

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.2 Reproduction in the Laboratory Mouse. . . . . . . . . . . . . . . . . . . . . 19
2.2.1 The Estrous Cycle and Pregnancy. . . . . . . . . . . . . . . . . 19
2.2.2 Inducing Ovulation in the Mouse (Superovulation). . . . 24
2.2.3 Artificial Insemination. . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.2.4 In Vitro Fertilization in the Mouse. . . . . . . . . . . . . . . . . 26
2.2.5 Ovary Transplantation . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2.6 Intra Cytoplasmic Sperm Injection . . . . . . . . . . . . . . . . 28
2.2.7 Cryopreservation of Mouse Embryos
and Spermatozoa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.2.8 Twinning in the Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.2.9 Cloning Laboratory Mice. . . . . . . . . . . . . . . . . . . . . . . . 30
2.2.10 Mosaics and Chimeras. . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.3 Basic Notions of Genetics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.3.1 Genes and Alleles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.3.2 Allelic Interactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
xi
xii Contents
2.3.3 Epistasis and Pleiotropy. . . . . . . . . . . . . . . . . . . . . . . . . 41

2.3.4 Penetrance and Expressivity. . . . . . . . . . . . . . . . . . . . . . 43
2.4 Phenotyping Laboratory Mice: The Mouse Clinics. . . . . . . . . . . . 45
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3 Cytogenetics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.2 The Chromosomes of the Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . 52
3.3 Identifying the Chromosome Pairs: The Normal Karyotype. . . . . 54
3.4 Meiosis and Gametogenesis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.5 Variations in Chromosome Number. . . . . . . . . . . . . . . . . . . . . . . . 62
3.5.1 The Euploid Heteroploidies. . . . . . . . . . . . . . . . . . . . . . 62
3.5.2 The Aneuploid Heteroploidies. . . . . . . . . . . . . . . . . . . . 63
3.6 Variations in Chromosome Structure. . . . . . . . . . . . . . . . . . . . . . . 68
3.6.1 The Structural Rearrangements Resulting from
a Single Break . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.6.2 The Structural Rearrangements Resulting from
Two Breaks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.6.3 Complex Structural Rearrangements. . . . . . . . . . . . . . . 80
3.6.4 Structural Rearrangements Created in Vitro . . . . . . . . . 81
3.7 Modeling Human Down Syndrome. . . . . . . . . . . . . . . . . . . . . . . . 82
3.7.1 Mouse Trisomy 16: A Model of Down Syndrome. . . . . 82
3.7.2 Ts(1716)65Dn: A Tertiary Trisomy Modeling
Down Syndrome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
3.7.3 Transgenic and Transchromosomic Models
of Down Syndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
3.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4 Gene Mapping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
4.1.1 The Discovery of Linkage Groups: A Historical
Perspective. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
4.2 From Linkage Groups to Genetic Maps. . . . . . . . . . . . . . . . . . . . . 91
4.2.1 Detecting Linkage and Measuring the Distances
Between Loci. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
4.2.2 Ordering the Genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
4.2.3 Establishing a Correspondence Between LGs
and Chromosomes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
4.2.4 Positioning the Centromere. . . . . . . . . . . . . . . . . . . . . . 101
4.3 Genetic Markers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
4.3.1 Markers Scored by Examination of the External
Phenotype. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
4.3.2 Electrophoretic Variant of Enzymatic Proteins . . . . . . . 104
4.3.3 Plasmatic Proteins and Cell Surface Antigens. . . . . . . . 104
4.3.4 Polymorphisms Detected at the DNA Level . . . . . . . . . 104
Contents xiii
4.4 High-Resolution, High-Density Genetic Maps . . . . . . . . . . . . . . . 110

4.5 Somatic Cell Hybrids and Radiation Hybrids as
Tools for Gene Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
4.6 Recombinant Inbred and Recombinant Congenic Strains. . . . . . . 112
4.7 Establishing Consensus Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
4.8 Positional Cloning of Mutations and QTLs. . . . . . . . . . . . . . . . . . 119
4.9 Physical Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
4.10 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
5 The Mouse Genome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
5.2 The Sequence of the Mouse Genome. . . . . . . . . . . . . . . . . . . . . . . 128
5.2.1 The Mouse Genome is Enormous in Size,
and its Structure is Complex . . . . . . . . . . . . . . . . . . . . . 128
5.2.2 How Was the Mouse Genome Sequenced? . . . . . . . . . . 130
5.3 The Structure of the Mouse Genome. . . . . . . . . . . . . . . . . . . . . . . 134
5.3.1 Finding the Coding and Related Sequences. . . . . . . . . . 134
5.3.2 The Canonical Architecture
of a Protein-Coding Gene. . . . . . . . . . . . . . . . . . . . . . . . 140
5.3.3 Finding the Regulatory Sequences. . . . . . . . . . . . . . . . . 144
5.3.4 Organization of Syntenic Regions at the
Chromosome Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
5.3.5 Gene Families and Pseudogenes . . . . . . . . . . . . . . . . . . 150
5.3.6 Copy Number Variations . . . . . . . . . . . . . . . . . . . . . . . . 155
5.3.7 Single Nucleotide Polymorphisms. . . . . . . . . . . . . . . . . 158
5.3.8 Tandem Repeated Sequences. . . . . . . . . . . . . . . . . . . . . 158
5.3.9 Interspersed Repeated Sequences: Transposable
Elements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
5.4 The Transcriptome: Coding and Non-coding RNAs. . . . . . . . . . . 166
5.4.1 ncRNAs Involved in Protein Synthesis . . . . . . . . . . . . . 168
5.4.2 The ncRNAs Functioning as Post-transcriptional
Regulators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
5.5 Ultraconserved Elements (UCE) and Long Conserved
Non-coding Sequences. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
5.6 Mitochondrial DNA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
5.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
6 Epigenetic Control of Genome Expression. . . . . . . . . . . . . . . . . . . . . . 187

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
6.2 X-Chromosome Inactivation in Mammals. . . . . . . . . . . . . . . . . . . 188
6.2.1 In Female Mammals Only One X is
Transcriptionally Active. . . . . . . . . . . . . . . . . . . . . . . . . 188
6.2.2 The Mechanisms Controlling
X-Chromosome Inactivation . . . . . . . . . . . . . . . . . . . . . 193
xiv Contents
6.3 Parental Imprinting of Autosomal Genes. . . . . . . . . . . . . . . . . . . . 196

6.3.1 Evidence of Genomic Imprinting in the Mouse. . . . . . . 196
6.3.2 Characterization of the Imprinted
Regions in the Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . . 203
6.3.3 What are the Molecular Mechanisms that Control
Genomic Imprinting?. . . . . . . . . . . . . . . . . . . . . . . . . . . 206
6.3.4 Genomic Imprinting Across Mammalian Species. . . . . 211
6.3.5 The Origin and Evolution of the Imprinting
Mechanisms in Mammals . . . . . . . . . . . . . . . . . . . . . . . 212
6.3.6 The Pathological Aspects Associated with
Genomic Imprinting. . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
6.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
7 Mutations and Experimental Mutagenesis. . . . . . . . . . . . . . . . . . . . . . 221

7.1 The Importance of Mutations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221
7.2 The Different Types of Mutations . . . . . . . . . . . . . . . . . . . . . . . . . 222
7.2.1 Mutations Resulting from Base-Pair Substitutions
in the Coding Sequences . . . . . . . . . . . . . . . . . . . . . . . . 223
7.2.2 Base-Pair Substitutions in the Non-coding Regions . . . 228
7.2.3 Insertions, Deletions, and Duplications. . . . . . . . . . . . . 231
7.2.4 Triplet Expansions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
7.2.5 Mutations Resulting from the Insertion
of Mobile Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
7.2.6 Mutations Due to Non-homologous Recombination
or Non-homologous End Joining. . . . . . . . . . . . . . . . . . 234
7.2.7 Copy Number Variations . . . . . . . . . . . . . . . . . . . . . . . . 234
7.3 Spontaneous Mutation Rates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
7.4 Mutagenesis in the Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
7.4.1 Gametogenesis and Experimental Mutagenesis. . . . . . . 238
7.4.2 The Induction of Mutations by Radiation . . . . . . . . . . . 240
7.4.3 The Induction of Mutations by Chemicals. . . . . . . . . . . 242
7.5 Protocols of Experimental Mutagenesis. . . . . . . . . . . . . . . . . . . . . 246
7.5.1 Phenotype-Driven, Genome-Wide Mutagenesis . . . . . . 247
7.5.2 The Induction of New Mutant
Alleles at Specific Loci . . . . . . . . . . . . . . . . . . . . . . . . . 250
7.5.3 The Induction of Mutations in Specific Regions
of the Genome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
7.5.4 A Gene-Driven Strategy for the Production
of Mutations at Specific Loci. . . . . . . . . . . . . . . . . . . . . 255
7.6 Other Techniques for the Production
of Mutations in the Mouse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
7.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260
Contents xv
8 Transgenesis and Genome Manipulations . . . . . . . . . . . . . . . . . . . . . . 267

8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
8.2 Transgenesis Resulting from Pronuclear Injection
of Cloned DNAs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
8.2.1 The Basic Experimental Protocol. . . . . . . . . . . . . . . . . . 268
8.2.2 Factors Influencing Transgenic Expression. . . . . . . . . . 271
8.2.3 Using Transgenic Mice for Studying Gene Function
and Regulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
8.2.4 The Use of Transgenic Technology to Generate
Tissue- or Cell-Specific Ablations. . . . . . . . . . . . . . . . . 276
8.2.5 Transgenic Complementation of a Mutant Allele
Identified by Positional Cloning. . . . . . . . . . . . . . . . . . . 276
8.2.6 Using Transgenic Mice for Modeling
Human Diseases. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
8.2.7 Transgenic Animals with Large DNA Inserts . . . . . . . . 279
8.2.8 Transgenic Knockdowns . . . . . . . . . . . . . . . . . . . . . . . . 280
8.2.9 Assessing the Mutagenic Activity of Chemicals
with Transgenic Mice. . . . . . . . . . . . . . . . . . . . . . . . . . . 281
8.2.10 Mutations Induced by Pronuclear Transgenesis. . . . . . . 281
8.3 Generating Alterations in the Mouse Genome Using
Embryonic Stem Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
8.3.1 Embryonic Stem Cells and their Advantages. . . . . . . . . 282
8.3.2 Targeted Mutagenesis in ES Cells. . . . . . . . . . . . . . . . . 285
8.3.3 Induction of Mutations in ES Cells with
Chemical Mutagens. . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
8.4 Inducible Transgenesis: The Tet-off and Tet-on
Expression Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
8.5 Other Techniques for the Production of Transgenic Mice. . . . . . . 304
8.5.1 Transgenesis by Retroviral Infection
of Early Embryos. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
8.5.2 In Vivo Genome Editing: The Production
of Targeted Alterations Using
Engineered Nucleases . . . . . . . . . . . . . . . . . . . . . . . . . . 306
8.6 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
9 The Different Categories of Genetically Standardized

Populations of Laboratory Mice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
9.2 Inbred Strains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
9.2.1 Inbred Mice are Isogenic and Homozygous
at All Loci. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
9.2.2 Inbred Mice are Genetically Stable
in the Long Term . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
xvi Contents
9.2.3 The Genetic Purity of Inbred Strains Must be

Regularly Monitored . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
9.2.4 Most Inbred Strains are Derived from a Small
Number of Ancestors. . . . . . . . . . . . . . . . . . . . . . . . . . . 333
9.2.5 Laboratory Inbred Strains have
a Polyphyletic Origin. . . . . . . . . . . . . . . . . . . . . . . . . . . 334
9.2.6 Inbred Strains Recently Derived
from Wild Specimens. . . . . . . . . . . . . . . . . . . . . . . . . . . 334
9.2.7 Phylogenic Relationships Between Inbred Strains . . . . 336
9.3 Interstrain F1 Hybrids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
9.4 Co-isogenic and Congenic Strains. . . . . . . . . . . . . . . . . . . . . . . . . 338
9.4.1 Co-isogenic Strains . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
9.4.2 Transgenic Strains are Equivalent but not Identical
to Co-isogenic Strains . . . . . . . . . . . . . . . . . . . . . . . . . . 339
9.4.3 Congenic Strains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
9.5 Consomic Strains. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
9.6 Recombinant Inbred Strains and Recombinant
Congenic Strains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
9.7 The Collaborative Cross. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
9.8 Outbred and Random-Bred Stocks. . . . . . . . . . . . . . . . . . . . . . . . . 353
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
10 Quantitative Traits and Quantitative Genetics. . . . . . . . . . . . . . . . . . . 361

10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
10.2 Mean and Variance: Two Essential Parameters
for the Characterization of a Population . . . . . . . . . . . . . . . . . . . . 362
10.3 Why Study the Genetics of Complex Traits
in Laboratory Mice?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
10.4 The Genetic Determinism of Quantitative Traits. . . . . . . . . . . . . . 364
10.5 The Concept of Quantitative Trait Locus (QTL). . . . . . . . . . . . . . 365
10.6 Positioning QTLs on the Genetic Map. . . . . . . . . . . . . . . . . . . . . . 366
10.6.1 Using Two-Generation Crosses
for the Detection and Positioning of QTLs . . . . . . . . . . 367
10.6.2 Point-by-Point Analysis of the Progeny. . . . . . . . . . . . . 369
10.6.3 The Concept of LOD Score. . . . . . . . . . . . . . . . . . . . . . 369
10.6.4 Threshold of Significance . . . . . . . . . . . . . . . . . . . . . . . 371
10.7 Assessing the Strength of a QTL on the Trait Studied. . . . . . . . . . 371
10.8 Interval Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
10.9 Searching for Multiple QTLs Simultaneously. . . . . . . . . . . . . . . . 374
10.10 Using Recombinant Inbred and Recombinant
Congenic Strains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
10.10.1 Recombinant Inbred Strains. . . . . . . . . . . . . . . . . . . . . . 374
10.10.2 Advantages and Disadvantages of RIS . . . . . . . . . . . . . 375
10.10.3 Recombinant Congenic Strains . . . . . . . . . . . . . . . . . . . 376
Contents xvii
10.11 Using Congenic Strains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377

10.12 Using Other Strains or Stocks for the Mapping of QTLs. . . . . . . 380
10.12.1 Consomic Strains (CS). . . . . . . . . . . . . . . . . . . . . . . . . 380
10.12.2 The Collaborative Cross (CC): A Novel,
Powerful Tool for Studying the Genetics
of Complex Traits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
10.12.3 Interspecific Recombinant Congenic Strains (IRCS). . . 381
10.12.4 Diversity Outbred (DO) Stock. . . . . . . . . . . . . . . . . . . 382
10.13 Cloning QTLs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
10.13.1 Analyzing the DNA Polymorphisms
in the QTL Region. . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
10.13.2 Quantitative Complementation. . . . . . . . . . . . . . . . . . . 384
10.14 The Analysis of Expression QTLs (eQTLs). . . . . . . . . . . . . . . . . 384
10.15 The Case of Modifier Genes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
10.16 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
Glossary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
Chapter 1
Origins of the Laboratory Mouse
1.1 Introduction
Because they are often closely associated with humans and sometimes “share”
food with them, zoologists consider the mouse as a commensal species (from the
Latin cum mensa, which means eating at the same table). For the same reason,
mice are often referred to as “house” mice in the English literature, as opposed to
wild or feral mice even though, in fact, they are the same species. Because they
have been described as “invasive,” “prolific,” “troublesome,” and “devastating,”
farmers consider mice to be pests. Physicians and epidemiologists don’t like mice
(and rodents more generally) because they are natural reservoirs of many patho-
gens, some of them deadly. For some other people, on the contrary, mice are cute
pets, easy to breed, cheap to buy, and with beautiful coat colors (Fig. 1.1).
Researchers, geneticists in particular, have the greatest respect for mice and
have even graded the species to the rank of domestic species, which they breed in
large numbers to fulfill their needs of experimental models. With so many diver-
gent opinions, the time has come to address a few basic questions about mice:
what are they actually? Where do they come from? Why and how have they
become such popular models for research in genetics over the last century?
1.1.1 Phylogenetic Relationships of Laboratory Mice

with Other Mammals
Figure 1.2 represents a phylogenetic tree of a sample of 28 vertebrate species,

including most domestic species (dog, cow, chicken, etc.) as well as species that
are commonly used in scientific research (mouse, frog, zebrafish, etc.). This evolu-
tionary tree is a useful tool for molecular biologists wishing to make comparisons
at the genomic level (sequence comparisons, origin of gene families, etc.). We will
frequently refer to it in this book.
© Springer-Verlag Berlin Heidelberg 2015 1

J.L. Guénet et al., Genetics of the Mouse, DOI 10.1007/978-3-662-44287-6_1
2 1 Origins of the Laboratory Mouse
Fig. 1.1 Some mouse mutations with effect on the coat color. Mice represented in this figure are
homozygous or heterozygous for mutations affecting coat color. Many phenotypes of this kind have
been collected over the years by fanciers and geneticists and are still for sale in many traditional
pet shops. Associated or not in the same individual, they have produced a large variety of beautiful
specimens that have captured the enthusiasm of many children. Even today, rare specimens are regu-
larly exchanged between members of many pet clubs. Coat color mutations, behavioral mutations,
and mutations affecting the fur or the skeleton were also used by mouse geneticists in the early days
as genetic markers for the detection of genetic linkage because they had little or no effect on viability
and fertility. The first linkage discovered in the mouse (linkage group I—now chromosome 7) was
between the coat color mutations pink-eyed dilution (p, now Oca2p) and chinchilla (cch, now Tyrc-ch)
Laboratory mice (together with rats and Guinea pigs) belong to the order
Rodentia, which is the largest group of mammals on earth, comprising around 40 %
of mammalian species (Fig. 1.3). Rabbits are not rodents sensu stricto but lago-
morphs1. However, due to their evolutionary proximity they are often merged with
the rodent family and together with them are referred to as the superclass Glires.
About two-thirds of rodent species belong to the superfamily Muroidea, a
superfamily that is itself composed of six families including the Muridae family,
which includes the “house” mouse Mus musculus2 and two species of rats, Rattus
norvegicus and Rattus rattus. This is a very large family of mammals with at least
1,300 species. Laboratory mice belong to the genus Mus that itself contains four
subgenera: Mus, Coelomys, Pyromys and Nannomys, and at least 40 different
1 There are two families in the order Lagomorpha: the Leporidae (hares and rabbits) and the
Ochotonidae (pikas).
2 Many rodent species carry the name “mouse”, meaning a mouse-like small furry creature,
hence the importance of the binominal nomenclature.

1.1 Introduction 3
Human
Chimpanzee
Macaque
Bushbaby
Tree shrew
Rat
Mouse
Guinea Pig
Rabbit
Shrew
Hedgehog
Dog
Cat
Horse
Cow
Armadillo
Elephant
Tenrec
Opossum
Platypus
Chicken
Lizard
Frog
Tetraodon
Fugu
Stickleback
Medaka
Zebrafish
Fig. 1.2 Phylogenetic tree representing the relationships between 28 vertebrate species. Branch

lengths are proportional to the number of base-pair substitutions at a number of specific sites.
The estimated time of divergence between humans and mice is approximately 75–80 Myr ago
(redrawn and modified from Blanga-Kanfi et al. 2009)
species. Figure 1.4 summarizes the phylogenetic relationships within the genus

Mus. Inside each subgenus the different species are, with few exceptions,
extremely similar in size and morphology.
Because of these similarities, the phylogenetic relationships between the differ-
ent species have been difficult to establish, especially when morphological char-
acteristics (tail length, body shape, coat color, habitat, etc.) were the only criteria
taken into account for the establishment of the systematics (Fig. 1.5). Nowadays,
with possible reference to the complete genomic sequence of many genes as well
as to the sequence of mitochondrial DNA, the situation has been much clarified.
The geographic distribution of the genus Mus encompasses all of Eurasia and
Africa. The presence elsewhere of a single of its species, the house mouse,
particularly in Australia and the Americas, results from human-mediated

introductions during recent centuries (Jones et al. 2013).3 Hence, the house mouse
3 The house mouse, which is now endemic in Australia and the Americas, was involuntarily
transported from Europe or from Asia by maritime traffic. Many genetic markers (endogenous cop-
ies of retroviruses inserted as proviral DNA, in particular) confirm the origin of these “stowaways”.
Fig. 1.3 Phylogenetic Sicista
relationships between Allactaga
32 species of rodents Dipus
representing 14 subfamilies
of the Muridae family. The Jaculus
estimated time of divergence Spalacinae
between the mouse and rat Rhizomyinae
species is approximately Macrotarsomys
12/15 Myr ago (redrawn from
Nesomys
Michaux et al. 2001)
Mystromys
Cricetomys
Saccostomus
Dendromus
Steatomys
Calomyscus
Clethrionomys
Dicrostonyx
Neotoma
Peromyscus
Cricetulus
Mesocricetus
Phodopus
Myospalax
Tatera
Gerbillus
Lophuromys
Deomys
Uranomys
Acomys
Rattus
Micromys
Otomys
Mus
is currently widely spread over the five continents, with the highest diversity in
Asia (with 3 subgenera and ~20 species), where this genus likely originated. Based
on recent observations, and if we consider that the habitat of some (still unknown)
species might be very limited, and possibly embedded in the wider habitat of other
species, it is likely that the number of species in the subgenus Mus will increase
further (Bonhomme et al. 2004).
The evolutionary divergence between humans (Homo sapiens) and mice of the Mus
genus probably occurred 70–75 million years (Myr) ago (Fig. 1.2) while the diver-
gence between humans and the other domesticated species (e.g., dog, cat, horse and
cow) is slightly greater (80–85 Myr) (Murphy et al. 2001). The divergence between
the Mus and Rattus genera probably occurred around 10–12 Myr ago. Finally,
1.1 Introduction 5
6 MYrs 5 4 3 2 1
domesticus
musculus
musculus
Mus
molossinus
castaneus
spretus
spicilegus
cypriacus
subgenus Mus
macedonicus
booduga
nitidulus
terricolor
fragilicauda
famulus
lepidoides
caroli
cooki
cervicolor
subgenera
Pyromys
other
Coelomys
Nannomys
Fig. 1.4 Consensus phylogenetic tree of the genus Mus issued from a compilation of all existing
studies. The estimated time of divergence of the different Mus species is indicated at the top of
the diagram
the individualization of the subgenus Mus sensu stricto occurred around 6 Myr ago
with the split from three other subgenera (Boursot et al. 1993; Musser and Carleton
1993; Chevret and Dobigny 2005; Chevret et al. 2005; Suzuki and Aplin 2012).
1.1.2 How the House Mouse Became a Domestic Species
The beginning of human/mouse commensalism is very ancient and probably dates

to some 12,000 years ago. Mouse remnants and paintings have been discovered in
the very large Neolithic settlement of Çatal Hüyük (southern Anatolia, Turkey),
which existed from approximately 7,500 to 5,700 B.C., suggesting that, during
this period, mice were considered (worshipped?) as holy creatures or living sym-
bols. Finally, numerous historical records (Keeler 1931; Morse 1978; Berry 1987;
Sage et al. 1993; Moriwaki et al. 1994) indicate that 3,000 years ago mice were
pet animals in Europe, Japan, and China.4
4 InJapanese traditional writing there is only one Kanji to define both rats and mice: nezumi.
This is a possible source of confusion.
1.1 Introduction 7
t Fig. 1.5 Some specimens of the order Rodentia. This panel represents eight specimens of the
order Rodentia. In spite of great similarities in size and body shape, some of these “mice” are
only weakly related species. Mus m. castaneus (b) and Mus spretus (c) can produce viable
and fertile hybrids with mice of the Mus m. domesticus species (a) or with laboratory strains.
Interspecific hybrids resulting from crosses between Mus spretus males and laboratory females
are fertile but only in the female sex (Haldane’s rule), and this sort of cross has been used exten-
sively for the development of the mouse genetic map. The reciprocal cross (laboratory males ×
Mus spretus females) is much less fertile and produces hybrids only in special conditions. The
possibility of obtaining hybrids between Mus cypriaticus (d) and laboratory strains has not yet
been tested. Hybrids generated by the artificial fertilization of laboratory females with sperm of
Mus caroli (e) complete fetal development, and a low percentage of them survive to maturity
but are stunted and do not reproduce. Embryonic cells of Mus caroli can participate in the for-
mation and development of a chimeric fetus when associated with cells of a laboratory inbred
strain. Hybrids between Coelomys pahari (f) and laboratory strains have never been produced
and would presumably not be viable. Rodents of the Arvicanthis ansorgei species, also known
as the Sudanian grass rat (g), are endemic in West African countries and do not produce hybrids
with mice of the genus Mus. Rodents of this species, unlike the other rodents presented here,
have essentially diurnal activity. Finally, Calomys callosus, the large vesper mouse, is a South
American rodent of the family Cricetidae. Despite their similarities to the other mice represented
in the picture, which all are of the family Muridea, these rodents are phylogenetically closer
to hamsters (Cricetulus griseus) and deer mice (Peromyscus maniculatus) than to mice of the
genus Mus. Several of these species and subspecies have been established as laboratory colo-
nies. One of the most diverse collections of wild-derived strains can be found at the Université
de Montpellier, Place Eugène Bataillon, France, c/o Dr. François Bonhomme. Six pictures in
this panel (a–f) are from the wild rodent repository of Dr. François Bonhomme. The picture of
Arvicanthis ansorgei (g) is from Dr. Sophie Reibel-Foisset, (Chronobiotron, Strasbourg, France).
The picture of Callomys callosus (h) is from Dr. Adriano Abbud (Instituto Adolfo Lutz, São
Paulo, Brazil)
All this evidence indicates that mice and humans have been in contact for a
very long time. It was then logical that these small mammals, as well as the rat,
were used by early scientists for performing their experiments, and if this choice
appears nowadays to be more opportunistic rather than based on scientific
considerations, it nevertheless appeared to be an excellent one in the context of
modern biomedical research.
Mice are easy to breed. As they are rodents, they eat a rather large quantity of
food but do not have very specific or expensive nutritional requirements. When
kept in laboratory facilities with stable environmental conditions (light and
temperature), they do not hibernate (meaning that they have a decreased physi-
ological activity) and breed all year round, with a short generation time. They
deliver relatively large progenies and tolerate inbreeding rather well compared to
other mammalian species. For all these reasons, but also because some ancestral
specimens were tame and easy to handle, mice have been used in biomedical
research since the beginning of the sixteenth century, when biology gradually
shifted from a descriptive to an experimental science. Herbert C. Morse (1978,
1981) reported that William Harvey (1578–1657) used mice for his fundamental
studies on reproduction and blood circulation while according to Richard J. Berry
(1981), the earliest record of the use of mice in scientific research seems to have
been in England, in 1664, when Robert Hooke (1635–1703) used mice to study the
biological consequences of an increase in air pressure. Much later, Joseph Priestley

(1733–1804) and his intellectual successor, Antoine Lavoisier (1743–1794), both
used mice repeatedly in their experiments on respiration. Subsequently, an ever-
increasing number of scientists used mice in their experiments, at least as long as
the small size of the rodents was compatible with the experimental project. This is
how the house mouse progressively and logically became “the laboratory mouse”.
1.1.3 How the House Mouse Became a Model for Geneticists
1.1.3.1 From Louis-Théodore Coladon to Gregor Mendel and Lucien

Cuénot
Over the past two centuries, fanciers in Europe, in Japan, and in the United States
were regularly breeding and exchanging pet mice with a wide variety of coat color
or amusing behavior (e.g., the famous “dancing mice” homozygous for the Cdh23v
or waltzer mutant allele). All these mice were crossed to produce new eye-catch-
ing phenotypes to supply the market, and it progressively became obvious that
many of these traits were inherited. The mouse then appeared to be an excellent
model for studying inheritance!
According to Hans Grüneberg (1952) and Jean Rostand (1957), some among
these fanciers played an important role. Louis-Théodore Coladon (sometimes
spelled Colladon; 1792–1862), a pharmacist established in Geneva, was prob-
ably the first to report the results from breeding experiments achieved between
1825 and 1829, which were in perfect agreement with what we now call
the Mendelian ratios. This, however, was 36 years before the publication of
Mendel’s own results on peas. The experimental results of Coladon have never
been published but were merely quoted by Jean-Baptiste Dumas, the famous
chemist, in the Dictionnaire classique d’Histoire naturelle (tome 7, p. 202)
and by William F. Edwards, a physiologist and ethnologist, in a book published
in 1829.
As revealed by Hugo Iltis (1932) and commented on by Kenneth Paigen in his
notes on the history of mouse genetics (2003a, b), we know that Gregor Mendel
also bred mice (grey and white mice) in his monastic room. These mice were bred
as pets, but it is likely that Mendel (who later proved to be a sagacious observer!)
had his attention focused on the segregation of coat color in the progenies and may
have had sufficient experimental data to make observations on the transmission of
these characteristics. However, Mendel never commented on these observations
and was asked by his hierarchy to stop breeding mice. At the time it was positively
indecent and even immoral to perform experiments dealing with animal reproduc-
tion (mating) and inheritance, particularly in a monastery. Accordingly, Mendel
changed his experimental model to garden peas and published his famous obser-
vations in 1866 in a botanical journal, where they had a rather low impact and
remained virtually ignored until the beginning of the twentieth century.
1.1 Introduction 9
Once rediscovered by H. de Vries, C. E. Correns, and E. von Tschermark-

Seysenegg, the three of them working independently with plants, it was tempting
to test whether the so-called Mendel’s laws were also valid for animals.5 Lucien
Cuénot (1902), a professor of biology at the University of Nancy (France),
crossed mice segregating for several common coat color markers including the
albino (Tyrc) and published the results of experiments indicating that this was
indeed the case. Cuénot’s observations were shortly confirmed and extended to
other species, as well as for other genetic traits by W. Bateson, E. R. Saunders,
A. Garrod, W. E. Castle and C. C. Little (Paigen 2003a, b). Cuénot was also able
to interpret correctly the unusual pattern of transmission (1/3–2/3 instead of the
classical 1/4–3/4) of the yellow dominant allele (Ay) at the Agouti locus
(chromosome 2 [Chr 2]), suggesting that Ay/Ay embryos died in utero at an early
stage of development, and accordingly that yellow mice (Ay/A) could not “breed
true” (Cuénot 1905).
1.1.3.2 The Origins of Laboratory Mouse Strains
The majority of albino mouse strains used today in experimental research are
derived from ancestral breeders bought in pet shops, which were bred either
by the researchers themselves or by amateurs as a source of income. For many
years, and even today, many of these albino mice bred for general purpose in
laboratories, were collectively designated “Swiss” mice to recall their Helvetian
origin (perhaps they were indeed distantly related offspring of Coladon’s
mice?). These mice were bred with no specific mating protocol, and the only
criteria for selecting the breeders, generation after generation, were docility
and good health. The breeding colonies were regularly decimated by outbreaks
of infectious diseases or sometimes reduced to a few breeding pairs as a con-
sequence of a lack of space (or of funding!). A consequence of this “bottle-
neck effect” was that the mice became progressively (and insidiously) inbred.
However, strict inbreeding was absolutely avoided based on the negative experi-
ence of livestock and dog breeders.
Strain DBA/2 (formerly dba, then DBA) is the most ancient of all inbred
strains. It was started by Clarence C. Little in 1909 (Russell 1978) by intercross-
ing mice homozygous for the coat color markers non-agouti (a), brown (for-
merly b, now Tyrp1b) and dilute (formerly d, now Myo5ad). About 10 years later,
Miss Abbie Lathrop of Granby, a retired school teacher from Massachusetts
(USA), established strain C57BL/6 by intercrossing the “black” offspring of
female 57 (Strong 1978). According to several historical records, Miss Lathrop
played an important role in the development of laboratory strains because she
was keeping excellent records of the pedigrees of her strains. In collaboration
with researchers on the East coast of the United States (in particular, Leo Loeb,
5 For
an interesting historical account, refer to The Monk in the Garden: The Lost and Found
Genius of Gregor Mendel, the Father of Genetics, by Robin Marantz Henig (2001).
a pathologist working at the University of Pennsylvania), she made interesting

observations about the occurrence of specific cancers. Miss Lathrop was well
aware of the importance of breeding top-quality mouse strains for the progress
of science, and her reputation was so good that the US government made an
appeal to her to supply mice (and guinea pigs) to research laboratories (Shimkin
1975; Steensma et al. 2010). Being close to the Bussey Institute at Harvard
University, she also collaborated with William Castle (considered the father of
mammalian genetics) and Clarence C. Little.
Strains C3H, CBA, and A were created during the same time period by Leonell
C. Strong, a cancer geneticist established at Cold Spring Harbor Laboratory
(Strong 1978). At this point it is interesting to note that, among the strains estab-
lished by Leonell C. Strong, strains CBA and C3H stemmed from the offspring of
an outcross with wild specimens trapped in a pigeon coop in Cold Spring Harbor.
This probably explains how the wild-type allele at the agouti locus (A) was rein-
troduced into laboratory strains.
With a few exceptions, historical records concerning the genealogy of most
laboratory inbred strains are well documented, and several interesting reviews
on this subject are available (Strong 1978; Festing 1979; Rader 2004; Artzt
2012). A chart describing the genealogy of these strains, including the recently
established ones, has been published, and regularly updated information is
available from The Jackson Laboratory website (Beck et al. 2000). In addition
to the chart published by Beck and co-workers, which was based mostly on
historical records, a mouse “family tree” was also published by Petkov and co-
workers (2004), which is based on a set of 1,638 informative single nucleotide
polymorphism (SNP) markers (see Chaps. 4 and 5), located 1.5 Mb apart and
tested in 102 mouse strains. These family trees have been documented further
and greatly enriched, and have become invaluable tools for researchers who
are willing to make interstrain comparisons because they make it possible to
select pairs of strains that are more or less distantly related before c omparing
specific phenotypic traits (Yang et al. 2007; Szatkiewicz et al. 2008). This
is extremely important, for example, for the analysis of quantitative traits
(see Chap. 10).
1.1.3.3 Mice Have Been Instrumental in Research in Biology

and Genetics
Laboratory mice have been at the origin of many important discoveries in biol-
ogy. To cite just a few, we could say that our understanding of the genetic deter-
minism underlying the success or failure of tissue transplantations is a direct
consequence of experiments performed with inbred mouse strains by Peter Gorer
(1948), then by George D. Snell and co-workers (1978). These researchers devel-
oped a series of congenic resistant strains that were all genetically identical to
the C57BL/10Sn background strain, with the exception of single short chromo-
somal regions determining graft rejection. These very clever experiments led to
1.1 Introduction 11
the establishment of the so-called “laws of transplantation” and opened the way
to what has become known as Immunogenetics. For this discovery “concerning
genetically determined structures on the cell surface that regulate immunologi-
cal reactions”, George D. Snell, from the Jackson Laboratory, was awarded the
Nobel Prize in Physiology or Medicine in 1980, jointly with Professors Baruj
Benacerraf and Jean Dausset.
The hypothesis, proposed by Mary F. Lyon, that one X-chromosome out of two
was inactivated in female mammals followed from the observation of variegations
in the coat color for some X-linked mouse mutations and was interpreted by using
X-autosome translocations (Lyon 1961). Chimeric organisms were produced for
the first time by A.K. Tarkowski in Warsaw (1961) and B. Mintz in Philadelphia
(1962) by merging in vitro independent mouse embryonic cells.6 The testicular
terato-carcinomas, which are common in strain 129, and the cell lines derived
from these tumors and cultivated in vitro, have been a material of choice for inves-
tigating the processes at work in tissue differentiation for almost a decade (Stevens
and Little 1954; Stevens 1970; Jacob 1983). This work undoubtedly opened the
way to the establishment of so-called embryonic stem cells (ES cells) by Evans
and Kaufman (1981) and Martin (1981). These ES cells paved the way for the
“discoveries of the principles for introducing specific gene modifications in mice”,
for which M.R. Capecchi, M.J. Evans and O. Smithies were awarded the Nobel
Prize in Physiology or Medicine in 2007.
The discovery of parental imprinting of some chromosomal regions was a con-
sequence of experiments performed by McGrath and Solter (1984) and Surani
and co-workers (1984), who demonstrated that a normal mouse embryo can only
develop from the fusion of a male and a female pronucleus, while Cattanach and
Kirk (1985) demonstrated that the parental origin of the two elements of a given
chromosome pair was not always genetically equivalent.
The first transgenic mammal created by pronuclear injection of cloned DNA
was a mouse (Gordon et al. 1980), as was the first mammalian organism geneti-
cally engineered in vitro (Kuehn et al. 1987). Only the first cloned mammal was
not a mouse, but this type of uniparental procreation has been achieved in the
mouse, although the efficiency of the procedure is very low, like in other mammals
(Wakayama and Yanagimachi 1999). The first mammal whose genome was com-
pletely sequenced was a mouse of the C57BL/6 inbred strain (Waterston et al.
2002).7 Finally, and to cite just another example among many others, we could say
that the discoveries made by Bruce Beutler about innate immunity, for which he
was awarded the Nobel Prize in 2011, were made possible by the existence of a
large number of mutations induced in the mouse genome by the chemical mutagen
Ethyl-Nitroso-Urea.
6 In the 1970s, these chimeric mice were sometimes called allophenic to recall their origin.
7 A draft sequence of the human genome was published 2 years (2000) before the draft sequence
of the mouse (2002), but the human sequence still has some gaps while the mouse sequence is
99.5 % complete.
1.1.4 The Community of Mouse Geneticists
As is often the case when independent researchers use the same experimental
“material” and the same logistics, a community of mouse geneticists formed over
the years at the international level with the mouse as a common denominator. The
community had its own journal called Mouse News Letters and its own meetings
organized at various places in the Northern hemisphere, alternately in Europe, in
the USA and sometimes in Japan.
Mouse News Letters, first issued in 1949, was published regularly every semes-
ter until 1997 (95 issues). This informal publication, edited by scientists from
the Medical Research Council (first at Edinburgh, then at Harwell), was distrib-
uted free of charge worldwide for several decades and was the best medium for
the dissemination of information among the community. The name Mouse News
Letters was changed to Mouse Genome in 1990, when this publication became a
peer-reviewed journal. Finally, in 1998, Mouse Genome merged with Mammalian
Genome—edited and published by Springer Verlag.
Mouse News Letters will forever remain the best place to find information
about the history of mouse genetics, and in particular about the history of most
traditional inbred strains, the progressive development and refinement of the link-
age map, and the discovery and initial description of hundreds of spontaneous
mutations. The scientific content of the successive issues of Mouse News Letters
will never be obsolete. On the contrary, it is the “memory” of the early days of
mouse genetics.
1.1.5 The Main Institutions Involved in Mouse Genetics
The Jackson Laboratory (internationally known as the JAX-Lab), which was

founded in 1929 by C.C. Little in Bar Harbor (Maine, USA), has played a cen-
tral role in the promotion of the mouse as a laboratory model and still is the
world’s largest center for mouse genetics. A second “JAX-Lab” opened recently
in Sacramento, California. The Jackson Laboratory is a non-profit organization
entirely and exclusively dedicated to basic research on mouse models of human
diseases. Its mission is to discover the genetic basis for preventing and treating
human disease, and to enable research and education for the global biomedical
community. It is, at the same time, a research institution, a meeting place where
courses and conferences are organized on various aspects of mouse genetics, and
the world’s largest genetic repository where a great variety of genotypes (around
6,000) and biological samples of all kinds are stored and distributed to the scien-
tific community either as living animals or, in most instances, in the form of frozen
embryos or sperm cells. The Jackson Laboratory is also the home of the Mouse
Genome Informatics database (MGI at http://www.informatics.jax.org/), an essen-
tial tool for mammalian geneticists.
1.1 Introduction 13
Several other institutions, like the Oak Ridge National Laboratory in Tennessee
(USA) and the MRC centre at Harwell in England, have also played (and still
play) a very important role in the development of the mouse as a laboratory model
for research in genetics, oncology, and immunology. These two centers were
founded after World War II when the British and US governments decided to eval-
uate the genetic hazards that might be associated with the use of radiation, and,
more generally, of nuclear energy. Thousands of mice have been experimentally
irradiated in these centers to assess the genetic damage of various types of radia-
tion distributed at various doses. Accordingly, a very large number of mutations
and chromosomal rearrangements have been induced, collected, and preserved.
Both the mutations and the chromosomal rearrangements have been invaluable
tools for the establishment of the mouse genetic map. Many of them are also
interesting models of human diseases.
Other research centers must also be mentioned for their contribution to the
development of mouse genetics in the second half of the twentieth century: the
MRC centre in Edinburgh, Scotland, the Deutsches Forschungszentrum für
Gesundheit und Umwelt, at Neuherberg, Germany (now Helmholtz Zentrum
München), and the Institute of Genetics at Mishima in Japan.
More recently, the European Union has decided to support the establish-
ment of a wide network of genetic repositories (the so-called European Mouse
Mutant Archive or EMMA), with major nodes in Italy (EMMA headquarters is
in Monteretondo, near Rome), England (Harwell), France (Orléans-la-Source),
Germany (Munich), and Spain (Madrid). Finally, and even more recently, Japanese
scientists have created and implemented a large bioresource centre at the RIKEN
Institute in Tsukuba, with teaching and research activities focused on mouse
embryology and genetics. More information about all these centers is available on
their websites.
1.1.6 Books and Other Sources of Information Concerning

the Mouse
Readers who are interested in the history of mouse genetics are invited to consult
the following books, which are available online at the MGI website.
• Biology of the Laboratory Mouse edited by Earl L. Green—Dover Publications
1966
• Origins of Inbred Mice edited by Herbert C. Morse III—Academic Press 1978
• Mouse Genetics—Principles and Applications by Lee Silver—Oxford Press
1995
Some parts of the book Biology of the Laboratory Mouse are obsolete, but many
others are still a rich source of information with many references. This book is an
excellent textbook for all issues related to linkage and gene mapping.
Four other books are also an important source of information from a historical
point of view:
• Genetics of the Mouse by Grüneberg—Martinus Nijhoff 1952
• Inbred Strains In Biomedical Research by Festing—Macmillan Press 1979
• Biology of The House Mouse Edited by R. J. Berry—Academic Press 1982
• Making Mice: Standardizing Animals for American Biomedical Research,
1900–1955 by Rader—Princeton University Press 2004.
Finally, a number of Websites are commonly used by mouse geneticists. The fol-
lowing list is not intended to be comprehensive and additional URLs will be given
in the subsequent chapters:
• Emouseatlas (http://www.emouseatlas.org/emap/home.html) encompasses a 3-D
anatomical atlas of mouse embryo development and a database of mouse gene
expression.
• Pathbase provides a searchable database of histopathology images derived from
experimental manipulation of the mouse genome or experiments conducted on
genetically manipulated mice. It is a reference/didactic resource covering all
aspects of mouse pathology.
• e!Ensembl (http://www.ensembl.org/Mus_musculus/Info/Index) produces a
genome database for the mouse and makes this information freely available
online.
• The International knockout consortium (http://www.knockoutmouse.org/)
provides information on conditionally trapped and targeted genes in mouse
embryonic stem (ES) cells.
• MouseMine (http://www.mousemine.org/mousemine/begin.do) is a powerful
system for online access to mouse data from Mouse Genome Informatics.
In the UK, the Sanger Institute Mouse Genetics Project has recently launched the
Mouse Resources Portal (http://www.sanger.ac.uk/mouseportal/) with extensive
genotyping and phenotyping resources.
1.1.7 The Future of Mouse Genetics
The state of mouse genetics, as expected, changed dramatically at the turn of

the millennium for two main reasons. First, because the complete sequence of
the genome was established and immediately made available to the community,
making possible all sorts of comparisons and predictions at the genome level. In
several chapters of this book we will discuss the consequences of this unprece-
dented achievement and the projects that followed from it. The second reason is
that geneticists have now, at their disposition, all the tools and strategies allow-
ing them to make, almost at will, any type of alteration in the genome, from large
segmental deletions or additions to single base-pair substitutions. Enthralling
projects planned for the years to come have a new dimension, and the “mouse
1.1 Introduction 15
community” now involves virtually all biologists on the planet using mammals in
their research. Many projects of great importance will be undertaken in the future,
in particular for understanding the determinism of complex traits. For these pro-
jects, no species can seriously compete with the mouse, and this is why we predict
a promising future for mouse genetics.
Acknowledgements The authors thank Doctor François Bonhomme, Université de
Montpellier, France for his contribution to this chapter as well as for Fig. 1.4.
References
Artzt K (2012) Mammalian developmental genetics in the twentieth century. Genetics

192:1151–1163
Beck JA, Lloyd S, Hafezparast M, Lennon-Pierce M, Eppig JT, Festing MF, Fisher EM (2000)
Genealogies of mouse inbred strains. Nat Genet 24:23–25
Berry RJ (1981) Biology of the house mouse. Academic Press, London
Berry RJ (1987) The house mouse. Biologist 34:177–186
Blanga-Kanfi S, Miranda H, Penn O, Pupko T, DeBry RW, Huchon D (2009) Rodent phylogeny
revised: analysis of six nuclear genes from all major rodent clades. BMC Evol Biol 9:71
Bonhomme F, Orth A, Cucchi T, Hadjisterkotis E, Vigne JD, Auffray JC (2004) Découverte
d’une nouvelle espèce de souris sur l’île de Chypre. C R Biol 327:501–507
Boursot P, Auffray JC, Britton-Davidian J, Bonhomme F (1993) The evolution of house mice.
Annu Rev Ecol Syst 24:119–152
Cattanach BM, Kirk M (1985) Differential activity of maternally and paternally derived chromo-
some regions in mice. Nature 315:496–498
Chevret P, Dobigny G (2005) Systematics and evolution of the subfamily Gerbillinae
(Mammalia, Rodentia, Muridae). Mol Phylogenet Evol 35:674–688
Chevret P, Veyrunes F, Britton-Davidian J (2005) Molecular phylogeny of the genus Mus
(Rodentia: Murinae) based on mitochondrial and nuclear data. Biol J Linn Soc 84:417–427
Cuénot L (1902) La loi de Mendel et l’hérédité de la pigmentation chez les souris. Arch Zool exp
gén 3e séries 3:xxvii-xxx
Cuénot L (1905) Les races pures et leurs combinaisons chez les souris (4éme note) Arch Zool exp
gén 4e séries 3:cxxiii-cxxxii
Evans M, Kaufman M (1981) Establishment in culture of pluripotent cells from mouse embryos.
Nature 292:154–156
Festing MF (1979) Inbred strains in biomedical research. The MacMillan Press Ltd, London
Gordon JW, Scangos GA, Plotkin DJ, Barbosa JA, Ruddle FH (1980) Genetic transformation of
mouse embryos by microinjection of purified DNA. Proc Natl Acad Sci U S A 77:7380–7384
Peter Gorer A (1948) The significance of studies with transplanted tumours. Br J Cancer 2:103–107
Grüneberg H (1952) The genetics of the mouse (2nd edn). Martinus Nijhoff, The Hague
Henig RM (2001) The monk in the garden: the lost and found genius of Gregor Mendel, the
father of genetics. Mariner Books
Iltis H (1932) Life of mendel. George Allen, London, p 105
Jacob F (1983) Expression of embryonic characters by malignant cells. Ciba Found Symp 96:4–27
Jones EP, Eager HM, Gabriel SI, Jóhannesdóttir F, Searle JB (2013) Genetic tracking of mice and
other bioproxies to infer human history. Trends Genet 29:298–308
Keeler CE (1931) The Laboratory Mouse: its origin, heredity, and culture. Harvard University
Press, Cambridge, 81 p
Kuehn MR, Bradley A, Robertson EJ, Evans MJ (1987) A potential animal model for Lesch-
Nyhan syndrome through introduction of HPRT mutations into mice. Nature 326:295–298
Lyon MF (1961) Gene action in the X-chromosome of the mouse (Mus musculus L.). Nature
190:372–373
Martin G (1981) Isolation of a pluripotent cell line from early mouse embryos cultured in
medium conditioned by teratocarcinoma stem cells. Proc Natl Acad Sci USA 78:7634–7638
McGrath J, Solter D (1984) Completion of mouse embryogenesis requires both the maternal and
paternal genomes. Cell 37:179–183
Michaux J, Reyes A, Catzeflis F (2001) Evolutionary history of the most speciose mammals:
molecular phylogeny of muroid rodents. Mol Biol Evol 18:2017–2031
Mintz B (1962) Formation of genotypically mosaic mouse embryos. Am Zool 2:432
Moriwaki K, Shiroishi T, Yonekawa H (1994) Genetics in wild mice: its application to biomedi-
cal research. Japan Scientific Societies Press, Tokyo
Morse HC 3rd (1978) Origins of inbred mice. Academic Press, New York
Morse HC, 3rd (1981) The laboratory mouse—a historical perspective. In: Foster HL, Small JD,
Fox JG (eds) The mouse in biomedical research, vol 1. Academic Press, New York, pp. 1–16
Murphy WJ, Eizirik E, Johnson WE, Zhang YP, Ryder OA, O’Brien SJ (2001) Molecular phylo-
genetics and the origins of placental mammals. Nature 409:614–618
Musser GG, Carleton MD (1993) Family muridae. In: Wilson DE, Reeder DM (eds) Mammalian
species of the world, 2nd edn. Smithsonian Institution Press, Washington, pp 501–755
Paigen K (2003a) One hundred years of mouse genetics: an intellectual history. I. The classical
period (1902–1980). Genetics 163:1–7
Paigen K (2003b) One hundred years of mouse genetics: an intellectual history. II. The molecular
revolution (1981–2002). Genetics 163:1227–1235
Petkov PM, Ding Y, Cassell MA, Zhang W, Wagner G, Sargent EE, Asquith S, Crew V, Johnson
KA, Robinson P, Scott VE, Wiles MV (2004) An efficient SNP system for mouse genome
scanning and elucidating strain relationships. Genome Res 14:1806–1811
Rader K (2004) Making mice: standardizing animals for American Biomedical Research, 1900–
1955. Princeton University Press, New Jersey
Rostand J (1957) Un précurseur de Mendel: le pharmacien Coladon. C R Hebd Seances Acad Sci
244:2973–2974
Russell ES (1978) Origins and history of mouse inbred strains: contributions of clarence cook
little. In: Morse HC 3rd (ed) Origins of inbred mice. Academic Press, New York, pp 45–68
Sage RD, Atchley WR, Capanna E (1993) House mice as models in systematic biology. Syst Biol
42:523–561
Shimkin MB (1975) A. E. C. Lathrop (1868–1918) mouse woman of Granby. Cancer Res
35:1597–1598
Snell GD (1978) Congenic resistant strains of mice. In: Morse HC 3rd (ed) Origins of inbred
mice. Academic Press, New York, pp 119–156
Steensma DP, Kyle RA, Shampo MA (2010) Abbie Lathrop, the “mouse woman of granby”:
Rodent Fancier and accidental genetics pioneer. Mayo Clin Proc 85:e83. doi:10.4065/
mcp.2010.0647
Stevens LC (1970) The development of transplantable teratocarcinomas from intratesticular
grafts of pre- and postimplantation mouse embryos. Dev Biol 21:364–382
Stevens LC, Little CC (1954) Spontaneous testicular teratomas in an inbred strain of mice. Proc
Natl Acad Sci USA 40:1080–1087
Strong LC (1978) Inbred mice in science. In: Morse HC 3rd (ed) Origins of inbred mice.
Academic Press, New York, pp 69–75
Surani MA, Barton SC, Norris ML (1984) Development of reconstituted mouse eggs suggests
imprinting of the genome during gametogenesis. Nature 308:548–550
Suzuki H, Aplin KP (2012) Phylogeny and biogeography of the genus Mus in Eurasia. In:
Macholán M, Baird SJE, Munclinger P, Piálek L (eds.), Evolution of the house mouse.
Cambridge studies in morphology and molecules: new paradigms in evolutionary biology.
Cambridge University Press, Cambridge, pp 35–64
References 17
Szatkiewicz JP, Beane GL, Ding Y, Hutchins L, Pardo-Manuel de Villena F, Churchill GA (2008)
An imputed genotype resource for the laboratory mouse. Mamm Genome 19:199–208
Tarkowski AK (1961) Mouse chimaeras developed from fused eggs. Nature 190:857–860
Wakayama T, Yanagimachi R (1999) Cloning of male mice from adult tail-tip cells. Nat Genet
22:1217–1218
Waterston RH, Lindblad-Toh K, Birney E, Rogers J, Abril JF, Agarwal P, Agarwala R et al (2002)
Initial sequencing and comparative analysis of the mouse genome. Nature 420:520–562
Yang H, Bell TA, Churchill GA, Pardo-Manuel de Villena F (2007) On the subspecific origin of
the laboratory mouse. Nat Genet 39:1100–1107
Chapter 2
Basic Concepts of Reproductive Biology and
Genetics
2.1 Introduction
This chapter brings together a variety of information and concepts that are impor-
tant for understanding the following chapters. The first section is an overview con-
cerning mouse reproductive biology and embryology. This topic is important
because, nowadays, many experiments in genetics require the manipulation of
embryos at different stages of development, either to study their phenotype or for
the production of chimeras with other embryos or with genetically engineered
embryonic stem (ES) cells. The second part is a compilation of concepts of gen-
eral or molecular genetics related to the phenotypic expression of mutations. More
information can also be retrieved from several websites, where books and manuals
are freely available online.1
2.2 Reproduction in the Laboratory Mouse
2.2.1 The Estrous Cycle and Pregnancy
Laboratory mice are polyestrous mammals. This means that, provided they are
raised and housed in a suitable environment, the animals can reproduce all year
round with only a small decline in fertility during the winter season.2 In females,
sexual maturity (puberty) takes place gradually from the age of 3–4 weeks. The
vaginal orifice, which is normally sealed at birth by an epithelial operculum, opens
1 The website http://informatics.jax.org/ is a fundamental database resource for the laboratory
mouse, providing integrated genetic, genomic, and biological data. It is a true “gold mine” for
mouse geneticists to which we will frequently refer. Several books dealing with some fundamen-
tal aspects of mouse biology are freely available at this website.
2 The reproductive activity of wild mice is interrupted or reduced during winter. This period is
called anestrus.

20 2 Basic Concepts of Reproductive Biology and Genetics
between 25 and 40 days. From 6 to 8 weeks after birth, and depending on the
strain, ovulation starts, and, in principle, all females older than 8 weeks are able to
reproduce, exhibiting a typical cyclic sexual activity. Male puberty occurs slightly
earlier, sometimes as early as 5 weeks, usually at 6–8 weeks.
The female reproductive cycle, the estrous cycle, lasts 4–6 days and is arbitrar-
ily divided into four stages with the following order: proestrus, estrus, metestrus,
and diestrus.3 Proestrus and metestrus last about one day each, while the estrous
period lasts only 12–16 h. Diestrus is the last and longest stage of the estrous cycle
(~2 days).
Based on vaginal cytology, embryologists have defined criteria that character-
ize the four stages of the mouse estrous cycle (Byers et al. 2012). According to
these criteria the estrus period is characterized by the presence of many flat and
keratinized epithelial cells that are obvious upon examination of vaginal swabs.
These cells are eosinophilic, meaning that they are stained deep red by the dye
eosin. These visible changes during the estrous cycle reflect the variations in pro-
gesterone and estrogen levels. Female mice copulate only during the estrous
period, which is often designated the “heat period” by analogy with the sexual
behavior of other domestic females. The heat period lasts about 12 h and mating
generally occurs during the first half of the night. In mice, matings are uncom-
mon during the day.4
By using the above-described cytological parameters it is possible to identify
and sort out the female mice that are in the estrous phase of the cycle, and, accord-
ingly, that are hormonally prepared to copulate. However, this procedure is tedi-
ous and labor-intensive, especially when many females are to be selected, and for
this reason it is not used very much. In practice, researchers prefer to select the
female mice that are in the best conditions to mate by examining the external vagi-
nal morphology (Byers et al. 2012). In this case, the vulva is slightly swelled and
the vagina is slightly open. This kind of selection requires some experience but it
is fast, quite reliable, and has the enormous advantage of not stressing the mice in
a critical period.
The proestrus and estrus phases of the cycle are often designated the follicu-
lar phase because it is at the end of this phase that a batch of mature oocytes is
released from the ovarian follicles. This generally occurs during or immediately
after copulation, but copulation is not a prerequisite for this to occur because mice
are spontaneous ovulators. If males are not present in the cage, ovulation will still
normally occur during estrus.
Shortly after copulation, the fluids secreted by the various sexual glands of the
males (in particular the seminal vesicles and the coagulating glands), which are
components of the male’s ejaculate, coagulate to form a vaginal plug. The plug in
3 Estrus, sometimes spelled oestrus (UK), is a noun; estrous (oestrous) is the corresponding
adjective.
4 For some precisely timed pregnancies, female mice must sometimes be bred in a “light-
reversed” environment.
2.2 Reproduction in the Laboratory Mouse 21
question tightly seals the vaginal lumen and prevents any further mating.5 The
vaginal plug is a relatively hard substance and remains in the female’s vagina for
several hours (up to 6–8 h or even more). During this time the vaginal plug pro-
gressively resorbs and the spermatozoa are released. Detection of a vaginal plug
means that mating occurred during the preceding hours, but does not guarantee
that pregnancy will ensue.6
By analogy with the follicular phase, metestrus and diestrus constitute the
luteal phase. During this phase the corpus luteum forms and replaces the follicle.
The corpora lutea secrete the hormone progesterone, the hormone of pregnancy,
and persist until the end of pregnancy—if pregnancy ensues. If not, the corpora
lutea degenerate and a new cycle starts. Corpora lutea are easy to recognize at the
surface of the mouse ovary because they are slightly protuberant and often stained
light orange. After fixation with formalin or Bouin’s fixative, their identification is
even easier.
When virgin or non-pregnant females are housed in groups and mated with
males without prior selection of the phase of the estrous cycle, the frequency of
natural mating is not evenly distributed over the following nights. On the contrary,
one generally observes a peak after the third night of mating, indicating that some
synchronization of the estrous cycle occurred. Synchronization of the estrous
cycle by the presence of a male has been reported and is called the Whitten effect
(Whitten 1956). It is a consequence of the dispersion in the environment of vola-
tile pheromones that are at high concentration in the urine of males; these phero-
mones interfere with the hormonal control of the female cycle.
Fertilization of the oocytes takes place 10–15 h after ovulation, in the upper
segment of the female reproductive tract, more precisely during their transit
through the Fallopian tubes or oviducts (sometimes called ampulla). When the
head of a sperm cell succeeds in penetrating the oocyte after passing through the
zona pellucida (also designated oolemma), the penetration of other sperm cells is
blocked and this triggers the completion of the second meiotic division. The sec-
ond polar body from the oocyte is ejected within two hours; the male pronucleus
expands, and finally the two haploid pronuclei (male and female) fuse, and the
oocyte becomes an egg (i.e., a diploid embryo that is not yet implanted).
Segmentation in the embryo begins slowly at first. 68–72 h after fertilization (i.e.,
at the beginning of the 4th day after mating), the embryos enter the uterus and
implant into the uterine wall at the late or expanded blastocyst stage.
5 Such a vaginal plug is specific to the Mus genus and does not exist, for example, in the rat.
Whether it confers a selective advantage to the species is an open question.
6 As mentioned, most matings occur during the night; this is why “plugging” must be achieved
preferably during the morning of the following day. Detection of a plug is sometimes very easy,
especially when it bulges out of the vagina. In other instances, a probe may be necessary to
detect resistance when gently inserted into the vagina. The type of probe used by ophthalmolo-
gists to unclog the tear ducts of human patients is a perfect tool for this task.
Embryologists date the different stages of pregnancy from the day the vaginal plug
is discovered—i.e., day E0.5 by convention.7
Starting at 12–14 days of gestation, it is possible to detect the fetuses implanted
inside the uterus, which feel like “rosary beads” to the touch. To do this, the
female must be held firmly by the skin of its neck and back, with its abdomen
overturned, and gently palpated with the fingers of the other hand once the abdom-
inal wall is relaxed. Around 12 days of gestation, the pregnant females start to gain
weight and will soon show abdominal bulging; this can be another way to confirm
pregnancy by comparison with age-matched non-pregnant females.
Matching the number of corpora lutea with the actual number of fetuses
implanted in the uterine horns allows one to compute the number of conceptuses
that were possibly lost before implantation. This may be important, for example,
when an embryonic lethal mutation is suspected to be responsible for the reduc-
tion in the size of the progeny. In normal conditions, the number of corpora lutea,
which can be counted directly under a magnifying glass corresponds to the number
of implanted fetuses (see Sect. 2.2.7 on twinning).
The gestation period ranges from 19 to 22 days but this depends upon a num-
ber of parameters. For example, females that are pregnant for the first time (primi-
parous) deliver their progeny up to 1 day before multiparous females of the same
strain. The duration of pregnancy also varies slightly from one strain to another.
For example, pregnancy is, on the average, 1 day longer in mice of strain DBA/2
than in mice of strain C57BL/6.
At the end of the gestation period, the corpora lutea degenerate (luteolysis),
inducing parturition.8 The pelvic girdle of the females relaxes and parturition
begins in the following 2–4 h.9 During the same period, the behavior of the female
changes dramatically. The female is hyperactive and appears to have only one
thing in mind: preparing a nest in a corner of the cage, preferably in a darker area.
Parturition generally occurs at night and may last up to 3 h, depending on the litter
size. The fetuses are expelled one after the other, giving the mother time to take care
of each of the pups. The fetal membrane and the placenta, as well as the dead
embryos, if any, are carefully removed and ingested by the mother.10 Embryos are also
stimulated for breathing by repeated gentle pressure of the mother’s paws on the tho-
rax of the newborns. Once the last pup has been delivered and carefully revived, the
mother lays over all the newborns gathered in the nest and lactation starts. Newborn
mice are hairless, deaf, and blind, and are unable to regulate their body temperature
7 Dating the different steps of mouse embryonic development has been a matter of controversy.
Some embryologists wanted the first day of pregnancy to be designated day 1; others argued that
it should be day 0. In fact, the most accurate dating takes into account that, when the vaginal plug
is discovered, the embryo is at 0.5 days of development. At this time it is a one-cell embryo just
after fertilization (E0.5) (based on Theiler 1972).
8 Resorption of the corpora lutea is triggered by prostaglandins secreted by the placenta.
9 A gentle pressure on the pelvis of the mouse allows one to detect the relaxation of the pelvic
girdle.
10 Making the observation of non-viable (stillbirth) phenotypes difficult.
for the first 2 days of life ab utero; this is why the mother leaves the nest for only brief
periods, only to feed, defecate, and drink. Lactation normally lasts 3–4 weeks depend-
ing on the number and degree of vigor of the pups. In the mouse, the number of neo-
nates is frequently greater than the number of nipples (10), but this is not a problem
and the pups are generally fed adequately.11 From the age of 12–14 days, the young
mice start eating solid food and the mother’s milk is only a complement to the diet. At
the end of the lactation period, in general at the end of the third week of life, the young
mice are weaned and separated according to their sex by the technicians.
The standard reproductive cycle we have just described is sometimes modified
to fit with practical contingences. For example, adoption and foster nursing are
common practices in laboratory mouse breeding colonies, especially when the
number of progeny is low or the mother is not particularly good at nursing. When
there are only one or two pups in a progeny, the mother frequently abandons it/
them, presumably because the stimulation of milk production is insufficient. If this
situation occurs, it is then wise to take no risk and to transfer the secluded pups as
early as possible into an age-matched (up to 1 day younger) litter.12 Mice dams,
unlike many other female mammals, generally accept adopted pups to nurse and
milk, especially when they are young. Newborns selected for adoption can be sim-
ply added or exchanged in equal numbers with pups of the foster mother. It is rec-
ommended, when possible, that the newborns to be adopted be put in contact with
some urine-soaked wood-shavings taken from the mother’s bedding prior to the
transfer, to expose them to the foster mother’s smell.
Female mice can deliver up to eight progenies in their sexual life, depending on
the strain. However, the progeny size decreases after the fourth progeny and, most
importantly, the time that elapses between two successive progenies increases after
the third progeny. The number of progeny one can expect from a group of female
breeders can be evaluated based on the breeding records.13 Males can breed for a
very long time, sometimes up to 2 years; however, they are normally replaced after
10–12 months, depending on the strain.
Although mice are legendary for having exceptional aptitudes with regard to
reproduction in the wild, the situation is different in laboratory conditions and
sometimes requires special care. Reproduction and sexual behavior can be influ-
enced by a number of parameters that are not always easy to control. Pheromones,
for example, which are true olfactory hormones, play a major role in this matter.
The mouse is probably more affected by pheromones than any other mammal,
because of the complexity of its olfactory functions. Pheromones are proteins
which are released into the urine, the skin secretions, and the saliva of males and
11 If this is not the case, the pups are left outside of the nest; they progressively cool, do not
move much, and have no milk in their stomachs. Foster nursing is then urgent.
12 Selecting a mother nursing a litter with a different coat color (albino/non-albino) is a clever
way to check the success of the adoption without perturbing the mother.
13 A useful and reliable criterion is the average number of mice weaned per mated female per
week.
which modify the behavior of females. We have already reported the Whitten effect
(synchronization of estrous cycle) that affects female mice when they are housed
in groups. In addition to this observation, when females are kept in the absence of
male pheromones (which is not easy to achieve in practice), this leads to a state
of anestrus (lack of a normal estrus cycle). This phenomenon is called the Lee–
Boot effect (Van der Lee and Boot 1956). Finally, it is sometimes observed that
females, although found with a vaginal plug, never get pregnant when housed in
close vicinity with some males. This phenomenon is known as the Bruce effect and
an explanation is that the pheromones of the males prevent embryo implantation.
The males in question are called “strange males” (Bruce 1959).
Nutrition is another major parameter that must be seriously taken into account
concerning mouse reproduction. Since laboratory animals are fed exclusively on
industrial (pelleted) diets, it is extremely important to make sure that the diet con-
stantly provides the optimal amount of nutrients and vitamins, even after steriliza-
tion by heat or gamma rays. Some vitamins (C, B1, B9 for example) are extremely
heat-sensitive but yet are essential to the function of reproduction; it is therefore
essential to frequently change the heat-sterilized food. Nutritional deficiencies are
difficult to diagnose but they are insidious and almost always have consequences
on fertility, even if the mice do not exhibit any other obvious signs.
Environmental conditions (temperature, ventilation, noise, light cycle) are other
parameters to be controlled with care. Noise and vibrations are probably the worst,
especially when discontinuous, because the animals cannot become familiar with
them and are in constant stress. When the airflow bothers the animals they gener-
ally protect themselves and their nest by building a bulwark with their bedding.
This is a good indication that something is wrong with the air-conditioning system
or the airflow inside the individually ventilated cage. Environmental enrichment
like nesting materials and igloos are highly recommended to improve the breeding
performance of a mouse colony.
Finally, infectious diseases are also extremely important and must be carefully
monitored. Some viruses that cause unapparent diseases have a strong influence
on fertility, either because they interfere with the production gametes or because they
result in abortions or stillbirths. For more details concerning husbandry and mainte-
nance of laboratory mice, consult the books by Fox et al. (2007) and Hedrich (2012).
2.2.2 Inducing Ovulation in the Mouse (Superovulation)
The information provided in the preceding section concerning mouse reproduction

will be useful for those scientists who are willing to run a breeding unit. However,
in many cases, geneticists are only interested in harvesting large quantities of fer-
tilized eggs, for example, for creating transgenic animals by pronuclear injection,
or for making chimeras, or simply for the preservation of embryos at low tem-
perature. In some other cases, researchers are only willing to collect unfertilized
oocytes for in vitro fertilization. In these cases, young females aged 3–5 weeks
(prepubertal females) are treated by injection of gonadotropin hormones and

subsequently mated either to fertile or to vasectomized studs depending on the
aim of the experiment. In practice, the females in question receive a first injection
of 2.0–5.0 international units (IU) of the gonadotropin PMS (pregnant mare
serum) in the afternoon of day 1, an injection that artificially induces a first estrus
in the young females. 42–50 h later, they receive another injection of 2.0–5.0 IU
of the gonadotropin HCG (human chorionic gonadotropin) that artificially induces
ovulation.
Responses to gonadotropin injections vary from one strain to another, and, for
this reason, the optimum doses and age of the mice to be injected must be deter-
mined, for each strain, by doing preliminary experiments14 (Luo et al. 2011).
Ovulation occurs approximately 12 h after the HCG injection, at which time the
eggs (fertilized or not) can be collected by flushing the oviducts with a syringe
filled with the culture medium. In the best experimental conditions the treated
females can produce up to 30–40 embryos (hence the name superovulation),
although the response to the hormonal treatment is highly variable between inbred
strains (BALB/c is known to be a poor responder, while FVB is a high-responder).
Female mice can also be superovulated after puberty, but in this case the produc-
tion of embryos is much less efficient, presumably because the gonadotropin treat-
ment interferes with the hormonal status of the treated female. If fertilized eggs
are to be collected, it is important to mate no more than three or four hormonally
treated females per male. The males should be older than 8 weeks and, ideally,
proven breeders. Looking for vaginal plugs the following morning is then neces-
sary to select the females that will be sacrificed to collect the embryos (at the
desired stage). In order to be ready for the transfer of the manipulated embryos, it
is essential to produce pseudopregnant females that will serve as recipients. This is
typically achieved by mating outbred females (see Chap. 9) to vasectomized
(sterile) males (created through a simple surgery). This mating is necessary for the
uterine environment to become receptive, since only pseudo-pregnant females will
allow the successful implantation and development of the fostered embryos. For
more information and detailed protocols on these techniques, refer to the excellent
manual by Nagy et al. (2003) and visit the webpage of the International Society
for Transgenic Technologies (ISTT) at http://www.transtechsociety.org/.
2.2.3 Artificial Insemination
Several techniques for artificial insemination (AI) in the mouse have been
described in the past (Wolfe 1967; Leckie et al. 1973). These techniques are
simple and do not require sophisticated or expensive equipment. The sperm is
taken from the vas deferens or the epididymis, mixed at room temperature in a
14 The response to gonadotropin injections may also vary from one batch of hormone to the next.
few milliliters of tissue culture medium, and injected directly into the uterus of
the recipient female (at least 3 × 106 spermatozoa) using an insulin-type
syringe, with a blunted needle, and a speculum to avoid harming the vaginal
walls.15 In this case, however, the vasectomized male must not be placed with
the female before insemination, because the vaginal plug would interfere with
the process. Capacitation of the spermatozoa does not seem to be a problem
in this case.
Another technique has been reported where the sperm cells are injected directly
into the upper uterine horns or the ampulla with a glass micropipette after laparot-
omy (uterine insemination) (De Repentigny and Kothary 1996). This second tech-
nique does not require such a high number of sperm cells, as compared to vaginal
insemination.
Whatever the technique used, the yield in terms of embryo produced per insem-
inated female is quite low compared to other species. In spite of this low effi-
ciency, artificial insemination has proven useful for obtaining hybrids between
laboratory mice and mice of different species of the Mus genus (Mus caroli or Mus
cervicolor, for example) because mice of some of these species do not copulate
spontaneously with laboratory mice (West et al. 1977). Artificial insemination was
also used for studying the possible mechanisms leading to segregation distortion in
the progeny of males heterozygous for t-haplotypes16 (Olds-Clarke 1989).
When given a choice, one must remember that F1 hybrids or outbred females
have higher levels of fertility when used for AI. In addition, successful insemi-
nation can only occur when the inseminated female is in the late proestrus/early
estrus stage.
AI will probably not be used very much in the future, because alternative tech-
niques exist that are more reliable and have a much better yield.
2.2.4 In Vitro Fertilization in the Mouse
In vitro fertilization (IVF) is the most frequently used technology for assisted
reproduction in humans. The technology was adapted to the mouse several years
ago but this has not been easy to achieve and many critical steps had to be over-
come (Whittingham 1968; Vergara et al. 1997). A major difficulty has been the
development of suitable culture media allowing for a good rate of survival for the
early mouse embryos. Another problem has been to optimize the timing of supero-
vulation regimens for the different strains.
15 An ear speculum is an ideal tool. The extremity of a 20-ml glass pipette would also fit perfectly
for this purpose.
16 The t-haplotype is a small chromosomal region of chromosome 17 that is highly polymorphic
among wild mice of the Mus m. domesticus species. Frequently, t-haplotypes of wild origin are
not transmitted by heterozygous males in compliance with Mendel’s laws (i.e., 50:50), but at a
much higher frequency (95:5 or even 99:1).
Nowadays, protocols for IVF are available for most of the strains, even though
some of them exhibit a higher rate of fertilization than others (Sztein et al. 2000;
Nakagata et al. 2014).
The IVF technique generally consists of four steps: (i) young prepubertal
females are injected with gonadotropins as described above; (ii) the morning fol-
lowing HCG injection (~8 h after), the oocytes are collected and gently washed;
(iii) the oocytes are mixed for 4–6 h in vitro with either fresh or recently thawed
frozen spermatozoa; and (iv) after inspection and selection, the fertilized eggs
are transferred into a 0.5-day post-coitum (pc) pseudopregnant female. It is rec-
ommended to prepare the sperm sample one or two hours before mixing with the
oocytes to allow capacitation to occur, although capacitation of mouse spermato-
zoa does not seem to be as crucial as it is in other mammalian species.
IVF is the technology of choice when it is desirable to rapidly expand a
strain (for example, a transgenic line) from a few males that carry a desired or
unique genotype, or for maintaining strains with poor breeding performance.
IVF has the advantage that it can be performed using frozen or fresh sperm.
The technique can also be used for the re-derivation of infected mouse colo-
nies, and is frequently used for the transfer of genotypes of interest between
laboratories.
2.2.5 Ovary Transplantation
When a mutant or transgenic female is potentially fertile (i.e., when it produces

viable oocytes) but is unable to breed because of some kind of handicap, an
ovarian transplantation is a good option. Some classic mutant mice such as dys-
trophia muscularis (Lama2dy), obese (Lepob), and dwarf (Pou1f1dw) were histor-
ically maintained by performing serial ovary transplantation. The technique
consists of the surgical removal of the ovaries of the infertile donor female
(even from very young females), and transfer into the ovarian bursa of an ova-
riectomized recipient female. Again, either freshly collected ovaries from the
donor female or stored frozen/thawed ovaries can be used. Use of a recipient
female with a different coat color from the donor is recommended to differenti-
ate pups accidentally generated from residual ovary tissue (genotype of the
recipient female).17 Although the ovarian bursa is an immunologically privi-
leged site, it is convenient to use recipient females that are histocompatible with
the donor female. Alternatively, immunodeficient females (e.g., nude and SCID
mutants) can be used as recipients.
17 It is for the rapid and safe identification of the origin of its progeny that mice of the strain
129/J segregate for the coat color alleles Tyrc and Tyrch.
2.2.6 Intra Cytoplasmic Sperm Injection
Intra cytoplasmic sperm injection (ICSI, also known as micro-insemination)

is another technology that is commonly used in humans to overcome persistent
male infertility problems (for example, oligospermia, teratozoospermia, incapac-
ity of the spermatozoa to pass through the zona pellucida, etc.). Here again, the
technology has been adapted to the mouse with roughly the same basic protocol
as in humans. In short, mature oocytes are held at the tip of a micropipette, by
gentle suction, while a sperm head is injected deep into the cytoplasm of the
oocyte by using a piezo-driven micromanipulator. This equipment and procedure
allow for the safe injection of sperm heads by making only a very small hole in
the zona pellucida that is promptly resealed once the needle is withdrawn (Ogura
et al. 2003; Ogonuki et al. 2011). After the procedure, the oocyte is placed into an
appropriate culture medium where its development is checked for a few hours.
ICSI has been adapted for use with immature (haploid) spermatogenic cells
(round spermatids or elongated spermatids), and high rates of offspring develop-
ment have been obtained (~30 % in some cases).
ICSI and ROSI (round spermatid injection) technologies have also demon-
strated some practical advantages in the mouse. ICSI, for example, allowed for the
recovery of normal pups from spermatozoa taken from the testes or epididymides
of dead mice whose bodies had been stored at low temperature (between −20 °C
and −80 °C) for a few years (Ogura et al. 2005; Ogonuki et al. 2006).
ROSI technology has also been cleverly used to reduce the time required for
the development of fully congenic mouse strains by using the nucleus of round
spermatids removed from young males (22–25 days of age) for the fertilization
of superovulated oocytes flushed from 3-week-old females. With this technology,
a backcross generation could be reduced to only 41–44 days, and a fully con-
genic strain (homozygous for 97.7 % of 176 tested markers) could be produced in
190 days (~6 months) (see Chap. 9) (Ogonuki et al. 2009).
2.2.7 Cryopreservation of Mouse Embryos and Spermatozoa
The mouse was the first mammal whose embryos were successfully frozen and
stored at very low temperature. The methodology, which was published in 1972
(Whittingham et al. 1972; Wilmut 1972), required slow cooling (0.3–2 °C/min)
and slow warming at some critical steps as well as the use of cryoprotectants to
prevent ice crystals from damaging the cells of the embryo. In these initial
experiments the cryoprotectants were either dimethyl sulfoxide (DMSO) or gly-
col. Since these pioneering experiments, the technique has been improved and
nowadays mouse embryos are routinely stored at very low temperatures (in
liquid nitrogen at –196 °C) for virtually unlimited periods and thawed when
requested with quite high rates of survival.18 Embryo freezing and banking is
achieved routinely in many laboratories, and is also available as a service from
several commercial institutions. Short courses and demos with tutorials are
available in several formats, for example as “webinars” or highly didactic mov-
ies, and are freely available through the internet.
Vitrification is another method of cryopreservation that has been developed
more recently. With this method the embryos are osmotically dehydrated and then
cooled by a rapid transfer into liquid nitrogen.
Cryopreservation of mouse spermatozoa has proved capricious for a long
time and its rate of success is still relatively strain-dependent; for example,
C57BL/6 sperm is difficult to freeze and the proportion of unviable sperm cells
after thawing is quite high. However, the technology is rapidly improving and it
is likely that most of the technical problems that still remain nowadays will be
adequately solved in the near future (Sztein et al. 2000; Nishizono et al. 2004;
Nakagata et al. 2014).
Freezing embryos and spermatozoa both represent a safe and (relatively) cheap
way of exchanging mouse strains between different laboratories across the world.
This practice has the advantage of reducing the risk of transmission of infectious
diseases, a great concern for most veterinarians in charge of laboratory animal
facilities.
Ovarian cryopreservation has been demonstrated to be another valid option for
banking mouse genetic resources; in particular, it is the only technique that can be
used to preserve oocytes from aged or problematic female breeders (Sztein et al.
2010).
Readers who are interested in the practice of cryopreservation technolo-
gies can refer to comprehensive reviews on the subject by highly experienced
authors (Glenister and Rall 2000; Sztein et al. 2010; Nakagata 2011; Mochida
et al. 2011). A didactic movie is also freely available on the internet: see
reference list.
2.2.8 Twinning in the Mouse
The existence of the spontaneous occurrence of identical twins in the mouse is still
debated. According to Grüneberg (1952), twinning occurs in the mouse as in many
other mammalian species, but extremely infrequently; and twins may experience a
disadvantage during their early embryonic life. Identical twins have been
18 Experiments performed at the Harwell (MRC) Research Centre have demonstrated that the
damage caused by radiation (cosmic rays) to mouse embryos when stored at low temperatures
for very long periods is practically negligible.
occasionally observed in utero at very low frequency, between embryonic days 8

and 10, but such embryos have not been recorded by embryonic day 16–17.19, 20, 21
McLaren and colleagues, looking for identity by DNA fingerprinting (using
human minisatellite probes) in litters segregating for ten genetic loci, did not find
any evidence of twinning in a population of 2,000 outbred mice. The authors con-
cluded that twins are either extremely rare in the stock of mice they studied, or
that they have such reduced viability that their chance of surviving to weaning is
low (McLaren et al. 1995).
Spontaneous twinning is uncommon in the mouse; however, the experimen-
tal production of monozygotic twins by embryo splitting has been successfully
achieved in several laboratories. Illmensee and colleagues demonstrated that in
vitro splitting of mouse embryos at the 2-, 4- and 8-cell stage, followed by their
transfer into empty zonae pellucidae, could be achieved with a relatively high rate
of success. Embryonic development was monitored after in vitro culture for a few
days and twin blastocysts from 2- and 4-cell splitting showed well-developed colo-
nies with trophoblastic cells and clusters of inner cell mass (ICM) cells (Kaufman
and O'Shea 1978; Illmensee et al. 2005).
2.2.9 Cloning Laboratory Mice
Cloning is an asexual method of reproduction that is commonly used in plants

(e.g., cutting or striking) as well as in some insects: it offers the possibility of
obtaining a potentially unlimited number of genetically identical individuals. In
mammals, clones have also been produced experimentally by embryo splitting.
More recently, cloning has been achieved by the experimental replacement of the
nucleus of an unfertilized oocyte by the nucleus of a specific somatic cell from the
same species, a process known as somatic cell nuclear transfer (SCNT). In most
species, these experiments have been very difficult to perform, with low rates of
success. Beyond these difficulties, many clones have developed severe pathologies
that in many cases have undermined the interest of the enterprise. Cloning is no easy
endeavor and many fundamental questions regarding possible modifications at the
genome level during the early stages of development still remain to be understood.
19 It is not easy to observe twins by the mere examination of the implants in the mouse uterus, as
placental fusion is frequent in this species.

20 Discordances between the number of implants (dead or alive) and the number of corpora lutea
does not support the idea that twinning commonly occurs in the mouse.
21 Twinning (sometimes called “polyembryony”) is the rule in nine-banded armadillos of the
South American species Dasypus novemcinctus. In this species, the females regularly deliver
progenies composed of four monozygotic twins. This regular production of genetically identical
offspring makes the species a valuable model for multiple births.
Cloning the laboratory mouse has also been relatively difficult to achieve for
technical reasons. Nonetheless, cloned mice were produced for the first time after
the transplantation of nuclei taken from cells of the cumulus oophorus, hence the
name of the first cloned female mouse: “Cumulina” (Wakayama et al. 1998).22
Since then, mice have been cloned from a variety of different donor cells, includ-
ing fibroblasts (tail skin), olfactory sensory neurons, ES cells, bone marrow cells,
and liver cells. Recently, live mice have also been obtained after transplantation of
the nucleus of peripheral blood leukocytes into enucleated oocytes from a drop of
blood (Kamimura et al. 2013). Mice cloned from cumulus cell nuclei have even
been themselves cloned in series for 25 generations, producing over 500 viable,
fertile, and healthy clones derived from the original (single) donor. These experi-
ments proved that serial recloning over multiple generations is possible in the
mouse (Wakayama et al. 2013).
Compared to the situation in other species, in particular domestic species, the
cloning of mice has relatively limited applications. This is because it is very easy
in this species to obtain large populations of mice with exactly the same geno-
type. For example, mice of an inbred strain or born from a cross between two
inbred strains are all genetically alike exactly as if they were cloned individuals
(same genotypes). In these conditions, cloning mice may only help to enhance our
understanding of the technical and biological factors that contribute to successful
cloning in a species of economical interest. Experimenting with mice, biologists
may be able to understand how the donor nucleus taken from a differentiated cell
becomes reprogrammed by the oocyte cytoplasm to enable it to give rise to the dif-
ferent cell types. Similarly, the cloning of mice may help in the understanding of
the reversibility of epigenetic changes occurring during tissue differentiation.
2.2.10 Mosaics and Chimeras
The terms mosaic and chimera are frequently incorrectly used in the scientific lit-
erature, even under the signature of professional geneticists. Mosaics are organ-
isms composed of cells with a different genetic constitution, although deriving
from one single conceptus (embryo). For example, an organism composed of cells
with a different karyotype is a typical mosaic when this results from the loss or
abnormal disjunction of a chromosome during the many mitoses that occur
throughout embryonic development. An abnormal disjunction generates daughter
cells with 2n − 1 chromosomes and cells with 2n + 1 chromosomes, and these
cells are themselves mixed with normal 2n cells in variable proportions.23 Such
“chromosomal mosaics” are often viable, especially if the mosaicism concerns the
22 Cells of the cumulus oophorus are ovarian (but somatic) cells. They surround the oocyte and
are shed with it upon ovulation.
23 Cells with 2n + 1 chromosomes (trisomic) are in general more viable than cells with 2n − 1
chromosomes (monosomic).
X-chromosome or a minority of cells (see Chap. 3). 2n/3n and 2n/4n mixoploid

mosaics have also been described in several mammalian species, including the
mouse.
Mosaic organisms composed of normal cells and cells carrying a point muta-
tion at a specific locus are probably very common (this point will be discussed
in Chap. 7), but this mosaicism remains unnoticed if it has no deleterious conse-
quences for the mutant cell. On the contrary, when spontaneous mutations accu-
mulate in a cell (or group of cells) that provide the cell with the potential to divide
indefinitely or to resist cell death, then these cells may become cancerous (malig-
nant). In this sense, a mammalian organism affected by a cancer can be considered
as a true mosaic since the malignant cells have indeed acquired a genetic constitution
different from the non-malignant ones although they share the same origin.
Somatic (or mitotic) crossing-overs are yet another way of generating mosaic
organisms, but only very few cases have been reported and documented in the
mouse (Panthier et al. 1990). To conclude the definition of a mosaic, we note that,
in general, mosaics are produced naturally, with no human intervention, while this
is not the case with chimeras.
Chimeras (or chimaeras) are organisms that are composed of two (or more) dif-
ferent populations of genetically distinct cells (originated from different embryos),
which generally result from human intervention. For example, an immunodeficient
mouse that survives because it has received allogeneic bone marrow transplan-
tation is a chimera, as is a mouse that results from the in vitro fusion of two or
more morulae of different genetic origins (for example, from two different inbred
strains). In this chapter, we will consider exclusively the case of chimeras resulting
in a single complete mouse organism with pluri- or multiparental origin.
Mouse chimeras were created almost simultaneously in the early 1960s by
Mintz, working at the Fox Chase Cancer Institute (Philadelphia, USA) and by
Tarkowski, working at the University of Warsaw (Poland) (Tarkowski 1961, 1998;
Mintz 1962). The first chimeric mice were produced by merging two independent
morulae in vitro whose zona pellucida (oolemma) had been previously removed
by a brief treatment with the enzyme pronase. These chimeras are referred to as
aggregation or allophenic chimeras.24 They developed in a chimeric animal of
normal size, easily recognizable by a dappled coat color if the parental strains
were appropriately selected (Mintz and Silvers 1967).
The aggregation technique developed by Mintz and Tarkowski was replaced
in the early 1970s by a microsurgical technique that consisted of the injection of
embryonic cells directly into the cavity of blastocysts (Gardner 1971).25 This
technique was later modified and improved in several ways (Brinster 1974;
Mintz and Illmensee 1975; Papaioannou et al. 1975; Bradley et al. 1984; Stewart
et al. 1994).
24 Sometimes called tetraparental chimeras.

25 This cavity is often called a blastocoel.
Chimeric mice have been and still are important tools in biological research,
as they allow us to answer questions related to cell lineage and cell potential with
regard to tissue differentiation. By studying the muscles of chimeric mice con-
structed from two partner strains with different isocitrate dehydrogenase alleles
(Idh1a and Idh1b), it was demonstrated that the in vivo origin of the muscular syn-
cytium is from myoblast fusion and not from repeated nuclear division in a non-
dividing cell body (Mintz and Baker 1967).
Studying a series of hepatomas, which occurred in C3H/He × C57BL/6 chimeric
mice, researchers found that most of these tumors were derived from cells of the
hepatoma-susceptible C3H/He strain. However, they also observed that rare hepato-
mas were derived from cells of both genotypes, suggesting that some intercellular
transmission of tumor information may have occurred or that the transformation
might have occurred concurrently in two or more cells, indicating that hepatomas
may therefore be genetically complex entities (Condamine et al. 1971).
Nowadays, chimeric embryos are produced routinely by injecting totipotent
embryonic cells of different types (for example, embryonic cells from another
embryo, embryonic stem (ES) cells that may or may not be genetically engineered,
embryonic germ (EG) cells, etc.) into the blastocoel of recipient embryos. After this
injection, the cells of the ICM of the recipient embryo merge with the transplanted
cells and a chimeric embryo eventually develops to term. Today, the technique is
mostly used for introducing a new genotype (that of the engineered ES cells) into the
germ line of a chimera, allowing it to be ultimately materialized in a living mouse.
Another technique consists of using tetraploid embryos (which are artificially
made by electrofusion of two 2-cell diploid embryos) as recipients for the engi-
neered ES cells. It has been observed that, in this case, only the diploid cells (the
ES cells) contribute to the formation of the neonates’ body, while the cells derived
from the tetraploid embryo will exclusively give rise to the trophectoderm and
primitive endoderm. This technique is known as tetraploid complementation and,
although not used extensively, it has been successfully used to create mice entirely
derived from induced pluripotent stem cells (iPSCs) (Kang et al. 2009).
Another very clever technique resulting in 100 % germline transmission from
competent injected ESCs has been developed. This technique consists of using a
F1 host embryo (designated the “perfect host” or PH) which selectively ablates
its own germ cells via tissue-specific induction of diphtheria toxin. This approach
allows competent microinjected ES cells to fully dominate the germline, elimi-
nating competition for this critical niche in the developing and adult animal (Taft
et al. 2013).
Although chimeras can be either male or female, in experience the majority
is male because most of the ES cell lines are XY. Having male chimeras is actu-
ally good because they generally have good germline transmission (Nagy et al.
2003). Tetraparental chimeras can breed if the two embryos at the origin of the
chimera are both of the same sex. If this is not the case, for example if one set
of cells is genetically female and the other genetically male, intersexuality (and
sterility) often results. Even when the two embryos that participate in the forma-
tion of the chimera are of the same sex, the fertility sometimes depends on which
cell line gave rise to the ovaries or testes. For this reason, the association of a
tetraploid (4n) partner with a diploid (2n) one, as explained above, is particularly
advantageous.
The production of allophenic chimeras has been used in various contexts
to answer biological questions that would not have been easily answered other-
wise. For example, chimeras have been produced to transmit lethal genes in the
mouse and to demonstrate allelism of two X-linked male-lethal genes, jp and msd
(Eicher and Hoppe 1973). In another example, viable aggregation chimeras have
been made by merging normal embryos with embryo homozygous for the reces-
sive lethal mutation Hairy ears (Eh-Chr 15), which indicated that the mutation in
question was not cell-lethal (we now know that it is a large deletion) (Guénet and
Babinet 1978). Finally, especially noteworthy is the production, by Kobayashi
et al. (2010), of the first viable rat–mouse chimeras. In this report, the authors also
demonstrated that rat iPS cells could rescue organ deficiency in mice, opening new
frontiers for tissue engineering.
2.3 Basic Notions of Genetics
2.3.1 Genes and Alleles

In his famous note reporting the results of his Experiments on Plant Hybridization
(1866), Mendel alluded to “factors” or “units of inheritance” that are transmit-
ted from one generation to the next and determine the phenotypic characteristics
of plants. Using such words, it is clear that Mendel was referring to the concept
of genes, but he did not coin any specific word to define these “units of inherit-
ance”. Several years later, in 1889, de Vries published a book entitled Intracellular
Pangenesis in which he led an interesting discussion concerning the “units” or
“particle bearers of hereditary characters”, and he recommended that the word
pangens be used to specify these particles (de Vries 1910). Finally, it was the
Danish biologist Johannsen who proposed, in 1909, that the (Danish) word “gen”
be used to describe the units of heredity. The same Johannsen also introduced the
terms phenotype and genotype, and almost at the same time Bateson proposed the
term genetics to describe the science dealing with gens (genes).
Shortly after the confirmation that DNA was the molecular basis for inheritance
(seminal work published by Avery, McCarty, and MacLeod in 1944), the definition
of the gene was translated into molecular terms and became “a segment of DNA of
variable size encoding an enzyme”. This was in compliance with the famous “one-
gene-one-enzyme” theory formulated by Beadle and Tatum.26 This definition was
reconsidered when it was recognized that some proteins are not enzymes. The
26 G.W. Beadle and E.L. Tatum were awarded the Nobel Prize in Physiology or Medicine in
1958 for their discovery of the “role of genes in regulating biochemical events within cells”.
2.3 Basic Notions of Genetics 35
motto was then changed to “one-gene-one-polypeptide” and the definition of the

gene was modified accordingly.
In 2002, once the sequencing of the mouse nuclear DNA was completed, fol-
lowed by the extensive analysis of the transcriptome27 and the confirmation that a
great number of genes were not translated into polypeptides, the definition of the
gene changed again. Nowadays a gene corresponds to a segment of DNA that is
transcribed into RNA. Some of these RNA molecules, the messenger RNAs
(mRNAs), are translated into polypeptides while many others are not translated
but have nevertheless important functions as RNAs (see Chap. 5).
Recently, information collected from the systematic analysis of a single tran-
scriptome revealed that mammalian DNA is pervasively transcribed from both
strands, and that the proportion of DNA transcribed into RNAs is much greater
than expected. The same analysis also revealed that mammalian genes are not
always clearly individualized in the DNA strands; on the contrary, their limits
are often difficult to define, with some small genes being nested into larger ones,
inserted for example in the introns. Thus, it is seems clear that the concept of the
gene will have to be reconsidered and its definition reformulated in the near future.
For the time being, we will stay with the idea that a gene is a functional unit mate-
rialized in a short segment of DNA, which is transcribed into RNA, and whose
inheritance can be followed experimentally generation after generation.
For decades, the genome was known as the collection of genes of a given spe-
cies. The word was coined at the beginning of the twentieth century, and at that
time it was used to exclusively define the collection of genes. Nowadays, the
concept includes both the genes (i.e., the coding sequences) and the sequences
of DNA that are intermingled with the genes and are themselves heterogeneous.
Thus, when they refer to the mouse genome sequence, geneticists in fact refer to
the sequence of the whole nuclear DNA.
The number of genes in the mouse genome is expected to be in the range of
25,000–30,000 but, for reasons that will be discussed further, this assessment is
not accurate and will probably never be. For some genes, the number of copies
in the genome varies across the different strains, or even individuals, and many
among these genes are non-functional (see Chap. 5 regarding CNVs and pseudo-
genes). It is also known that some genes are present in some strains (or species)
and absent in others. All these variations, of course, hamper the accurate evalua-
tion of the number of genes.
In addition to these inter-strain quantitative variations in the number of genes,
we know that several different RNAs (coding and non-coding) can be transcribed
from the same gene by the mechanism of alternative splicing (detailed in Chap. 5),
and this tremendously increases the number and diversity of the molecules poten-
tially encoded in the genome. Obviously, it is the inventory and classification
of all these transcripts that would be important to make, rather than an accurate
27 The transcriptome corresponds to the full set of RNA molecules that are transcribed from the
genome. This point will be extensively discussed in Chap. 5.

assessment of the number of genes. This goal is certainly in the minds of many
geneticists, but it is a serious challenge and is difficult to achieve.
Whatever the actual number of genes in the mouse genome, once a gene is bio-
logically defined either in terms of function or structure, it can be precisely local-
ized on a specific chromosome of the species using a variety of techniques. The
position of such a gene defines its locus (plural loci, the Latin word for “place”)
and we will extensively discuss the strategies used for the localization of the genes
in Chap. 4.
Many genes exist in several versions (variants) called alleles. The word “allele”
is an abbreviation of the ancient word allelomorph, which was used in the past to
describe the different forms of a gene, detected as different phenotypes. Formerly,
the concept of alleles was tightly associated with the concept of mutation pro-
ducing a phenotypic variant different from the wild type (i.e., the version most
commonly found in wild animals). In this case, the new version of the gene was
defined as a mutant allele and was identified in mice, for example, by a different
coat color, a heritable skeletal defect, or a debilitating neurological disease.
The concept of the allele has also progressively changed and nowadays one can
say that any change at the DNA level that translates into a phenotype different
from the previously known phenotypes defines a new allele, regardless of whether
the phenotype associated with the new allele is deleterious. The substitution of a
nucleotide in a coding sequence that leads to a change in the global electrical
charge of a protein characterizes a new allele because, even if the function of the
protein is not affected, one can distinguish by electrophoresis the new protein from
the other proteins encoded by the same gene: it is a different phenotype.28 If the
nucleotide substitution modifies the activity of the protein, with deleterious conse-
quences, in this case the new allele is either a hypomorphic or null allele (see
Chap. 7).
Other types of structural variations at the DNA level (for example, the so-called
single nucleotide polymorphisms or SNPs) can also be used to distinguish allelic
variants (DNA variants in this case), even if these allelic variants do not confer any
phenotypic change on the animal. In these conditions the reader may appreciate
how the definition of the word allele has evolved with time. In the past, the func-
tion of the protein, assessed by its effect on the phenotype of the animal, was cru-
cial to define a new allele. Nowadays, any structural change that can differentiate a
gene from another at the same locus defines an allele, regardless of the phenotypic
consequences. We will come back to this point when discussing the genetic mark-
ers used for gene mapping (Chap. 4).
According to the Mouse Genome Database (as of November 2014), 10,425
genes of the mouse have at least a mutant allele and the mouse genome comprises
40,713 alleles altogether. The whole collection of alleles that are segregating in
a given population represent what geneticists call the genetic polymorphism. This
28 The word electromorph has been coined to define the alleles characterized by a different
global electrical charge.

notion of polymorphism applies to the series of alleles at a specific locus or to all

loci of a strain or species.
In the mouse, the gene encoding tyrosinase (Tyr), an enzyme that is instru-
mental in the synthesis of the pigment melanin, was one of the first (if not the
first) to be identified based on a variation in coat color. At the Tyr locus, one allele
encodes a normal, functional tyrosinase, and the other encodes a non-functional
enzyme resulting in albinism. Nowadays, over 120 different mutations have been
collected at the Tyr locus, some of them having a phenotype affecting coat color
(for example, chinchilla-Tyrc-ch, extreme-dilution-Tyrc-e, hymalaya-Tyrc-h, to cite a
few). However, it is likely that sequencing will identify many others that are not
yet detected because they have no obvious phenotype. Such a collection of a series
of alleles always represents an interesting resource for geneticists.
The Mouse Genome Database specifies rules and guidelines for mouse and rat
gene nomenclature (http://www.informatics.jax.org/mgihome/nomen/gene.shtml),
with which it is extremely important to comply because genetic nomenclature is
a universal language. We recommend that readers frequently refer to these guide-
lines, which are presented in a very didactic form with many different examples.
In case of doubt, information may also be requested directly from the staff of cura-
tors, as explained on the website.
2.3.2 Allelic Interactions
When the alleles at a given locus are the same on both chromosomes, the mouse is
homozygous and the phenotype that characterizes the allele in question is fully
expressed: the situation is simple. When the two alleles are different, the mouse is
heterozygous and the phenotype depends upon the interactions between the two
alleles. To illustrate the situation, we will again consider a gene we are already
familiar with: the gene encoding tyrosinase (Tyr-Chr 7). As we already mentioned,
this gene has several alleles, among which some are non-functional; this is the
case with Tyrc. When a mouse has the Tyrc allele on both chromosomes 7
(homozygous), it is albino. In contrast, when the mouse has a non-functional allele
on one chromosome 7 and a functional allele on the other chromosome, it is hete-
rozygous and is pigmented just like a wild mouse. The Tyrc allele is said to be
recessive and the normal allele, or wild-type allele (Tyr+ or sometimes only +), is
dominant. In this case, the lack of functional tyrosinase is completely compen-
sated for at the cellular (melanocyte) level by the presence of a single copy of the
normal (wild-type) allele.29
29 When an allele is fully dominant, geneticists often write the genotype Mut/–, indicating that
the allele in question completely determines the phenotype.

Some other alleles at the Tyr locus have less dramatic effects than Tyrc on the
synthesis of the pigment melanin and in many cases the mice are pigmented,
although always less than or differently from the wild type. For example, mice
homozygous for the extreme dilution Tyrc-e allele appear “slightly stained or dirty
black-eyed white” (Detlefsen 1921). They have a light grey coat color, almost
white, but their eyes are solid black, unlike albino mice. Mice homozygous for the
chinchilla allele Tyrc-ch have a diluted coat color (they really look like chinchil-
las—hence the name of the mutant allele) but their coat color is much darker than
mice homozygous for Tyrc-e. Finally, mice homozygous for the Himalayan allele
Tyrc-h/Tyrc-h have a remarkable pattern of pigmentation with a mainly white body
and light-ruby eyes and only the tip of the nose, tip of the ears, and the tail are
normally pigmented (black). This is because the tyrosinase encoded by the Tyrc-h
allele is active only in the colder parts of the body, where the temperature is below
35 °C (the enzyme is heat-labile or thermo-labile). This is the same phenotype
observed in Siamese cats.
With so many alleles at our disposition, we could breed a wide variety of
mice heterozygous or homozygous for different alleles and we would then dis-
cover that the normal allele (Tyr+) is dominant over all other alleles. However,
if we grade the phenotypes of the mice based on the decreasing intensity of the
coat color for all the possible combinations of the four alleles at the Tyr locus-
Tyr+; Tyrc-ch; Tyrc-e and Tyrc we observe that they display an almost continu-
ous gradient of pigmentation from type to albino (i.e. Tyr+/− > Tyrc-ch/Tyrc-ch
> Tyrc-ch/Tyrc-e > Tyrc-ch/Tyrc > Tyrc-e/Tyrc-e > Tyrc-e/Tyrc > Tyrc/Tyrc) (from
Silvers 1979). The observation of intermediate phenotypes such as Tyrc-ch/Tyrc-e
or Tyrc-e/Tyrc allows for the definition of another kind of allelic interaction
that is called incomplete dominance or intermediate dominance, or sometimes
partial dominance. In these cases one allele is not completely dominant over
another, and the expressed physical trait is in between the dominant and reces-
sive phenotypes. In this context, the phenotype of mice homozygous for the
Himalayan allele Tyrc-h cannot be considered as “intermediate”; they are simply
different and unique.
The series of alleles that we described at the Tyr locus is common in plants and
vertebrate species, and many other examples are available in the mouse. As we
already said, in most cases the wild-type allele, the one that is most frequently
found in wild mice, is often dominant over all other alleles at the same locus; but
this is far from being a rule. At the Agouti locus (A-Chr 2), where there is another
long series of alleles (over 400) affecting coat color, the wild-type allele agouti (A)
has an intermediate position: it is dominant over some alleles like black-and-tan
(at), non-agouti (a), or extreme non-agouti (ae), but it is recessive to yellow (Ay),
viable yellow (Avy) and a few other A alleles. By the way, it is interesting to note
that the yellow allele (Ay) in question, although dominant over A if we consider the
coat color, is nevertheless a recessive lethal when homozygous (see Fig. 1.1). Ay/A
mice have a beautiful yellow coat color but Ay/Ay embryos display characteristic
abnormalities at the blastocyst stage and die on the sixth day of gestation.30 This
observation means that the notion of dominance and recessivity must be consid-
ered only in the context of a specific phenotype.
True dominant mutations, i.e., mutations for which the phenotype of the het-
erozygote (Mut/+) is indistinguishable from the phenotype of the homozygous
mutant (Mut/Mut), are rare in the mouse and in mammals in general. In most
instances, the dominant alleles behave just like the yellow (Ay) allele and are lethal
when homozygous. Among the few exceptions are some keratin mutant alleles
such as Rex (Krt25Re), Caracul (Krt71Ca), and possibly a few others such as the
coat color mutation Sombre (Mc1rE-so) and the neurological mutation Trembler
(Pmp22Tr).
Another type of allelic interaction that is extremely common in mammals is
co-dominance. Co-dominance is when the two alleles at a given locus are both
expressed in the phenotype of the heterozygote, which has a phenotype of its own.
In most genetics textbooks the concept of co-dominance is exemplified by the AB
blood groups in humans, where the AB heterozygotes have a phenotype in which
both the A and B antigens are expressed on the red blood cells. Blood groups
homologous to the human AB system do not exist in the mouse, but practically
all the genes expressed in the form of proteins with different electric charges are
co-dominantly expressed. Glucose-6-phosphate isomerase (symbol Gpi1-Chr 7) is
an enzyme that is expressed in most cells; it catalyzes the conversion of glucose-
6-phosphate into fructose-6-phosphate. Several alleles at the Gpi1 locus have been
characterized, of which four are common, viable and functional: Gpi1a and Gpi1b
are found in laboratory inbred strains, Gpi1c is a spontaneous mutation of recent
occurrence in the BALB/c inbred strain, and Gpi1d was discovered in wild mice.
It is likely that many more alleles (electrophoretic variants) exist in wild mice and
have not (yet) been identified. All these alleles are co-expressed in mice heterozy-
gous at the Gpi1 locus.
When the phenotypes of the different alleles at a given locus are carefully ana-
lyzed, interesting observations can be made concerning the allelic interactions.
A good example is the case of the locus encoding the enzyme argininosuccinate
synthetase (ASS). At this locus, several mutant alleles have been identified in the
mouse that are potentially interesting models for the human disease citrullinemia
type I (CTLN1, OMIM# 215700). Among all the hypomorphic alleles, two are
more interesting than others: Ass1bar and Ass1fold, because they faithfully repli-
cate the pathology observed in human patients suffering from CTLN1, with vari-
ations in terms of survival rate, developmental delay, and neurological phenotype.
Homozygous and compound heterozygous combinations of the two alleles create
30 These yellow mice posed a problem to Cuénot while he was trying to demonstrate that
Mendel’s laws also apply to mammals. When intercrossing Ay/A mice, he did not find the
expected 1:2:1 proportions of phenotypes for a single gene with two alleles, but instead found a
1:2:0 ratio. However, Cuénot provided the correct explanation for these “unusual” proportions.
a spectrum of severe (Ass1bar/Ass1bar), intermediate (Ass1fold/Ass1fold), and mild

(Ass1bar/Ass1fold) phenotypes. However, the observation that the compound het-
erozygotes, carrying one severe allele (Ass1bar) and one mild allele (Ass1fold),
exhibited a milder phenotype (including residual activity of liver ASS and less
pronounced plasma ammonia levels) than mice carrying two copies of the mild
allele (Ass1fold/Ass1fold) was quite unexpected. Obviously, this warrants further
investigation concerning the molecular interactions between the different ASS1
mutant proteins (Perez et al. 2010).
Dominance, recessivity, and co-dominance are the most common forms of
allelic interactions but there are also a few others that deserve, at least, a short
comment. Semi-dominance has been used to characterize mutant alleles where
the phenotype of heterozygotes is different (and often intermediate) from both
kinds of homozygotes. A typical example is the KitW-f allele at the Kit locus (Chr
5), where KitW-f/+ heterozygous mice have a light grey coat with a spot on the
belly and a small blaze on the forehead, while heterozygous KitW-f/KitW-f mice are
extensively spotted (Guénet et al. 1979). Amazingly, the tails of these mice per-
fectly characterize the situation: the tail is normally (i.e., completely) pigmented
from the base to the tip in wild-type mice; it is half pigmented in heterozygotes
and it is completely unpigmented (white) in homozygotes. Another example
is the semi-dominant spontaneous mutation called Naked (N) on distal chromo-
some 15 (Hogan et al. 1995). The semi-dominant allelic expression is common in
the mouse but it is sometimes used for alleles that would be best characterized as
incompletely dominant.
Overdominance is a rather rare condition in which the heterozygotes (M/m)
have a phenotype that is more pronounced than that of either homozygote (M/M
and m/m). We report such a case of overdominance in Chap. 6 with the Callypyge
mutation in sheep. No similar mutation has ever been reported in the mouse, but
some may exist. Overdominance is sometimes used as an alternative word for super-
dominance. Superdominance characterizes a situation where the heterozygotes
have a selective advantage over homozygotes. This is the case, for example, with
the human allele that encodes sickle-cell anemia (HBBs) or drepanocytosis. In the
countries where malaria is endemic the gene encoding the defective hemoglobin,
although lethal when homozygous, is present in over 40 % of the population, while
we would expect it to be strongly counter-selected. This is because the HBBs allele
confers resistance to malaria in the heterozygotes. Homozygotes with the normal
allele get sick and sometimes die when infected by Plasmodium; homozygotes for
the mutant allele also get sick from drepanocytosis but the heterozygotes survive
Plasmodium infection and do not develop severe anemia. This selective advantage of
a specific combination of alleles is probably also found in wild mouse populations
but, to date, it has never been described in any laboratory mouse or rat population.
In this review of the different forms of allelic interactions, one must not forget
the case of genes that are X-linked. In mammalian species, the males have only
one X-chromosome and, in these conditions, the individuals are hemizygous for
all the genes carried by this chromosome and all are fully expressed. In females,
the situation is more complex and the situation will be discussed in full detail in
Chap. 6. Without going into detail, one can say that due to the phenomenon of
X-inactivation, which is a mechanism of dosage compensation operating in female
mammals, most X-linked genes are functionally haploid and only one copy of every
gene is transcribed, while the other copy is switched off. The inactivation of one
allele over the other is, in most cases, a random process. In the mouse, a few genes
are in the so-called pseudo-autosomal region of the X-chromosome and behave as
autosomal genes. The gene encoding steroid sulfatase (Sts) is one example.
When a mutation occurs in a mouse population, the allelic interactions exhib-
ited by the novel allele is important information to take into account in the process
of genome annotation. If the novel allele is dominant or semi-dominant, it makes
sense to guess that the observed phenotype is the consequence of a structural
defect of the protein encoded by the mutant allele. On the contrary, when the novel
allele is fully recessive, this would indicate a loss-of-function (or hypomorphic)
mutation for the protein encoded by the mutant allele. For example, mutations in
the genes encoding collagens or fibrillins, which generate a structural defect in the
proteins in question, are almost always dominant or semi-dominant.31 On the con-
trary, mutations that cause an “inborn error of metabolism”, as Garrod used to des-
ignate some metabolic diseases, are usually recessive. In fact, there is some logic
in these observations: the genes encoding metabolic enzymes are in general haplo-
sufficient (50 % of normal levels are sufficient to complete the metabolic func-
tion), while the situation is radically different if the encoded polypeptide is
involved in the differentiation of a specific tissue.
2.3.3 Epistasis and Pleiotropy
Many phenotypic traits are controlled by more than one gene, and, conversely, it
is relatively common to observe that a given gene contributes to the phenotypic
expression of one or several other genes. In the forthcoming chapters (in particular
in Chap. 10, which is devoted to quantitative genetics) this point will be consid-
ered in detail. For the time being, we will just discuss a few examples that will
help introducing two fundamental notions in genetics: epistasis and pleiotropy.
2.3.3.1 Epistasis
Epistasis characterizes a situation where the phenotypic expression of a gene (or

allele) A depends on the presence, at some other loci in the genetic background
(B, C, D), of one or several specific alleles that modify or suppress the classical
31 A mutation that leads to the synthesis of a mutant protein that interferes or disrupts the activ-
ity of the wild-type protein in the multimer is called a dominant-negative mutation. A typical
example is found in the syndrome of osteogenesis imperfecta (O.I. Type III) in which structurally
defective type I collagen is formed.
phenotype of gene A. To put it in other words: epistasis is an interaction between

non-allelic genes in which one gene suppresses or enhances the expression of
another. The gene that is expressed is epistatic over the others genes, which are
themselves hypostatic. Once more, the genes that are involved in the development
of mouse coat color offer simple and didactic examples.
Exploiting the variety of alleles at the five major loci governing the mouse
coat color (Agouti-A; Tyrosinase-Tyr; Brown-Tyrp1; Dilute-Myo5a; and Pink-
eyed dilution-Oca2) one can generate a large collection of mice with a wide
array of coat colors. However, sometimes it happens that the effects of a given
mutant allele cannot be observed if another allele is present in the genome of
the same animal. A mouse with a non-agouti, brown coat color (genotype a/a;
Tryp1b/Tyrp1b) would appear “chocolate”, unless the Tyrc allele (which is at the
Tyr locus on a different chromosome) is homozygous. In this latter case, the
mouse would appear albino—i.e., completely white, and this is because the Tyrc
allele exhibits epistatic interaction with all other coat color genes. We know the
reason: it is because the Tyrc allele is non-functional. Thus, there is no tyrosinase
synthesis, no melanin, and no pigment, be it black or yellow.
Another example of epistatic interaction in the mouse is between the Mc1re
allele (recessive yellow-Chr 8) at the locus encoding the melanocortin 1 recep-
tor and the Mlphln allele (leaden-Chr 1): mice with a Mc1re/Mc1re; Mlph+/−
genotype have a deep yellow coat color, and mice with a Mc1r+/−; Mlphln/Mlphln
genotype have a light gray coat color, like diluted, but these mice are indistin-
guishable from the Mc1re/Mc1re; Mlphln/Mlphln mice. In short: Mlphl is epistatic
to Mc1re and the phenotypic effects of the recessive yellow allele are entirely
suppressed by the phenotypic effects of leaden (Hauschka et al. 1968). Many
mutations affecting enzymes of cellular metabolism exhibit epistatic effects,
especially when they are in the same metabolic pathway.
Epistatic interactions are common with traits governing quantitative inherit-
ance: the quantitative trait loci (QTLs). A heritable quantitative trait can be under
the control of several genes with additive effects, the genes in question having dif-
ferent strengths in the determinism of the phenotype. When two alleles at different
loci have a stronger effect on the phenotype than each allele individually, this is
referred to as synergistic epistasis. The opposite situation also exists and is called
negative or antagonistic epistasis.
When we described the epistatic effects of the Tyrc (albino) allele on the
expression of all other genes involved in coat color determinism, we assumed that
this effect was the direct consequence of the expression (or non-expression, in
this case) of the protein encoded by the gene (tyrosinase). The situation is some-
times much more subtle. For example, the mutant allele ApcMin (at the adenoma-
tous polyposis coli gene-Chr 18) is a dominant allele (although recessive lethal)
that predisposes mice to the development of multiple intestinal neoplasia (Moser
et al. 1990). However, some mouse strains are much more susceptible to this syn-
drome than others. Mice of the strain C57BL/6, for example, are severely affected
and develop many intestinal tumors, while mice of the AKR strain, with the same
ApcMin allele (congenic mice), develop only a few tumors. This dramatic phe-
notypic difference between the two inbred strains has been found to be the con-
sequence of an epistatic interaction between the ApcMin allele and another gene
called Modifier of Min encoding a phospholipase A2 (Pla2g2a-Chr 4), itself with
two alleles: Pla2g2aMom1-r and Pla2g2aMom1-s. However, the Pla2g2a alleles have
a phenotypic effect only when the ApcMin allele is in the same genome. In other
words, Pla2g2a is a modifier gene whose phenotypic expression is conditional to
the presence of the ApcMin allele. Such situations are very common in the mouse,
and the ApcMin allele has several other independent modifiers (Dietrich et al.
1993). The identification and study of modifier genes opens interesting avenues
for unraveling the networks that determine robustness and resistance to certain dis-
eases. Hence, we emphasize the importance of the use of pure inbred backgrounds
in mouse models (see Chap. 9).
2.3.3.2 Pleiotropy
Pleiotropy describes a situation that is in fact extremely common in genetics: it

simply means that a mutant allele has an effect on different phenotypic traits. In
fact, if we carefully analyze most of the mutants with a deleterious phenotype, we
would then discover that almost all of them in fact exhibit a panel of different phe-
notypes. The yellow allele (Ay) was identified because of its eye-catching pheno-
type, with a beautiful yellow coat color, but the mutant mouse exhibits many other
phenotypes. For example, the mice are slightly diabetic, exhibit liver hypertrophy,
and many become obese and sterile after the first few months of life. They are
also more susceptible to several kinds of tumors than normal mice and are more
aggressive.
If we consider that the products of most genes are involved in several cellu-
lar functions, pleiotropy seems to be more the rule than the exception. It simply
means that the gene in question codes for a product that is used by various cells,
or has a signaling function on various targets, or that the protein is an enzyme or a
transcription factor that is involved in several pathways.
2.3.4 Penetrance and Expressivity
2.3.4.1 Penetrance
Penetrance is a term used to express the fraction of individuals of a given gen-
otype that effectively exhibit the expected phenotype. Penetrance is usually
expressed as a percentage. For example, if a particular dominant mutation has
80 % penetrance, then 80 % of the mice carrying the mutant allele will develop the
phenotype and 20 % will look normal (Fig. 2.1).
Fig. 2.1 Penetrance and expressivity. The figure illustrates the concepts of penetrance and
expressivity. In this example, the mutation brachyury (T-Chr 17), affecting six out of the seven
mice, exhibits great variations in expressivity; some mice have a tail longer than others, even if
they all are clearly short-tailed. When a mouse with a short tail (genotype certainly T/+) is crossed
with a normal mouse (+/+), the proportion of affected offspring is often lower than 50 %. Some
mice have an extremely severe reduction of the tail, exhibit a spina bifida, and die at birth while
others have an almost normal tail (normal overlaps). The penetrance characterizes the frac-
tion of individuals of a given genotype that actually exhibit the phenotype typical of the mutant
allele, irrespective of the degree of its expression. The expressivity characterizes the phenotypic
variation among individuals having the same genotype. It is now well established that modifier
genes influence the phenotypic expression of many mutant alleles. However, the action of these
modifiers cannot explain all types of variations since phenotypic variations are also observed in
inbred strains—as in the case illustrated here, where all the mice are from the same inbred strain.
Variations in penetrance and expressivity are common for skeletal and eye mutations in all species
2.3.4.2 Expressivity
A genotype exhibits variable expressivity when individuals with that genotype dif-
fer in the extent to which they express the phenotype normally associated with
that genotype. The best example illustrating the concept of expressivity and dif-
ferentiate it from the concept of penetrance (which is not always easy) was pro-
vided by Danforth regarding a population of cats in Key West Island (a population
also known as “Hemingway’s cats”), in which a dominant mutation resulting in
polydactyly is highly prevalent. Observing the cats in question, Danforth wrote,
“the polydactyly phenotype shows good penetrance, but variable expression”. This
simply meant that a high percentage of cats indeed had extra toes, but the num-
ber and size of the extra toes varied from one animal to the next (Danforth 1947).
Another example is the case of spotting in cattle. Observing a herd of cows of the
Frisonne breed one may notice that, although all the cows are spotted (penetrance
is 100 %), the ratio black/white is highly variable from one cow to the next.
The spotting is highly variable in shape (no surprise) but also in extension (which
is more surprising). These are variations in expressivity of the spotting allele.
Although the examples we selected were from cats or cows, similar situations
can be easily found in mice where mutations yielding extra digits and white spot-
ting are common. In short, variable expressivity means that there is a large amount
of phenotypic variation among individuals with the same genotype (Miko 2008).
The causes of penetrance and expressivity are not well understood. In the
mouse, as well as in the rat, one can study the phenotypic expression of the same
mutation in different genetic backgrounds and note more or less consistent differ-
ences, indicating the existence of a genetic component (modifier genes). However,
in the same species, one can also observe phenotypic variations in animals having
exactly the same mutation in exactly the same genetic background—meaning that
genetic factors are not the only factors involved in the variation of penetrance or
expressivity. In these conditions, it makes sense to consider that epigenetic factors
or stochastic events are probably also at work. In Chap. 6, dealing with the epi-
genetic control of genome expression, we will discuss a situation where the coat
color of mutant mice is strongly influenced by environmental factors.
Having control of the factors that determine expressivity is of major importance
in human medicine, because many diseases with a genetic determinism (for exam-
ple, cancers, neurological diseases, and skeletal abnormalities) often exhibit great
variations in expressivity (Nadeau 2003).
2.4 Phenotyping Laboratory Mice: The Mouse Clinics
As we will explain in the chapters to come, researchers now have all the means and
tools to create a great variety of alterations in the mouse genome; for example, they
can switch off any gene they wish, and at any time. They can interfere (transitorily
or not) with gene regulation, they can make all sorts of genetically engineered mice
with cloned DNAs of their choice, etc. Of course, all of these alterations induced at
the genome level are expected to result in changes at the phenotypic level in geneti-
cally modified animals, and the careful analysis of these phenotypic changes is
obviously fundamental for the process of genome annotation.32 However, the prob-
lem is that, though it is relatively easy to localize and precisely characterize a DNA
sequence, especially nowadays, it is much more difficult to unambiguously estab-
lish the link between a DNA alteration and an abnormal phenotype. Examples are
numerous where the knockout allele of a theoretically important gene was initially
reported as having “no detectable phenotype,” and this was to the great surprise
(and sometimes to the disappointment) of its creator (Colucci-Guyon et al. 1994).
32 Genome annotation consists of attaching biological information to a particular DNA
sequence, or of establishing a link between a gene (or a small chromosomal region) and a given
phenotype by any possible means.
Phenotyping has become one of the main concerns of mouse geneticists over the
last 10 years and, mainly for this reason, many laboratories and institutions have
developed what is now called a “Mouse Clinic”. In these clinics, mouse mutants
or strains are thoroughly analyzed for the greatest possible number of parameters
using a panel of highly standardized phenotyping protocols. In most cases the basic
protocols are focused on behavior, bone and cartilage development, neurology, clin-
ical chemistry, eye development, immunology, allergy, steroid metabolism, energy
metabolism, lung function, vision and hearing, pain perception, molecular pheno-
typing, cardiovascular analyses, and gross pathology. For example, the International
Mouse Phenotyping Resource of Standardised Screens (IMPReSS) contains stand-
ardized phenotyping protocols, which are essential for the characterization of
mouse phenotypes (see https://www.mousephenotype.org/impress). The use of
standard procedures and defined protocols allows data to be comparable and share-
able, and even allows interspecies comparisons to be performed, which may help in
the identification of mouse models of human diseases.
References
Bradley A, Evans M, Kaufman MH, Robertson E (1984) Formation of germ-line chimaeras from
embryo-derived teratocarcinoma cell lines. Nature 309:255–256
Brinster RL (1974) The effect of cells transferred into the mouse blastocyst on subsequent devel-
opment. J Exp Med 140:1049–1056
Bruce HM (1959) An exteroceptive block to pregnancy in the mouse. Nature 184:105
Byers SL, Wiles MV, Dunn SL, Taft RA (2012) Mouse estrous cycle identification tool and
images. PLoS ONE 7:e35538
Colucci-Guyon E, Portier MM, Dunia I, Paulin D, Pournin S, Babinet C (1994) Mice lacking
vimentin develop and reproduce without an obvious phenotype. Cell 79:679–694
Condamine H, Custer RP, Mintz B (1971) Pure-strain and genetically mosaic liver tumors histo-
chemically identified with the -glucuronidase marker in allophenic mice. Proc Natl Acad Sci
USA 68:2032–2036
Danforth CH (1947) Heredity of polydactyly in the cat. J Heredity 38:107–112
De Repentigny Y, Kothary R (1996) An improved method for artificial insemination of mice–
oviduct transfer of spermatozoa. Trends Genet 12:44–45
de Vries H (1910) Intracellular pangenesis (trans from German: Stuart Gager C). The Open Court
Publishing Co., Chicago
Detlefsen JA (1921) A new mutation in the house mouse. Amer Nat 55:469–476
Dietrich WF, Lander ES, Smith JS, Moser AR, Gould KA, Luongo C, Borenstein N, Dove W
(1993) Genetic identification of Mom-1, a major modifier locus affecting Min-induced
intestinal neoplasia in the mouse. Cell 75:631–639
Eicher EM, Hoppe PC (1973) Use of chimeras to transmit lethal genes in the mouse and to dem-
onstrate allelism of the two X-linked male lethal genes jp and msd. J Exp Zool 183:181–184
Fox JG, Barthold SW, Davisson MT, Newcomer CE, Quimby FW, Smith AL (2007) The mouse
in biomedical research, 2nd edn. Elsevier, New York
Gardner RL, Lyon MF (1971) X chromosome inactivation studied by injection of a single cell
into the mouse blastocyst. Nature 231:385–386
Glenister PH, Rall WF (2000) Cryopreservation and rederivation of embryos and gametes. In:
Jackson IJ, Abott CM (eds) Mouse genetics and transgenics: a practical approach. Oxford
University Press, Oxford
Grüneberg H (1952) The genetics of the mouse. Martinus Nijhoff, The Hague
References 47
Guénet JL, Babinet C (1978) The hairy ear mutation (Eh) is not cell lethal. Mouse News Letter
58:67
Guénet JL, Marchal G, Milon G, Tambourin P, Wendling F (1979) Fertile dominant spotting in
the house mouse. J Hered 70:9–12
Hauschka TS, Jacobs BB, Holdridge BA (1968) Recessive yellow and its interaction with belted
in the mouse. J Hered 59:339–341
Hedrich HJ (2012) The laboratory mouse, 2nd edn. Elsevier Academic Press, Amsterdam
Hogan ME, King LE Jr, Sundberg JP (1995) Defects of pelage hairs in 20 mouse mutations. J
Investig Dermatol 104(5 Suppl):31S–32S
Illmensee K, Kaskar K, Zavos PM (2005) Efficient blastomere biopsy for mouse embryo splitting
for future applications in human assisted reproduction. Reprod Biomed Online 11:716–725
Kamimura S, Inoue K, Ogonuki N, Hirose M, Oikawa M, Yo M, Ohara O, Miyoshi H, Ogura A
(2013) Mouse cloning using a drop of peripheral blood. Biol Reprod 89:24
Kang L, Wang J, Zhang Y, Kou Z, Gao S (2009) iPS cells can support full-term development of
tetraploid blastocyst-complemented embryos. Cell Stem Cell 5:135–138
Kaufman MH, O’Shea KS (1978) Induction of monozygotic twinning in the mouse. Nature
276:707–708
Kobayashi T, Yamaguchi T, Hamanaka S, Kato-Itoh M, Yamazaki Y, Ibata M, Sato H, Lee YS,
Usui J, Knisely AS, Hirabayashi M, Nakauchi H (2010) Generation of rat pancreas in mouse
by interspecific blastocyst injection of pluripotent stem cells. Cell 142:787–799
Leckie PA, Watson JG, Chaykin S (1973) An improved method for the artificial insemination of
the mouse (Mus musculus). Biol Reprod 9:420–425
Luo C, Zuñiga J, Edison E, Palla S, Dong W, Parker-Thornburg J (2011) Superovulation strate-
gies for 6 commonly used mouse strains. J Am Assoc Lab Anim Sci 50:471–478
McLaren A, Molland P, Signer E (1995) Does monozygotic twinning occur in mice? Genet Res
66:195–202
Miko I (2008) Phenotype variability: penetrance and expressivity. Nature Education 1:137
Mintz B (1962) Formation of genetically mosaic mouse embryos. Am Zool 2:432
Mintz B, Baker WW (1967) Normal mammalian muscle differentiation and gene control of isoci-
trate dehydrogenase synthesis. Proc Natl Acad Sci USA 58:592–598
Mintz B, Illmensee K (1975) Normal genetically mosaic mice produced from malignant terato-
carcinoma cells. Proc Natl Acad Sci USA 72:3489–3585
Mintz B, Silvers W (1967) “Intrinsic” immunological tolerance in allophenic mice. Science
158:1484–1487
Mochida K, Hasegawa A, Taguma K, Yoshiki A, Ogura A (2011) Cryopreservation of Mouse
Embryos by Ethylene Glycol-Based Vitrification. J Vis Exp 57:e3155. doi:10.3791/3155
Moser AR, Pitot HC, Dove WF (1990) A dominant mutation that predisposes to multiple intesti-
nal neoplasia in the mouse. Science 247:322–324
Nadeau JH (2003) Modifier genes and protective alleles in humans and mice. Curr Opin Genet
Dev 13:290–295
Nagy A, Gertsenstein M, Vintersten K, Behringer R (2003) Manipulating the mouse embryo, a
laboratory manual, 3rd edn. Cold Spring Harbor Press, New York
Nakagata N (2011) Cryopreservation of mouse spermatozoa and in vitro fertilization. Methods
Mol Biol 693:57–73
Nakagata N, Takeo T, Fukumoto K, Haruguchi Y, Kondo T, Takeshita Y, Nakamuta Y, Umeno
T, Tsuchiyama S (2014) Rescue In Vitro Fertilization Method for Legacy Stock of Frozen
Mouse Sperm. J Reprod Dev 60(2):168–171
Nishizono H, Shioda M, Takeo T, Irie T, Nakagata N (2004) Decrease of fertilizing ability of mouse
spermatozoa after freezing and thawing is related to cellular injury. Biol Reprod 71:973–978
Ogonuki N, Inoue K, Hirose M, Miura I, Mochida K, Sato T, Mise N, Mekada K, Yoshiki A, Abe
K, Kurihara H, Wakana S, Ogura A (2009) A high-speed congenic strategy using first-wave
male germ cells. PLoS ONE 4:e4943. doi:10.1371/journal.pone.0004943
Ogonuki N, Inoue K, Ogura A (2011) Birth of normal mice following round spermatid injection
without artificial oocyte activation. J Reprod Dev 57:534–538
Ogonuki N, Mochida K, Miki H, Inoue K, Fray M, Iwaki T, Moriwaki K, Obata Y, Morozumi K,

Yanagimachi R, Ogura A (2006) Spermatozoa and spermatids retrieved from frozen repro-
ductive organs or frozen whole bodies of male mice can produce normal offspring. Proc Natl
Acad Sci USA 103:13098–13103
Ogura A, Ogonuki N, Inoue K, Mochida K (2003) New microinsemination techniques for labora-
tory animals. Theriogenology 59:87–94
Ogura A, Ogonuki N, Miki H, Inoue K (2005) Microinsemination and Nuclear Transfer Using
Male Germ Cells. Int Rev Cytol 246:189–229
Olds-Clarke P (1989) Sperm from tw32/+ mice: capacitation is normal, but hyperactivation is
premature and nonhyperactivated sperm are slow. Dev Biol 131:475–482
Panthier JJ, Guénet JL, Condamine H, Jacob J (1990) Evidence for mitotic recombination in
W(ei)/+ heterozygous mice. Genetics 125:175–182
Papaioannou VE, McBurney MW, Gardner RL, Evans MJ (1975) Fate of teratocarcinoma cells
injected into early mouse embryos. Nature 258:70–73
Perez CJ, Jaubert J, Guénet J-L, Barnhart KF, Ross-Inta CM, Quintanilla VC, Aubin I, Brandon
J, Otto N, DiGiovanni J, Gimenez-Conti I, Giulivi C, Kusewitt DF, Conti CJ, Benavides F
(2010) Two hypomorphic alleles of mouse Ass1 as a new animal model of citrullinemia type
I, and other hyperammonemic syndromes. Am J Pathol 177:1958–1968
Silvers WK (1979) The coat colors of mice: a model for mammalian gene action and interaction.
Springer, Berlin
Stewart CL, Gadi I, Bhatt H (1994) Stem cells from primordial germ cells can reenter the germ
line. Dev Biol 161:626–628
Sztein J, Vasudevan K, Raber J (2010) Refinements in the cryopreservation of mouse ovaries. J
Am Assoc Lab Anim Sci 49:420–422
Sztein JM, Farley JS, Mobraaten LE (2000) In vitro fertilization with cryopreserved inbred
mouse sperm. Biol Reprod 63:1774–1780
Taft RA, Low BE, Byers SL, Murray SA, Kutny P, Wiles MV (2013) The perfect host: a mouse
host embryo facilitating more efficient germ line transmission of genetically modified embry-
onic stem cells. PLoS ONE 8:e67826. doi:10.1371/journal.pone.0067826
Tarkowski A (1998) Mouse chimaeras revisited: recollections and reflections. Int J Dev Biol
42:903–908
Tarkowski AK (1961) Mouse chimaeras developed from fused eggs. Nature 190:857–860
Theiler K (1972) The house mouse. Springer, New York
Van der Lee S, Boot LM (1956) Spontaneous pseudopregnancy in mice II. Acta Physiol
Pharmacol Neerl 5:213–215
Vergara GJ, Irwin MH, Moffatt RJ, Pinkert CA (1997) In vitro fertilization in mice: Strain dif-
ferences in response to superovulation protocols and effect of cumulus cell removal.
Theriogenology 47:1245–1252
Wakayama S, Kohda T, Obokata H, Tokoro M, Li C, Terashita Y, Mizutani E, Nguyen VT,
Kishigami S, Ishino F, Wakayama T (2013) Successful serial recloning in the mouse over
multiple generations. Cell Stem Cell 12:293–297
Wakayama T, Perry AC, Zuccotti M, Johnson KR, Yanagimachi R (1998) Full-term development
of mice from enucleated oocytes injected with cumulus cell nuclei. Nature 394:369–374
West JD, Frels WI, Papaioannou VE, Karr JP, Chapman VM (1977) Development of interspecific
hybrids of Mus. J Embryol Exp Morphol 41:233–243
Whitten WK (1956) Modification of the oestrous cycle of the mouse by external stimuli associ-
ated with the male. J Endocrinol 13:399–404
Whittingham DG (1968) Fertilization of mouse eggs in vitro. Nature 220:592–593
Whittingham DG, Leibo SP, Mazur P (1972) Survival of mouse embryos frozen to −196 degrees
and −269 degrees C. Science 178:411–414
Wilmut I (1972) The effect of cooling rate, warming rate, cryoprotective agent and stage of devel-
opment on survival of mouse embryos during freezing and thawing. Life Sci II 11:1071–1079
Wolfe HG (1967) Artificial insemination of the laboratory mouse (Mus musculus). Lab Anim
Care 17:426–432
References 49
Didactic movie
http://www.jove.com/video/3155/cryopreservation-mouse-embryos-ethylene-glycol-based
Reference Book
Papaioannou VE, Behringer R (2005) Mouse phenotypes: a handbook of mutation analysis.

CSHL Press, New York, p 235
Chapter 3
Cytogenetics
3.1 Introduction
Cytogenetics, as the name indicates, lies at the intersection between cell biology
and genetics. It came into being as an independent discipline after the advent of
the chromosomal theory of heredity at the beginning of the twentieth century,
when W.S. Sutton and T.H. Boveri (then T.H. Morgan) identified the chromosomes
as the physical structures on which the genes were anchored (1902–1915).1
Cytogenetics deals with all aspects of chromosomes biology: their morphology
and structure, their number, and their behavior during mitosis and meiosis. It also
deals with the pathology and functional changes associated with accidental varia-
tions in chromosome number or structure.
For a long time cytogenetics remained a rather descriptive discipline, but in
recent years, particularly after the development of highly sophisticated staining
techniques and because of the availability of enormous collections of chromo-
somal rearrangements, it has contributed to the development of genetic maps, to
a better understanding of the developmental consequences of chromosomal aber-
rations in humans (Down syndrome, for example), and to the elucidation of some
fundamental biological questions such as genomic imprinting (see Chap. 6).
With the recent advent of fluorescence in situ hybridization (FISH) techniques,
cytogenetics has changed profoundly in the sense that it is now possible to visu-
alize the chromosomes, or a specific region of them, in the interphasic nucleus.
Taking advantage of the possibilities offered by such “interphasic cytogenetics”,
it has thus become possible to obtain information about chromosome number and
structure in all tissues, at any time, independent of the cell cycle. In turn, inter-
phasic cytogenetics has made possible the development of “functional cytoge-
netics”, providing information relative to some epigenetic mechanisms of gene
1 Thomas H. Morgan was awarded the Nobel Prize in Physiology or Medicine in 1933 for his
discoveries concerning the role played by the chromosomes in heredity. Morgan proposed that
each chromosome contains a collection of small units called “genes” that were linearly arranged
on the chromosomes.

52 3 Cytogenetics
regulation; for example, in relation to imprinting. Finally, over the past few years,
cytogenetics has benefited from the exuberant development of the molecular tech-
niques of genetic engineering in embryonic stem cells (ES cells) (see Chap. 8). As
a result, virtually any chromosomal rearrangement (deletion, duplication, trans-
position, inversion, etc.) of interest to researchers can now be designed in silico,
engineered in vitro, and ideally analyzed in the context of a living mouse.
3.2 The Chromosomes of the Mouse
A simple way to observe mouse chromosomes is to collect a bone marrow sample

and disperse it into tissue culture medium, to which is added 1–2 drops of a
0.025 % colchicine solution. After 20–30 min of such treatment, all actively divid-
ing cells of the bone marrow are blocked at the metaphase stage because the drug
prevents the synthesis of spindle fibers and, therefore, stops mitosis in metaphase.
If these arrested cells are transferred into a hypotonic solution (for example,
0.56 % KCl) for a few minutes and then placed on a glass slide, they swell and
burst open, and the chromosomes appear spread out on the slide. They can then be
fixed with a simple fixative medium (for example, a 3:1 cold mixture of methanol
and glacial acetic acid) and observed under the microscope, either in phase con-
trast or after staining, and counted. Such a simple technique allowed it to be
shown, almost a century ago, that the normal laboratory mouse has 40 chromo-
somes (Cox 1926).2 The word “chromosome” comes from the Greek chroma
meaning color and soma meaning body, referring to their tendency to be strongly
stained by stains like orcein or Sudan black (Fig. 3.1).
Fig. 3.1 Spermatogonial metaphase of the mouse (Sudan black squash). In this figure, which
was published in the book Biology of the Laboratory Mouse (Dover Publications, 1966, Chap. 7),
the 40 chromosomes of the mouse are clearly visible but it is very difficult to specifically identify
the elements of the different pairs
2 A technique for rapid chromosome staining is provided at: http://www.jax.org/cyto/marrow_pr
eps_alt.html.
3.2 The Chromosomes of the Mouse 53
For many years, chromosomes were represented as a core of proteins, mostly

but not only histones, to which a filament of supercoiled DNA was tightly
attached, giving each chromosome its specific shape. Nowadays, this folding of
DNA into coils and loops of increasing diameter is being reconsidered and sci-
entists have several alternative descriptions of the intimate organization of the
DNA molecule inside the mitotic chromosomes (Uhlmann 2013). Histones serve
as a frame to hold the association in a compact structure and also to control DNA
transcription, as we will discuss later. This association of DNA with histones,
and some other proteins of the polycomb group, is given the generic name of
chromatin. Two types of chromatin can be distinguished at the cellular level: (i)
euchromatin, whose structure is open and moderately condensed, allowing the
transcription of a variety of RNAs, and (ii) heterochromatin, whose structure is
highly condensed, impeding DNA transcription. Heterochromatin contains repeti-
tive sequences and a special form of it is constitutive heterochromatin, which is
located mostly at or near the centromeres. Cell biologists hypothesize that consti-
tutive heterochromatin probably has only a structural function, while euchromatin
has a function in gene expression and regulation.
When cells enter the process of mitotic division the chromatin becomes pro-
gressively condensed, the chromosomes become shorter in size and clearly visible.
They then split into two chromatids that remain attached for a period of time by
their centromere. The chromatids bind to the microtubule of the mitotic spindle
by a special protein structure: the kinetochore. Then, the centromeres themselves
split, the mitotic process goes on, and the chromatids become individualized in
chromosomes that are pulled apart to the opposite poles of the cell. Finally, this
results in two daughter cells, each with its own replica of the original set of chro-
mosomes. After mitotic division, the cells return to interphase, the chromatids
uncoil, and DNA can again be transcribed.
The chromosomes vary in size according to the stage of the cell cycle in which
they are observed and the degree of chromatin condensation. They also vary in
shape and exist either as single linear strands (unduplicated chromosomes) or as
duplicated chromosomes, just before anaphase, when the two chromatids are still
joined by their centromere.
As long as it is not duplicated, each chromosome contains a single and unique
molecule of supercoiled DNA. The mammalian genome is thus fragmented into
discrete elements that consist of a linear array of genes encoding RNAs and pro-
teins, interspersed with noncoding DNA. They also contain a special DNA seg-
ment that constitutes the centromere and two special stretches at their ends: the
telomeres.
In the mouse, the centromeres (symbol Cen) consist of a relatively large array
of repetitive heterochromatin containing satellite DNA (see Chap. 5), where the
sequence within the individual repeats is similar but not identical. The normal his-
tone H3 is replaced by a variant (CENP-A) that is believed to be important for the
assembly of the kinetochore.
The telomeres (symbol Tel) consist of a variable number of the tandemly
repeated motif–5′-TTAGGG-3′–bound to specialized proteins and measuring
54 3 Cytogenetics
Fig. 3.2 Telomere length. Telomere length of BALB/c (M. musculus), M. spretus, and
(BALB/c × M. spretus) F1 somatic cells. Genomic DNA was subjected to restriction digestion
with the enzymes HinfI and RsaI, and analyzed by pulse field electrophoresis. BALB/c mice have
“long” telomeres, while M. spretus mice have “short” telomeres (from Zhu et al. 1998)
up to 40 kb in the laboratory mouse. For many years, the telomeres had been con-
sidered as mere insulators at the end of the chromosomal DNA filament, playing
a role similar to the role played by the plastic caps that are molded at the ends of
our shoelaces; to prevent their ends from being damaged or accidentally tied to
each other by the DNA repair enzymes. Nowadays, some geneticists hypothesize
that the enzyme telomerase, whose function is, in particular, to control the length
of telomeres by adding new–5′-TTAGGG-3′–monomers, plays a crucial role in
the control of cell senescence, and that telomere length is a way of monitoring
the cell’s replicative potentialities, with long telomeres being indicative of greater
potentialities (Blasco 2005; Sahin and De Pinho 2010). However, this point is still
debated because the comparison of telomere sizes in mice from different species
but of the same genus Mus does not support the hypothesis that telomere shorten-
ing is correlated with senescence in mice (Kim et al. 2003). Laboratory mice (Mus
musculus), for example, have long telomeres, while wild mice of the species Mus
spretus have short ones (i.e., like human) but a very similar lifespan and behavior
(Zhu et al. 1998). Similarly, mice homozygous for a knockout allele of Tert (the
gene encoding telomerase reverse transcriptase) do not exhibit any phenotypes
related to accelerated senescence (Fig. 3.2).
3.3 Identifying the Chromosome Pairs: The Normal

Karyotype
In most mammalian species, including human, one can generally observe three
kinds of chromosomes, depending on the position of the centromere. When the
centromere is roughly in the middle of the chromosome, the latter is said to be
metacentric. When it is slightly shifted and divides the chromosome unequally,
3.3 Identifying the Chromosome Pairs: The Normal Karyotype 55
(a) (b) (c) (d)
Fig. 3.3 Sorting out the different mouse chromosomes. Chromosomes are generally classified
according to their size and shape. a Metacentric; b Sub-metacentric; c Acrocentric; d Telocentric.
Telocentrics are an extreme form of acrocentrics. The centromeric index is computed based on
the ratio p/p + q, where p is the size of the short arm and q the size of the long arm. Metacentric
chromosomes have a centromeric index of 0.5 (p/p + q = 0.5). Acrocentric chromosomes have a
centromeric index <0.5. Traditional laboratory mice have acrocentric chromosomes only
with long arms and short arms, the chromosome is said to be sub-metacentric.
Finally, when the centromere is subterminal, the chromosome is said to be acro-
centric. In literature from the 1960s some chromosomes were depicted as telo-
centric with their centromere completely shifted to one end. Specialists now
consider that telocentric chromosomes are unstable structures and probably do
not exist in reality. Chromosomes are also classified according to two criteria:
first, their global size, and second, their centromeric index. The centromeric
index is computed based on the ratio p/p + q, where p (from the French petit
meaning small) is the size of the short arm and q the size of the long arm (from
index of 0.5 (p/p + q = 0.5), while acrocentrics have a centromeric index ≪0.5

the French queue meaning tail). Metacentric chromosomes have a centromeric
(Fig. 3.3).
The chromosome set of a given species is generally presented with the chro-
mosomes being displayed according to their size, from the largest to the smallest,
and, when of the same size, according to the centromeric index. Once arranged
as described, the chromosome set of a given species represents its karyotype. The
karyotype is a fundamental parameter that is generally unique to a species. Thus,
the karyotype of the normal laboratory mouse (MMU or sometimes Mmu for M.
musculus) consists of 40 acrocentric chromosomes, i.e., 19 autosomes plus an X
and a Y (in short 40,XY). Unfortunately, unless these chromosomes are stained
with a special technique (see below), it is difficult to individually identify the
members of the different pairs. Chromosome 1 represents 7.5 % of the mouse hap-
loid genome, the X chromosome is the fifth in size and represents 6.60 % of the
haploid genome, and chromosome 19 is the smallest entity and represents approx-
imately 2.3 % of the haploid genome. The Y chromosome has the same size as
chromosome 18 (3.5 %) (Fig. 3.4).
56 3 Cytogenetics
Fig. 3.4 Ideogram of the mouse chromosomes. Ideogram of the metaphasic chromosomes of a

female mouse (from the book Biology of the Laboratory Mouse, Dover Publications, 1966). The
two chromosome 19s are clearly identifiable by their short size on this preparation
Wild mice sometimes exhibit variations in their karyotype concerning the total
number of centromeres but not the total number of arms.3 For example, some wild
mice living in Western Europe (Switzerland, southern Germany, northern Italy) and
North Africa have a karyotype with a variety of metacentric chromosomes. All these
variations are considered normal and do not affect the fertility and/or viability of the
animals as long as they are homozygous. However, when these mice are crossed
with laboratory strains, whose karyotypes are composed of acrocentric chromo-
somes only, they produce normal, healthy F1, but the latter are almost always
completely sterile because they produce gametes with abnormal (unbalanced) sets of
chromosomes (≠n). We will come back to this point later in this chapter.
During the early 1970s, considerable progress was made concerning the tech-
niques used to stain the chromosomes. The first of these techniques was reported
for human chromosomes by T. Caspersson and colleagues. It makes use of the flu-
orescent dye quinacrine (Caspersson et al. 1970, 1972; Miller et al. 1971; Miller
and Miller 1975) and yields a series of lightly and darkly stained bands on the
chromosome arms. This banding pattern is characteristic of each and every pair of
chromosomes, or nearly so. These bands are called the Q bands and the technique
is still in use, although infrequently.
Another popular technique was developed almost simultaneously (Sumner et al.
1971), and was based on the controlled enzymatic digestion of chromatin with
either trypsin or chymotrypsin, followed by conventional staining with the Giemsa
dye. The banding pattern characteristic of this technique rapidly became popular
3 The total number of chromosome arms per set of chromosomes is called the fundamental num-
ber (nombre fondamental) after Matthey.

Fig. 3.5 a The mouse karyotype. The mouse karyotype stained by the Giemsa dye after trypsin
digestion. This G-banding pattern allows the individual identification of each chromosome
pair with only a few ambiguities (8–12; 9–13). (This figure is courtesy of Dr. Heinz Winking,
Insitut für Biologie, Medizinische Universität zu Lübeck, Germany). b The mouse ideogram is
a schematic and standardized representation of the different chromosome bands (Figure from
Dr. David Adler, University of Washington Department of Pathology). http://www.pathology.
washington.edu/research/cytopages/idiograms/mouse
and is referred to as G-banding: dark regions are heterochromatic and AT-rich,

while light regions are euchromatic and GC-rich (Sawyer et al. 1987). This tech-
nique using trypsin/Giemsa is still used today. The G-banding pattern is highly
reproducible and the schematic representation of the bands in the mouse is called
the ideogram. The Q-banding pattern is very similar to the G pattern, and, being
technically more difficult to achieve, this explains why it has been progressively
replaced by the G-banding technique.
These methods will normally produce around 350 bands in a normal mouse
karyotype (Nesbitt and Francke 1973) (Fig. 3.5).
Using the same Giemsa staining with some technical variations, other banding
patterns have been described in the past:
• R-banding, which is the reverse of G-banding (the R stands for “reverse”). The
dark regions are euchromatic (GC-rich regions), while the bright regions are
heterochromatic (AT-rich regions).
• C-banding: Giemsa binds to constitutive heterochromatin and stains the
centromeres.
• T-banding that visualizes the telomeres.
All these techniques, except for G-banding, have now been replaced by more
modern methods that have been transferred from human cytogenetics. These
58 3 Cytogenetics
techniques make use of different fluorochromes and a variety of (human or mouse)

chromosome-specific DNA probes and they are, at the same time, more reliable
and easier to standardize. They can stain specifically an entire mouse chromo-
some, if the DNA probe is specific for this mouse chromosome, or the fragment
of a mouse chromosome that is homologous to a human chromosome, if the
DNA probe is specific for a human chromosome. These techniques for “chromo-
some painting” are now standardized and are very helpful for the transposition of
information from one species to the next, and can be applied to any domestic spe-
cies if necessary. Chromosome painting is also useful for analyzing chromosome
rearrangements.
Fluorescence in situ hybridization (FISH) is an interesting technique when
used for the localization of a specific gene or transgene in the mouse. One can,
for example, stain the karyotype with either the G-banding technique or any other
chromosome painting technique, and look for the localization of a specific DNA
sequence (a gene or a transgene, for example) with another probe labeled either
with a fluorochrome or a radioactive isotope. The localization of a transgene,
which has always been problematic, has been greatly simplified by the use of such
staining techniques even though, in many cases, the localization is only roughly
assessed (Fig. 3.6).
Spectral karyotyping is a molecular technique used to visualize the differ-
ent pairs of chromosomes of a given species in different colors (Liyanage et al.
1996). This is achieved by labeling chromosome-specific DNA with different fluo-
rophores and then using these probes for hybridization with the chromosomes.
Fig. 3.6 Mapping a transgenic insertion. Observation of a transgene in the mouse by fluores-

cence in situ hybridization with a molecular probe (stained in red by rhodamine). The transgenic
insertion is on mouse chromosome 12 that appears labeled green by chromosome-specific paint-
ing. The technique is of great help for the derivation of a mouse strain homozygous for the trans-
genic insertion. (This figure is courtesy of Dr. M.G. Mattei, Hôpital de la Timone, Marseille)
Fig. 3.7 Spectral karyotyping. The figure represents the karyotype from a normal male mouse
(strain C57BL/6) generated by using the multi-color FISH cytogenetic method called spectral
karyotyping or SKY. The colors to the left of each black and white (G-banded) chromosome
(derived from inverted-DAPI staining) are the RGB (red–green–blue) display of the fluoro-
chromes. The pseudo-color to the right of the Giemsa-banded chromosomes are referred to as the
classified colors and are derived from a mathematical algorithm that translates the wavelengths
of each chromosome, for each pixel, and converts the wavelength chromosome-specific assign-
ments into these classified colors. With this technique, cytogeneticists are now able to visual-
ize complex karyotypes which involve multi-chromosomal rearrangements in an unambiguous
manner. Contrary to traditional black-and-white karyotypes, visualization of chromosomal rear-
rangements with spectral karyotyping is straightforward, as one or more colors will show within
a single chromosome. (This figure is courtesy of Drs. Hesed M. Padilla-Nash and Thomas Ried,
Genetics Branch, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA)
(Liyanage et al. 1996; Green and Ried 2011)
Because there is a limited number of spectrally distinct fluorophores, a combi-

natorial labeling method is used to generate different colors. Spectral differences
generated by combinatorial labeling are captured and analyzed by using an inter-
ferometer attached to a fluorescence microscope and then processing the pictures
with a computer program that assigns a pseudo-color to each spectrally different
combination, allowing the visualization of the individually colored chromosomes
(Fig. 3.7).
Finally, taking advantage of the numerous variations in mouse karyotypes that
are easily available nowadays, mouse chromosomes have been partially sorted by
using flow cytometry. This has enabled the preparation of chromosome-specific
DNA probes useful for mapping projects or for labeling (Baron et al. 1990).
3.4 Meiosis and Gametogenesis
Meiosis is a fundamental event in the life cycle of organisms reproducing sexu-

ally because, with the production of gametes, it marks the end of diplophase and
the beginning of haplophase. Meiosis has been extensively studied and appears to
60 3 Cytogenetics
be remarkably similar across different species. In some aspects it resembles mito-

sis, another crucial step in the somatic cell cycle, but it has several fundamental
differences.
1. While mitosis occurs in all types of cells and tissues, at least when the embryo
differentiates and develops, meiosis occurs exclusively in the gonads, and more
precisely in certain cells of the germinal lineage.
2. The end products of mitosis are two diploid (2n) daughter cells, which are abso-
lutely identical from a genetic point of view, except in rare cases. These cells have
chromosomes that are similar in number and structure to those in the mother cell
and carry the same genes with the same linear ordering. In meiosis the situation is
radically different: there are four daughter cells instead of two, and these cells are
genetically unique in the sense that they have a unique, de novo assortment of the
parental alleles and their chromosome number is halved (they are haploid).
3. During meiotic prophase (diplonema—diakinesis) the chromosomes are dupli-
cated, creating two exact copies (or sister chromatids), but remain attached by
their centromere. The maternal and paternal chromatids then pair and exchange
segments by homologous recombination, leading to a patchwork between the
maternal and paternal versions of the chromosome. The pairing between homolo-
gous chromosomes (between the two pairs of sister chromatids) is mediated by
a protein structure known as the synaptonemal complex. This structure is tem-
porary and disappears during late prophase. This inter-chromatid recombination
(chiasmata) is specific to the meiotic process and is an efficient mechanism for
the generation of diversity.
In male mice, meiosis occurs in spermatocytes I, producing spermatocytes II that later
differentiate progressively into spermatids and finally become mature sperm cells
(spermatozoa). Spermatogenesis in the mouse starts at about 3 weeks of age, puberty
is reached by 6 weeks, and by 8 weeks of age a male is normally fully fertile. The
duration of spermatogenesis is shorter in the mouse than in most other mammalian
species and the time taken by spermatogonia to become mature spermatids, which are
released into the lumen, is only 5 weeks. In theory, a single A1 spermatogonium gives
rise to 256 sperm cells, but there is some attrition of cells during spermatogenesis and
the actual number is smaller than this theoretical maximum (Fig. 3.8).
In female mice, meiosis begins immediately after the migration of the pri-
mordial germ cells into the ovary while the young female is still in utero. It
is interrupted at the diplotene stage of meiosis I and the cells remain resting
for a very long period. Then, stimulated by a burst of gonadotropin hormones
released by the pituitary gland, groups of oocytes resume meiosis I and stop
again at meiosis II until fertilization. At fertilization, when a spermatozoon
penetrates the oocyte, meiosis is completed with the expulsion of the second
polar body. In female mice, unlike in males, meiosis is a discontinuous process
and its end-products are very unequal with two very small cells, a pronucleus I
(first polar body with 2n chromosomes) and a pronucleus 2 (second polar body
with n chromosomes), and one enormous cell, the oocyte with n chromosomes.
However, it is important to note that, in the female, haploidy of the gamete is
3.4 Meiosis and Gametogenesis 61
2n 1
n 7
Fig. 3.8 Schematic representation of the meiotic process. 1. The oogonia or spermatogonia stem

cells are diploid (2n chromosomes). 2. After one round of DNA synthesis, the chromosomes
duplicate and create two exact copies of both the paternal and maternal chromosomes. These
two copies, the chromatids, remain attached by their centromere. 3. The maternal and paternal
chromosomes, although divided in chromatids, pair and exchange parts by homologous recom-
bination (crossing over). Chiasmas (*) can be observed at this stage. 4. The centromere of each
chromosome pair does not divide, but binds to the spindle fibers. They then split and the spindle
fibers pull the chromatids to the opposite pole of the cell. This is the first disjunction. 5. Sister
chromatids remain together and form another equatorial plaque. 6. The centromeres split and
the individual chromosomes migrate to the opposite pole of the cell. This is the second disjunc-
tion. 7. Four haploid cells are formed with only n chromosomes. Some of these chromosomes
are recombinant and have genes from both parental chromosomes. In the male, the four products
of meiosis are equivalent (spermatocytes II). In the female, one of the daughter cells at step 5
degenerates and forms the first polar body (with 2n chromosomes). Another one degenerates at
step 7, when the spermatozoon penetrates the oocyte, and forms the second polar body (n chro-
mosomes). The formation of the polar body is a random event and does not depend upon the
genetic makeup of the cell (with a possible exception for XO females—see text)
only virtual because the expulsion of the second polar body (and its n chromo-
somes; late anaphase II) is triggered by the penetration of the spermatozoon
into the oocyte.
62 3 Cytogenetics
Meiosis is a complex process involving several critical steps. Errors can occur
at many of the steps, often leading to severe or even lethal abnormalities in the
embryo. This will be the topic of the next sections in this chapter.
3.5 Variations in Chromosome Number
Following defective gametogenesis or abnormal fertilization, variations in chro-

mosome number may occur accidentally and an embryo may then start its devel-
opment with a set of chromosomes different from the fundamental diploid number
2n. These aberrations, called heteroploidies, are common to all mammalian spe-
cies including human and are, in most instances, incompatible with normal devel-
opment and often result in abortions. In the mouse, some heteroploidies represent
interesting models for understanding the homologous human pathologies.
Geneticists sort these heteroploidies into two different subcategories accord-
ing to the number of chromosomes involved. Euploid heteroploidies designate the
case where the number of chromosomes in the conceptus is a multiple of the hap-
loid number (n) (i.e., n, 3n, 4n, etc.). Aneuploid heteroploidies correspond to all
other cases where there is a deviation from the normal 2n (for example, 2n + 1,
2n − 1, 2n + 1 + 1 etc.).
3.5.1 The Euploid Heteroploidies
3.5.1.1 Haploidy (n)
In natural conditions, haploid embryos occur spontaneously but, as a rule, they are
lethal at a very early stage of development. As we will discuss later (Chap. 6), a
mammalian embryo cannot develop to term unless some of its chromosomes come
from a male parent and others from a female parent. This, of course, represents a
serious constraint on the development of haploid embryos.4
However, considering that haploid organisms have been created in several spe-
cies, including vertebrates such as the medaka fish (Oryzias latipes), experiments
have also been undertaken in the mouse. Haploid embryonic stem cells (ESCs) have
been produced in vitro that were derived from parthenogenetic or androgenetic hap-
loid embryos of several inbred mouse strains, collected at the blastocyst stage.
These haploid ESCs have been proven capable of a differentiation potential simi-
lar to that of diploid ESCs, and some have been used for genetic screening as well as
for the production of homozygous mutant animals. These cells, however, are unsta-
ble and often spontaneously return to the diploid state. Two publications can be rec-
ommended concerning this subject (Leeb and Wutz 2011; Zhang and Teng 2013).
4 UpD are exceptions that will be discussed later in this chapter and in Chap. 6.
3.5 Variations in Chromosome Number 63
3.5.1.2 Triploidy (3n)
Triploidy represents 2–3 % of human pregnancies and culminates in early sponta-

neous miscarriages. These heteroploidies result either from digyny (the basic 2n
plus an extra haploid set of maternal origin), and originate through errors in meio-
sis II, or, in the majority of cases, from diandry (2n plus an extra haploid set of
paternal origin) and originate from dispermy. Births of living triploid infants have
been recorded, but these neonates suffer from multiple developmental defects and
die shortly after birth.
In mice, micromanipulatory techniques have been used to produce digynic
(2nm + np) and diandric (nm + 2np) triploid embryos (Niemierko 1981). These con-
ceptuses revealed an ability to implant once transplanted into the uterus of suit-
able recipients and some developed up to the 15- to 25-somite stage. However,
here again, aneuploid embryos appeared considerably smaller than euploids ana-
lyzed at the same stages of development. In the mouse, in contrast to what was
described in human, digynic triploid conceptuses showed poorer embryonic devel-
opment than the diandrics, but both were morphologically abnormal (Kaufman
et al. 1989). Chimeric embryos created from triploid cells and normal diploid cells
can progress to term, and some can even survive, depending on the ratio of the
3n/2n cells.
3.5.1.3 Tetraploidy (4n)
The generation of tetraploid embryos by electrofusion of 2n blastomeres of mouse

embryos at the 2-cell or 4-cell stages of development has been reported by Kubiak
and Tarkowski (1985). By applying a pulse of current, the authors succeeded in gen-
erating tetraploid embryos that could develop to the blastocyst stage but not further.
Since these early experiments, tetraploid blastocysts have been used as recipients
for the production of mouse chimeras derived from either genetically engineered ES
cells or from diploid embryonic cells. In these cases there is complete segregation
of the descendants of the ES cells. The tetraploid cells do not contribute at all to the
formation of the embryo proper, but instead create the primitive endoderm deriva-
tives and the trophectoderm (Naumann 2008). This method has also been success-
fully used for analyzing genes known to be heterozygous embryonic lethal as well as
to rescue embryonic lethality caused by an additional maternally inherited X chro-
mosome in the mouse (Carmeliet et al. 1996; Goto and Takagi 1998).
3.5.2 The Aneuploid Heteroploidies
Aneuploid heteroploidies are very common in the human species, where they repre-
sent up to 50 % of miscarriages. These heteroploidies are of two kinds, depending
on whether the aberration results from the loss or from the gain of one (or more)
64 3 Cytogenetics
chromosome(s). Monosomies and nullisomies5 characterize respectively the situa-

tion where a single chromosome or the two chromosomes of a given pair is (are)
missing. Trisomies or tetrasomies represent the opposite situation, where the karyo-
type displays one or two extra copies of the same chromosome pair. All these
numerical abnormalities result from errors occurring either during gametogenesis or
at fertilization, and this explains why a karyotype with more than four copies of the
same chromosome is virtually never observed in practice.
Aneuploid heteroploidies have all been modeled and studied systematically in
the mouse because, as we will explain later they can be produced experimentally
almost at will, with high frequency (see section referring to Robertsonian translo-
cation later in this chapter). A number of conclusions have been drawn from these
experiments and observations that we summarize here, but it must be kept in mind
that the phenotype of the aneuploid heteroploidies, the trisomies in particular,
depends upon the chromosome involved in the primary defect.
3.5.2.1 Nullisomies and Monosomies
As a rule, autosomal nullisomies and monosomies for any of the mouse autosomes
are lethal in utero at an early stage. Nullisomics (2n − 2) are so severely affected
that the condition is incompatible with egg segmentation: the conceptuses degen-
erate shortly after fertilization and are resorbed. Monosomics (2n − 1; symbol Ms)
for an autosome can develop for a few hours, but most die prior to or during the
implantation period and only rare survivors can be detected 6 days after fertili-
zation. This early lethality probably indicates that, for many loci scattered over
the autosomes, a 50 % reduction in gene expression is insufficient to assure nor-
mal embryonic development (Magnuson et al. 1985; Beechey and Searle 1988).
Genomic imprinting is also likely involved.
3.5.2.2 Trisomies, Tetrasomies, Double Trisomies etc
Trisomies (2n + 1) result from chromosomal non-disjunction during gametogenesis

or during the early stages of development. When the two chromosomes of a given
pair, instead of migrating to the opposite poles during meiotic anaphase I or ana-
phase II, migrate to the same pole and go into the same daughter cell, this is called
non-disjunction. This accident, which is relatively rare, results in a gamete being dis-
omic while the complementary gamete is nullisomic. When such a disomic gamete
fuses with a normal gamete (n chromosomes), this generates a trisomic embryo
(2n + 1), while the nullisomic gamete when it fuses with a normal gamete results in
a monosomic embryo (2n − 1). When chromosomal non-disjunction occurs during
5 Sometimes called nullosomies.

embryonic development (i.e., in the somatic cells—during mitosis), the result is sim-
ilar: a euploid mother cell, with 2n chromosomes, produces two daughter cells: one
with a 2n + 1 complement and the other with a 2n − 1 complement. In this case, the
embryo is a mosaic6 of euploid and aneuploid cells. The aneuploid cells are some-
times counter-selected compared to the normal euploid cells, especially if they do
not divide exactly at the same pace.
Trisomic embryos (symbol Ts) are often affected by severe specific defects. In
the mouse, all individual trisomies, including those of the X chromosome, have
been observed and studied in detail and, unlike in the case of the monosomic
embryos, the morphology of affected trisomics is highly variable (Gropp et al.
1975). At one extreme are trisomics for chromosome 19 (symbol Ts19), the short-
est autosome, which exhibit an almost normal morphogenesis up to 10 days in
utero and then appear slightly delayed until birth. Some Ts19 trisomic mice sur-
vive for a few days after birth, but many have a cleft palate and die. Mice trisomic
for chromosome 12 (Ts12) also survive for quite a long time in utero, but all die
at birth because they suffer from exencephaly. Mice trisomic for chromosome 14
and 16 also die at birth with rather characteristic pathological features. At the other
extreme, trisomics for chromosome 2, 7, 8, and 15 have extremely severe pheno-
types, with growth retardation and death occurring by the time of implantation or
shortly after (Beechey and Searle 1988) (Fig. 3.9).
The viability of the trisomic conceptuses is not correlated with the size of the
chromosome but probably with the density of genes, and this seems quite logical. In
humans, for example, trisomy 21 is the only viable trisomy, probably because chro-
mosome 21 is a small chromosome with only ~270 genes. Correlations between the
origin of the extra chromosome (paternal or maternal) and the severity of the pheno-
type have not been clearly documented in the mouse but probably exist if one takes
into account the phenomenon of parental imprinting (developed in Chap. 6).
Mouse primary trisomies, those involving a complete intact chromosome,
are, unfortunately, not good models for studying human trisomies, for two rea-
sons. First, the syntenic assortments of mouse and human genes on the different
chromosomes are so different that no human chromosome has its faithful, com-
plete, orthologous replica in the mouse species, and vice versa. Second, even if
the mouse genes were distributed along the mouse chromosome exactly as they
are in human, the phenotypes resulting from differences in gene dosage (3/2)
in one species may be expressed differently in another. In other words, and for
many reasons in addition to this one, mice are definitely not humans in reduc-
tion. However, analysis of mouse trisomies, in combination with human stud-
ies, sometimes provides a powerful system for understanding aneuploidy in both
6 Mosaics refers to organisms whose cells have a different genetic makeup, although they are all
derived from the same egg. Mice composed of both XO and XX cells, because one X was lost
during development, for example, are mosaics. Chimeras are organisms whose cells do not have
the same genetic makeup because they are derived from different embryonic cells. Mosaicism is
natural, while chimerism is, in most instances, the result of experimental manipulation.
66 3 Cytogenetics
Fig. 3.9 Trisomies. a A mouse trisomic for chromosome 12 (day 18 p.c.) and its age-matched
control. Note the exencephaly that is characteristic of this trisomy. (Courtesy of Dr. Heinz
Winking, Medizinische Hochschule, Lübeck, Germany). b A mouse trisomic for chromosome
16 (day 15 p.c.) and its age-matched control. (Courtesy of Dr. Muriel Davisson, The Jackson
Laboratory, Bar Harbor, Maine, USA). Ts16 embryos are slightly retarded and edematous
species (Hernandez and Fisher 1999). In addition, mouse trisomies are excellent
tools for studying the effect of variations in gene copy numbers.
Chimeric mice resulting from the association of trisomic cells with normal
euploid cells (Ts ↔ 2n) have been produced experimentally and have revealed
some interesting aspects of early tissue differentiation. It was repeatedly observed,
for example, that in Ts12 ↔ 2n chimeras, cells trisomic for chromosome 12 were
able to participate in the formation of most tissues, including the ovary, but were
never found in lymphocyte populations, presumably as a consequence of early
negative selection in this particular cell lineage (Fundele et al. 1985). Other chi-
meric mice with a trisomic partner (Chr 16 for example), have been produced
and have also been found to be fully viable, indicating that trisomic cells (at least
some of them) can be successfully integrated in a developing chimeric embryo
and, accordingly, that they are not cell-lethal. This sort of experiment might still
be used in the context of parental imprinting to analyze the consequences of gene
dosage effects in some chromosomal regions (see Chap. 6).
Tetrasomies (2n + 2) and double trisomies (2n + 1 + 1) are extremely rare
anomalies even when produced experimentally, and have not been studied in detail.
3.5.2.3 Aneuploidies of the Sex Chromosomes
Aneuploid heteroploidies concerning either the X or the Y chromosome have been

frequently reported in laboratory mice for two reasons. First, these heteroploidies,
unlike autosomal heteroploidies, are viable, even if the affected mice are often ster-
ile. Second, mouse geneticists have excellent X-linked markers that allow the detec-
tion of sex chromosome numerical anomalies at a glance, simply by observing that
the sex of the mouse does not match with the presumptive genotype for the sex chro-
mosome (or gonosomes). The X-linked marker Tabby (Ta), for example, has been
extensively used in this context and has allowed the detection of XXY or XO indi-
viduals because of an unexpected striped or non-striped coat color.7
The viability of mice with extra X or Y chromosomes is no surprise if we con-
sider that all X chromosomes but one in a karyotype are functionally inactivated
(Chap. 6), and that the Y chromosome is a relatively gene-poor element.
In the mouse, monosomy of the X chromosome (39,X0) is compatible with
normal survival and behavior. Female mice, unlike women who are affected by the
homologous syndrome, Turner syndrome (45,X0), can breed but they produce
many less X0 offspring than theoretically expected (i.e., 1/3). To explain this
shortage, it has been suggested that, during gametogenesis, X0 females preferen-
tially segregate the set of chromosomes lacking the X into the polar body (Morris
1968).8 In contrast with X0 mice, women affected by Turner syndrome exhibit
some phenotypic differences with normal women (short stature, sterility, etc.).
Geneticists think that these differences are attributable to the functional haploidy
of some X-linked genes that are not normally inactivated in human females. For
example, the SHOX gene, which maps to the human pseudo-autosomal region
(PAR), might be responsible for the short stature characteristic of Turner syn-
drome (X0) in human females, while its mouse ortholog is not X-linked and
accordingly is not affected by the monosomy.
7 Tabby (Ta) is an X-linked coat color and fur marker. XTaX+ females are striped; XTaY males
have a typical coat color with bare patches behind the ear, greasy fur, and a “sticky” tail. A Ta-
striped male is then unexpected unless it is XXY. A female with a Tabby [Ta] phenotype is
expected to be X0.
8 The theoretically expected 25 % of mice with a 39,0Y karyotype die at a very early stage of
development because of the X nullisomy. X0 females seldom produce more than 10 % X0 off-
spring and have a reduced stock of oocytes, resulting in a much shorter breeding period than
normal XX females.
68 3 Cytogenetics
Mice with a XXY constitution (41,XXY) have been found but their frequency is
very low (approximately 0.04 % among laboratory males and 0.08 % among wild-
caught males in some populations). These males, equivalent to human Klinefelter
syndrome, have a normal body mass and appearance, but significantly smaller testes
than normal, and no visible germ cells (Cattanach 1961; Hauffe et al. 2010).
The XYY sex-chromosome constitution, which is relatively common in human,
has also been described in the mouse (Cattanach and Pollard 1969). These males
are sterile probably because of the combined deleterious effects of two Y chromo-
somes acting prior to meiosis, and pairing abnormalities leading to meiotic break-
down (Hunt and Eicher 1991).
3.6 Variations in Chromosome Structure
A great variety of structural variations has been described concerning the

mouse karyotype. Some of these variations are relatively minor and are char-
acterized, for example, by moderate uncoiling of chromatin in the peri-centro-
meric regions, imparting a more or less elongated appearance to one or a few
chromosome pairs (Forejt 1973; Vig et al. 1994). In other instances, scientists
observed that in a group of wild mice from a certain geographic area, a spe-
cific band is slightly enlarged, indicating local chromatin amplification (often
described as homogeneous staining regions—HSR). These morphological vari-
ations of the karyotype have proved interesting for their cladistic value but
they have in general little or no influence on the survival and behavior of the
affected mice.
In contrast, some other structural changes resulting from chromosome break-
ages have more or less severe consequences, either on the survival of the embryo
or on the fertility of the affected animals. We will review the most important of
these structural abnormalities.
To understand the nature of these structural alterations, let us imagine that we
are sitting in front of a panel on which the individual mouse chromosomes are dis-
played, assorted in pairs as they appear in the ideogram. Let us now imagine that
we are handed a pair of scissors and requested to make cuts (one, two, three…)
randomly in the chromosome arms, then to pick up the fragments and glue them
back, also randomly, but always associating a telomeric fragment with a centro-
meric fragment, regardless of whether this was regenerating the original picture
or whether this created a “recombinant” association. In fact, what we were dem-
onstrating with the scissors analogy is exactly what happens in reality, either nat-
urally or after exposition of the post-meiotic germ cells to X- or γ-rays or after
injection of a chemical mutagen a few weeks before mating.
To complete this rather simple scenario we must make some additional comments:
• Because they are sticky, the ends of the broken chromosomes have a spontane-
ous tendency to reunite with other damaged ends (Schulz-Schaeffer 1980).
3.6 Variations in Chromosome Structure 69
• Structural rearrangements resulting from a single break have severe conse-

quences if the break is not repaired, because they split the chromosomes in two
pieces: one that remains attached to the centromere and the other that segregates
randomly during the mitotic or meiotic process. Such structural rearrangements
are almost always lethal for the cell because, by definition, the genetic material
is no longer distributed evenly after cell division. This explains why structural
rearrangements resulting from a single break, paradoxically, are rare.
• Breaks can occur at any time, in the adult or in the embryo, in somatic or germ
cells. If the alteration occurs in the germ line, it may interfere with the meiotic
process, resulting in reduced fertility or even complete sterility of the affected
individuals. Structural reshuffling occurring in the germ line may also lead to
the production of gametes with an abnormal set of chromosomes (unbalanced
gametes).
3.6.1 The Structural Rearrangements Resulting from a

Single Break
These structural rearrangements are called deletions, or more precisely ter-

minal deletions to distinguish them from the interstitial deletions that require
two breaks (see below). These deletions are uncommon because they result in
partial (or tertiary) monosomies for the telomeric fragment and accordingly
have highly deleterious consequences. As mentioned above, a chromosome
that has lost its telomere is unstable and is accordingly strongly counter-
selected. Conversely, a fragment with no centromere (acentric) is rapidly lost
when the cell divides.
3.6.2 The Structural Rearrangements Resulting from Two

Breaks
These two breaks can occur on the same chromosome but, in most cases, they
involve two different chromosomes.
3.6.2.1 The Structural Rearrangements Resulting from Two Breaks in

Two Different Chromosomes
Reciprocal Translocations
Reciprocal or balanced translocations, as the name indicates, result from a r eciprocal

exchange between the telomeric ends of two non-homologous chromosomes. They
70 3 Cytogenetics
2 2 8 8 2 28 8 82
Fig. 3.10 Reciprocal translocations. Reciprocal translocations are the most common kind of

chromosomal rearrangement (one break in two different chromosomes with reciprocal fusion).
When heterozygous they cause reduced fertility (semi-sterility) or sometimes complete sterility,
in one sex or the other. Some reciprocal translocations are viable and fertile when homozygous
are, by far, the most common form of structural rearrangement of the mouse karyo-
type and around 150 such translocations are listed in the Mouse Genome Database.9
The standard symbol used to define these reciprocal translocations is T. When
the chromosomes involved in the translocation are identified, the symbol contains
this information: T(2;8)26H, for example, is the 26th reciprocal translocation
recorded at Harwell; it involves chromosomes 2 and 8. When the positions of the
breakpoints relative to the G-banded karyotype are known, this can also be indi-
cated by adding the band numbers after the corresponding chromosome numbers:
the same T26H would then be designated T(2H1;8A4)26H, since the breakpoints
are respectively in band H1 of Chr 2 and band A4 of Chr 8 (Beechey and Evans
1996) (Fig. 3.10).
Mice heterozygous for reciprocal translocations, in most cases, have no visible
external phenotype,10 which is in keeping with the fact that these structural rearrange-
ments do not quantitatively alter the genetic makeup of the affected animals. However,
some heterozygous mice are sterile in one sex or the other, sometimes in both.
Gametogenesis in mice heterozygous for a reciprocal translocation is always
strongly perturbed, leading to the production of a high percentage of abnormal
gametes. To explain this, we will consider the meiosis of a heterozygous mouse
T(2;8)26H/+ (de Boer and de Maar 1976) and as a simplification we will con-
sider exclusively the chromosomes involved in the structural rearrangement.
9 Chromosome 2 appears to be more frequently involved in reciprocal translocations than

expected based on its size (27 occurrences). The reason for this bias is unknown.
10 The reciprocal translocation T26H is one of the very few exceptions because it is associated
with a coat-color change visible only in homozygous (T26H/T26H) animals. This change is
probably a consequence of an alteration at the Agouti locus (Chr 2) generated by the structural
rearrangement.
In these conditions, the genotype of the T26H/+ mice could be symbolized as

2 + 8 + 28 + 82, where 2 and 8 are the normal chromosomes, and 28 and 82 the
recombinant (or translocated) chromosomes, the superscript being a reference to
the origin of the telomeric segments. A T26H/+ heterozygous mouse produces dif-
ferent kinds of gametes depending on the segregation of the chromosomes during
meiosis. The first class is represented by the gametes with the intact chromosomes
2 and 8 (2 + 8). Another class corresponds to the translocated chromosomes 28
and 82. In these two chromosomes there is a redistribution of the genetic mate-
rial, but there is no loss or gain and for this reason these gametes are called bal-
anced. When a gamete carrying the two intact chromosomes (2 + 8) fuses with
another normal gamete (2 + 8), this restores a fully normal mouse karyotype and
the translocation is lost. When a gamete carrying (28 + 82) fuses with a normal
gamete (2 + 8), this generates a mouse heterozygous for the translocation, like the
heterozygous parent of the mating.
In addition to the gametes mentioned above, (2 + 8) and (28 + 82), mice het-
erozygous for the translocation T(2;8)26H/+ also produce (28 + 8), (2 + 82),
(2 + 28), and (82 + 8) gametes. All these gametes are unbalanced and when they
fuse with a normal gamete, the resulting embryos are all unviable because some
chromosomal segments are duplicated, while others are missing. For example,
a conceptus with a karyotype (28 + 8 + 2 + 8) is, at the same time, mono-
somic for a (telomeric) piece of chromosome 2 and trisomic for a telomeric
(distal) piece of chromosome 8. Geneticists say that conceptuses with such an
unbalanced karyotype are tertiary monosomic for the telomeric segment of Chr
2 and tertiary trisomic for the telomeric portion of Chr 8. Embryos with this
type of karyotype die in utero at an early stage, probably because of the tertiary
monosomy. This wastage of conceptuses explains why the progenies of mice
heterozygous for a reciprocal translocation (T/+) are always reduced in num-
ber. Mouse geneticists designate this reduced fertility by the term semi-sterility.
Semi-sterility is a phenotype common to all reciprocal translocations that can be
objectivized by looking at the uterine content of pregnant mice at day 13/16 of
pregnancy, and observing that about 50 % of the implants are in the process of
resorption (Fig. 3.11).
So far we have considered only the cases of crosses between mice heterozygous
for a reciprocal translocation and normal mice (for example, T26H/+ x +/+).
However, intercrosses between mice heterozygous for the same translocation (for
example, T26H/+ x T26H/+) are also possible. In most instances, the offspring of
such crosses are abnormal, with an unbalanced karyotype, and the living progenies
are extremely reduced. However, two cases are noteworthy:
• The first case is when, for example, a (28 + 82) balanced gamete from one part-
ner merges with another similar gamete (28 + 82) of the other partner. In this
case, an embryo homozygous for the translocation (28 + 28 + 82 + 82) results.
In the specific case of T26(2;8)H, the mouse is viable and fully fertile, but this
is not a rule, and mice homozygous for some other reciprocal translocations are
unviable or, less frequently, sterile. Carefully selected mice of this kind, with a
72 3 Cytogenetics
8 82
2 82
A A' B B' C C'

2 8 2 8
8 2 2 8
8 2 8 2 2 2 8 8 82
Alternate segregation Adjacent-1 segregation Adjacent-2 segregation
Fig. 3.11 Meiosis in a mouse heterozygous for the reciprocal translocation T(2;8)26H. Mice

heterozygous for a reciprocal translocation (T/+), when fertile, have to form unusual mei-
otic structures in order for all chromosomal segments to pair with their homologous regions.
Disjunction of the centromeres can be either alternate (A and A′), adjacent 1 (B and B′), or adja-
cent 2 (C and C′). Adjacent 2 is rare and is sometimes referred to as non-disjunction. Gametes
resulting from alternate segregation are either normal (A) or balanced (A′), with one (and only
one) copy of each gene. When such gametes merge with normal gametes of the opposite sex,
this generates normal embryos, with either a normal karyotype or a karyotype with the reciprocal
translocation. Gametes resulting from adjacent-1 segregation (B and B′) or adjacent-2 segrega-
tion (C and C′) are unbalanced and have either 0, 1, or 2 copies of each gene, depending on the
chromosomal segment. When these gametes merge with a normal gamete, aneuploid embryos
result. Reciprocal translocations have been very helpful for the establishment of genetic maps as
well as for the analysis of the epigenetic mechanisms at work in parental imprinting
karyotype different from that of normal laboratory mice, have been used as a
source of cells (for example, T or B lymphocytes) for performing transplanta-
tions or grafts because the different morphology of the chromosomes allows for
tracking of the transplanted cells in the chimeric organism.
• Another interesting situation is when, unbalanced gametes with a complemen-
tary karyotype fuse together. To explain the situation, let us consider the case of
another reciprocal translocation: T(2;11)30H. As in the case described above,
mice heterozygous for this reciprocal translocation produce six kinds of gametes
with the chromosomal constitution: (2 + 11); (211 + 112); (2 + 112); (211 + 11);
(11 + 112) and (211 + 2). However, when a non-balanced gamete with the con-
stitution (211 + 2) fuses with the complementary gamete (11 + 112) contributed
by the sexual partner, this generates a euploid embryo (2 + 11 + 211 + 112),
which is heterozygous for the translocation and has a balanced karyotype.
However, in this embryo, the centromeric part of chromosome 2 comes from
the same parent, while the other parent contributes the centromeric segments of
chromosome 11. This situation is also known in the human species and is desig-
nated as double non-disjunction or uniparental disomy (UpD).
Experiments focusing on the developmental potentialities of mouse embryos
resulting from such double non-disjunctions have been achieved by scientists at
the Harwell MRC Research Centre using a variety of reciprocal translocations and
a variety of genetic markers, allowing the unambiguous identification of the paren-
tal origin of the chromosomal segments. The conclusion of these experiments is
that, unexpectedly, euploid embryos resulting from complementary UpD are not
always viable. Sometimes they are viable when the UpD is of maternal origin, but
lethal when it is of paternal origin or vice versa, depending on the chromosomes
involved. In some instances, the embryos are viable but smaller sized (or larger
sized) than their littermates, depending on the crosses. This clearly indicates that
the genetic contribution of one parent is not equivalent to the contribution of the
other parent. We will come back extensively to this point in Chap. 6, which is
devoted to epigenetics and parental imprinting.
All these peculiarities of reciprocal translocation have been extremely useful at
several crucial steps in the development of mouse genetics. Because they disrupt
the linkage relationships between the genes on the same chromosome and simulta-
neously create new linkage groups by associating genes that were originally non-
linked, they were extensively used from the late 1970s to the early 1980s to assign
each and every linkage group to a particular chromosome and to determine the
position of the centromere for the linkage groups (Searle et al. 1971). The idea
behind this strategy is that reciprocal translocations have, at the same time, a phe-
notype that one can observe with a microscope (i.e., a reshuffled karyotype) asso-
ciated with semi-sterility and, when crossed, they exhibit new linkage relationships
between their genes while the original ones are disrupted (see Chap. 4).11
Reciprocal translocations have also provided essential tools for the localization
of genes associated with a variety of human cancers and hereditary diseases.
Another interesting point when crossing translocation carriers (T/+) is that,
among the aneusomic embryos that are produced, some are, by accident, tertiary
trisomics, i.e., trisomics for a small piece of chromosome or even for a complete
“recombinant” autosome. The reciprocal translocation T(14;15)6Ca, for exam-
ple, is characterized by a very unequal reciprocal exchange with a relatively long
chromosome 1415 (actually longer than Chr 1) and a very small chromosome 1514
(shorter than Chr 19). When these mice are intercrossed they occasionally produce
aneuploid conceptuses with an extra chromosome 1514. These mice are viable,
they are tertiary trisomics for a small piece of mouse Chr 15 (the centromeric end)
and a small piece of Chr 14 (the telomeric end), and they have been used in very
clever experiments to map the position of the centromeres in Chr 14 and 15 and to
11 Experiments involving crosses between translocation carriers (T/+) are difficult to achieve
because semi-sterility dramatically reduces progeny sizes. In addition, and as commented, some
reciprocal translocation carriers are infertile, impeding many experiments.
74 3 Cytogenetics
clarify the cytological identification of linkage group III (Eicher and Green 1972).
A procedure for genetic mapping, making use of the reciprocal translocations
T(X;7)1Ct and T(7;19)145H, called the duplication-deficiency method, has also
been reported (Eicher and Washburn 1978). Finally, and as we will explain fur-
ther in this chapter, Ts(1716)65Dn tertiary trisomic mice have been used to model
human trisomy 21, or Down syndrome.
Very unequal reciprocal translocations producing a long chromosome and a com-
plementary very short chromosome are not common, but some have been described
and are known as tandems. In some cases the very small chromosome is lost dur-
ing cell division with no consequences, since, as we already mentioned, it consists
mostly of heterochromatin. The consequence of this type of translocation is an irre-
versible reduction in the number of chromosome arms and centromeres. Such a tan-
dem has been reported as a derivative of the reciprocal translocation T(7;15)33Ad,
with breakpoints in bands 7A1 and 15F3. Outcrossing the original semi-sterile
T(7;15) mice generated monosomic mice for the short marker 715. By intercross-
ing these mice, viable nullisomic progeny for chromosome 715 were obtained that
could be intercrossed to produce a breeding stock with 38 chromosomes (Schriever-
Schwemmer and Adler 1993).
Robertsonian Translocations
Robertsonian translocations, named after the American biologist W. Robertson,

who described them first, are the last case of structural rearrangement resulting
from two breaks on different chromosomes. In fact, as for the tandem described
above, they are a special kind of reciprocal translocation resulting in the fusion
of two independent acrocentric chromosomes into a single unit that looks like a
metacentric chromosome. For this reason, they are also designated centric fusions
or whole-arm translocations.
The mechanism of formation of these reciprocal translocations is not com-
pletely elucidated and may not be the same in all cases. However, based on obser-
vations made after specific staining, cytogeneticists believe that most fusions, as
for the tandems presented above, result from reciprocal translocation, with the
chromosomal breakpoints being very close to the centromeres of two different
acrocentric chromosomes (Fig. 3.12). Translocated chromosomes may then have
either one or possibly two centromeres. If they have only one centromere, the
structural rearrangement is irreversible whereas, if they have two, it is theoretically
reversible. The complementary short chromosome, which in most cases contains
nonessential genes and possibly no centromere, is usually lost after a few cell divi-
sions, as in the case of tandems. This loss of centromeres is clearly an evolutionary
mechanism leading to the reduction in chromosome number. However, if the kar-
yotype is reduced (or virtually reduced) by one centromere, the fundamental num-
ber (the number of chromosome arms) remains the same and the global genetic
information remains unaltered although it is distributed differently.
The common symbol for Robertsonian translocation is Rb and, when the
arms (i.e., the acrocentric chromosomes) are identified, this is integrated into the
8
(a)
3 3 8 8 8
? ?
3
Robertsonian translocation Rb(3.8)68Lub
(b)
Fig. 3.12 Robertsonian translocations. Robertsonian translocations are an extreme case of

reciprocal translocation. a In some cases (probably rare), one of the rearranged chromosomes
has 2 centromeres and the other 0. In every case, when one of the two reciprocally rearranged
chromosomes is very small, it is often lost. Such translocations (or centric fusions) are observed
in many mammalian species and in particular in wild mice of the Alpine valleys. b The figure
represents the G-banded karyoptype of a wild mouse trapped in Italy. Most chromosome pairs
are involved in a Robertsonian translocation. Mice with such a karyotype are fully fertile, but
their F1 with a fully acrocentric laboratory strain are nearly sterile due to a high frequency
of meiotic non-disjunction. (Figure courtesy of Dr. Heinz Winking, Insitut für Biologie,
Medizinische Universität zu Lübeck, Germany.)
symbol. Rb(16.17)7Bnr, for example, is a Robertsonian translocation resulting

from the fusion of the acrocentric mouse chromosomes 16 and 17; this was the
seventh Robertsonian translocation (Rb7) discovered by Alfred Gropp in a wild
population of mice of the Poschiavo valley in Switzerland, and was bred in his
laboratory in Bonn am Rhein (Bnr).
76 3 Cytogenetics
Robertsonian translocations are common polymorphisms in nature, especially in

humans and in mice of the species Mus m. domesticus. In humans, only the acro-
centric chromosomes (group D) are involved in these structural rearrangements;
the three most important translocations (about one in a thousand newborns) being
between chromosomes 13q and 21q, 14q and 21q, and 15q and 21q. t(14q21q) is
frequently involved in Down syndrome and t(13q21q) in Patau syndrome.
In the mouse, Robertsonian translocations have been observed in laboratory pop-
ulations (Rb(9.19)163H and Rb(6.15)1Ald, for example), and there are many isolates
of wild animals in the Alpine valleys in Switzerland and Italy in which almost all
chromosomes are metacentric. Isolates of this type are numerous, and many differ-
ent associations of acrocentric chromosomes in centric fusions have been observed.
It is even likely that the complete inventory for this kind of rearrangement has not
yet been achieved in wild mice. Based on the observations made, and unlike what
is observed in humans, there seems to be no restrictions on these acrocentric asso-
ciations, even for the X chromosome. In contrast, the Y chromosome has never been
found associated with any autosome in the form of a Robertsonian translocation.
Mice heterozygous for a Robertsonian translocation, unlike mice heterozygous
for reciprocal translocations, are relatively fertile but exhibit a high percentage of
chromosomal non-disjunctions during meiosis. This peculiarity has been exploited
for the experimental production of trisomic and monosomic mice; we will briefly
describe the experimental protocol.
When crossing a mouse homozygous for a Robertsonian translocation involv-
ing chromosome 17 (for example, Rb(16.17)7Bnr) with a mouse homozygous
for another Robertsonian translocation involving the same chromosome 17
(for example, Rb(8.17)Rma), the F1 are heterozygous for both Rbs. If we
ignore the chromosomes that are structurally identical in both partners of the
cross, the genotype of these F1 can be symbolized as Rb7Bnr+/+Rb8Rma
(i.e., 8 + 16 ↔ 17 + 8 ↔ 17 + 16). Mice with this type of karyotype (dou-
ble heterozygous for Rb with monobrachial homology-17) are normal and
fertile, but they generate a high percentage of unbalanced gametes due to
non-disjunction at anaphase I. The gametes of these mice are either normal;
for example, (8 ↔ 17 + 16) or (8 + 16 ↔ 17) or unbalanced; for example,
(8 ↔ 17 + 16 ↔ 17) or (8 + 16). When these F1 mice are crossed with a nor-
mal mouse (8 + 8 + 16 + 16 + 17 + 17), up to 15 % of the embryos are tri-
somic for Chr 17 (Ts17) with a constitution (8 + 16 + 17 + 8 ↔ 17 + 16 ↔ 17).
Similarly, the same percentage of monosomic embryos (Ms17) is also produced
with a karyotype (8 + 16 + 17 + 8 + 16). Both these aneuploid embryos are
easy to recognize based on the examination of their karyotypes (2 or 0 metacentric
chromosomes). They both die at or shortly after implantation.
This situation is similar to the situation we reported for the reciprocal transloca-
tions, although, here, the resulting aneuploid offspring are primary trisomics or mon-
osomics and not tertiary. Since there is a very large number of theoretically possible
combinations for the production of double heterozygotes with monobrachial homol-
ogy, all trisomic (or monosomic) animal models have been created and studied in the
laboratory, for all chromosome pairs except X and Y (Gropp et al. 1975).
3.6.2.2 The Structural Rearrangements Resulting from Two Breaks

in the Same Chromosome
Interstitial Deletions
As we mentioned earlier, when chromatids (or chromosomes) are broken, the cel-
lular repair mechanisms are immediately activated and, depending on the breaks,
the event may or may not result in a loss of genetic material. When there is a loss
of genetic material, the structural alteration is called a deletion or interstitial dele-
tion with the symbol Del.12 When the deletion is cytologically visible in the karyo-
type, its designation takes this into account. Del(5B1), for example, designates a
deletion of the band B1 of chromosome 5. Depending on their size, these deletions
generally behave like dominant mutations with pleiotropic effects, less frequently
as recessive ones. They are often lethal when homozygous.
Hundreds of deletions of this type were produced in Harwell (UK) and Oak
Ridge and Argonne National Laboratory (USA) during the 20 years following
World War II, while health physicists were studying the effects of X-rays, γ-rays,
and neutrons on the genetic material of mammals. Many of these mutations have
contributed to the development of the mouse linkage map, although some of them,
because they are in fact small-sized chromosomal rearrangements rather than
true point mutations, have been difficult to use, due to their suppressing effect on
recombination leading to confounding results (compressions in the genetic maps)
(Fig. 3.13).
The interest of deletions in genetics is well illustrated in the case of the analy-
sis of the albino region (Chr 7-around the Tyr locus) by geneticists at the Albert
Einstein College of Medicine (Gluecksohn-Waelsch 1979) and at Oak Ridge
(Klebig et al. 1992). Deletions have been and still are of importance for mouse
geneticists because, when they are numerous and overlapping, they allow the
study, in great detail, of some regions of the genome and possibly the identifi-
cation of new alleles after mutagenesis (see Chap. 7 regarding mutations and
mutagenesis). Carefully selected deletions also allow the study of some regions
of the mouse genome in the haploid state (see Chap. 6, devoted to the analysis
of parental genomic imprinting). If all deletions that have been described in the
mouse could be gathered into a single animal, they would make up about 1/4 of
the total genome in the haploid state.
Finally, it is important to keep in mind that deletions frequently occur in vitro,
in cell cultures, with (apparently) little or no consequences on cell growth and pro-
liferation. For this reason, it is necessary to carefully and regularly check the kar-
yotypes of embryonic cell lines (ES cell lines), making sure that they are always
able to participate in the formation of viable chimeras with germinal transmission
and to differentiate into all types of tissues. The presence of even small deletions
12 A deletion (Del) is different from a deficiency (symbol Df) by its origin. Deficiency for a chro-
mosome segment is generally associated with a duplication (Dp) of the same segment and results
from the abnormal (unbalanced) segregation of a structurally rearranged chromosome.
78 3 Cytogenetics
Fig. 3.13 Schematic (a) (b)

representation
of chromosomal deletions.
a Terminal deletion (common a a a a
in vitro and rare in vivo).
b Interstitial deletion e
b b b
c c c
d d d
e e
in the genome of such ES cells may insidiously prevent their use for the produc-
tion of genetically modified mice by homologous recombination in vitro.
Inversions
When a segment of a chromosome is isolated by two breaks, flipped over by 180°,

and reintegrated into the same chromosome, the rearrangement is called an inver-
sion. Rearrangements of this type, like reciprocal translocations, are relatively
common events in all diploid species and participate in evolution at the chromo-
some level. They have an impact on the karyotype, because they generally alter the
banding pattern characteristic of the affected chromosome.
When the chromosome segment generated by the two breaks includes the cen-
tromere, the inversion is said to be pericentric. When the centromere is outside
the inverted segment, the inversion is called paracentric. In the normal laboratory
mouse, pericentric inversions are virtually nonexistent, since the chromosomes
are all acrocentrics and the centromeres are sub-terminal. The two types of inver-
sions share the same symbol In, and when the affected chromosome is identified,
here again the designation of the inversion takes this into account. In(2)5Rk, for
example, designates an inversion of chromosome 2, which is the fifth chromosome
anomaly of this kind collected by T.H. Roderick and colleagues.
Paracentric inversions have been induced experimentally in the mouse, with rel-
atively high efficiency, by submitting male mice to either X- or γ-ray irradiation or
by treating with alkylating mutagens, and by analyzing the offspring conceived in
the following four weeks (Roderick and Hawes 1974).13 Confirmation of the
induction of an inversion in the offspring was based on the observation of so-
called anaphase bridges in biopsies or sections of the seminiferous epithelium of
the presumptive carriers (Fig. 3.14).
Paracentric inversions interfere with the normal meiotic process, because
homologous chromatids cannot pair and come into close contact, as they generally
13 In this case, the targeted cells were the late spermatids or spermatozoa.
(a)
a a
d
b
c c
b
d
e e
(b)
(c)
N Rec
Rec
(d)
Fig. 3.14 Paracentric inversions. a Schematic representation of a paracentric inversion. b In

mice heterozygous for a paracentric inversion, the homologous inverted chromatids form a loop
that allows correct pairing. c When a crossing over occurs within the inverted segment (the loop),
this generates acentric fragments (with no centromere) and the reciprocal dicentric fragments. d
Since the two ends of a same dicentric chromatid are pulled apart, at the opposite poles of the
cell, this results in an anaphase bridge that can be observed in a histological section of the testis
(Figure courtesy of Dr. T. Roderick, The Jackson Laboratory, Bar Harbor, Maine, USA)
do during synapsis. To get around this problem, the inverted chromatids form a
loop that allows the correct orientation for pairing, but when a recombination
event (a crossing over) occurs in heterozygous mice within the inverted segment,
this generates an acentric fragment (with no centromere) and a reciprocal dicen-
tric fragment, with a centromere at both ends of the same chromatid. When these
centromeres are pulled apart to the opposite poles during anaphase of the divid-
ing cell, this causes an anaphase bridge that is often visible under the microscope
(Torgasheva and Borodin 2001). In his initial description of the protocol for induc-
ing inversions in the mouse, Roderick reported that several inversions induced
by the alkylating agent tri-ethylene melamine (TEM) could yield up to 70 %
of the cells exhibiting anaphase bridges (Roderick and Hawes 1974). In theory,
the longer the inverted segment, the higher the observed frequency of anaphase
bridges. In practice, however, this prediction is not confirmed, especially with the
longer inversions, yielding a lower than expected percentage of anaphase bridges.
Since the chromosomes affected by paracentric inversions cannot easily
pair with their normal homologous chromosomes during the pachytene stage
of meiosis and, taking into account the fact that all crossing overs occurring
within the inverted segments lead to a defective gamete, one can consider that
80 3 Cytogenetics
paracentric inversions act as virtual “crossing over suppressors” along the length
of the inverted segment, and probably a little beyond the borders. For this rea-
son, inversions are quite useful genetic tools for the recovery and maintenance
of mutations in model organisms; in fact, they recreate a situation in the mouse
analogous to the famous ClB (or Muller 5) condition designed by H. Muller
for the induction and collection of X-linked mutations in the X chromosome of
Drosophila (Muller et al. 1954).
In his experiments, Roderick noted that some hybrid males between the sub-
species Mus m. molossinus and Mus m. musculus displayed a high frequency of
first meiotic anaphase bridges, and sometimes double bridges, suggesting that the
chromosomes of these subspecies may differ from those of the normal laboratory
mouse by at least two paracentric inversions. This observation must be kept in
mind because, if they indeed exist, such inversions may generate some difficulties
in the analysis of mapping data when the subspecies Mus m. molossinus is a part-
ner of the cross (as is often the case).
In contrast with deletions, inversions generally do not change the overall
amount of the genetic material, and for this reason most of them are viable when
homozygous. In some cases, one of the chromosome breaks is inside or in the
close vicinity of a gene of essential function, and this sometimes generates a
mutation with a visible phenotype. For example, the mouse mutation hairy ears
(Eh-Chr 15), identified after neutron irradiation of post-meiotic germ cells due to
the presence of a tuft of hair on the outside of the ear in heterozygotes, was later
found to be at the breakpoint of an inversion spanning ~30 cM at the distal end of
Chr 15. Eh is lethal at an early stage when homozygous (Davisson et al. 1990a).
The case of Eh is uncommon and very few inversions have been found associated
with a phenotype. In most instances, the inversions collected in specially designed
experiments have been bred to homozygosity, indicating that the breakpoints are
not frequently in essential regions (Katayama et al. 2009).
Inversions have some influence on gametogenesis of In/+ carriers, since many
of the gametes recombinant within the inverted segment are wasted. However, in
most instances, this does not produce more than a slight reduction in fertility.
3.6.3 Complex Structural Rearrangements
The same game we began to play when making cuts in the karyotype and then
re-associating the fragments in all possible positions could be pursued to the
three-cut step, yielding increasingly complicated situations. In practice, very
few structural rearrangements resulting from three chromosome breaks exist in
the repositories, but at least two are worthy of comment. The first is the famous
“Cattanach transposition” discovered in Harwell. The Cattanach transposition
corresponds to the insertion (symbol Is) or transposition (symbol Tp) of a frag-
ment of chromosome 7 in the middle of the mouse X chromosome (two breaks
in chromosome 7 and one in the middle part of the X chromosome). The full
symbol of this rearrangement is (Is(In7;X)1Ct or XCt). Because the transposed/
inserted segment contains the wild-type gene encoding tyrosinase (Tyr), the
Cattanach transposition has been a useful tool for the study of X-chromosome
inactivation (see Chap. 6). Albino mice heterozygous for Is(In7;X)1Ct appear
“patchy,” having a coat with pigmented patches on an otherwise albino back-
ground, depending on the active X-chromosome in the melanocytes. The other
insertion is Is(7;1)40H, which corresponds to the insertion of part of Chr 7 into
Chr 1. This insertion is male-sterile and has been used for the purpose of gene
assignment.
3.6.4 Structural Rearrangements Created in Vitro
The great majority of the chromosomal rearrangements discussed in this chapter

were found by chance, either among the offspring of mice submitted to muta-
genic treatment resulting in chromosome breakages (X-rays, γ-rays, sometimes
neutrons) or in wild mice. As we said, these chromosomal rearrangements have
been very helpful, for example, for the assignment of mouse genes to specific
chromosomes or for the orientation of linkage groups with respect to the cen-
tromeres (see Chap. 4). More recently, they have also been extremely useful for
the discovery and genetic analysis of imprinted regions in the mouse (Chap. 6)
and the discovery of homologies in the human species. Unfortunately, the enor-
mous collection of chromosomal rearrangements has not been as useful as
geneticists would have wished for the genetic analysis of trisomies for the sim-
ple reason that the mouse genes, even if they have high homologies (orthologies)
with human genes at the molecular level, are nevertheless distributed very dif-
ferently in the genome of each species, making it difficult to faithfully model a
human trisomy. We have already noted that no mouse trisomy is viable for more
than a few hours ab utero, while humans affected by Down syndrome can live
for decades. Confronted with this intrinsic and unavoidable difficulty, geneticists
have tried to develop better and more refined models by transposing the refined
techniques they use for making alterations in the genome of ES cells to the field
of cytogenetics. The cuts we made virtually in the former paragraphs, with a pair
of scissors cutting the mouse chromosomes, can now be made extremely pre-
cisely (in fact, to the base pair) in the mouse genome by the molecular tech-
niques of homologous recombination. This means that any kind of reciprocal
translocation or inversion can now be “tailor-made”. Similarly, extra pieces of
mouse (or human) chromosomes can also be added to the mouse karyotype by
trangenesis, simulating tertiary trisomies. We will summarize all these possibili-
ties in the next section by presenting, as an example, the progress made in mod-
eling Down syndrome.
82 3 Cytogenetics
3.7 Modeling Human Down Syndrome
Among the trisomies that affect the human species,14 Down syndrome (DS—a tri-
somy of HSA21—or 47,XY + 21), is by far the most important for two reasons:
(i) because of its relatively high frequency (approximately one newborn in 750 is
affected) and (ii) because the syndrome is complex with highly variable and often
severe pathologies including mental retardation, congenital heart defects, dysmor-
phic features, early-onset Alzheimer disease, increased risk of specific leukaemias,
immunological deficiencies, and some other health problems.15
3.7.1 Mouse Trisomy 16: A Model of Down Syndrome
When human geneticists realized that Down syndrome (DS) was the consequence
of an imbalance in gene dosage for some of the ~268 genes linked to human chro-
mosome 21, and considering that a great number of the human genes on HSA21
have an orthologous copy on mouse chromosome 16 (MMU16), they had the logi-
cal idea to model DS by producing mice trisomic for this autosome (41,XY + 16)
or (41,XX + 16).16 Such mice can be easily produced, for example, by crossing
mice double heterozygous for the Robertsonian translocations Rb(16.17)7Bnr and
Rb(6.16)9Rma with normal laboratory mice (40,XX or 40,XY). Among the off-
spring of such crosses, most mice inherit only one of the two metacentric chromo-
somes, either Rb7Bnr or Rb9Rma, plus a complementary set of acrocentric
chromosomes (Cox et al. 1984). However, in some instances (up to 10 %), non-
disjunctions occur, the two metacentric chromosomes stay together to form a dis-
omic gamete, and a trisomic offspring results when the gamete in question merges
with a normal one. As expected, such trisomic mice exhibit some features charac-
teristic of human trisomy 21 (edema, cardiac anomalies, etc.) but, unfortunately,
the model had serious drawbacks (Epstein et al. 1985; Epstein 1990) (see Fig.
3.9b). First, the mice did not survive ab utero but died at a late stage of pregnancy,
blocking some experiments, in particular behavioral tests. Second, and most
importantly, although these mice were trisomic for the segment harboring the
mouse genes orthologous to the genes on HSA21, they were disomic (i.e., normal)
for some other genes of the same HSA21 that have homology with a segment of
MMU10 or MMU17. Reciprocally, because MMU16 has syntenies with regions of
14 Several autosomal primary trisomies have been described in the human species, but only
three, trisomies for chromosome 13 (Patau syndrome), for chromosome 18 (Edwards syndrome),
and for chromosome 21 (Down syndrome), affect live born children. Patau and Edwards syn-
dromes are extremely severe. The relatively low gene density on chromosome 21 is consistent
with the observation that trisomy 21 is one of the only viable human autosomal trisomies.
15 HSA21 = abbreviation for Homo sapiens chromosome 21; MMU16 = abbreviation for Mus
musculus Chr 16.

16 This estimation of the number of genes on human chromosome 21 (HSA21) is from
S. Scherer, A short guide to the human genome, Cold Spring Harbor Laboratory Press, 2008, p21.
3.7 Modeling Human Down Syndrome 83
HSA3, HSA8, and HSA16, many genes triplicated in Ts16 mice were not involved
in the etiology of human DS. For these reasons, primary trisomics for MMU Chr 16
have been abandoned as models of DS.
3.7.2 Ts(1716)65Dn: A Tertiary Trisomy Modeling Down

Syndrome
A more refined model of DS was developed by M.T. Davisson from The Jackson
Laboratory, USA (Davisson et al. 1990a, 1993) and was extensively studied by R.H.
Reeves and colleagues from Johns Hopkins University, Baltimore USA (Reeves
et al. 1995). This model is commercially available under the name of Ts(1716)65Dn.
These mice are tertiary trisomics, meaning that, in addition to the normal set of 40
chromosomes, they have in their karyotype an extra small chromosome resulting
from a radiation-induced reciprocal translocation between Chr 16 and Chr 17 and
comprising the centromere and proximal end of Chr 17 (~9.5 Mb) and the distal end
of mouse Chr 16 (~34 Mb or ~100 genes, with an orthologous copy on HSA21).
Ts65Dn mice survive to adulthood and exhibit many of the features of humans with
Down syndrome. For example, they have spatial learning and memory defects and
show some developmental delay. They also exhibit locomotor hyperactivity, lack of
behavioral inhibition, and stereotypic behavior.
Ts65Dn mice are considered good models of DS, but they nonetheless also
have some imperfections. The first and most important one is that only a segment
of Chr 16 with orthology in the HSA21 segment 21q21-21q22.3 is triplicated in
T65Dn. Another imperfection is that, here again, some genes that are triplicated in
Ts65Dn mice are on Chr 17 and have no orthologous counterpart on HSA21. This
variation in copy number probably interferes with the phenotype because some
genes of MMU17, triplicated in the Ts65Dn mice, are known to play an important
role in regulating neuronal functions.
In spite of these imperfections, T65Dn mice have the great advantage of exhib-
iting highly reproducible phenotypes, with clear similarities to DS, indicating that
dosage imbalance for a gene or group of genes in the triplicated region definitely
has a major contribution to this pathology. It is then likely that the corresponding
dosage imbalance for the human orthologous copies of these genes also contribute
to cognitive deficits in DS. These genes are part of the so-called DS critical region.
3.7.3 Transgenic and Transchromosomic Models of Down

Syndrome
Many other models of DS have been developed over the past few years. Most of
these models were created by pronuclear transgenesis (see Chap. 8) with cloned
DNA (yeast artificial chromosomes or bacterial artificial chromosomes) of various
84 3 Cytogenetics
Tc1
p12
Ts1Yey
Ts65Dn
Lipi
TgYAC (152F7, 230E8,…)

Dp/Df(Tiam1-Kcnj6)
MMU16
Mrpl39
Ts1Cje, Ms1Cje
q21.1
App
Ts1Rhr, Ms1Rhr
Dp/Df(Abcg1-Rrp1b) Ts2Yey
q21.2
TgBAC Dyrk1a
Sod1
TgBAC PCP4
q21.3
Runx1
Dp/Df(Prmt2-Pdxk)Ts3Yey
Ts1Yah, Ms2Yah
Dopey2
q22.11
Dyrk1a
Pcp4
Zfp295
Umodl1
Abcg1
MMU17
Ms4Yah
q22.2
U2af1
Rpr1b
Ms1Yah
Pdxk
q22.3
MMU10
Cstb
Col6a1
Prmt2
sizes from the relevant mouse chromosome, and assembled in the same individual
by sexual reproduction. Depending on the genes in the transgenic segments, these
models exhibited a variety of phenotypes more or less reminiscent of DS—and
were considered “partial” models. Among these models, one results from the addi-
tion, in the same genome, after several rounds of crossing and selection, of three
duplications (symbol Dp) of chromosomal regions of the mouse that are homol-
ogous to HSA21. These duplications are 2.3 Mb of MMU10 (Dp(10)1Yev/+)
containing 41 genes, 1.7 Mb of MMU17 (Dp(17)1Yev/+) containing 19 genes,
and 22.9 Mb of MMU16 (Dp(16)1Yev/+) containing 115 genes, all orthologous
to HSA21 (Yu et al. 2010). The production of these models (around a dozen as
of today) with copy number variation for regions homologous to HSA 21, has
contributed to a better understanding of the individual influence of the different
regions of human chromosome 21 on the brain alterations and a better definition
of several DS critical regions.
3.7 Modeling Human Down Syndrome 85
 Fig. 3.15 Mouse models of Down syndrome. This figure represents some of the different mod-
els of Down syndrome that have been described. On the left-hand side is a diagrammatic repre-
sentation of human chromosome 21 (HSA21) with its three homologous regions in the mouse
(MMU10, MMU17 and MMU16). Tc1 is a diagrammatic representation of the transchromo-
somic mouse model indicating the segments of HSA21 that have been retained, lost, or dupli-
cated. Ts65Dn is a tertiary trisomic mouse model with two segments of mouse chromosome, one
of MMU16 (in red), containing a “Down syndrome critical region” of human chromosome seg-
ment 21q22 and another smaller one of MMU17 (grey dotted line). Ts65Dn mice are trisomic
for ~13.4 Mb of the HSA21 syntenic region, containing approximately 99 orthologs of HSA21
genes, and are one of the most popular models of DS. Several other segmental trisomies for a
shorter region of MMU16 (T1Cje, etc.) or transgenic strains for yeast artificial chromosomes
or bacterial artificial chromosomes of MMU16 have also been published, but all these models
exhibit a less severe phenotype than T65Dn. On the right-hand side is a model described by
Yu et al. (2010) in which the regions of mouse chromosome 10, 16, and 17, syntenic to human
chromosome 21 (boxed), have been duplicated in vitro, in ES cells, then assembled in the same
genome by sexual reproduction. All the models represented here exhibit more or less faithfully
some of the DS-related neurological defects and will certainly be very useful for understanding
the cognitive disability associated with DS. Unfortunately, due to the complexity of the genetic
interactions involved in DS cognitive phenotypes, it is likely that no mouse model will ever reca-
pitulate the whole spectrum of intellectual disabilities observed in DS. (The background of this
picture is courtesy of Dr. Yann Herault, Institut Clinique de la Souris, Strasbourg, France)
Finally, a truly original model has been created by genetic engineering in ES

cells (see Chap. 8), leading to the production of a “transchromosomic” line that
stably transmits a freely segregating, almost complete human chromosome 21
(HSA21) (O’Doherty et al. 2005). This model exhibits phenotypic alterations in
behavior, synaptic plasticity, cerebellar neuronal number, heart development, and
mandible size that relate to human DS.
Two comprehensive reviews of these models have been published explaining
their peculiarities, advantages, and drawbacks (Herault et al. 2012; Rueda et al.
2012). None of these mouse models faithfully replicate human Down syndrome in
all of its details, but many of them allow the identification of brain regions affected
by homologous trisomies in the two species (Fig. 3.15).
3.8 Conclusions
The aim of this chapter was to provide an overview of the most important aspects
of mouse cytogenetics, describing the most common chromosome aberrations and
anomalies and their phenotypes or consequences on reproduction. We realize that
this presentation is rather superficial and greatly simplified, and for this reason we
provided references to the most important publications on the subject. The chro-
mosome aberrations and anomalies we have listed here have proved to be invalu-
able tools for the establishment of the genetic map, for unraveling some aspects
of genomic imprinting, and, more recently, for modeling Down syndrome. In the
future, they may still prove useful in experimental contexts where gene dosage is
an important issue.
86 3 Cytogenetics
Acknowledgments The authors are appreciative of the critical comments on the present
chapter provided by Drs. Marie Geneviève Mattei (Hôpital de la Timone, Marseille) and Yann
Herault (Institut Clinique de la Souris, Strasbourg).
References
Baron B, Métézeau P, Kiefer-Gachelin H, Goldberg ME (1990) Construction and characterization

of a DNA library from mouse chromosomes 19 purified by flow cytometry. Biol Cell 69:1–8
Beechey CV, Evans EP (1996) Numerical variants and structural rearrangements. In: Lyon MF,
Rastan S, Brown SDM (eds) Genetic variants and strains of the laboratory mouse, vol 2, 3rd
edn. Oxford University Press, Oxford
Beechey CV, Searle AG (1988) Effects of zero to four copies of chromosome 15 on mouse
embryonic development. Cytogenet Cell Genet 47:66–71
Blasco MA (2005) Mice with bad ends: mouse models for the study of telomeres and telomerase
in cancer and aging. EMBO J 24:1095–1103
Carmeliet P, Ferreira V, Breier G, Pollefeyt S, Kieckens L, Gertsenstein M, Fahrig M,
Vandenhoeck A, Harpal K, Eberhardt C, Declercq C, Pawling J, Moons L, Collen D, Risau
W, Nagy A (1996) Abnormal blood vessel development and lethality in embryos lacking a
single VEGF allele. Nature 380:435–439
Caspersson T, Lindsten J, Lomakka G, Moller A, Zech L (1972) The use of fluorescence tech-
niques for the recognition of mammalian chromosomes and chromosome regions. Int Rev
Exp Pathol 11:1–72
Caspersson T, Lindsten J, Lomakka G, Wallman H, Zech L (1970) Rapid identification of human
chromosomes by TV-techniques. Exp Cell Res 63:477–479
Cattanach BM (1961) XXY mice. Genet Res 2:156–158
Cattanach BM, Pollard CE (1969) An XYY sex-chromosome constitution in the mouse.
Cytogenetics 8:80–86
Cox DR, Smith SA, Epstein LB, Epstein CJ (1984) Mouse trisomy 16 as an animal model of
human trisomy 21 (Down syndrome): production of viable trisomy 16 diploid mouse chime-
ras. Dev Biol 101:416–424
Cox EK (1926) The chromosomes of the house mouse. J Morphol Physiol 43:45–54
Davisson MT, Roderick TH, Akeson EC, Hawes NL, Sweet HO (1990a) The hairy ears (Eh)
mutation is closely associated with a chromosomal rearrangement in mouse chromosome 15.
Genet Res 56:167–178
Davisson MT, Schmidt C, Akeson EC (1990b) Segmental trisomy of murine chromosome 16: A
new model system for studying Down syndrome. Prog Clin Biol Res 360:263–280
Davisson MT, Schmidt C, Reeves RH, Irving NG, Akeson EC, Harris BS, Bronson RT (1993)
Segmental trisomy as a mouse model for Down syndrome. Prog Clin Biol Res 384:117–133
de Boer P, de Maar PHMD (1976) A histological study of embryonic death caused by heterozy-
gosity for the T26H reciprocal mouse translocation. J Embryol Exp Morph 35:595–606
Eicher EM, Green MC (1972) The T6 translocation in the mouse: its use in trisomy map-
ping, centromere localization, and cytological identification of linkage group 3. Genetics
71:621–632
Eicher EM, Washburn LL (1978) Assignment of genes to regions of mouse chromosomes. Proc
Natl Acad Sci U S A 75:946–950
Epstein CJ, Cox DR, Epstein LB (1985) Mouse trisomy 16: an animal model of human trisomy
21 (Down syndrome). Ann N Y Acad Sci 450:157–168
Epstein C (1990) Genetic control of survival of murine trisomy 16 fetuses. Teratology
42:571–580
Forejt J (1973) Centromeric heterochromatin polymorphism in the house mouse. Evidence from
inbred strains and natural populations. Chromosoma 43:187–201
References 87
Fundele R, Jägerbauer EM, Kolbus U, Winking H, Gropp A (1985) Viability of trisomy 12 cells
in mouse chimaeras. Wilhelm Roux’s Arch Dev Biol 194:178–180
Gluecksohn-Waelsch S (1979) Genetic control of morphogenetic and biochemical differentia-
tion: lethal albino deletions in the mouse. Cell 16:225–237
Goto Y, Takagi N (1998) Tetraploid embryos rescue embryonic lethality caused by an additional
maternally inherited X chromosome in the mouse. Development 125:3353–3363
Green J, Ried T (eds) (2011) Genetically Engineered mice for cancer research: design, analysis,
pathways, validation and pre-clinical testing. Springer
Gropp A, Kolbus U, Giers D (1975) Systematic approach to the study of trisomy in the mouse II.
Cytogenet Cell Genet 14:42–62
Hauffe HC, Giménez MD, Garagna S, Searle JB (2010) First wild XXY house mice.
Chromosome Res 18:599–604
Herault Y, Duchon A, Velot E, Maréchal D, Brault V (2012) The in vivo Down syndrome
genomic library in mouse. Prog Brain Res 197:169–197
Hernandez D, Fisher EM (1999) Mouse autosomal trisomy: two’s company, three’s a crowd.
Trends Genet 15:241–247
Hunt PA, Eicher EM (1991) Fertile male mice with three sex chromosomes: evidence that infer-
tility in XYY male mice is an effect of two Y chromosomes. Chromosoma 100:293–299
Katayama K, Miyamoto S, Furuno A, Akiyama K, Takahashi S, Suzuki H, Tsuji T, Kunieda T (2009)
Characterization of the chromosomal inversion associated with the Koa mutation in the mouse
revealed the cause of skeletal abnormalities. BMC Genet 10:60. doi:10.1186/1471-2156-10-60
Kaufman MH, Lee KKH, Speirs S (1989) Influence of diandric and digynic triploid genotypes on
early mouse embryogenesis. Development 105:137–145
Kim SH, Parrinello S, Kim J, Campisi J (2003) Mus musculus and M. spretus homologues of the
human telomere-associated protein TIN2. Genomics 81:422–432
Klebig ML, Kwon BS, Rinchik EM (1992) Physical analysis of murine albino deletions that dis-
rupt liver-specific gene regulation or mesoderm development. Mamm Genome 2:51–63
Kubiak JZ, Tarkowski AK (1985) Electrofusion of mouse blastomeres. Exp Cell Res
157:561–566
Leeb M, Wutz A (2011) Derivation of haploid embryonic stem cells from mouse embryos.
Nature 479:131–134
Liyanage M, Coleman A, du Manoir S, Veldman T, McCormack S, Dickson RB, Barlow C,
Wynshaw-Boris A, Janz S, Wienberg J, Ferguson-Smith MA, Schröck E, Ried T (1996)
Multicolour spectral karyotyping of mouse chromosomes. Nat Genet 14:312–315
Magnuson T, Debrot S, Dimpfl J, Zweig A, Zamora T, Epstein CJ (1985) The early lethality of
autosomal monosomy in the mouse. J Exp Zool 236:353–360
Miller OJ, Miller DA (1975) Cytogenetics of the mouse. Annu Rev Genet 9:285–303
Miller OJ, Miller DA, Kouri RE, Allderdice PW, Dev VG, Grewal MS, Hutton JJ (1971)
Identification of the mouse karyotype by quinacrine fluorescence, and tentative assignment of
seven linkage groups. Proc Natl Acad Sci U S A 68:1530–1533
Morris T (1968) The XO and OY chromosome constitutions in the mouse. Genet Res
12:125–137
Muller HJ, Herskowitz IH, Abrahamson S, Oster II (1954) A nonlinear relation between x-ray
dose and recovered lethal mutations in drosophila. Genetics 39:741–749
Naumann R (2008) Production of tetraploid mouse embryos by electrofusion. Biocompare arti-
cle, Monday, 04 Aug 2008
Nesbitt MN, Francke U (1973) A system of nomenclature for band patterns of mouse chromo-
somes. Chromosoma 41:145–158
Niemierko A (1981) Postimplantation development of CB-induced triploid mouse embryos. J
Embryol Exp Morphol 66:81–89
O’Doherty A, Ruf S, Mulligan C, Hildreth V, Errington ML, Cooke S, Sesay A, Modino S, Vanes
L, Hernandez D, Linehan JM, Sharpe PT, Brandner S, Bliss TV, Henderson DJ, Nizetic D,
Tybulewicz VL, Fisher EM (2005) An aneuploid mouse strain carrying human chromosome
21 with Down syndrome phenotypes. Science 309:2033–2037
88 3 Cytogenetics
Reeves RH, Irving NG, Moran TH, Wohn A, Kitt C, Sisodia SS, Schmidt C, Bronson RT,
Davisson MT (1995) A mouse model for Down syndrome exhibits learning and behaviour
deficits. Nat Genet 11:177–184
Roderick TH, Hawes NL (1974) Nineteen paracentric chromosomal inversions in mice. Genetics
76:109–117
Rueda N, Flórez J, Martínez-Cué C (2012) Mouse models of Down syndrome as a tool to
unravel the causes of mental disabilities. Neural Plast 2012:584071, Article ID 584071.
doi:10.1155/2012/584071
Sahin E, De Pinho RA (2010) Linking functional decline of telomeres, mitochondria and stem
cells during ageing. Nature 464:520–528
Sawyer JR, Moore MM, Hozier JC (1987) High resolution G-banded chromosomes of the
mouse. Chromosoma 95:350–358
Schriever-Schwemmer G, Adler ID (1993) A mouse stock with 38 chromosomes derived from
the reciprocal translocation T(7;15)33Ad. Cytogenet Cell Genet 64:122–127
Schulz-Schaeffer J (1980) Cytogenetics plants—animals—humans. Springer, New York
Searle AG, Ford CE, Beechey CV (1971) Meiotic disjunction in mouse translocations and the
determination of centromere position. Genet Res 18:215–235
Sumner AT, Evans HJ, Buckland RA (1971) New technique for distinguishing between human
chromosomes. Nat New Biol 232:31–32
Torgasheva AA, Borodin PM (2001) Synapsis and recombination in inversion heterozygotes.
Biochem Soc Trans 38:1676–1680
Uhlmann F (2013) Open questions: chromosome condensation—why does a chromosome look
like a chromosome? BMC Biol 11:9
Vig BK, Latour D, Frankovich J (1994) Dissociation of minor satellite from the centromere in
mouse. J Cell Sci 107:3091–3095
Yu T, Li Z, Jia Z, Clapcote SJ, Liu C, Li S, Asrar S, Pao A, Chen R, Fan N, Carattini-Rivera S,
Bechard AR, Spring S, Henkelman RM, Stoica G, Matsui S, Nowak NJ, Roder JC, Chen C,
Bradley A, Yu YE (2010) A mouse model of Down syndrome trisomic for all human chromo-
some 21 syntenic regions. Hum Mol Genet 19:2780–2791
Zhang S, Teng Y (2013) Powering mammalian genetic screens with mouse haploid embryonic
stem cells. Mutat Res 741–742:44–50
Zhu L, Hathcock KS, Hande P, Lansdorp PM, Seldin MF, Hodes RJ (1998) Telomere length
regulation in mice is linked to a novel chromosome locus. Proc Natl Acad Sci U S A
95:8648–8653
Chapter 4
Gene Mapping
4.1 Introduction
Now that the sequence of the mouse genome is completely known, the position
of any gene of the species can be accurately and rapidly established by searching
the appropriate database. In this context, a chapter devoted to gene mapping and
genetic maps might appear somewhat outdated, not to say useless. However, we
thought that it might be interesting to reconsider this subject for at least three rea-
sons. The first is that gene mapping has been a major component of the activities
of mouse geneticists during most of the twentieth century; it is then interesting, if
only from a historical point of view, to briefly describe the techniques and meth-
ods that have made the genetic map of the mouse the richest and most documented
map of all mammals, including humans, for nearly 50 years. The second reason
is more fundamental and refers to the many mutations that occur spontaneously
in the breeding nuclei of inbred strains or those that are induced by mutagenic
agents. All these mutations are initially characterized by an abnormal pheno-
type and some of them may appear of potential interest, for example, as models
of human diseases. However, annotating and characterizing all these mutations
requires that they be first carefully located on a chromosome and analyzed at the
molecular level, when relevant. Finally, and as we will discuss in Chap. 10, under-
standing the determinism and mechanisms at work in the transmission and expres-
sion of quantitative traits requires that the genetic determinants of these traits be
accurately identified, and this always begins with a mapping experiment.
4.1.1 The Discovery of Linkage Groups:

A Historical Perspective
After the initial observations of Sutton, Boveri, and Morgan (reported in Chap. 3), it
was recognized that the genes in the mouse nuclear genome were all physically

90 4 Gene Mapping
associated with one or other of the 19 autosomes or X–Y chromosomes.1 Under

these conditions, it was implicitly accepted that two (or more) genes, once on the
same chromosome, would have a tendency to remain associated or “linked” together
in the same parental association, generation after generation, while all other genes,
those that were each on different chromosomes, would segregate independently. It
was also known that this physical association between genes on the same chromo-
some was not permanent or absolute, since recombination (crossing-over or chias-
mata) was regularly observed during meiotic prophase, when homologous
chromatids paired. It was then logical to guess that the probability for such an event
to occur between any two loci would depend upon the physical distance between the
loci in question. To express it differently, two genes that are distant by only a few
kilobases (kb) on the same chromosome would almost always segregate together
because the probability for a crossing-over to occur and split the association is very
low. In contrast, two other genes that are, for example, 50 Mb apart will have a much
greater chance of being separated by a recombination event. As a consequence of this
principle, when two genes are relatively close to each other on the same chromosome
they no longer segregate independently and this, by definition, generates distortions
from the expected Mendelian proportions. This fundamental concept of genetic link-
age was introduced in 1906, by Bateson and Punnett, after a series of experiments on
the inheritance of comb shape in chickens (Bateson and Punnett 1906).
This situation, which nowadays may appear obvious, was not so simple to
unravel. In 1904, Darbishire, lecturer in Genetics at the University of Edinburgh,
reported the results of crosses involving two recessive alleles, chinchilla (cch, now
Tyrc-ch) and pink-eyed dilution (p, symbolized e in the original publication and now
Oca2p) and concluded that, in this particular cross, Mendel’s law did not apply
(Darbishire 1904). This statement was in contradiction with Cuénot’s observations,
reported in Chap. 1, demonstrating that Mendel’s law indeed applied to mammals
(Cuénot 1902). A few years later, Haldane et al. (1915), re-examining Darbishire’s
observations, interpreted the latter as resulting from “reduplication” or linkage, as
we would now say. This linkage between two loci was the first to be reported in any
vertebrate species. Darbishire’s results were replicated in many other laboratories,
including in Haldane’s, using the same chinchilla allele (Tyrc-ch) or another recessive
allele (extreme dilution Tyrc-e) at the same albino locus (C or Tyr).2 The discovery of
this linkage indicated that the loci for c and p were presumably at some distance on
the same chromosome, but the chromosome in question was not known and the two
genes were simply considered as the first members of “linkage group I”.
After this first observation, many other coat color and phenotypic markers were
used in a variety of crosses, and more linkages were progressively discovered. In
1927, Gates (1927) could not obtain any mouse that was simultaneously dilute and
1 The structure of the mitochondrial genome (or mtDNA) is discussed in Chap. 5; here we con-
sider only the genes in the nuclear genome.

2 In this type of cross, the classical recessive allele Tyrc (albino) cannot be used because it has an
epistatic interaction with pink-eyed dilution, affecting eye and coat color, which makes pheno-
typing difficult.
4.1 Introduction 91
short-eared (i.e., d/d; se/se) among an enormous F2 progeny of 1,312 offspring.

He got three categories of offspring: 321 dilutes (d/d and +/se?); 653 wild-type
(+/d? and +/se?),3 and 338 short-eared (se/se and +/d?), compatible with a 1:2:1
ratio, and accordingly he concluded in “absolute linkage”.4
Two years later Lord and Gates (1929) reported that shaker-1, a mouse mutation
with a head-shaking behavior (the old symbol was sh-1 and is now Myo7ash1), was
also linked to both the c and p loci. This was confirmed by Grüneberg (1935) after
analysis of a large cross (1,144 mice). After these observations the notion of “linkage
group” was clearly established, with linkage group (LG) I now consisting of three
genes: p, c, and sh-1. Linkage group II had only two genes, dilute (d) and short-ears
(se), but, progressively, other genes were also reported linked to these genes. At the
end of World War II, 30 genes were found to be members of one or the other of the
10 identified LGs. They were 72 in 1955, with 15 LGs, 162 in 1965 (Eicher 1981)
and 217 in 1971 with a set of 20 LGs.5 Twenty, by the way, was exactly the theoreti-
cally expected maximum number of LGs, since there are 19 autosomes and one X
and one Y chromosome in the mouse genome.6, 7 At this point several questions
could then be addressed: (i) what is the gene order in a given LG?; (ii) to which chro-
mosome should a given LG be assigned?; and (iii) at what end of a LG is the cen-
tromere of the corresponding chromosome? The last two questions, although they
have required a lot of work and a lot of skillfully designed crosses, were solved in
less than 4 years (1971–1975). It took much longer to finally integrate the individual
mapping data into a single consensus map. We will briefly review the different steps
of the establishment of the mouse genetic (or linkage or meiotic) map.
4.2 From Linkage Groups to Genetic Maps
4.2.1 Detecting Linkage and Measuring the Distances

Between Loci
For a gene to be mapped, a fundamental prerequisite is that at least two allelic

forms of this gene be available and that these two allelic forms be involved in the
same cross with other allelic forms at the neighboring loci on the same linkage
3 +/d? indicates that the actual genotype of the mouse is not known. It may be either +/+ or
+/d.
4 Reading the original publications is sometimes difficult due to the use of a nomenclature sys-
tem different from the one in use nowadays. The dilute locus, for example, was designated “den-
sity” with two alleles: D and d. Nowadays, the same gene is symbolized Myo5ad.
5 A review by Dr. Eva Eicher from The Jackson Laboratory is a rich source of information con-
cerning the historical aspects of mouse gene mapping and the progressive development of the
genetic map in this species (Eicher 1981).
6 It is now established that there are very few genes on the Y chromosome.
7 The presumption that the twenty LGs identified at that time were each located on different
chromosomes was shown to be wrong. In fact, a few genes were still mis-assigned and one LG
was not yet identified.
92 4 Gene Mapping
group (or chromosome). Nowadays, as we will discuss further, the situation is dif-
ferent and much simplified because many genes are characterized at the molecular
(DNA) level and the genotype can be considered to be merged with the phenotype.
Let us, however, stay for another few pages at the pre-molecular era to outline as
simply as possible the basic principles for the detection of genetic linkage.
The notion of genetic linkage, as we said, means that the parental allelic associa-
tions have a tendency to remain unchanged in the successive progenies unless a
recombination event occurs to split the association in question. To make this clear,
let us imagine two genes at two different loci on the same chromosome: A and B.
Alleles A and B are fully dominant over the recessive forms a and b, and we will
assume that all four alleles are fully viable and fully penetrant. The male parent is
homozygous for the dominant allele A and for the recessive allele b while the other
parent, the female, is homozygous for the recessive allele a, and homozygous for the
dominant allele B. As a convention, such genotypes will be denoted Ab/Ab for the
male and aB/aB for the female.8 The F1 offspring of this cross will all have the same
genotype Ab/aB and, when intercrossed, these F1 will produce an F2 in which one
expects to get a variety of genotypes. If these genes are distant although still on the
same chromosome, non-parental (or recombinant) allelic associations will be com-
mon and we will observe four phenotypes: [AB], [Ab], [aB], [ab], with proportions
close to the expected Mendelian proportions 9/16, 3/16, 3/16, 1/16 (Table 4.1a–c).
If the two loci are tightly linked (as in the case reported above for the dilute
(d) and short ears (se) loci on LG II), then only three phenotypic classes will be
observed: [AB], [Ab], and [aB], with the proportions 1/2, 1/4, and 1/4, while the
phenotype resulting from two recombinant chromosomes [ab] will virtually never
occur. Finally, Table 4.1c represents an intermediate situation in which one third
(1/3) of the gametes are recombinant and the rest (2/3) non-recombinant. In this
case we will still observe the expected four phenotypic classes, but the one result-
ing from the fusion of two recombinant chromosomes will be less frequent.
In the example we just described, we mated a male with the genotype Ab/Ab
and a female with the genotype aB/aB, then we mated the F1 (Ab/aB × Ab/aB)
to produce the F2 offspring. This sort of cross, an intercross or F2 cross, is com-
mon because in most instances the two recessive alleles, a and b, were discov-
ered independently and in different populations or strains, and accordingly there
is a very low probability of finding them associated on the same chromosome (ab/
ab) just by chance. However, if such a genotype occurs, either spontaneously or
among the offspring of a cross, then it would be possible to cross a mouse with the
genotype AB/AB with mice with the genotype ab/ab. The F1 mice would then have
the genetic constitution AB/ab and the F2 would be of the same kind as above,
although the recombinant genotypes would be different and the phenotypic pro-
portions would also be different in the case of linkage. Finally, if the mating can
be set up between a male AB/ab and a female ab/ab, then the situation would be
much more advantageous for the detection of linkage and somewhat simpler to
8 If the alleles at the A and B locus were not linked or if the linkage was not known, the symbols
for the genotypes would be: A/A; b/b for the male and a/a; B/B for the female.
4.2 From Linkage Groups to Genetic Maps 93
Table 4.1 The upper part of this table (a, b and c) represents the expected proportions of the
different phenotypes [AB], [Ab], [aB] or [ab], in the progeny of an intercross for two alleles A-a
and B-b in repulsion (Ab/aB)
a when the two loci are not or only very loosely linked, in this case 50 % of the gametes are
recombinant; b when they are tightly linked, in this case there are virtually no recombinant gam-
etes; and c when they are moderately linked, in this case one third of the gametes are recombi-
nant and the other 2/3 are non-recombinant. The proportion of mice homozygous for the two
recessive alleles (ab/ab) varies from 1/16 (absence of linkage) to 0 (absolute or complete link-
age). The lower part of the table (d, e and f) specifies the different genotypes (and phenotypes) of
the offspring of a testcross or backcross: AB/ab x ab/ab. The female partner, being homozygous
for both the a and b alleles, produces only one type of gamete. This sort of cross allows one to
assess easily if the loci A and B are linked or not. If they are linked (e and f), the proportions
of the different genotypes are different from the Mendelian proportions and vary from 0 to a
value statistically lower than 1/4 (or 25 %). Computing the recombination frequency allows one
to measure the distance between loci A and B. In the cases where the recessive alleles a and b are
not fully penetrant or if some genotypes are unviable, the phenotypic class may be under-repre-
sented and a correction must then be applied.
94 4 Gene Mapping
analyze, since the recombination events occurring in the heterozygous parent (the
male in our case) would all be informative and easy to score just by looking at
the phenotypes of the offspring. The expected theoretical proportions in this case
would be: 1/4 [AB], 1/4 [Ab], 1/4 [aB], and 1/4 [ab] and any deviation from these
proportions would be suggestive of linkage, the fraction of recombinant genotypes
being [Ab] + [aB]/total number of offspring (Table 4.1d–f).
Crosses of this second type, with a male AB/ab and a female ab/ab or the reverse,
are called testcrosses (because one can test for linkage directly by counting the differ-
ent categories of phenotypes in the offspring population) or backcrosses because the
cross involves the F1 and a partner whose genotype is like the one of the ab/ab parent.
The genetic constitution AB/ab is designated double heterozygotes in coupling,
while the reciprocal genotypic constitution Ab/aB is called double heterozygotes in
repulsion.9
Intercrosses or F2 have, at the same time, a drawback and an advantage. The
drawback is because the detection of linkage requires the phenotyping of more
mice than in a testcross to reach the same level of significance. In a testcross one
has to check whether the four phenotypic classes observed in the progeny match
with the theoretically expected proportions 1/4, 1/4, 1/4, and 1/4, while in an F2
one has to check whether the four phenotypic classes match the classical propor-
tions 9/16, 3/16, 3/16, and 1/16. On the other hand, the intercrosses or F2 have the
advantage that, by genotyping a single individual, in fact we analyze the results
of two meioses—one in each parent—not just a single one. This advantage will
become more obvious when discussing molecular or co-dominant markers.
Once a situation of linkage is established between any two loci, a second step
must then be envisaged: assessing the strength of this linkage; in other words,
estimating the distance between the two loci. To explain this point we will take
another simple example: we will mate a male mouse heterozygous in coupling for
two dominant and two recessive alleles at the C and D loci (genotype CD/cd) with
many cd/cd female mice. This is a simple testcross—it will produce four classes
of offspring: CD/cd [CD], cd/cd [cd], Cd/cd [Cd], and cD/cd [cD]. Mice with a
[CD] or [cd] phenotype are those resulting from non-recombinant gametes pro-
duced by the male, while the other mice, those with a [Cd] or [cD] phenotype,
result from recombinant gametes. If we breed 317 offspring and observe, for
example, 38 mice with either a [Cd] or [cD] phenotype, the ratio of recombinant
offspring would be 38/317 = 0.11987 (i.e., 11.98 %). This result is an estimation
of the actual linkage between C and D and if we were to repeat the experiment
many times, we would get different results fluctuating around the value mentioned
above. It is also intuitive that if we had raised ten times more offspring (3,170
instead of only 317) we would have obtained a more reliable estimate of the actual
recombination frequency between the C and D loci. This is basic statistics, and
formulas are available to compute the most likely estimate of the recombination
9 When one of the mutant alleles (M) is dominant over wild type (+): the phase is +M/am for
coupling and aM/+m for repulsion. In other words, the dominant alleles are associated on the
same chromosome when in coupling.
frequency based on the number of mice scored in the progeny. For example, in the
case reported above the confidence interval (at the 5 % risk level) for the recombi-
nation frequency is given by the formula:

po qo
p = po ± 1.96
N
where po is the observed ratio of recombinant offspring (in our example it is

38/317 = 0.11987), qo = (1 – po) = 0.88013, and N is the total number of mice scored
for the phenotype (317 in our case). This yields 0.11987 ± 0.0356 (11.9 ± 3.5 %).
With a number of 3,170 offspring (and, say, 387 recombinants), the recombina-
tion frequency would be 0.1220 ± 0.0113 (12.2 ± 1.1 %) at the same 95 % confi-
dence level.
Many computer programs are available for analyzing these data and in most
instances it is sufficient to define the sort of cross, to introduce the number of mice
in the different phenotypic classes and the required degree of confidence for the
test (in general 5 %, sometimes 1 %), and the result is computed almost
instantly.10
For geneticists, a 1 % recombination frequency is equivalent to one unit of map
distance and is expressed as one centiMorgan (cM), a unit coined in homage to
T.H. Morgan.11
It is important to clearly distinguish between recombination fractions and
genetic distances. Recombination fractions are not additive measures. In fact, if
one considers three loci A, B, and C, with B located between A and C, the recom-
bination fraction between A and C is often less than the mere sum of the distances
between A and B, and between B and C. This is due to the fact that any individ-
ual recombinant between A and B and between B and C will be counted as non-
recombinant between A and C. To overcome this non-additivity of recombination
fractions, a number of geneticists, starting with J.B.S. Haldane, designed mapping
functions that convert recombination fractions into genetic distances, which are
additive. The various mapping functions make different assumptions on the inde-
pendence of crossing-overs that occur close to one another. Effectively, when a
crossing-over occurs at a given position, there is a highly reduced probability that
another one will occur nearby. This phenomenon is called interference, and varies
in strength and extent between species. It is quite strong in the mouse, so that dou-
ble recombinants are rare and recombination fractions can be reliably added for
distances lower than 25 cM.
10 Many of these software programs are listed at: http://www.jurgott.org/linkage/ListSoftware.h
tml. The most popular are MAPMAKER, MAPMANAGER and GENE LINK.
11 When the computed genetic distances are short or very short (<3 cM), it is recommended to
express them with the lower and upper limits of the exact 95 % confidence interval calculated
from the binomial distribution, as they appear in Table D5 and D6 (pp. 303–304) of Silver’s book
Mouse Genetics: concepts and application, Oxford University, 1995. This textbook is freely
available at the Mouse Genome Informatics website.
96 4 Gene Mapping
When the loci C and D are distant the recombination fraction reaches 50 %,
meaning that every other gamete is recombinant, and the two loci in question seg-
regate independently, just as if they were located on different chromosomes. This
point is confirmed by experimental data.
Another important point to consider is related to the sex of the heterozygous
progenitor. In the testcross reported above, the male was the heterozygous part-
ner with the genetic constitution CD/cd, while females were all homozygous for
the recessive allele at this locus (cd/cd), producing only one sort of gamete (cd).
However, it is now well established that the genetic distances computed from male
meiosis are generally not the same as those estimated from female meiosis (Petkov
et al. 2007). If our cross had been set up with the heterogametic sex (the male)
being cd/cd, and the homogametic sex (the female) CD/cd, the estimation of the
genetic distance would certainly have been different. On average, the recombina-
tion rate is higher in the homogametic sex than in the heterogametic sex. In some
chromosome regions, however, such as subtelomeric or imprinted regions, this ratio
is inverted. This means that a computed genetic distance is no more than an esti-
mation of the actual physical distance, but these distances become more and more
accurate as data from independent crosses, involving the two sexes, accumulate.
The genetic distances computed in cM by definition have an equivalent in DNA
units (i.e., in kb or Mb), even if we know that this equivalent is not uniform as a
consequence of the variations we just mentioned and others that will be discussed
later. This equivalence has been estimated in the mouse by several means (to be
discussed elsewhere), and a rough estimate is that one cM of mouse genome
equals ~1.70 Mb of genomic DNA on average.12, 13 Once the distances between
loci have been established, linkage maps can then be constructed, in which loci are
positioned according to the distance between them. However, before we can draw
a map involving several genes, we must order all these genes linearly on their
respective LGs.
4.2.2 Ordering the Genes
As we said above, 20 linkage groups were characterized in the late 1960s, mean-
ing that, within each group, any individual gene had been found linked to at least
one other gene of the same group. However, no information was available con-
cerning the order of all the genes in the same LG. If it was established, for exam-
ple, that the three genes F, G, and H were all members of the same linkage group
we could not decide about the order of the three loci: it could be F, G, H or F, H,
G or G, F, H.
12 The physical (or DNA) size of the genome is estimated to be 2.7 Gb, and the mouse genetic
(meiotic) map is estimated to span ~1,600 cM.
13 This equivalence between cM and kb/Mb applies to the mouse only. In the human species,
1 cM is equivalent to ~0.7–1 Mb of DNA.

Knowing the genetic distances between the genes taken by pairs sometimes
gives an indication. If, for example, we find that genes F, G, and H have the
respective distances F–8 cM–G; G–11 cM–H; and F–18 cM–H, it is likely that
the G locus is between the two loci for F and H. But such an ideal situation is
not common, and distances such as F–0.5 cM–G; G–16 cM–H; and F–15 cM–H
do not allow the loci to be ordered. Taking into account the experimental errors
that always occur, the order is ambiguous and may be F–G–H or G–F–H. In
such a situation, the best procedure is to set up what geneticists call a three-point
backcross.
Such a cross consists of mating, for example, a male heterozygous at the three
loci F, G, and H with a group of females, all homozygous for the recessive alleles
at the same loci: F G H/f g h × f g h/f g h, and then to carefully phenotype all the
offspring at all three loci. Since we have no a priori knowledge of the actual gene
order, we will tentatively choose the alphabetical order F–G–H and classify the
eight phenotypic groups.
Let us then assume that the experimental results are as follows, for a total of
1,078 mice:
[FGH] = 360 non-recombinant
[fgh] = 372 non-recombinant
[Fgh] = 66 recombinant between F and G loci
[fGH] = 68 recombinant between F and G loci
[FGh] = 13 recombinant between G and H loci
[fgH] = 17 recombinant between G and H loci
[fGh] = 93 recombinants between F and G and G and H (double recombinant)
[FgH] = 89 recombinants between F and G and G and H (double recombinant)
In this case, we can immediately observe that, taken two by two, the differ-
ent reciprocal classes of phenotypes (for example, [fGH] and [Fgh] or [FGh] and
[fgH]) are of the same order of magnitude (66/68 for the former group, 93/89 for
the latter).
We also note that the phenotypic classes [fGh] and [FgH], resulting from gam-
etes with a (supposed) double crossing-over, are quite large, whereas we would
have expected them to be less frequent (two recombination events). This simply
means that we were wrong when guessing a priori the order to be the alphabetical
order F–G–H. In fact, if we modify the order and put the H locus between the F
and G loci (the order now being F–H–G), then the phenotypic classes [FhG] (13
mice) and [fHg] (17 mice) are both double recombinant and less numerous.
We now have coherent data and the right order, and can then undertake the
computation of the distances between the loci by applying the methods described
above for only two loci. We will find that there are 66 + 68 + 13 + 17 = 164
mice out of a total of 1,078 whose genotype is recombinant between the F and
H loci, yielding an estimated distance of 15.2 ± 2.1 cM (at the 5 % risk). For
the other two loci we will find that 93 + 89 + 13 + 17 = 212 mice are recombi-
nant between the H and G loci on a total of 1,078, giving an estimated distance of
19.6 ± 2.4 cM (at the 5 % risk).
98 4 Gene Mapping
With these distances and gene order, we can now draw a map with the locus
F being at one extremity, the locus G at the other extremity, and the locus H in
between. At this point, however, we have no idea of the centromere position but
we know that it must be before F or after G given that the mouse chromosomes are
all acrocentric (the centromere is near one end) (Fig. 4.1a, b).
When displaying the results of the three-point testcross in the example described
above, we noted that the phenotypic classes were coherent in number and concluded
that the three recessive alleles f, g, and h were fully viable and fully expressed in
the homozygotes. Indeed, if we compute the number of mice with either a [f..] or
[.g.] or [..h] phenotype, the three classes will be close to 50 %. Unfortunately, this
is not a common situation. Frequently, one phenotypic class of offspring exhibits
a shortage because a recessive allele is less viable than its dominant counterpart or
because some mice are misclassified as a consequence of a non-expressed or mildly
expressed phenotype (lack of expressivity or penetrance). This is quite common,
for example, with mutations affecting either the eye or the skeleton, and when this
(a) Haplotypes (b)

Locus F
Locus F 15.2 ± 2.1 cM
Locus H Locus H
Locus G 19.6 ± 2.4 cM
Numbers
66
89
93
68
13
17
360
372
Locus G
Fig. 4.1 A three-point testcross. A three-point backcross or testcross (FHG/fhg × fhg/fhg)

allows one to order the genes linearly and estimate the genetic distances between the different
loci. a The different haplotypes (grey heterozygous; white homozygous) for markers F/f, H/h,
and G/g, respectively. b The genetic map drawn from the analysis of the genotyping data. The
class of offspring whose phenotype is the least numerous corresponds to a double cross-over.
Positioning the centromere requires other experiments
occurs the two classes of reciprocal recombinants are not equivalent. In this case, a
correction must be made before computing the distances, and geneticists generally
consider that the class with the largest number of recombinant offspring is probably
a better estimate for computing the recombination frequency. Not taking this into
account would result in an underestimation of the genetic distances.
Both interference and expressivity or penetrance are parameters that must be
seriously taken into account in mapping experiments. Corrections for interference
are extensively discussed in the book by Silver (1995) that is freely available at the
Mouse Genome Informatics website. We strongly recommend this excellent book
to readers with interest in issues related to genetic mapping.14
In the experiment reported above, we ordered the genes by setting up a three-
point backcross. In fact, the strategy can equally apply to more genes, and setting
up a four-point or five-point backcross is perfectly conceivable. Such crosses
would be fairly difficult to prepare with classical phenotypic markers, but they are
commonly performed with molecular markers, as we will explain later.15
Finally, one must remember that when linearly ordering the genes of one spe-
cies it is always interesting to have a look at the situation in other species where
the gene is known. As we will discuss in Chap. 5, homologies in the linear
arrangement of genes are highly preserved across the different mammalian spe-
cies, especially when they are short. In this case, geneticists speak of homology of
synteny or conservation of synteny.
4.2.3 Establishing a Correspondence Between LGs

and Chromosomes
Once the genes in a linkage group are linearly ordered, the next step consists of
identifying the chromosome that encompasses the LG in question. Nowadays, the
problem would be easily solved by selecting a molecular probe corresponding to
one of the genes of the LG in question, labeling it with a fluorescent dye and per-
forming fluorescence in situ hybridization (FISH) on a chromosome preparation as
described in Chap. 3.16 The localization could be confirmed by using another
14 http://www.informatics.jax.org/silverbook/frames/frame9-1.shtml.
15 Preparing a four-point backcross involving traditional or classical genetic markers, i.e., those
that are scored by scrutinizing the mice one after the other, requires a lot of crosses because the
markers in question are almost always in independent stocks or strains and first have to be gath-
ered in the same stock by sexual reproduction.
16 Most of the genes used for mapping in the past are now cloned and their DNA sequence is
known. For LG I, for example, Tyr is the gene encoding tyrosinase and Oca2 (oculocutaneous
albinism II) is a gene encoding a transmembrane transporter essential for normal pigmentation.
Both genes are involved in the production of melanin, and both are cloned and sequenced. The
two genes can then be considered under two aspects: either as a mouse with a specific coat color,
or as fluorescent dots on a mouse chromosome (see Chap. 3). In this particular case, it is chromo-
some 7 (bearing the whole of LG I).
100 4 Gene Mapping
cloned gene, belonging to the same LG as the previous one, and then labeling it
with another fluorescent dye for another in situ hybridization. In this latter case,
not only would the location of the LG be confirmed but, in favorable conditions,
the position of the centromere would also be revealed (see Chap. 3, Fig. 3.6).
Unfortunately, FISH was not available when the LGs of the mouse were await-
ing allocation to a specific chromosome, and in these conditions geneticists had
to use other strategies. The basic principle of these strategies consisted of match-
ing some morphological characteristics observed at the karyotype level [for exam-
ple, the size of the chromosome, the specific G-banding pattern, the presence of
a secondary constriction or of a negative-staining heteropycnotic (NHR) region,
etc.] with the mapping data collected independently and concerning the recipro-
cal translocation breakpoints. The reciprocal translocations, because they result
from two breaks on two different chromosomes with reciprocal exchange of the
telomeric segment, split the LGs that are usually associated with the normal chro-
mosomes and create two new ones. In these conditions, the chromosomal break-
points themselves can be mapped like ordinary genetic markers since they have a
phenotype (semi-sterility of the heterozygotes as a consequence of chromosome
reshuffling) and a genotype (they induce differences in linkage). For example,
from karyotypic observations performed at Harwell, the translocation Rb163H
was demonstrated to involve a very short autosome, in fact the shortest of the kar-
yotype (Chr 19), and a medium-sized acrocentric partner (Chr 9). From crosses
involving phenotypic markers the same Rb163H fusion was found to involve both
LG II and LG XII, meaning that one of the two LGs was on chromosome 19. The
ambiguity was resolved after the observation that the reciprocal translocation
T145H involved the same short autosome as Rb163 (Chr 19), and the two linkage
groups LG I and LG XII. The logical conclusion was that chromosome 19 encom-
passed LG XII (Lyon 1969; Eicher 1971).
Other similar experiments were conducted in different labs, involving different
chromosomal rearrangements (mainly reciprocal translocations) and the same
methodology.17
Other strategies have also been used for establishing a correspondence between
LGs and chromosomes, for example by analysis of the expression profile of a spe-
cific gene in cells of teratocarcinomas from the LT/Sv strain, or in cells of trisomic
embryos, or simply by looking for small morphological differences (Davisson
et al. 1976; Eicher 1978; Eicher and Washburn 1978).
17 The breakpoint of the reciprocal translocation T(2;8)26H on chromosome 2 is within the
Agouti locus and inactivates this gene. Mice homozygous for the translocation, which are easy to
identify by analysis of the karyotype, are also non-agouti (with a black coat color). This obser-
vation (and others) allowed it to be established that LG V (including the Agouti locus) was on
Chr 2.
4.2.4 Positioning the Centromere
Once the different loci on a given linkage group were unambiguously ordered and
assigned to a specific chromosome, the last ambiguity to be considered was the
position of the centromere. This was relatively difficult to resolve if we recall that,
in the mouse species, at least in the laboratory strains, all chromosomes are acro-
centric with no genetic markers available on the short arm, at least at that time.
In these conditions, positioning the centromere was impossible just by using clas-
sical mapping techniques. To achieve this, geneticists used a variety of differ-
ent approaches that are discussed in detail in the book by Silver (1995). Among
these approaches, the most popular and efficient consisted of using Robertsonian
+ + + c b a
9 19 x 9 19
c b a c b a
Rb + + + +abc
+ abc + abc
+ + + c b +
9 19 9 19
c b a c b a
Rb + + + Rb + b c
+ abc + abc
c b a c + +
9 19 9 19
c b a c b a
Rb a b c Rb + + c
+ abc + abc
Fig. 4.2 Positioning the centromere. Robertsonian translocations have been useful tools for
positioning the centromere once a linkage group is assigned to a specific chromosome. The fig-
ure represents a backcross between a mouse heterozygous for several recessive markers of chro-
mosome 9 (a, b and c) and the Robertsonian translocation Rb(9.19)163H (left), and a mouse
homozygous for the same three recessive markers and a normal karyotype (right). Analysis of
the phenotypes of the offspring at the a, b, and c loci and for the presence/absence of the centric
fusion Rb163 in the karyotype indicates the relative position of the centromere. If mice homozy-
gous for the c marker and heterozygous for the Rb163 metacentric chromosome are more fre-
quently observed than mice homozygous for the a marker and heterozygous for the same Rb163
marker, this means that the genetic distance to the centromere is shorter for a than for c and,
accordingly, that a is closer to the centromere of chromosome 9 than c. In the same sort of cross,
it is also possible to assess the distance between the centromere and the most proximal marker (a
in this case). These distances, however, are relatively unreliable
102 4 Gene Mapping
translocations. The centric fusion can be identified in a karyotype, and its presence
or absence is used as phenotypic information that can be integrated into a map-
ping experiment with the phenotype of other mutations on the same chromosome.
When, for example, a mutant allele segregating in a cross has a tendency to stay
associated with the metacentric chromosome in the same parental configuration
(i.e., in coupling or repulsion), this means that the centromere is very likely in the
close vicinity of the locus for the mutant allele in question (Fig. 4.2).
In the sort of cross we just described, the distance between the centromere and
the nearest genetic locus could in principle be estimated by counting the recombi-
nant/non-recombinant genotypes. Unfortunately, it has been demonstrated that in
crosses where this sort of translocation segregates, the distances to the centromere
are frequently underestimated due to structural differences that interfere with mei-
otic pairing (Davisson and Akeson 1993). The positioning of the centromeres was
rapidly established. It took a little longer to compute the distance between the cen-
tromere proper and the first (proximal) locus.
4.3 Genetic Markers
Any gene or short DNA sequence whose chromosomal location is precisely

known and whose structure exhibits variations among the individuals of the same
strain or species can be considered a genetic marker. The ancestral mutation
albino is a good example to explain this concept. The mutation has been cloned
and found to be the consequence of a G-to-C transversion at nucleotide 308 of
the gene encoding tyrosinase (Tyr-Chr 7) (Yokoyama et al. 1990). The nucleotide
substitution results in the replacement of a cysteine residue by a serine (a missense
mutation), making the product of the Tyr gene unable to cooperate in the produc-
tion of the pigment melanin, thus explaining why the affected mice are albino.
Looking in more detail at the sequence of the Tyrc gene, it was also noticed that
the nucleotide substitution introduced a novel DdeI restriction site. In these con-
ditions, the coat color (albino), the G-to-C transversion and the DdeI restriction
site are all genetic markers whose polymorphism can be used to track the same
Tyr allele (i.e., the same DNA sequence of Chr 7), generation after generation.
The coat color of the mouse is a phenotypic marker, while the restriction site and
the nucleotide substitution are molecular markers (DNA markers, in this case) of
exactly the same locus.
As we explained at the beginning of this chapter, when mouse genetics began,
the markers were exclusively phenotypic, i.e., were identified as visible changes
such as an alteration of the coat color or fur texture, or skeletal anomalies, or
abnormal behavior, etc. With the progress of biochemistry, molecular mark-
ers were developed that proved to be more abundant and easier to characterize.
In addition, most of these molecular markers are co-dominantly expressed, rather
than dominant/recessive like phenotypic markers. Nowadays, most of the genetic
markers are scored at the DNA level, and are abundant and easy to characterize
4.3 Genetic Markers 103
with highly reliable techniques. These DNA markers represent small structural
changes at the DNA level, distributed over the whole genome, including the
coding and non-coding regions. They have allowed the very rapid expansion of
genetic maps in all mammalian species.
4.3.1 Markers Scored by Examination of the External

Phenotype
The mouse genetic (or linkage) map originated from the observation that the
albino (Tyrc) locus was linked to the pink-eyed dilution locus (p—now Oca2p).
After this initial observation, the linkage map expanded for over 60 years as a con-
sequence of the continuous discovery of new mutant alleles, dispersed at different
loci throughout the mouse genome, and meticulously mapped, one at a time.18, 19
All these mutant alleles, with an obvious phenotypic effect, could be used
as genetic markers, for example for the mapping of a new heritable trait.
Unfortunately, they have two major drawbacks. The first and most important is
that they often impair the viability and/or the fertility of the affected animals in
one sex or in both, and for this reason it is extremely difficult to set up crosses
(especially testcrosses) involving more than three (maximum four) markers of
this kind. This would make the chromosomal assignment of a new trait time-
consuming and expensive, not taking into account the sacrifice of many animal
lives. Second, even if all these new mutant phenotypes, by their abundance, were
revealed to be important for establishing the frame (or scaffolding) of the genetic
map, they would nevertheless remain insufficient for developing the high-density
map that would be indispensable for the analysis of the whole genome. For these
reasons, genetic markers detectable by examination of the external phenotype are
rarely used.
18 Accumulation of these new mutant alleles was, in part, a direct consequence of the use of the
mouse as a model organism for the evaluation of the effects of radiation or of chemical mutagens
on the genome, and an indirect consequence of inbreeding as a mating system. Inbreeding has no
effect on the mutation rate, but it increases the chance that individuals will be homozygous for
recessive mutations, which, accordingly, are more easily identifiable.
19 During the years 1965–1975, when the genetic map of the mouse was expanding, researchers
at Harwell MRC and The Jackson Laboratory were using the so-called linkage testing (or linkage
tester) stocks. These stocks were homozygous for up to seven carefully selected recessive, fully
viable, and fully penetrant coat color markers mapping to different chromosomes. The phenotype
of each of these markers could be detected independently, with no interference from the other
markers. One of these stocks was the famous PT stock, which was extensively used at Oak Ridge
by W.L. & L.B. Russell for estimating the rate of induced mutations either with chemicals or
radiation. The PT stock was homozygous for seven markers on five chromosomes: a/a; b/b;
cch-p/cch-p; d-se/d-se; wa1/wa1.
104 4 Gene Mapping
4.3.2 Electrophoretic Variant of Enzymatic Proteins
From the late 1960s onwards, the development of gel electrophoresis and the
concomitant discovery of techniques for staining the product(s) of enzymatic
reactions allowed the identification of polymorphisms resulting from discrete
variations in the electrical charge of enzymatic proteins. These types of molec-
ular markers, referred to as electromorphs or electrophoretic variants, were
found to be very convenient for the purpose of mapping because they have
three advantages over the phenotypic markers described above: (i) they are co-
dominantly expressed and can thus be independently typed in heterozygotes;
(ii) they are in general compatible with a normal function of the enzyme and
accordingly do not impair the viability or fertility of the animals; (iii) they are
relatively abundant, probably because they are selectively neutral, and therefore
allow a wide coverage of the genome. Well over 100 markers of this kind are
available and new variants are regularly discovered, particularly in wild mice.
These biochemical markers permitted a rapid expansion of the mouse linkage
map in the mid-1970s (Bonhomme and Selander 1978). Unfortunately, they
have the drawback that their characterization requires relatively sophisticated
techniques, making large-scale linkage experiments based on this approach
quite expensive to run. Nowadays, these markers have been virtually abandoned
(Fig. 4.4a).
4.3.3 Plasmatic Proteins and Cell Surface Antigens
Many plasmatic proteins (albumin, globulins, etc.) or tissue-specific proteins (sem-

inal vesicle proteins, lens proteins, urinary proteins, etc.) and cell surface antigens
have also been used as genetic markers, with the same advantages and drawbacks
as the electrophoretic variants. They can still be used, but they do not exhibit any
obvious advantages over the molecular (DNA) markers geneticists now have at
their disposition.
4.3.4 Polymorphisms Detected at the DNA Level
The progress in DNA technology accomplished in the early 1980s has permit-
ted the development of two new kinds of genetic markers: (i) those that are
generated by restriction endonucleases and detected by Southern blotting: the
so-called restriction fragment length polymorphisms (RFLPs), and (ii) those
that are detected by DNA amplification using the polymerase chain reaction
(PCR).
Fig. 4.3 a Linkage map of the mouse as it appeared in 1971. This map was compiled and kept
up-to-date by Dr. Margaret Green from The Jackson Laboratory, Bar Harbor, Maine, USA. The
only LG that was assigned to a chromosome was LG XX (assigned to Chr X due to sex linkage).
b Genetic map of the mouse as it appeared in the Mouse News Letter in 1975. This map was also
compiled by Dr. Margaret Green. The genes of LG XVI were incorrectly assigned to Chr 12 and
should have been assigned to Chr 3. All other assignments were correct and the centromere was
also correctly positioned
106 4 Gene Mapping
Fig. 4.3 (continued)
4.3.4.1 Restriction Fragment Length Polymorphisms
In 1982, Botstein et al. (1980) reported that the restriction fragments generated
by digestion of DNA samples from different individuals, with one of the vari-
ous restriction endonuclease, often exhibited size polymorphisms when observed
with the technique of Southern blotting. These restriction fragment length
(a) (c)
s s/f f
Hb
(b) (d)
FVB/N
C57BL
DBA/2
129X1
BALB
CAST
129S1
Mouse
NZW
NOD
C3H
A/J
Strains
C/T SNP C C C C C C T C C T C
a/a a/b b/b
Fig. 4.4 A sample of co-dominant molecular markers. The picture represents four kinds of
genetic markers, selected from those that have been (or still are) the most frequently used. a
represents a gel discriminating two electrophoretic variants of the enzyme adenosine deaminase
(Ada-Chr 2). The slow variant (s) exists in mice of the Mus spretus species (strain SEG/Pas).
The fast variant (f) is common in classical laboratory strains. The F1 (s/f) between SEG and the
laboratory strain synthesizes the two forms. b represents the size polymorphism observed with
microsatellite D19Mit1 and DNA samples from two laboratory inbred strains. Microsatellites are
extremely abundant in mammalian genomes. They consist of a tandemly repeated, short-sized
motif, usually 2–6 bp long. Genotyping of this class of marker is achieved by PCR amplification
using primers designed from the flanking sequences, while the size differences of the amplifica-
tion products are assessed in agarose or polyacrylamide gels. Heterozygous genotypes are clearly
visible. c represents a single strand conformation polymorphism or SSCP assay. Once denatured
by heat (~90 °C), single-stranded DNAs exhibit a three-dimensional folding that is influenced
by their sequence. The spatial conformation of the single-stranded DNA determines its ability
to move in the gel (acrylamide). SSCP is an extremely sensitive technique but is now being sup-
planted by sequencing techniques on account of efficiency and accuracy. d represents a single
nucleotide polymorphism (SNP). This type of polymorphism is extremely abundant
polymorphisms or RFLPs also behave as co-dominant markers, and accordingly

have been extensively used for linkage analysis. Restriction endonucleases are
extremely abundant, and every event that suppresses or creates a restriction site in
the DNA sequence can be regarded as a creator of a new genetic marker. RFLPs
can be detected independently of their position in the DNA chain, be it within or
outside the coding sequences. The RFLP method was a true breakthrough because
it allowed mapping of virtually any locus encoding for a protein once its DNA
sequence was known. For this reason, it permitted the very rapid expansion of
linkage maps in the 1990s (Minty et al. 1983). Nowadays, the detection of RFLPs
by Southern blotting has been virtually abandoned, but checking for the presence/
absence of a restriction site in a PCR amplification product may still be used.
108 4 Gene Mapping
4.3.4.2 Markers Detectable by PCR
Several other techniques have been reported that are also based on the analysis
of structural variations at the DNA level. Among the most interesting are those
taking advantage of PCR, because they require very small amounts of template
DNA and because the genotyping can be completed within only a few hours at a
relatively low cost. The most popular of these techniques consists of the ampli-
fication of short sequences (usually less than 300 bp) whose polymorphisms are
either in size [simple sequence length polymorphisms (SSLPs or microsatellites)]
or in the sequence itself [single strand conformation polymorphisms (SSCPs),
denaturing gradient gel electrophoresis (DGGE)]. Many other DNA markers
have been described and used in the mouse and found to be interesting, either
because of their simplicity or because they provided geneticists with an almost
unlimited number of markers at very low cost, for example the RAPDs (Serikawa
et al. 1992). Unfortunately, these markers had the drawback of not being repeat-
able from one experiment to the next, and ultimately they did not present definitive
advantages over the microsatellites. For this reason they were only used briefly.
Microsatellites
Microsatellites [also known as simple sequence repeats (SSRs), short tandem

repeats (STRs), or simple sequence length polymorphisms (SSLPs)], correspond to
the repetition in tandem of relatively short motifs, usually 2–6 bp long.20 They are
co-dominant markers. They are commonly found in the mammalian genomes and
have been extensively used for the purpose of mapping in a variety of species.
Their origin is debated but it makes sense to think that they result from errors
occurring during DNA replication, a sort of stuttering or slippage of the DNA pol-
ymerase, or from unequal recombination events. The size polymorphisms are gen-
erally assessed by analysis of the migration of the amplification products after
electrophoresis in agarose or polyacrylamide gels. The primers used for amplifica-
tion are designed to target the unique sequences flanking the repeats (Love et al.
1990; Hearne et al. 1991; Montagutelli et al. 1991) (Fig. 4.4b).
These microsatellites represent almost ideal molecular markers because: (i) they
are highly polymorphic among strains; (ii) they are usually found in the non-coding
regions and the polymorphisms therefore are less likely to alter phenotype of the
mouse; (iii) they are abundant (in the range of 105 copies in the mouse genome, at
least for CA repeats); and (iv) they are relatively stable generation after generation.21
20 Repeated units such as T, CA, CT, and CAG are among the most common.
21 In fact, the microsatellites are more mutable than most other molecular markers previously
described, but their mutation, in general, generates a novel allele that is not identical to any of
the parental alleles. In these conditions the structural instability does not result in a confusion. It
may, however, be a problem when microsatellites are used for the genetic monitoring of inbred
strains (see Chap. 9).
Nowadays, primer sequences for thousands of SSLP markers are deposited in

public databases, and sometimes even commercially available. These markers can
be used either for the mapping of new polymorphic loci on the mouse linkage map
or for the identification of a DNA clone in a library.
When using these markers for the genotyping of DNA samples from inter-
specific or inter-subspecific crosses, one must be careful and make sure that the
selected primers prime amplification equally well with DNA of the two parental
strains or species. Not taking this warning into account may lead to confounding
results, with heterozygous genotypes being incorrectly classified as homozygous.
This warning, of course, applies to all PCR assays performed on DNAs of differ-
ent species and is not specific to the SSLPs, even though it is frequently observed
in this particular case.
Single Strand Conformation Polymorphisms
Sequence polymorphisms of PCR products can also be evaluated by comparing

their electrophoretic mobility in a polyacrylamide gel, after denaturation into
a single-stranded form (SSCP) or by measuring their migration abilities within
a gel containing a gradient of denaturing compounds such as urea or formamide
(DGGE). In these two cases, the method is so sensitive that a difference of only a
single base-pair in the sequence of any DNA molecule 80–250 bp long is gener-
ally detectable (Fig. 4.4c).
Single Nucleotide Polymorphisms
After the sequencing of the mouse genome, a new kind of genetic marker has
been developed: the single nucleotide polymorphisms (SNPs—see Chap. 5 for a
full description). These polymorphisms are single base-pair changes occurring
throughout the genome in all sorts of sequences (coding and non-coding). They
are extremely abundant and stable, and SNP genotyping is available on different
platforms including real-time PCR (TaqMan®), DNA microarrays, and competi-
tive allele-specific PCR coupled with fluorescence resonance energy transfer tech-
nology (Livak 1999; Nijman et al. 2008; Yang et al. 2009). However, sequencing
short DNA stretches (e.g., using pyrosequencing) in search of SNPs is still an
alternative approach to small-scale projects. SNPs are generally bi-allelic, mean-
ing that there are generally no more than two alleles across the different strains
or species of the genus Mus. Petkov and coworkers from The Jackson Laboratory
have described the allelic distribution of 235 SNPs in 48 mouse strains and
selected a panel of 28 such SNPs, enough to characterize hundreds of strains
(Petkov et al. 2004a). The same laboratory developed a new set of 1,638 inform-
ative SNPs selected from the publicly available databases and tested 102 inbred
strains (Petkov et al. 2004b). For those interested in the allele distribution of
SNPs in different inbred strains, the Mouse Phenome Database (MGD) presents
110 4 Gene Mapping
the most comprehensive collection of SNPs, with more than 8 million unique loci
and numerous inbred strains genotyped. In short, one can say that nowadays, with
microsatellites and SNPs, genetic markers are very abundant and relatively cheap
to characterize (Fig. 4.4d).
4.4 High-Resolution, High-Density Genetic Maps
When discussing the strategies used for evaluating the genetic distances between
genes or markers, we explained that the precision of these distances depended
upon the number of mice scored. Every mouse in a backcross progeny represents a
certain number of independent recombination events, randomly distributed over
the whole genome, and the greater the number of mice scored, the more informa-
tion is collected. Theoretically, and not taking statistical variations into account,
genotyping 100 offspring of a backcross progeny should be sufficient to find, on
the average, one mouse whose genome is recombinant between two genes or
markers distant by 1 cM (or ~1.7 Mb of DNA). If we increase the number of back-
cross progeny, for example 10 times, and score 1,000 mice instead of only 100,
then the theoretical resolution of our mapping would rise up to the milliMorgan
level (equivalent to 170 kb of DNA). In this case our map would be considered
a high-resolution genetic (or meiotic) map, meaning that it is very precise. Such a
map would be very useful for anchoring overlapping cloned DNAs covering
a chromosomal region. However, and unless we are able to precisely localize all
the thousands of cryptic recombination events, such a map will not be very help-
ful. In other words, by genotyping 1,000 backcross mice we would be able to
compute with great precision the distances between the few markers that are poly-
morphic in the backcross, but nothing else. In contrast, if we collect a few hundred
mice from a backcross set up between two remotely related inbred strains belong-
ing to two sub-species of the Mus genus (Mus m. domesticus × Mus m. castaneus,
for example), we may then establish a high-resolution and high-density map
because, as explained earlier, thousands of markers of all sorts would be polymor-
phic in such a cross.22
Several high-resolution/high-density maps were developed in Europe and in the
USA in the late 1990s (Dietrich et al. 1994; Rhodes et al. 1998). Among the most
important resources of this kind is the European Collaborative Interspecific
Backcross (the EUCIB resource), which was established from a collection of 982
DNA samples prepared from the progeny of two large backcrosses involving mice
of Mus spretus species and of the C57BL/6 inbred strains. This resource, which
incorporates 3,368 microsatellite markers distributed among 2,302 genetically
22 Around 70 % of SSLPs (microsatellites) or SNPs have been found to be polymorphic between
any two strains derived from progenitors of independent (wild) origins of the same Mus genus.
Altogether, this means that around 30,000 SNPs or SSLPs could (potentially) be used for the pur-
pose of mapping in a cross involving two inbred strains derived from two different subspecies.
4.4 High-Resolution, High-Density Genetic Maps 111
separated bins, with 1.46 markers per bin on average, allowed mapping any DNA
with a genetic resolution of 0.3 cM at the 95 % confidence level (approximately
600 kb in the mouse genome).23, 24
High-resolution/high-density maps have been used for assembling the cloned
DNAs in a physical map, and accordingly have logically contributed to the estab-
lishment of the mouse genome sequence.25
4.5 Somatic Cell Hybrids and Radiation Hybrids as Tools

for Gene Mapping
In the period spanning 1975–1985, geneticists used hamster/mouse somatic cell

hybrids to assign mouse genes to a particular chromosome. These cell hybrids
were obtained by merging in vitro somatic cells (fibroblasts) of the two spe-
cies, the fusion being triggered by small amounts of polyethylene glycol (PEG)
(Pontecorvo 1976). These cell hybrids were unstable, progressively and randomly
losing the chromosomes of the mouse. After a few generations of in vitro culture,
the situation stabilized and some of the clones remained stable, with a full set of
intact hamster chromosomes plus some extra mouse chromosomes or fragments of
mouse chromosome in their karyotype. The set of mouse chromosomes retained in
the individual clones was identified by one of the techniques described in Chap. 3,
and was in most instances unique for each clone. In these conditions, the cell
hybrids could be used to perform the rapid chromosomal assignment of mouse-
specific markers accessible in vitro (protein or a DNA polymorphism), by match-
ing the results of typing for the different clones, expressed in terms of presence (1
or +) or absence (0 or –), with the karyotype of the same clones and the presence
or absence of a specific mouse chromosome. This method of somatic cell hybrids
was relatively global, and provided only a chromosomal assignment.
The method was greatly improved in the early 1990s by replacing the
mouse cells with cells of the same species irradiated with a sub-lethal dose
of X- or γ-rays. In these conditions it was fragments of mouse chromosomes
23 A bin is a group of syntenic genetic markers that have not been separated (ordered) by meiotic
recombination in a given cross.
24 The mapping of these microsatellites has been achieved in two successive steps. First, all
the 982 DNA samples of the backcross progeny were initially typed for 78 primary anchor loci
spanning the entire genome, with 3–6 anchors per chromosome. In a second step, only the DNA
samples demonstrated to be recombinant in one or the other of these intervals tagged by the 78
primary anchor loci were typed for the greatest possible number of markers located (or presumed
to be located) in these regions.
25 In fact, and as we shall see in Chap. 5, the mouse genome has been sequenced by using a
global strategy known as whole-genome sequencing or WGS. However, the physical map of the
mouse genome, established by anchoring a variety of DNA clones on the genetic map, has been
helpful in many experiments of positional cloning, and is still used for the analysis of quantita-
tive traits.
112 4 Gene Mapping
instead of complete chromosomes that were retained in the genome of the non-
irradiated hamster cells. The size of these fragments was inversely correlated
to the dose of irradiation. In the case of the mouse–hamster T31 panel, the
panel of radiation hybrids (RH) that was mostly used during this period, the
dose of irradiation was 30 grays, generating fragments measuring 10 Mb on
average. In these conditions, when several hybrid clones were discovered to
exhibit the same pattern of presence/absence for a specific marker, this was
suggestive of linkage to the same fragment of chromosome inserted some-
where in a hamster chromosome. Compared with the other mapping strategies,
the use of RH for high-resolution mapping had some undisputable advantages.
The first is that it did not depend upon meiotic recombination and accord-
ingly did not require any cross between animals. Another advantage is that the
method could be used even in the cases where the cloned DNA to be mapped
did not exhibit any polymorphism. Finally, the results gathered from one
experiment contributed to the enrichment of the database associated with the
panel of RH cells used.
The method has been extremely helpful for defining gene order and distances
in the range of 4–8 Mb (i.e., equivalent to ~2–5 cM), especially in the later phases
of the development of the mouse genetic map, for the establishment of the high-
resolution/high-density consensus maps. At the end of 2001, the map established
with the help of the T31 RH panel contained up to 11,109 markers, positioned rel-
ative to a reference map containing 2,280 genetic markers. It included 3,658 genes
homologous to the human genome sequence. Nowadays, given that the mouse
genome is completely sequenced, the method no longer has an application. Indeed,
no map can have a better resolution than the sequence itself. For more informa-
tion concerning this non-sexual mapping strategy, the following references are
recommended: Cox et al. 1990; McCarthy et al. 1997; Flaherty and Herron 1998;
Hudson et al. 2001.
4.6 Recombinant Inbred and Recombinant Congenic

Strains
The nature of recombinant inbred strains (RIS) and their importance in mouse
genetics are described in detail in Chap. 9. Here we will only discuss the interest
of these strains as a tool for gene mapping.
RIS are derived from two unrelated parental inbred strains by systematically
intercrossing (brothers × sisters) the successive offspring of pairs of F1s for at
least 20 generations and often many more (Bailey 1971; Taylor 1978; Williams
et al. 2001). Figure 4.5 illustrates the genetic structure of such strains. RIS derived
from the same parental strains go by sets, or panels. At the present time, the BXD
panel, derived from the inbred strains C57BL/6 and DBA/2, is the largest panel
with ~90 strains available for research, but several other smaller-sized panels have
also been developed.
4.6 Recombinant Inbred and Recombinant Congenic Strains 113
͙͘͘
Strain A Strain AXB1 Strain AXB2 Strain AXB5
͙͘͘ ͙͘͘
Strain AXB9 Strain AXB13 Strain AXB19 Strain B
Fig. 4.5 The genetic structure of recombinant inbred strains. RIS are derived from two parental
inbred strains (here Strain A and Strain B) and propagated by strict inbreeding of several pairs of
inter-strain (A × B) F1. After 20 or more generations, these strains are totally inbred and each
chromosome of their genome is a patchwork of the two parental components. A given locus has
either the A or the B allele, as exemplified by the arrow. These RIS have been useful for the pur-
pose of mapping molecular markers, and are still extremely helpful for the analysis of the genetic
determinism of the quantitative traits (QTLs) that are different in strains A and B
Just like the parental strains they are derived from, RIS are also inbred, mean-
ing that within a given strain all individuals are genetically identical. However,
each of them has a unique combination of the parental alleles. For example, any
strain of the BXD panel carries either the C57BL/6J (B) or the DBA/2J (D) allele
(in homozygosity) at any given locus of its genome, in a 50:50 ratio. By typing
all of these allelic forms at every locus and for each of the strains, one establishes
what geneticists call a strain distribution pattern (SDP). The SDP is a permanent
source of information that is progressively implemented after every new genotyp-
ing. It is also a basic characteristic of the panel of RIS.
RIS have proven to be excellent tools for mapping mouse genes, and nowadays
the strategy is in expansion for the mapping of quantitative traits (Fig. 4.5). The
reasons for this success are two-fold. (i) Each strain in a panel of RIS represents
a collection of individuals with identical genomes that can be bred in unlimited
numbers. These strains remain stable generation after generation, with the only
exception of possible rare new mutations. Samples of animals of a given strain
can then be phenotyped and genotyped very accurately, for all sorts of characteris-
tics including quantitative traits, molecular markers, etc. A great advantage of RIS
is that large homogeneous samples of animals can be prepared for genotyping.
Another advantage is that recurrent phenotypings are also possible. (ii) Having
origins in two unrelated progenitor strains, the genome of each strain looks like
114 4 Gene Mapping
a patchwork made up of chromosomal fragments derived, randomly, from one or

the other of the two progenitor strains. In these conditions it is clear that the genes
linked on a given chromosome have a tendency to remain associated in the same
parental configuration through the successive generations of inbreeding, except
when a crossing-over splits the association. While inbreeding progresses, crossing-
over events occur at every generation, in each sexual partner, which modify the
genotype of the strain as long as the chromosomal segment stays heterozygous.
The chromosomes are thus progressively sliced into smaller-sized segments. In the
end, the genetic structure of the strain stabilizes and each individual strain of the
panel is homozygous for, on average, 50 % of the alleles coming from one paren-
tal strain and 50 % of the alleles coming from the other parental strain, but each
strain has a unique patchwork.
As we already mentioned, the phenotyping and genotyping information col-
lected in a panel of RIS can be used additively. The results collected for markers A
and B, for example, can be used for the mapping of markers C, D, etc..
The most popular panels are BXD (inbred parents C57BL/6J and DBA/2J),
AXB-BXA (inbred parents C57BL/6J and A/J), AKXD (inbred parents AKR/J and
DBA/2J), and AKXL (inbred parents AKR/J and C57L/J). Other panels are also
available with only a few strains (CXB, BXH, etc.), but they are also useful.
When a new marker or a new phenotype exhibits differences between the
parental strains, a novel strain distribution pattern (SDP) can be established and
matched to the SDPs already stored in the databases, in general with the help of
a computer program. The position of the new marker or phenotype and its link-
age with flanking markers is then calculated. The principle for positioning a new
marker or phenotype is to generate the smallest possible SDP discordances with
the flanking markers. As we already said above, when the marker density is very
high in a chromosomal region, ordering of the different loci is sometimes impos-
sible and, in this case, the group of unordered, syntenic markers represents a bin
(Fig. 4.6).
If the panel of RIS is large enough, the genetic distances can be calculated
based on standard formulas published in the genetic literature (Bailey 1971; Taylor
1978; Silver 1985). For example, if the number of discordant strains for a pair of
adjacent (linked) loci is i and the total number of RIS is N, one has:

R=i N
Once R is established from the analysis of the SDPs, the recombination frequency
r can be estimated from the formula:

r = R (4−6R)
Tables are available that provide an estimation of the percentage recombination
r (equivalent to the genetic distance) between two loci, and the upper and lower
95 % and 99 % confidence limits for r, when 6 ≤ N ≤ 45) (Silver 1985).
For example, if 5 RIS, out of a total of 23, are found to be discordant in their
SDP for markers A and B, two genes on the same chromosome, then i = 5,
N = 23, the formula returns a recombination fraction (r) of 0.0806 (=8.06 %), and
4.6 Recombinant Inbred and Recombinant Congenic Strains 115
2 5 6 8 9 11 12 13 14 15 16 18 19 20 21 22 2 5 6 8 9 11 12 13 14 15 16 18 19 20 21 22
Locus 1 B B B D B B B B B D B B D B B D Locus 1 B B B D B B B B B D B B D B B D
x x x x
Locus 2 D B B D B B D B B D B B D B B D Locus 2 D B B D B B D B B D B B D B B D
x x x x
Locus 3 D B B D B D D B B D B B B B B D Locus 3 D B B D B D D B B D B B B B B D
x x x x x x
Locus 4 D D B D B D D B B D D B B B D D Locus 4 D D B D B D D B B D D B B B D D
x x x x x x
Locus 5 D D D D D D D B B D D B B B D B Locus 5 D D D D D D D B B D D B B B D B
x x x x x x
Locus 6 D D D B D D D B B B D B B D D B Locus 6 D D D B D D D B B B D B B D D B
x x x x x x
Locus 7 D D D B D B B B D B D D B D D B Locus X D D D B D B D B D B D B B D D B
x x x x x
Locus 8 B D D B B B B B D B D D D D D B Locus 7 D D D B D B B B D B D D B D D B
x x x x x
Locus 9 B B D B B B B D D B D D D D D B Locus 8 B D D B B B B B D B D D D D D B
x x x x
Locus 10 B B D B B D B D D B D D D B D B Locus 9 B B D B B B B D D B D D D D D B
x x x x x x
Locus 11 B B D B D D B D D D D D D B B D Locus 10 B B D B B D B D D B D D D B D B
x x x x x x x x
Locus 12 B B B D D D B D D D B B D B B D Locus 11 B B D B D D B D D D D D D B B D
x x x x x x
Locus 13 D B B D D D B D D D B B B B B D Locus 12 B B B D D D B D D D B B D B B D
x x
Locus 13 D B B D D D B D D D B B B B B D
Locus X D D D B D B D B D B D B B D D B
Fig. 4.6 Mapping a trait with RIS. This table represents the (theoretical) results of the genotyp-
ing (or phenotyping) for each strain of a panel of 16 RIS (top row). This panel is commonly des-
ignated as the Strain Distribution Pattern or SDP. All strains are homozygous for one or the other
allelic forms present in the parental strains (in this case, B for C57BL/6 or D for DBA/2). This
SDP is permanent information that can be easily found in the public databases. When the paren-
tal strains are discovered to differ for a particular phenotype (or genotype), the panel can then be
used for mapping this new characteristic. Each strain is typed as B (identical to parent C57BL/6)
or D (identical to parent DBA/2), and the new SDP (Locus X) is plotted with the existing data
looking for the highest possible concordances. This is generally achieved very rapidly by using
simple, publicly available software. In the case shown, the best position for Locus X (or pheno-
type X) is between locus 6 and 7. Nowhere else could the SDP be more similar to the neighbor-
ing loci (two discordances with locus 6 and two with locus 7)
the table gives 2.1 and 31.7, respectively, as lower and upper confidence limits at
the 5 % risk. With the same number of discordant strains (i = 5) but a larger num-
ber of RIS (N = 44, for example), the same table would give 0.0342 (= 3.4 %) for
r, with 1.0 and 9.7 as lower and upper confidence limits at the same 5 % risk.
The strategy making use of the RIS “expands” the map over short distances,
and is more efficient than a backcross population for estimating recombination
when map distances are relatively small (<12.5 cM).26 On the other hand, cross–
intercross or cross–backcross protocols are more appropriate for the detection of
linkages over distances in the range of 20–30 cM (Silver and Buckler 1986).
RIS have proven extremely helpful for the rapid regional assignment of micros-
atellites on a given chromosome when these markers were cloned by the thousands
for the establishment of high-density genetic maps (Dietrich et al. 1994). They
have also been used for the mapping of chromosomal regions (QTLs) involved in
the genetic determinism of several behavioral characteristics (for example, taster/
non-taster for chemical compounds, alcohol intake, etc.) or for the mapping of
some immunological responses, or susceptibility to pathogens. They will very
likely be of great help in many other experiments where the phenotype is meas-
ured on a group of animals rather than on individuals (Zou et al. 2005).
26 This is sometimes referred to as a “magnifying glass effect.”.

116 4 Gene Mapping
Recombinant congenic strains (RCS) are similar to RIS in their genomic struc-
ture, except that the proportion of the parental alleles in a given panel is not 50:50
but 75:25 or even 87.5:12.5, depending on the panel (Demant and Hart 1986).
These RCS are produced by crossing mice of the first or second backcross genera-
tion to one of the parental inbred strains (the background strain), followed by strict
inbreeding. As we will explain in Chap. 10, RCS are helpful for identifying genes
or QTLs, especially when the latter are numerous. RCS with a small percentage
of introgressed genome in a background strain have a greater power of resolution,
and their use increases the likelihood of no or only a single locus (or QTL) gov-
erning the phenotype being isolated in a given strain. For example, RCS have been
very helpful for unraveling the genetic determinism of colon cancer in the mouse
(Demant 2003). Interspecific recombinant congenic strains (IRCS) have also been
developed from the parental strains C57BL/6JPas and SEG/Pas (Mus spretus)
(Burgio et al. 2007). This set of strains has permitted the analysis of the genetic
architecture of some anatomical traits (Burgio et al. 2012).
4.7 Establishing Consensus Maps
In the previous sections of this chapter, we explained that the genetic localization
of genes or cloned DNAs could be achieved by different methods. Two of these
methods are based on meiotic recombination (linkage maps and RIS), while the
third is based on the analysis of the retention of mouse chromosome fragments
of various sizes in hybrid cells. These methods were extensively used by mouse
geneticists at the end of the twentieth century and have contributed to the con-
struction of rich and dense maps. However, problems arose when the point came
to merge the mapping data collected through one of the three methods into one
and the same consensus map (Fig. 4.7).
When embarking on such a project, a few points must be taken into account:
• The construction of a consensus map based on independent primary maps is
possible only when a number of markers (designated anchor markers) are com-
mon to all primary maps.
• The distances computed between loci are not the same for meiotic and for RH
maps, but the distances computed by RH mapping are closer to the physical dis-
tances than those provided by linkage maps.
• The genetic distances computed from meiotic recombination depend upon the
crosses. As we already said, meiotic distances computed from male meiosis are
in general not the same as distances computed from female meiosis. We must
also mention that crosses involving progenitors from different strains, or from
different subspecies, sometimes result in distortions in the genetic distances.
This is generally the consequence of small differences in the chromosomal
structure (inversions, deletions, etc.) and must be taken into account (Paigen and
Petkov 2010).
4.7 Establishing Consensus Maps 117
D2Mit1
D2Mit2, D2Mit3
D2Mit4
Vim
Sd
Etl4 Etl4
Surf
D2Mit7 Pax8, D2Mit7
Mdk
Chr 2
Fig. 4.7 Consensus map. The observation of degenerative changes of the notochord in mice

homozygous for Danforth’s short tail dominant mutation (Sd) at day 10 of development, and the
concomitant localization of this mutation in the proximal region of Chr 2 were strong arguments
for making the gene encoding for paired box 8 (Pax8) a likely candidate for Sd. Merging the
data of two genetic maps, one involving the mutant allele Sd and the molecular markers Etl4 and
D2Mit7 (left), the other involving several molecular markers including Pax8 (right) into a con-
sensus map, ruled out this hypothesis because Sd was found to be proximal to Etl4 while Pax8
is distal to the same marker. This consensus map could be established only because two markers
(Etl4 and D2Mit7) were common to the primary maps (Redrawn from Koseki et al. 1993)
• The maps established by analysis of the SDP of RIS have gaps. These gaps are
inherent to the origin of the set of RIS and result from an absence of genetic
polymorphism between the parental strains in some chromosomal regions.
• The mapping information collected from one set of RIS (order and distances)
can be merged with the data collected from another set, provided that this refers
to the same markers or genes.
• Data relative to gene order in the syntenic regions of other mammalian spe-
cies are important to consider, especially when the species are closely related.
118 4 Gene Mapping
However, they must be used only as an indication. The genetic distances cannot
be compared because the recombination frequencies are species-specific.
In the years 1991–1999, mouse geneticists established consensus maps by gather-
ing the largest possible amount of mapping data from the literature. These maps
were mostly molecular maps excluding most of the mutant phenotypes, unless
the mutant alleles were cloned. These consensus maps, compiled by chromosome
committees, have been published in several successive special issues of the journal
Mammalian Genome between 1991 and 1999 (Fig. 4.8).
Meiotic map RH map
Fig. 4.8 Merging two independent maps in a consensus map. These pictures represents two dif-
ferent maps of the regions flanking the locus encoding 2′-5′ oligoadenylate synthetase (mouse
Chr 5). The map on the left is a meiotic map established by using polymorphic molecular mark-
ers (mostly microsatellites) ordered linearly after analysis of the 982 haplotypes of an inter-
specific backcross progeny (the EUCIB resource). The distances are in cM, with 95 % upper
and lower confidence limits. Some markers could not be ordered and represent a bin. The map
on the right was established by using the T31 radiation hybrids panel (RH map) and molecular
markers. The distances are in centiRays (cR). For most of the markers that are common to the
two maps, it can be seen that the order is the same. However, some markers have been ordered
in the RH mapping that could not be ordered in the meiotic map (e.g., D5Mi319-D5Mit407;
D5Mit369-D5Mit95). These maps are then complementary and allow the establishment of a
refined consensus map of the region (Redrawn from Mashimo et al. 2003)
4.8 Positional Cloning of Mutations and QTLs 119
4.8 Positional Cloning of Mutations and QTLs
The mutations that appear spontaneously in the breeding colonies of inbred strains
or those that occur after mutagenic treatment are all interesting, either because
they represent potential models of human diseases or, simply, because they can
help in the annotation of the mouse genome. For this, however, they must be
accurately phenotyped (we discussed this issue in Chap. 2) and at the same time
precisely characterized at the molecular level: this is precisely the objective of
positional cloning.
Positional cloning is a forward genetic (from phenotype to genotype) approach
whose aim is to characterize the structural alteration at the genome level that is
responsible for a specific mutant phenotype.27 A good historical example of posi-
tional cloning is the identification of the gene responsible for the obese mutation
(ob, now Lepob) on mouse Chr 6 (Zhang et al. 1994). To achieve this goal, an effi-
cient strategy is to build a high-resolution/high-density molecular map encompass-
ing the mutant locus, then to align and anchor this map on the sequence of the
mouse genome that is stored in the databases, with the help of the molecular mark-
ers whose sequences are known. This approach is now routine in many laborato-
ries, and it is greatly simplified once a couple of closely linked molecular markers
flanking the mutant locus have been identified.
The first step in a positional cloning experiment always consists of setting up a
cross in which a great number of polymorphic DNA markers segregates in addi-
tion to the mutant allele of interest characterized by a specific phenotype. To set up
this cross, it is recommended to select inbred strains that are as distantly related as
possible, because this increases the genetic polymorphism segregating in the cross.
Similarly, performing an intercross (or F2) between these distantly related strains
would be a better choice than setting up a backcross because, in these conditions
and as we already commented, two meioses are screened when genotyping each
offspring instead of only one in the case of a backcross.
Genotyping a first sample of 50–60 F2 offspring of the cross with the mutant
phenotype (equivalent to 100–120 meiosis) for a set of ~80 microsatellites or SNP
markers, evenly distributed over the whole genome, is generally sufficient to
assign the locus of the mutation into a 20-cM interval (equivalent to 35 Mb),
allowing the identification of two markers at the edges of the interval (Fig. 4.9).28
Once this first step is achieved, it is then necessary to genotype a much larger
progeny (i.e., around 600–800 mice) to yield a greater resolution.29 Of course,
among this large progeny only the mice whose genotype is recombinant in the
27 Reverse genetics is the opposite approach: its aim is to characterize the function of a gene
by analyzing the consequences at the phenotypic level of alterations occurring spontaneously or
engineered by researchers at the DNA level.
28 5–6 markers for the largest chromosomes, 4–5 for the medium-sized and 3 for the smallest is
ideal.
29 The first sample of 50–60 mutant mice generally consists of the first offspring of the larger
population.
120 4 Gene Mapping
Locus -------------------- haplotypes --------------------------
D9Mit64
D9Mit208
D9mit233
tbl, D9Mit303
Rora, D9Mit302
D9Mit165
D9Mit54
D9Mit308
Number 24 1 1 1 1 1 1 1
+/+
Fig. 4.9 Positional cloning. Using an inter-subspecific cross (F2), the mouse mutation tam-
baleante (Herc1tbl-Chr 9) was found to map to a 33-cM region of Chr 9 flanked by the DNA
markers D9Mit64 and D9Mit308. Analyzing the offspring of a very large inter-subspecific F2,
recombinant in the 33-cM interval, allowed separation of the tbl locus from the Rora locus, a
possible candidate gene for tbl. The mouse in the right column was instrumental for this map-
ping in the sense that it is recombinant between Rora and tbl, and proved to be homozygous for
the wild-type allele of tbl after breeding. This observation was sufficient to reduce the critical
interval hosting the mutant allele tbl to a 1.6-cM genomic region (Figure redrawn from Mashimo
et al. 2009)
interval defined in step 1 have to be genotyped. All others, by definition, are not
informative and can be discarded.30 Once the mapping data are collected, a new
molecular map can then be drawn and new molecular markers can be identified
that accurately delimit a much smaller critical region where the mutation definitely
maps. When the critical interval is in the range of 0.2–0.3 cM (~350–600 kb) or
less, no more mapping is necessary and the region can then be inspected in detail
for candidate genes. Within an interval of 350–600 kb, one expects to find, on
average, between 5 and 15 genes whose sequence is available in the different
databases.31
The last step of positional cloning consists of a functional analysis of can-
didate genes, for example by asking questions such as: where (in which tis-
sues) are they transcribed? Is the pattern of expression for each of these genes
in agreement with the observed phenotype? Are the genes in the interval equally
well expressed in normal animals and mutant mice? These basic questions
can be answered, for example, by looking at the transcriptional activity of the
30 This is why it is best to perform genotyping as early as possible, before weaning.

31 Aswe will explain in the next chapter, gene density is extremely variable from one genomic
region to the next. Accordingly, these estimations must be considered only as indications.
4.8 Positional Cloning of Mutations and QTLs 121
different genes. Another question might be related to the structural integrity of

the region: for example, are the PCR amplification products the same size in the
mutant and wild-type mice? Many other questions of this kind may be asked,
and if we refer to the data collected for the many mutations that have been
already cloned by a positional approach, it is likely that two or three genes out of
the 5–15 would become top-ranked candidates after this screening. At this step it
may then be wise to perform sequencing of the whole region followed by analy-
sis of the data.
Sequencing must be considered as early as possible in a positional clon-
ing experiment because it is a straightforward approach and it is in general suf-
ficient to discover the structural defect causative of the mutant phenotype.
Next-generation sequencing (NGS) technology currently offers the possibility
of establishing the whole-genome sequence of a mutant genotype for a moderate
cost. The defect in question can range from a simple base-pair substitution in an
exon or a splicing site to a more extensive rearrangement involving a few kilo-
bases. N-Ethyl-N-Nitrosourea (ENU)-induced mutations are base-pair changes in
most cases. Spontaneous mutations are sometimes more complex, with deletions
of various sizes.
Difficulties sometimes arise when the sequence comparison points to a mis-
sense mutation leading to an amino acid substitution, for example. In this case,
which is common after ENU mutagenesis, it is necessary to compare the mutation
with the sequence of the wild-type locus or haplotype in the same strain where
the mutation arose, because a missense mutation is not automatically correlated
with protein dysfunction. In fact, an experiment of positional cloning looks very
much like a police investigation, and sometimes it requires time and many redun-
dant clues to unmask the “culprit”. There are also cases, fortunately rare, in which
the positional cloning does not lead to a conclusion. For example, in the case of
the neurological mutation Purkinje cell degeneration (pcd), a member of a series
of alleles at the Agtpbp1 locus (Chr 13), its positional cloning ended up incon-
clusive with no obvious structural change in the coding region or splicing sites
(Fernandez-Gonzalez et al. 2002).
4.9 Physical Maps
The genomic DNA of several strains or subspecies of the genus Mus has been
cloned into a variety of vectors to build genomic libraries. Yeast artificial chromo-
somes (YACs) have been used because they have the advantage of featuring large
inserts (500–1,000 kb on average) allowing a reduction in the number of clones in
the library. Unfortunately, these vectors have the major drawback of being rela-
tively unstable and unreliable, with chimeric and deleted clones. Bacterial artificial
chromosomes (BACs) have been preferred as cloning vectors and have commonly
been used in practice. With these vectors, the insert size varies from 80 to 250 kb
(Osoegawa et al. 2000).
122 4 Gene Mapping
Using different cloning protocols, different BAC libraries have been prepared
with inserts of different sizes, and this has optimized the coverage of the genome
by eliminating the gaps due to the cloning protocol. For most of these libraries, the
coverage of the genome is high enough to guarantee that any genomic segment has
greater than 90 % chance of being represented in at least one clone of the library
(×8 to ×12 coverage).
The extremities of the BAC clones are sequences allowing one to organize the
library into groups of head-to-tail overlapping units called contigs.32 Finally, these
contigs can be anchored to specific regions of the mouse genome by using the
molecular markers (mostly microsatellites) of the high-resolution/high-density
map of the whole genome: this results in a physical map.
Such a map of the mouse genome was constructed in mid-2002. It was com-
posed of 296 contigs of overlapping BAC clones that were aligned to the human
genome sequence on the basis of 51,486 homology matches (Gregory et al. 2002).
As we will discuss in the next chapter, this collection of ordered clones encom-
passing a large part of the mouse genome has provided a framework that has been
very helpful for the assembly of the whole-genome sequence. A variety of mouse
DNA BACs have been used for making transgenic mice by in ovo injection, and it
is likely that more mice will be created in the future, with the aim of complement-
ing some of the QTL candidate regions.
4.10 Conclusion
The genetic localization of mouse genes, which started in 1915 with Haldane’s
discovery of the first linkage group, has been an enthralling enterprise that kept
researchers busy for most of the twentieth century and ended with the integral
sequencing of the genome. This being accomplished, geneticists have now under-
taken the functional annotation of all the genes, including the non-protein-cod-
ing sequences. Concurrently, they are associating a biological function to some
genomic regions that are not translated into proteins but are nevertheless highly
preserved across species. Finally, another important project will be to understand
the inheritance of quantitative traits and the structure of the so-called quantitative
trait loci (QTLs). All these projects are ambitious and challenging but, here again,
the mouse will probably appear to be a privileged model, and many of the tools
and strategies that were developed for the purpose of gene mapping (RIS, RCS,
molecular markers, etc.) will certainly prove useful.
32 Because there is contiguity in their sequence.

References 123
References
Bailey DW (1971) Recombinant-inbred strains. An aid to finding identity, linkage, and function
of histocompatibility and other genes. Transplantation 11:325–327
Bateson W, Punnett RC (1906) Comb characters. Rep Evol Comm R Soc Lond II:11–16
Bonhomme F, Selander RK (1978) Estimating total genic diversity in the house mouse. Biochem
Genet 16:287–297
Botstein D, White RL, Skolnick M, Davis RW (1980) Construction of a genetic linkage map in
man using restriction fragment length polymorphisms. Am J Hum Genet 32:314–331
Burgio G, Baylac M, Heyer E, Montagutelli X (2012) Nasal bone shape is under complex epi-
static genetic control in mouse interspecific recombinant congenic strains. PLoS One
7:e37721. Epub 2012 May 25
Burgio G, Szatanik M, Guénet JL, Arnau MR, Panthier JJ, Montagutelli X (2007) Interspecific
recombinant congenic strains between C57BL/6 and mice of the Mus spretus species: a pow-
erful tool to dissect genetic control of complex traits. Genetics 177:2321–2333
Cox DR, Burmeister M, Price ER, Kim S, Myers RM (1990) Radiation hybrid mapping: a
somatic cell genetic method for constructing high resolution maps of mammalian chromo-
somes. Science 250:245–250
Cuénot L (1902) La loi de Mendel et l’hérédité de la pigmentation chez les souris. Arch Zool exp
gén 3e séries 3:xxvii–xxx
Darbishire AD (1904) On the result of crossing Japanese waltzing with albino mice. Biometrika
3:1–51
Davisson MT, Akeson EC (1993) Recombination suppression by heterozygous Robertsonian
chromosomes in the mouse. Genetics 133:649–667
Davisson MT, Eicher EM, Green MC (1976) Genes on chromosome 3 of the mouse. J Hered
67:155–156
Demant P (2003) Cancer susceptibility in the mouse: genetics, biology and implications for
human cancer. Nat Rev Genet 4:721–734
Demant P, Hart AA (1986) Recombinant congenic strains–a new tool for analyzing genetic traits
determined by more than one gene. Immunogenetics 24:416–422
Dietrich WF, Miller JC, Steen RG, Merchant M, Damron D, Nahf R, Gross A, Joyce DC, Wessel
M, Dredge RD, Andre Marquis A, Stein LD, Goodman N, Page DC, Lander E (1994) A
genetic map of the mouse with 4,006 simple sequence length polymorphisms. Nat Genet
2:220–245
Eicher EM (1971) The identification of the chromosome bearing linkage group XII in the mouse.
Genetics 69:267–271
Eicher EM (1978) Murine ovarian teratomas and parthenotes as cytogenetic tools. Cytogenet
Cell Genet 20:232–239
Eicher EM (1981) Foundation for the future: formal genetics of the mouse. In: Mammalian
genetics and cancer: the Jackson laboratory fiftieth anniversary symposium. Alan R. Liss, Inc
New York, p 7–49
Eicher EM, Washburn LL (1978) Assignment of genes to regions of mouse chromosomes. Proc
Natl Acad Sci USA 75:946–950
Fernandez-Gonzalez A, La Spada AR, Treadaway J, Higdon JC, Harris BS, Sidman RL, Morgan
JI, Zuo J (2002) Purkinje cell degeneration (pcd) phenotypes caused by mutations in the
axotomy-induced gene, Nna1. Science 295:1904–1906
Flaherty L, Herron B (1998) The new kid on the block—a whole genome mouse radiation hybrid
panel. Mamm Genome 9:417–418
Gates WH (1927) Linkage of Short Ear and Density in the House Mouse. Proc Natl Acad Sci
USA 13:575–578
Gregory SG, Sekhon M, Schein J, Zhao S, Osoegawa K, Scott CE, Evans RS, Burridge PW,
Cox TV, Fox CA, Hutton RD, Mullenger IR, Phillips KJ, Smith J, Stalker J, Threadgold GJ,
Birney E, Wylie K, Chinwalla A, Wallis J, Hillier L, Carter J, Gaige T, Jaeger S, Kremitzki
124 4 Gene Mapping
C, Layman D, Maas J, McGrane R, Mead K, Walker R, Jones S, Smith M, Asano J, Bosdet

I, Chan S, Chittaranjan S, Chiu R, Fjell C, Fuhrmann D, Girn N, Gray C, Guin R, Hsiao
L, Krzywinski M, Kutsche R, Lee SS, Mathewson C, McLeavy C, Messervier S, Ness S,
Pandoh P, Prabhu AL, Saeedi P, Smailus D, Spence L, Stott J, Taylor S, Terpstra W, Tsai
M, Vardy J, Wye N, Yang G, Shatsman S, Ayodeji B, Geer K, Tsegaye G, Shvartsbeyn A,
Gebregeorgis E, Krol M, Russell D, Overton L, Malek JA, Holmes M, Heaney M, Shetty
J, Feldblyum T, Nierman WC, Catanese JJ, Hubbard T, Waterston RH, Rogers J, de Jong
PJ, Fraser CM, Marra M, McPherson JD, Bentley DR (2002) A physical map of the mouse
genome. Nature 418:743–750
Grüneberg H (1935) A three-factor linkage experiment in the mouse. J Genet XXXI:157–162
Haldane JBS, Sprunt AD, Haldane NM (1915) Reduplication in mice. J Genet 5:133–135
Hearne CM, McAleer MA, Love JM, Aitman TJ, Cornall RJ, Ghosh S, Knight AM, Prins JB,
Todd JA (1991) Additional microsatellite markers for mouse genome mapping. Mamm
Genome 1:273–282
Hudson TJ, Church DM, Greenaway S, Nguyen H, Cook A, Steen RG, Van Etten WJ, Castle AB,
Strivens MA, Trickett P, Heuston C, Davison C, Southwell A, Hardisty R, Varela-Carver A,
Haynes AR, Rodriguez-Tome P, Doi H, Ko MS, Pontius J, Schriml L, Wagner L, Maglott D,
Brown SD, Lander ES, Schuler G, Denny P (2001) A radiation hybrid map of mouse genes.
Nat Genet 29:201–205
Koseki H, Zachgo J, Mizutani Y, Simon-Chazottes D, Guénet JL, Balling R, Gossler A (1993)
Fine genetic mapping of the proximal part of mouse chromosome 2 excludes Pax-8 as a can-
didate gene for Danforth’s short tail (Sd). Mamm Genome 4:324–327
Livak KJ (1999) Allelic discrimination using fluorogenic probes and the 5′ nuclease assay. Genet
Anal 14(5–6):143–149
Lord EM, Gates WH (1929) Shaker, a new mutation of the house mouse (Mus musculus). Am
Nat 63:435–442
Love JM, Knight AM, McAleer MA, Todd JA (1990) Towards construction of a high resolu-
tion map of the mouse genome using PCR-analysed microsatellites. Nucleic Acids Res
18:4123–4130
Lyon MF (1969) Mapping data. Mouse News Lett 40:26
Mashimo T, Glaser P, Lucas M, Simon-Chazottes D, Ceccaldi PE, Montagutelli X, Desprès P,
Guénet JL (2003) Structural and functional genomics and evolutionary relationships in the
cluster of genes encoding murine 2′,5′-oligoadenylate synthetases. Genomics 82:537–552
Mashimo T, Hadjebi O, Amair-Pinedo F, Tsurumi T, Langa F, Serikawa T, Sotelo C, Guénet JL,
Rosa JL (2009) Progressive Purkinje cell degeneration in tambaleante mutant mice is a con-
sequence of a missense mutation in HERC1 E3 ubiquitin ligase. PLoS Genet 5(12):e1000784
McCarthy LC, Terrett J, Davis ME, Knights CJ, Smith AL et al (1997) A first-generation whole
genome-radiation hybrid map spanning the mouse genome. Genome Res 7:1153–1161
Minty AJ, Alonso S, Guénet JL, Buckingham ME (1983) Number and organization of actin-
related sequences in the mouse genome. J Mol Biol 167:77–101
Montagutelli X, Serikawa T, Guénet JL (1991) PCR-analyzed microsatellites: data concerning
laboratory and wild-derived mouse inbred strains. Mamm Genome 1:255–259
Nijman IJ, Kuipers S, Verheul M, Guryev V, Cuppen E (2008) A genome-wide SNP panel for
mapping and association studies in the rat. BMC Genom 9:95
Osoegawa K, Tateno M, Woon PY, Frengen E, Mammoser AG, Catanese JJ, Hayashizaki Y, de
Jong PJ (2000) Bacterial artificial chromosome libraries for mouse sequencing and functional
analysis. Genome Res 1:116–128
Paigen K, Petkov P (2010) Mammalian recombination hot spots: properties, control and evolu-
tion. Nat Rev Genet 11:221–233
Petkov PM, Cassell MA, Sargent EE, Donnelly CJ, Robinson P, Crew V, Asquith S, Haar RV,
Wiles MV (2004a) Development of a SNP genotyping panel for genetic monitoring of the
laboratory mouse. Genomics 83(5):902–911
KA, Robinson P, Scott VE, Wiles MV (2004b) An efficient SNP system for mouse genome
scanning and elucidating strain relationships. Genome Res 14(9):1806–1811
References 125
Petkov PM, Broman KW, Szatkiewicz JP, Paigen K (2007) Crossover interference underlies sex
differences in recombination rates. Trends Genet 23:539–542
Pontecorvo G (1976) Polyethylene glycol (PEG) in the production of mammalian somatic cell
hybrids. Cytogenet Cell Genet 16:399–400
Rhodes M, Straw R, Fernando S, Evans A, Lacey T, Dearlove A, Greystrong J, Walker J, Watson
P, Weston P, Kelly M, Taylor D, Gibson K, Mundy C, Bourgade F, Poirier C, Simon D,
Brunialti AL, Montagutelli X, Guénet JL, Haynes A, Brown SD (1998) A high-resolution
microsatellite map of the mouse genome. Genome Res 8:531–542
Serikawa T, Montagutelli X, Simon-Chazottes D, Guénet JL (1992) Polymorphisms revealed by
PCR with single, short-sized, arbitrary primers are reliable markers for mouse and rat gene
mapping. Mamm Genome 3:65–72
Silver J (1985) Confidence limits for estimates of gene linkage based on analysis of recombinant
inbred strains. J Hered 76:436–440
Silver J, Buckler CE (1986) Statistical considerations for linkage analysis using recombinant
inbred strains and backcrosses. Proc Natl Acad Sci USA 83:1423–1427
Silver LM (1995) Mouse genetics—concepts and applications. Oxford University Press, Oxford
Taylor BA (1978) Recombinant inbred strains: use in gene mapping. In: Morse HC III (ed)
Origins of inbred mice. Academic Press, NY, pp 423–438
Williams RW, Gu J, Qi S, Lu L (2001) The genetic structure of recombinant inbred mice: high-
resolution consensus maps for complex trait analysis. Genome Biol 2(11):RESEARCH0046.
Epub 2001 Oct 22
Yang H, Ding Y, Hutchins LN, Szatkiewicz J, Bell TA, Paigen BJ, Graber JH, de Villena FP,
Churchill GA (2009) A customized and versatile high-density genotyping array for the
mouse. Nat Methods 6(9):663–666
Yokoyama T, Silversides DW, Waymire KG, Kwon BS, Takeuchi T, Overbeek PA (1990)
Conserved cysteine to serine mutation in tyrosinase is responsible for the classical albino
mutation in laboratory mice. Nucleic Acids Res 18(24):7293–7298
Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, Friedman JM (1994) Positional cloning of
the mouse obese gene and its human homologue. Nature 372(6505):425–432
Zou F, Gelfond JA, Airey DC, Lu L, Manly KF, Williams RW, Threadgill DW (2005)
Quantitative trait locus analysis using recombinant inbred intercrosses: theoretical and empir-
ical considerations. Genetics 170(3):1299–1311
Chapter 5
The Mouse Genome
5.1 Introduction
In Chap. 4 we explained how mouse geneticists were able to develop high-density

and high-resolution genetic maps of the mouse genome by taking advantage of the
unequaled strategies and tools they had at their disposition: i.e., inter-subspecific
crosses, recombinant inbred strains, radiation hybrids and a wealth of polymorphic
molecular markers of all kinds. We also explained how the same geneticists could
develop physical maps by anchoring virtual (i.e., in silico) DNA fragments cloned
into BACs, YACs or cosmids onto the molecular markers previously ordered along
each chromosome. It is clear that, while building these maps and associated librar-
ies of cloned DNAs, geneticists were in fact gathering the essential ingredients for
undertaking the logical next step: the sequencing of the whole mouse genome.
The decision to undertake such an ambitious (and, at the time, expensive) pro-
ject was made at the turn of the millennium and was strongly influenced by the
decision to sequence the human genome, made a few years earlier (International
Human Genome Sequencing Consortium 2001; Venter et al. 2001). A first draft
of the mouse genome sequence was released in 2002, only a few months after
the release of the first draft sequences of the human genome (Mural et al. 2002;
Mouse Genome Sequencing Consortium, Waterston et al. 2002) and 2 years before
the publication of the rat sequence (Gibbs et al. 2004).
The completion of these projects, as we will see in this chapter and the fol-
lowing chapters, had an enormous impact in many areas of genetics and biology.
Making these genome sequences available to the community provided a wealth
of information about genome structure and evolution through the identification of
similarities and differences across species. As Robert Waterston and his colleagues
wrote in the conclusion of their publication: “The mouse provides a unique lens
through which we can view ourselves […]. With the availability of [its genome]
sequence, it […] provides a model and informs the study of our genome as well”
(Mouse Genome Sequencing Consortium, Waterston et al. 2002).

128 5 The Mouse Genome
Nowadays, geneticists have direct and free access to a variety of high-quality

genomic sequences through the Internet, and most of them would probably find it
difficult to work without having these tools at hand.
5.2 The Sequence of the Mouse Genome
The availability of the mouse genome sequence represented such an impor-

tant piece of information for the development of the genetics of this species that
it would certainly have become available sooner or later, for example, as a con-
sequence of the continuous addition of the ever-increasing number of sequence
fragments released by independent laboratories. However, such a disorganized
approach would have inevitably resulted in delay, in a sequence with plenty of
gaps and redundancies, and finally in a higher cost. Retrospectively, the decision
to give support and priority to the complete and systematic sequencing of the
mouse genome, and to make it a concerted project completed by a team of special-
ists, should be considered very wise. This decision was also very altruistic because
the laboratories that did not have easy access to sequencing facilities, for whatever
reasons, can now benefit from this resource, entirely free of charge, for designing
their experiments. Further evidence of this achievement is provided by the enor-
mous and ever-increasing number of scientific papers that have been published
since the release of the initial draft sequences of the mouse genome and make
direct reference to it. This trend will certainly grow in the years to come with the
progress made in sequencing technologies and the associated dramatic reduction
in cost.
The sequence of the rat genome has also turned out to be a valuable piece of
information for geneticists, because it has allowed three-way comparisons with the
other two species (human and mouse). These comparisons have provided details
about how evolution proceeds over a relatively short timescale. As mentioned in
Chap. 1, the human and rodent lineages split around 75–80 Myr ago, while the
mouse and rat lineages split around 12–14 Myr ago.
5.2.1 The Mouse Genome is Enormous in Size, and its

Structure is Complex
Measurements of the intensity of the brilliant purple color performed on mouse

cell nuclei (early spermatids, for example), after a Feulgen reaction, indicated
that the DNA content of the mouse haploid genome corresponds to approximately
3 × 10−12 g (= 3 pg), which translates into a molecular weight of ~1.8 × 1012
daltons (Da). Since the average molecular weight of a double-stranded DNA
base-pair (bp) is ~600 Da, this means that one expects to find ~3 × 109 bp or 3.0
5.2 The Sequence of the Mouse Genome 129
Giga-base-pairs (Gb) of DNA in a mouse haploid genome (Silver 1995). This

is ~650 times more than in the genome of the bacterium Escherichia coli K-12,
which comprises 4,639,221 bp. To translate this into more palpable terms, we
computed that, if the haploid mouse DNA sequence was printed as a single line
using the 11-point Courier font, all in uppercase, to symbolize the four bases
(A, T, G, C), the length of this line would be roughly equal the distance from
London to New York City (5,600 km or 3,480 miles). To express this still differ-
ently, the printed transcription of the message in 12-point Times font would rep-
resent around 3,500 books with a size similar to the one you have in your hands.
However, although obviously enormous, this sequence can be stored on the hard
disk of a personal computer (Silver 1995). Finally, mouse nuclear DNA has an
A + G/C + T ratio of 49.99/50.01 (~1), as in human.
Aside from its large size, the mouse genome is also heterogeneous.
Biophysicists who studied the thermodynamics of nucleic acid denaturation/rena-
turation had already recognized this peculiarity, over 40 years ago, by measuring
the C0t1/2 value, a parameter reflecting the structural heterogeneity of a DNA sam-
ple that is based on the speed of reconstitution of double-stranded DNA (dsDNA)
from previously denatured single-stranded DNA (ssDNA). The same biophysicists
also demonstrated that some fractions of the mouse genomic DNA renatured much
faster than others as a consequence of a high proportion of repeated sequences.
Another interesting comparison is between the physical size of the mouse and
pufferfish (Takifugu rubripes) genomes, leading to the observation that the genome
of the fish is about nine times smaller (0.35 pg of DNA or 340 Mb) than that of
the mouse. Considering that all vertebrates presumably have a similar number of
protein-coding genes (between 20,000 and 30,000, as we will discuss further), it
has been suggested that the difference in size between the two genomes is prob-
ably due to the presence, in the mouse but not in the fish, of non-protein-coding
DNA sequences of unknown function.
The mouse genome also contains sequences that are repeated many times. This
was revealed by the observation that, if we use a randomly cloned 1–2-kb DNA
segment as a probe and label it with a fluorescent dye, in most cases this probe will
hybridize with several chromosomal regions, indicating extensive redundancies.
Finally, if we consider that there are between 20,000 and 30,000 genes in a
mouse genome (which is a reasonable guess) and only 4,377 genes in E. coli, this
indicates that the average gene density in the mouse is much lower than in the
bacterium (~1/100 kb in the mouse versus roughly ~1/1 kb in the bacterium). All
these observations support the idea that a large proportion of the mouse genome
does not code for proteins and may represent what Susumu Ohno called “junk”
DNA (Ohno 1972)—unless we find that part of the DNA in question serves other
functions that might be important.
Considering all these issues (i.e., a genome with a large size, with a heteroge-
neous structure, with many redundancies and a large amount of possibly “junk”
DNA), scientists were then warned from the beginning that unraveling the com-
plete sequence of a mammalian genome would be a long and difficult enterprise.
5.2.2 How Was the Mouse Genome Sequenced?
There are basically two strategies for sequencing a complete mammalian genome.
The first one, known as hierarchical shotgun sequencing (HSS), makes use of
cloned DNA with large inserts such as bacterial artificial chromosomes (BACs—
with 150–250 kb DNA inserts), P1 phages or, less frequently, yeast artificial
chromosomes (YACs—200–1,000 kb). As explained in Chap. 4, these clones of
DNA are assembled into a series of overlapping elements known as contigs (from
contiguous DNA segments), which altogether make a physical map encompass-
ing chromosomal segments of the greatest possible dimension. The DNA clones
mentioned above are generally selected once they have been thoroughly checked
for structural integrity, rejecting those that are chimeric or have deletions (a situa-
tion that is common in YACs but less common with BACs). The assembly of these
cloned DNAs into contigs is achieved by careful fingerprinting of each and every
clone. When the contigs are established, in general from several individual clones
ranging from 100 to 1,000 kb, a subset of minimally overlapping clones is cho-
sen and each of its elements is sequenced several times to minimize the effect of
sequencing errors (this minimal set is sometimes called the “Golden Tiling Path”
or simply the “Golden Path”). The primary sequence is called a read and the
released genome sequence, or draft, results from the integration of several inde-
pendent reads (in general 10–15, sometimes more). After computerized process-
ing of these independent reads, and if we suppose that the sequencing errors occur
randomly, the final rate of errors in a given consensus sequence is very low, in
general less than one error per 105 bp.
The HSS strategy is relatively slow and tedious, but it is systematic, progres-
sive and highly reliable. The use of clones with large DNA inserts is also a way to
circumvent, at least to a certain extent, the difficulties associated with the sequenc-
ing of DNA repeats and variations in copy number, which are true nightmares
for sequencers. However, the HSS strategy has the disadvantage that only long
DNA fragments cloned in a vector can be sequenced. Unfortunately, it is virtually
impossible to clone the whole of a mammalian genome in BAC or YAC vectors,
for reasons that are associated with both the structure of the DNA in some chro-
mosomal regions and with the cloning technology.
A second strategy, called whole-genome shotgun (WGS), consists of the
mechanical fragmentation (e.g., by sonication) of the mammalian DNA into seg-
ments measuring 100–400 bp, which are sequenced from both ends using the
chain termination method. Multiple reads of the targeted DNA are obtained by
performing several independent rounds of this fragmentation, each followed by
sequencing. Once the sequence of the targeted DNA is achieved, computer pro-
grams are then used to assemble the pieces of the puzzle, ordering the individual
fragments into virtual contigs, then in super- or hypercontigs and finally in ultrac-
ontigs based on the overlapping sequences of the different reads.
The WGS method is fast and (in theory) does not require the pre-existence of a
physical map. Unfortunately, it does not allow the sequencing of certain genomic
segments such as highly repeated regions. Combining the two strategies (WGS
first, then HSS) allows for the correction of most of these difficulties. In short,
the two strategies are complementary: WGS provides rapid and relatively good
coverage early in a project, while HSS is more systematic and more efficient
for the sequencing of regions with repeated sequences. The human genome was
sequenced by using mostly the HSS strategy, while the mouse and all other mam-
malian genomes were sequenced by using mostly the WGS strategy, with the help
of HSS only for finishing some regions.
In fact, technical and methodological difficulties emerge when the objective
is to sequence the genome of a species for the first time (the human genome in
this case), but the situation is greatly simplified when the project is to sequence
the genome of evolutionary related species. This is because it is possible to take
advantage of the existence of the many interspecific structural homologies that
exist at the chromosomal level. Thus, the mouse and rat genomes were sequenced
mostly by WGS, and accordingly were completed much faster than the sequencing
of the human genome (Fig. 5.1).
Sequencing techniques have progressed enormously recent years and many
steps are now fully computerized, reducing human intervention and cost. The lat-
est assembly released by the Mouse Genome Sequencing Consortium (MGSC) has
a length of 2,730,871,774 bp (Golden Path from Ensembl—September 2013).
Curators of the database consider that at least 99 % of the mouse genome sequence
is established, with the exception of only a few small gaps (~180) scattered in
between a total of 750 contigs, with less than one sequencing error per 105 bp. All
of the chromosomes have been entirely sequenced, including the X and the Y,
allowing comparisons with homologous regions of the human and other mamma-
lian genomes to be performed at a very high resolution.1
Such comparisons, revealing similarities and differences, are a rich source of
information. Similarities (i.e., sequence conservation), as we will discuss later,
allow us to detect regions that are very likely under selective pressure and which,
for this reason, have remained unchanged or nearly so for millions of years, indi-
cating that they are presumably genetically important and, accordingly, have
resisted random drift. Differences at the sequence level may be even more inter-
esting a priori, because they may contain information explaining how speciation
proceeds. It will be obviously interesting to discover both the mechanisms govern-
ing these processes and the consequences of these differences at the phenotypic
level. We will come back to this point several times, which is well exemplified in
the case of variations in gene or DNA copy numbers (copy number variations or
CNVs, see Sect. 5.3.6.).
The mouse sequencing project was undertaken by the MGSC, an organization
that consisted originally of three laboratories: the Whitehead Institute for Biomedical
Research at the Massachusetts Institute of Technology (USA), the Washington
University Genome Sequencing Center (USA), and the Wellcome Trust Sanger
1 The mitochondrial DNA has also been sequenced. See Sect. 5.6.

(a)
(b)
Fig. 5.1 Strategies used for sequencing mammalian genomes. Two strategies have been used
for sequencing the mammalian genomes: hierarchical shotgun sequencing (HSS) and whole-
genome sequencing (WGS). HSS (Fig. 5.1a, b) has been used for sequencing the human genome.
It works in two successive steps and makes use of bacterial artificial chromosomes (BACs, ~150–
300 kb) or yeast artificial chromosomes (YACs, ~500–2,000 kb) that have been previously used
for the establishment of the physical map or “contig map”. In the first step (a), the integrity and
quality of these cloned DNAs is carefully checked (absence of mosaicism, absence of deletion).
Then the most interesting elements (b) of these contigs (those representing the “golden path,”
with minimum overlap) are completely sequenced and the sequence ordered. The HSS strategy
is systematic and reliable, but it is slow and does not allow the sequencing of regions with repeti-
tive DNA. The whole-genome sequencing strategy (WGS) (Fig. 5.1c, d, e) has been used for
sequencing most of the mouse genome. This strategy completely bypasses the BAC/YAC step
and consists of the direct mechanical fragmentation of DNA samples to obtain a mixture of inde-
pendent, randomly cut stretches of DNA 100–400 bp long (c). These stretches are then cloned
using adaptors, labeled, and sequenced end-to-end (d). In a third step (e), sequence overlaps are
looked for by using appropriate computer software and the clones are then arranged in a head-
to-tail manner to form virtual contigs of non-redundant, top-quality sequences. In the final step,
the contigs are anchored to the specific chromosome they belong to. The process is generally
repeated several times to reduce the number and size of the unsequenced regions and strengthen
the quality of the sequence. The gaps in the sequence resulting from the WGS strategy are filled,
where possible, by HSS. In the current mouse sequence, the number of gaps is extremely reduced
(c)
(d)
(e)
AATGTAGCCTGACTCCCTAGTATGCTTCTCC
AATGTAGCCTGACTCCCTAGTATGCTTCTCCCTAGTACCTAGTAAGG
TGCTTCTCCCTAGTACCTAGTAAGGCTCCTC
ACCTAGTAAGGCTCCTCCCTTCCCTAGT
TCTCCCTAGTACCTAGTAAGGCTCCTCCCTTCCCTAGTAAGTACTAGTACTGTAGCCT
GCTTCTCCCTAGTACCTAGTAAGGCTCCTCCCTTCCCTAGTAAGTACTAGTACT
TCCCTAGTATGCTTCTCCCTAGTACCTAGTAAGGC
AATGTAGCCTGACTCCCTAGT
CTTCTCCCTAGTACCTAGTAAGGCTCCTCCCTTCC
TTCCCTAGTAAGTACTAGTACTGTAGCCTAGTCTAATGCA
AATGTAGCCTGACTCCCTAGTATGCTTCTCCCTAGTACCTAGTAAGGCTCCTCCCTTCCCTAGTAAGTACTAGTACTGTAGCCTAGTCTAATGCA
Fig. 5.1 (continued)
Institute (UK). Based on discussions with the scientific community, MGSC investi-
gators decided to sequence, first, the genome of a female from the C57BL/6 inbred
strain. At the same time, four other inbred strains (A/J, DBA/2J, 129X1/SvJ, and
129S1/SvImJ) were being sequenced by the CELERA firm in another independent
WGS project. Here again, interstrain comparisons have been of great interest when
matched with particular phenotypes. Nowadays, the original projects are finished,
even though molecular biologists at the MGSC keep working on some specific
regions. The Mouse Genomes project from The Wellcome Trust Sanger Institute
recently completed the sequencing of an additional 17 inbred mouse strains: 129P2,
129S1/SvImJ, 129S5, A/J, AKR/J, BALB/cJ, C3H/HeJ, C57BL/6NJ, CAST/EiJ,
CBA/J, DBA/2J, LP/J, NOD/ShiLtJ, NZO/HiLtJ, PWK/PhJ, SPRETUS/EiJ, and
WSB/EiJ (see http://www.sanger.ac.uk/resources/mouse/genomes/). These strains
were very carefully selected after extensive discussions via the Internet among the
members of the community of mouse geneticists. The genome of the FVB/N inbred
strain, popular for the production of transgenic animals and for skin carcinogenesis
studies, is now also available (Wong et al. 2012).
These genome sequencing projects are now benefiting from new, ultra-efficient
sequencing technologies known as next-generation sequencing (NGS). It is likely,
for example, that many genome sequences from highly informative strains (strains
from the Collaborative Cross, for example—see Chaps. 9 and 10) or even some
carefully selected individual mice will become available, contributing efficiently
to the analysis of complex traits. Even if the development of bioinformatics
resources for the interpretation of the tremendous and ever-increasing amount of
data remains a challenge, we can say that the mouse genome-sequencing project
is, without any qualification, a complete success from an analytical point of view.
However, from now on scientists will have to consider a new challenge, at least as
important: the annotation of all sequences in this genome. No doubt they will be
kept very busy for another few years.
5.3 The Structure of the Mouse Genome
Once a genome is entirely sequenced and the sequence stored in a database, sci-
entists can then start looking at it in more depth. This structural analysis, run in
parallel with a functional analysis, is part of the so-called genome annotation
process, and one of the first challenges in this matter is to identify and charac-
terize as accurately as possible the DNA regions containing the genes proper
(i.e., the DNA coding for proteins or RNAs), the regulatory elements, and
some other potentially important structures. This is a real challenge because,
if we recall what we said earlier when discussing gene density in mammalian
genomes, the protein-coding and related sequences represent only a very small
proportion of the mammalian DNA. However, if we consider that this func-
tionally important fraction of mouse DNA, because it is under the constraint
of purifying (i.e., negative) selection during evolution, is likely to be highly
preserved across different species, we already have outlined a strategy to iden-
tify and estimate it. This estimation has been achieved, shortly after the release
of the first draft of the mouse sequence, by cross-comparing several regions
of the human genome with various short sequences of the mouse genome, and
the answer was that there is indeed great interspecific homology (over 95 %)
for around 3.5–5 % of the genomic DNA sequences. There are good reasons
to believe that the genes encoding proteins and other important sequences are
gathered in this fraction.
5.3.1 Finding the Coding and Related Sequences
The first step in the process of genome annotation generally consists of check-
ing for the presence or absence in the newly sequenced genome of some specific
sequences previously characterized in other species (the exons, for example), and
5.3 The Structure of the Mouse Genome 135
evaluating the number of copies, their organization and flanking sequences, etc.
The geneticist may also wish to make an inventory of all the genes of a given spe-
cies: those encoding proteins and those transcribed only into RNAs. These ques-
tions have triggered a multitude of intensive studies, many of which have now
resulted in more or less precise answers.
5.3.1.1 Retrieving Specific Sequences
Nowadays, finding a particular sequence in a genome is relatively easy and several

software packages have been designed for this purpose. The most popular is
BLAST. BLAST allows similarity searches to be performed against any databases
of recently sequenced organisms. BLAST will rapidly identify and retrieve a
sequence in the human or rat genome that resembles a mouse sequence based on
similarity of sequence. These software packages work, roughly, like the sub-pro-
grams that are activated when, working on a text file, one selects the command
“Find” to specifically retrieve a chain of characters, with the important difference
that BLAST can retrieve sequences that are not 100 % identical to the queried one.
ROSETTA2 and SEQUENCHER® sequence analysis softwares are other packages
useful for finding genes (and not only coding sequences) by comparisons, for
example, between human and mouse DNAs. ROSETTA performs sequence align-
ments and compares the exon sizes, splicing sites, etc., and finally makes gene
predictions.3
When a coding sequence (a mouse exon, for example) is used as a template
for retrieving the most closely related sequences in the human or rat genomic
sequence, in ~95 % of the cases BLAST retrieves a sequence with high similarity
and 90 % of these sequences are on the homologous chromosomal segment in all
three species. Geneticists say that they share the same syntenic location (from the
Greek, meaning “on the same ribbon”) and these genes are called 1:1 orthologs.
This indicates that most of the genes in a given mammalian genome are part of
an ancestral heritage and do not vary much among other mammalian species even
if, sometimes, there are variations in terms of copy numbers, as we will discuss
further. Differences in terms of presence versus absence are rare but occasionally
occur. For example, approximately one hundred predicted mouse genes identified
in the initial mouse draft sequence were reported as missing (having no homolo-
gous counterpart) in the human genome. The reverse of course is also true, and
some human genes are absent in both the mouse and rat genomes. A good example
of such a situation is the gene encoding human interleukin 8 (IL8), which cannot
be found in the rat and mouse regions of homology for HSA-Chr 4 (see Fig. 5.2).
2
https://www.rosettacommons.org/.
3 Sequencherversion 5.1 sequence analysis software, Gene Codes Corporation, Ann Arbor, MI
USA http://www.genecodes.com.
ALB
Human
?
Mouse
Rat ?
Fig. 5.2 Sequence comparisons between mammalian genomes. The orthologous copy of the

human gene encoding interleukin-8 (Il8) is missing in the mouse and rat genomes. The fig-
ure shows the region of human chromosome 4 (HSA4) where the IL8 gene is located, with
the homologous regions in mouse chromosome 5 (MMU5) and rat chromosome 14 (RNO14).
The rat chromosome is affected by a paracentric inversion when compared with the human and
mouse homologous regions. Such rearrangements are extremely common in the mammalian
genomes and are very useful (with other methods) for establishing the phylogenetic relationships
between species. The images are from the Ensembl Genome Browser database (May 2013)
These qualitative differences are not easy to explain and can result either from
true deletions, with no consequences at the phenotypic level, or from the fact that
the supposedly deleted genes in fact still exist elsewhere in the genome but have
evolved so rapidly, in one or the other lineage, that they are no longer recogniz-
able as orthologs based on sequence comparisons. The first hypothesis is the most
likely, since such segmental deletions of recent origin have been discovered, by
chance, in the genome of several inbred strains while others were reported normal
(undeleted). For example, mice of the C57BL/6JOlaHsd substrain (also known as
C57BL/6S) are homozygous for a deletion encompassing the entire α-synuclein
gene (Snca-Chr 6) (Specht and Schoepfer 2001). These mice are fertile and have
a normal lifespan, but they have at least one gene inactivated compared with most
other C57BL/6 substrains. Examples of this kind have been reported in many other
laboratory inbred strains and also exist in the human and rat species (Perez et al.
2013).
Finding genes in the genome of one species, once the orthologous versions of
these genes are known and already identified in the genome of another closely
related species (such as human, rat, and mouse), is then relatively straightforward
and many computer programs can do this, even if surprises and difficulties occa-
sionally occur, as we will see later.
5.3.1.2 Identification of the Coding Sequences
The situation is more complicated when the objective is to identify all the coding
sequences (all the exons, for example) in a freshly sequenced genome.
A first and relatively efficient technique, known as exon trapping, was pub-
lished in 1991 (Buckler et al. 1991). With this technique, a cloned genomic DNA
was inserted, by genetic engineering, into an intron of the human immunodefi-
ciency virus 1 (HIV-1) tat gene (Trans-Activator of Transcription), contained within
the plasmid pSPL1. COS-7 cells were then transfected with these constructs, and
the resulting RNA transcripts were processed in vivo. The splice sites of exons
contained within the inserted genomic fragment were put in phase with the splice
sites of the flanking tat intron. The mature RNA collected from the COS-7 cells
contained the potential exons, which could then be amplified via RNA-based PCR
and ultimately cloned.
Exon trapping has been a very helpful technique, especially in the projects
whose aim was the positional cloning of a gene identified only by a mutant allele.
However, compared to more recent techniques, it has two major drawbacks: (i)
it does not trap faithfully the large or very small exons, and (ii) it is relatively
expensive because it requires a significant amount of bench work and in vitro cell
culture.
5.3.1.3 Using Expressed Sequence Tags (ESTs) for the Detection

of Transcribed Sequences
Taking into account the fact that several mammalian genomes are now entirely
sequenced, strategies have been developed that are based on the identification at
the genome level, by all possible techniques, of sequences deduced from tran-
scribed products. One of the first strategies consists in using so-called Expressed
Sequence Tags (ESTs). ESTs are short sub-sequences of cDNA corresponding to
a few hundred (~350–500) base-pairs of a cDNA, starting from the 3' end, some-
times from the 5' end. Millions of such ESTs (from several mammalian species)
are available in public databases, and the sequence of each of these ESTs can be
used as a molecular probe to retrieve the complete sequence of the gene the EST
belongs to (or is related to), simply by “pulling on” the flanking sequences. Since
the ESTs stored in a given database were in general prepared from a specific tissue
(brain, blood, skin, neoplastic tissue, etc.) at a certain step of development (embry-
onic, 10 days, adult, senescent, etc.), using these ESTs for gene identification has
the additional advantage of providing information concerning the transcriptional
level and the gene expression pattern for the annotation process. ESTs have been
instrumental for the initial identification of many genes in the mouse as well as in
the human genome, and still are. In addition, the sequence alignments can be per-
formed entirely in silico, which means rapidly and at virtually no cost. The major
drawback of these ESTs is that only a fraction of the genes are expressed simulta-
neously, and consequently the EST collection in a particular database represents
only a fraction of the genes of a given species. Finally, some genes are transcribed
only in particular circumstances, at very low levels, or transiently and, by defini-
tion, they are poorly represented in EST libraries or databases.
5.3.1.4 Using Strategies Based on Artificial Intelligence
Other strategies, requiring sophisticated informatics, rely on the identification

of some transcription-related motifs that are part of most protein-coding genes
(Blanco and Guigo 2005; Harrow et al. 2009) (see also next section). These
motifs have been successfully used for gene detection with software systems like
GENSCAN, developed by Burge and Karlin (1997). In addition to the strategies
mentioned above, more refined prediction programs, often referred to as de novo
or ab initio gene finders, have also been developed by geneticists and computer
scientists. These programs are based on the existence of subtle differences at the
sequence level that can be used to sort out putative coding regions from non-cod-
ing regions by making use of the so-called hidden Markov chain models. These
prediction models are based on the fact that biases and dependences exist in cod-
ing sequences that are not observed in non-coding regions. This means, for exam-
ple, that the five preceding bases influence the probability of finding a particular
base at the sixth position of a new sequence if, and only if, the sequence in ques-
tion is a coding sequence. When scanning a novel nucleotide sequence, the pro-
gram computes a coding likelihood score, based on a Markov chain model of order
5, and makes an assessment as to whether the sequence is more likely to be from
an intron, exon or intergenic region (Harrow et al. 2009).
All these sequence prediction algorithms are being constantly improved based
on the experience acquired from training with DNA samples whose sequence is
fully annotated. These programs work more or less like the software designed for
language translation. Years ago, the meaning of “computer-translated” sentences
was only remotely related to the meaning in the original sentence and sometimes
limited to an unordered set of key words. Nowadays, the quality of the translation
is very good (at least for certain languages). Based on their encouraging results,
researchers consider that, as of today, around 85 % of genes should be rapidly
and easily detected in any new mammalian genomic sequence by using software
of this kind. Most of these newly discovered genes must, however, be validated
by other approaches because the discovery of a gene-like structure does not auto-
matically mean that an authentic, indisputable, and functional protein- or RNA-
encoding gene has been “fished”. This validation is very important work, whose
aim is to create a gold-standard reference for gene annotation. A program of this
kind has been undertaken by the Human and Vertebrate Analysis and Annotation
(HAVANA) team at the Sanger Institute, where the human, mouse, and zebrafish
genomes are carefully annotated manually.
Making sequence comparisons (or alignments) with other genomes (human,
rat, zebrafish) has allowed a rather rapid identification of a great number of mouse
genes. However, from now on, the identification of novel genes in the mouse will
probably progress at a somewhat slower pace because the situations researchers
face are sometimes difficult. Some genes, for example, are very large and exten-
sively fragmented, while others are very small with only one intron or even no
intron at all (for example, the intronless genes encoding RNAs and histones).
Since neither of these two categories of genes correspond to the “canonical” repre-
sentation of most mammalian genes, they have to be annotated manually and this
takes much more time. Another very common situation is that, although they share
a syntenic location as expected, orthologous genes are not always in a 1:1 ratio but
rather in 1:2, 1:3, and so on. We will describe situations of this kind, where the
“pseudo-orthologous” copies are sometimes slightly altered or incomplete, but are
still transcribed and accordingly annotated as a true gene.
Finally, overlapping and nested genes have been shown to exist in mammals
just like in Drosophila, with various imbrications of their structure with their
neighboring genes. Nested genes were generally described as genes with a rel-
atively short size, consisting in general of only one exon and entirely nested
within a single intron of a host gene. The situation has recently changed dra-
matically as a consequence of more in-depth analysis of the mouse transcrip-
tome, as we will discuss further in this chapter, and many RNAs are transcribed
from the mouse genome whose function is not yet established. In the same
way, genes have been found that are transcribed in the opposite orientation
to their neighboring host genes, and sometimes negatively influence the tran-
scription of these genes via antisense-mediated inhibition (see below—Chap.
6 on X-inactivation). Identification of nested genes is difficult but, fortunately,
approximately 60 % of nested genes are conserved in mouse and human in the
same genomic context.
The ENCODE project (ENCyclopedia Of DNA Elements), which is essentially
the next step for the Human Genome Project, has set as its major aim the estab-
lishment of all the structural and functional elements of the genome. It is defi-
nitely an ambitious project but it makes a lot of sense and is really necessary if we
consider its potential applications. Here again, just like for the sequencing of the
mouse genome, we can say that this meticulous analysis conducted at the DNA
level would have to be achieved one day because the general feeling of the com-
munity is that it is a crucial endeavor, if not simply the essence of genetics: then
why not do it right now, as rapidly as possible, on a systematic basis?
The preliminary results of the ENCODE project, although still fragmentary,
have already changed our understanding of the mammalian genome by demon-
strating that the mammalian DNA hitherto labeled “junk” might not be junk after
all.
5.3.2 The Canonical Architecture of a Protein-Coding Gene
As discussed in the preceding section, many points remain to be elucidated con-

cerning the structural organization of the mouse genome. However, as of today,
hundreds of genes have been entirely sequenced in several species including
mouse, rat, human, and domestic animals. As a result, it is now possible to outline
the classical or canonical architecture of the “average” mammalian gene.
A gene is a segment of DNA that encodes an RNA molecule that may or may
not be translated into a protein. For this reason, geneticists formally distinguish
two types of genes: the protein-coding genes and the non-protein-coding genes.
For many years, and up to relatively recently, molecular geneticists considered that
the two strands of the DNA molecule were not equivalent: one of them was the
coding strand while the other was the template or anticoding strand. However, and
unexpectedly, it has recently been demonstrated that mammalian DNA is perva-
sively transcribed from both strands. We will come back to this important point
later in this chapter when discussing the transcriptome and the non-protein-coding
RNAs. Here, we will simply discuss the organization of a classical protein-coding
gene as it has been established as the result of thirty years of careful positional
cloning, sequencing, and annotation.
The transcription of a protein-coding gene into a primary mRNA proceeds
from the 5' to the 3' end and starts ~50–60 bp upstream of the first AUG codon,
encoding a Methionine. The ~50 bp between the transcription initiation site and
the initial AUG is part of the so-called 5'-untranslated region (5'-UTR) or leader
sequence. This sequence usually contains a ribosome binding site (RBS), known
as the Kozak sequence (gcc)gcc(A/G)ccAUGG (Kozak 1987), that includes the
AUG initiation codon.
Upstream of the transcription start site at the 5' end, several consensus
sequences have been identified that are part of the promoter sequence of the
gene, as we will see in the next section of this chapter. Opposite to the 5'-UTR
is the 3'-UTR or trailer sequence, required for the processing of mRNA, the size
and canonical sequence of which is not as precisely known as that of the leader
sequence. The end of a structural gene is called the transcription termination site.
Some specific sequences are also found in the 3'-UTR. First is a polyadenylation
signal composed of sequences like AAUAAA or a slight variant. The polyadenyla-
tion signal indicates that transcription will be terminated approximately 30 base-
pairs downstream of it, while a tail composed of a few hundred adenine residues
(the poly-A tail) will be added to the transcript. The poly-A tail is important for
the nuclear export, translation, and stability of the mRNA.
Since 1977, it has been established that many (around 60 %) mammalian
genes have a heterogeneous structure: some parts are included in the final protein
or RNA product, while others have another destination or are merely degraded.
Hence, the coding sequences of most mammalian genes are composed of an alter-
nation of exons (expressed region) and introns (intragenic region). Introns are
spliced off during RNA processing or maturation when the pre-mRNA becomes
a mature mRNA, ready to be translated into a protein product. RNA splicing is

a complex and very precise procedure that is regulated and controlled at the cel-
lular level, at the base-pair level of precision (Fig. 5.3). This process requires
several highly specific tools: at least five small nuclear RNAs and around 150 pro-
teins, collectively known as the spliceosome (Hoskins and Moore 2012). Among
the most important are the small nuclear ribonucleic acids or snRNAs, the small
nucleolar RNA or snoRNAs, and specific enzymes including the ribonucleopro-
teins or “snurps”.
Splicing sites can be identified at the DNA level because they have a consensus
sequence: the first two bases at the beginning of an intron (at the 5' end) are almost
always GT and the last two, at the 3' terminus of the same intron, are almost
always AG. The sequences immediately upstream of the AG and downstream of
the GT are also conserved, although to a lesser degree. For example, the intronic
region upstream from the AG is usually a region rich in C and T. The regions at
the 5' boundaries of the introns are called the donor sites and those at the 3' end
are called the acceptor sites. We already mentioned earlier that these splicing sites
have been used for the identification of exons with the exon trapping technique.
Most sequence identification software can also identify these sites in a mouse
DNA sequence and label them as “candidate splicing sites” (Fig. 5.4).
Not all exons in a gene are spliced and subsequently assembled to form the
final RNA product. In fact, if we consider that the exons correspond to “functional
units” and the introns are “spacers” between these functional units, we observe
that the exons can be assembled into different combinations to produce different
polypeptides. This is known as alternative splicing and it is estimated that ~95 %
of multi-exonic genes are alternatively spliced in mammalian genomes. From
numerous observations, it is also known that the exons in a gene are of two types:
(i) those that are always present in all transcripts, which are often referred to as
constitutive or major forms of transcripts; and (ii) those that are optional or alter-
native. Exons of the second type, those that are only included in some spliced
ŝŶƚƌŽŶϭ
Fig. 5.3 Canonical (and simplified) representation of a protein-coding mammalian gene. The

enhancers represented in the figure are not always present and are sometimes distant from the
promoter region by several Mb. Many sequences in the promoter region are important for gene
regulation, but not all of them have been identified and they probably vary from one gene to
another. Not all genes have a CAAT box or a TATA box. Finally, not all genes have intronic
sequences, and not all exons are represented in the final product
introns exons introns
acceptor site donor site
Fig. 5.4 Splicing sites. The figure represents the splicing sites found in the sequence of the gene
encoding the mouse leptin receptor (Lepr-Chr 4). The intronic sequence of the donor (GT) and
acceptor (AG) sites are highly preserved (this is known as the GT-AG rule)
forms, as opposed to the major transcript forms, are mostly not conserved across
species and are probably of recent origin (Modrek and Lee 2003) (Fig. 5.5).
Alternative splicing can generate a large variety of proteins from the same
DNA coding sequence by modifying the exonic contribution of the mature mes-
senger RNAs. It is clear that the actual number of genes in a species has only a
relative meaning, since the splicing machinery can tremendously increase genetic
diversity. In this context, the number of exons is certainly much more informa-
tive for researchers than the number of genes. Alternative splicing is considered to
be a very important mechanism for resolving the discrepancy between actual gene
number and organismal complexity.
The mechanisms regulating alternative splicing and leading to the incorpo-
ration, or lack thereof, of a specific exon into the final product (sometimes des-
ignated the splicing code) are not yet completely unraveled. These processes
probably involve trans-acting proteins (repressors and activators) encoded else-
where in the genome that pair with cis-acting regulatory targets on the pre-mRNA.
It is also likely that the secondary structure of the pre-mRNA transcripts plays a
role in the regulation of splicing (Barash et al. 2010).
As we will discuss in Chap. 7 and as demonstrated by hundreds of positional
cloning experiments performed in the mouse and rat, splicing sites are common
targets for the occurrence of mutations. These sites are not always in frame with
the sub-modulation of the mRNAs in triplet, and can also occur within codons.
(a) Promoter 1 2 3 4
1 2 3 4
Protein A
(b) Promoter 1 2 3 4
1 2 4
Protein B
(c) Promoter 1 2 3 4
1 3 4
Protein C
Fig. 5.5 Alternative splicing. In mammals, around 95 % of multi-exonic genes are alternatively

spliced to produce different proteins (A, B, C …). Some exons are present in all transcripts
(constitutive or major forms of transcripts), while others are optional or alternative. Exons of
the second type are mostly not conserved across species and are probably of recent origin. For
orthologous genes, the number of exons is sometimes variable among species, and the presence
of recently captured exons is sometimes observed
In these cases, of course, the two contiguous exons are inseparable and are jointly
incorporated into the transcript or skipped. The mRNA transcript, once adequately
spliced, receives a cap of a methylated guanine nucleotide that is added to its 5'
end to protect it.
The enormous amount of information collected by mouse geneticists indicates
that the average size of a mouse gene is approximately 30–40 kb at the DNA level,
while the average mature or processed mRNA molecule (mRNA mature tran-
script) is approximately 2 kb. The average gene density is in the range of 1 gene
per 95 kb of DNA, i.e., very close to the predictions. The smallest (known) gene
is 0.1 kb and encodes the t-RNATyr. The largest gene is Titin (Ttn-Chr 2), with
2.8 Mb of genomic sequence and 363 exons producing a spliced mRNA larger
than 100 kb. The introns are also of various sizes, ranging from around 0.5 kb
for the short ones to 30 kb for the longest (dystrophin-Dmd), with an average
intron size of 4.7 kb. For the exons, the shortest consists of only 9 bp (exon 29 of
Myo5a), and the largest is 7.6 kb long (exon 26 of Apob), with an average exon
size of approximately 290 bp. Altogether, when added up, the exons represent
1.2 % of the total mouse DNA, the introns 26.7 %, and the intergenic regions
69.3 %. The number of exons per multi-exon gene varies from 1 to 363 with an
average of 8.4. Finally, around 4,000 genes have only one exon.
The configuration of the “typical” mouse structural gene, as we just outlined
it, is probably very similar to the average mammalian gene, and this is a blessing
for the establishment of comparative maps; in short, the DNA sequence of two
(not only one) mammalian genomes is an invaluable tool for making predictions
about a third one. Many examples could be obtained from the cross-compari-
sons of mouse, rat, and human sequences. As of today, 17,054 mouse genes have
an orthologous copy in the human genome, while 18,458 mouse genes have an
orthologous copy in the rat. Finally, a total of 20,388 mouse genes have orthology
annotations with at least one other species.
The classical gene we just described corresponds to a protein-coding gene. In
fact, we now know that this category of genes represents only a proportion of the
genes in the mouse genome that specialists consider to be in the range of 25–30 %.
Most other genes encode RNA molecules of various sizes: some have an open
reading frame (ORF) but most do not. Some are spliced, others are not, and the
majority of these transcripts are processed further in smaller molecules. Most of
the RNAs stay in the nucleus, suggesting that they have a function. Finally, all
these RNAs exhibit a rather low degree of interspecific homology, indicating
that the selection pressure they experience is of a different type. We will discuss
this point more extensively at the end of this chapter when discussing the mouse
transcriptome.
In November 2014, the Mouse Genome Informatics database estimated the
number of mouse genes with nucleotide sequence data at 34,628 and the number
of genes with protein sequence data at 24,553. This information seems reliable
when compared with other species. Out of these genes, only 16,345 have experi-
mentally based functional annotation.
Finally, we must point out that the distribution of genes in the mouse genome
is very uneven. Mouse chromosome 11, for example, has twice the gene density of
chromosomes 10 or 12, and the Y chromosome has only a few genes in an “ocean”
of repeated DNA.
5.3.3 Finding the Regulatory Sequences
One of the biggest challenges of genome annotation is to identify gene regulatory

regions. These comprise proximal and distal regulatory elements, according to
their distance from the transcription starting point. Proximally are the promoters
and associated promoter elements. Distal elements are enhancers, silencers, insu-
lators and locus control regions (Fig. 5.3). Proximal and distal elements are usu-
ally composed of clusters of short intermingled transcription factor-binding DNA
motifs referred to as modules or cis-regulatory modules (CRMs) (Hardison and
Taylor 2012).
DNA sequence and local chromatin landscape act jointly to determine tran-
scription factor (TF) binding intensity profiles. As a result, a regulatory module is
defined by its sequence, since it binds transcription factors, and is thus expected
to contain specific binding sites for these. It is further defined by its accessibility
to TFs, which is linked to chromatin structural specificities such as histone modi-
fications and local occupancy by nucleosomes. These are highly dynamic events
which reflect the history of the cell and which are responsible for differential
gene expression in animal development and cell differentiation. This implies that
canonical binding sites for transcription factors are seldom sufficient to define
a regulatory module, and methods relying on binding site identification usually
have a high rate of false positives. For example, out of 132 regulatory modules
predicted by algorithm analysis to bind TCF4 (a key transcription factor in the
WNT1 signaling pathway), only 10 were validated using chromatin immunopre-
cipitation (ChIP)—little more than a random representation (Hatzis et al. 2008).
This further implies that most CRMs will be difficult to identify until the chro-
matin landscape around them is defined. As a result, whereas the transcriptional
apparatus reads the regulatory elements in the genome very efficiently, we still
lack a universal syntax to decipher them, and this is quite critical: for regions that
are defined by an unequivocal syntax, such as the coding exons, mutations can
be characterized by just sequencing the whole mutated genome, together with
low-resolution meiotic mapping, using no more than two dozen F2 mice (Xia
et al. 2010; Arnold et al. 2011). Reaching the same level of power for regula-
tory regions would change the face of gene regulation analysis. Fortunately, this
field is developing at a rapid pace, following the systematic reliance on strategies
that directly measure sequence occupancy by Transcriptional Regulatory Factors
(TRFs) within the living cell, such as chromatin immunoprecipitation followed by
DNA sequencing (ChIP-seq) or DNase I digital genomic footprinting, which are
currently performed or compiled by ENCODE (the ENCODE Project Consortium
2011—see above). Most of these results to date have been obtained for human but
major conclusions also apply to the mouse, as demonstrated by results already
obtained in this species.
Proximal regulatory modules (PRMs) at and around transcriptional start sites
(TSSs) are the most straightforward regions to identify, since the TSS is acces-
sible from the transcription product, the RNA. Cap-analysis of gene expression
(CAGE) and RNA sequencing (RNA-seq) have contributed to the definition of
TSS and consequently of PRMs. From these analyses, it appears that mamma-
lian promoters can be separated into two classes: evolutionarily conserved pro-
moters bearing a TATA box, and more plastic, evolvable CpG-rich promoters.
The latter are by far the most frequent promoters since the TATA box (with a
core DNA sequence 5'-TATAAA-3') is found in only one quarter of all promot-
ers in a mammalian cell, usually around 30 bp upstream of the transcription
start site. The TATA box, the first core promoter element identified in eukaryotic
protein-coding genes (Goldberg 1979), is an anchoring site for the pre-initia-
tion complex of transcription involving RNA polymerase II. The CpG sequence
works similarly via the Sp1 factor. A CAT (or CCAAT, or CAAT) box, with a
consensus sequence GGCCAATCT, is inserted upstream of the TATA box,

75–80 bp from the transcription start site. Some genes with relatively ubiqui-
tous expression do not have this GGCCAATCT sequence. In CpG-rich promot-
ers, the start sites are usually multiple and organized in clusters at the 5' end
of the gene, whereas TATA-box-bearing promoters have a single or at least a
predominant start site. As of 2006, 729,504 potential mouse TSS sites were
defined, organized in 159,075 clusters, a figure that far exceeds the number of
genes identified (see above) (Carninci et al. 2006). Furthermore, mapping tech-
niques such as CAGE are quantitative and provide a measure of the amount of
transcription initiation in any given genomic region or for a given gene, in dif-
ferent tissues.
The situation is much less clear for distal regulatory elements such as silencers,
enhancers or locus control regions. Enhancer elements, which can be located at
some distance from the core promoter elements, where the transcription initiation
apparatus is bound, are sites for fixation of transcription factors. The enhancer-
bound transcription factors bind co-activators such as Mediator and p300, which
in turn bind the transcription initiation apparatus, thus providing a link between
enhancers and promoters. Non-coding RNAs may be associated with Mediator in
this process (Lai et al. 2013).
Constraints on distal regulatory elements appear rather loose. Enhancers have
been located at the 5' or 3' ends of coding regions, within introns and even within
coding exons (Birnbaum et al. 2012), where they impose a further layer of con-
straint on the coding sequence. They can be close to the transcription start site
or, in contrast, extremely remote (one to several megabases)—not to mention the
possibility of them lying on a separate chromosome, from which they act in trans
on a gene-coding region (Savarese and Grosschedl 2006). Furthermore, there is
no evidence that the closest enhancer to a gene is the one likely to be active on
this gene (Li et al. 2012). In cases when regulatory modules are remote, muta-
tions that affect them may lie within another gene. For example, the CRM driving
Sonic hedgehog (Shh) expression in the limb lies within the intron of another gene,
Lmbr1, which for some time puzzled geneticists (Hill 2007). Similarly, a Gremlin1
(Grem1) CRM lies within the Formin (Fmn1) gene, such that the latter was long
considered as responsible for a limb defect (its original name was Limb deform-
ity), whereas it does not have any known function in limb development, contrary
to Gremlin.
These difficulties will be overcome when most regulatory regions have
been defined according to transcription factor occupancy using strategies such
as ChIP-seq. There is still a long way to go: according to experiment matri-
ces recently published by UCSC Genome Bioinformatics, only 13 out of about
60 known histone modifications and 120 out of the estimated 1,700–1,900 tran-
scription factors have been examined to date in the human genome by ChIP-
seq. These, furthermore, have been analyzed in a number of cell lines in culture
(which often bear little similarity to cells within organisms) or in readily acces-
sible adult cells, such as blood cells, but many tissues in the adult, not to men-
tion embryonic stages, have not been investigated—and the mouse genome lags
behind. Tissues, especially embryonic tissues, provide only sparse material,

and methods will have to be miniaturized before they can be extensively ana-
lyzed. Nevertheless, the power of these new strategies is such that we can be
confident that, in the near future, the regulatory syntax of the genome will be
worked out. One major difficulty that may remain is attributing a given CRM
to a specific gene in a defined physiological or developmental context, since,
as we have seen, enhancers may be very remote and there is evidence that the
closest enhancer to a gene is not necessarily active on that gene. Assessing the
correlation of the chromatin state at enhancers and RNA-PolII occupancy at
promoters, for each possible enhancer–promoter pair of elements in a chromo-
somal domain, may help define enhancer–promoter organization (Shen et al.
2012). This may be insufficient, due to the properties of enhancers discussed
above. We see that the regulatory sequences of a gene can hardly be circum-
scribed a priori. At this stage, genetic approaches may prove very helpful, since,
following mutagenesis, a phenotype attests to an alteration that affects one
gene with no a priori hypothesis on the regulatory mechanisms for this gene.
Unfortunately, ENU mutagenesis is much more efficient at mutating coding
sequences than regulatory sites, for reasons that are not entirely clear. It may
be because regulatory regions are often redundant (Lagha et al. 2012), and
there may be multiple TRF binding sites within an enhancer, making it unlikely
that a single mutagenesis experiment will abolish all the binding sites. In con-
trast, exceptions to this rule have proven highly educational. This is the case
for the limb-specific regulatory module of Shh, which is located nearly 1 Mb
(~0.6 cM) upstream of the coding region and has been extensively characterized
via genetic strategies.
These strategies take advantage of several assets of genetic tools. First, they
allow a fine mapping of the genetic alteration. This may be very valuable in the
case of distant regulatory sequences. It should nonetheless be kept in mind that
CRMs are often too remote for molecular walking strategies along the chromo-
some, but too close for genetic segregation and localization. While a huge number
of polymorphisms have been defined in the mouse genome (SNPs), the preci-
sion of mapping still depends on the possibility of getting them to segregate in a
cross—i.e., the number of meioses that can be analyzed (with 1,000 meioses yield-
ing a 0.1 cM precision). In a historical attempt to localize Hx, a limb mutation that
turned out to affect the distant CRM of Shh, analysis of more than 2,000 meio-
ses in a cross involving Mus m. castaneus reduced the candidate region to a little
more than 400,000 bp—a genetic tour-de-force, but still insufficient to identify a
causative point mutation. At a minimum, genetic mapping based on segregation
defines boundaries within which the regulatory sequence can be sought by other
approaches.
To characterize the affected sequence in a mutant, an essential strategy is the
reliance on multiple alleles for the mutation. It is even better to have alleles of
a different nature in addition to point mutations (insertions, translocations),
to allow easier entry points into the mouse genome. Thus, for the Shh CRM, as
for Gremlin 1 (Limb deformity—Grem1ld, another limb mutation), a transgene
insertion provided an entry point to the CRM (Lettice et al. 2002). This illus-
trates the value of mutagenesis strategies that generate chromosomal accidents
(deletions, translocations, transposon insertions—see Chap. 3) to locate regula-
tory modules. Examples include PiggyBac, Sleeping beauty or Tol2 transposons,
and ethyl methane sulfonate (EMS)-induced deletions in ES cells (Munroe and
Schimenti 2009).
It has been shown over the past few years that many CRMs are active on more
than one gene, defining so-called “regulatory landscapes”. Thus, many genes in
the landscape show the same expression profile as the gene of interest and may
be suspected to encode proteins acting in trans as regulatory factors. Examples
such as Shh (CRM within the Lmbr1 gene) and Grem1 (CRM within the Fmn1
gene) are illustrative in this respect. In such cases, it is essential to define whether
regulation occurs in cis or trans, and, up to very recently, only genetic tests could
unambiguously settle the issue. The principle of the test is straightforward, but
requires that two allelic forms of the regulatory region and its target, respectively,
can be discriminated in a genetic cross. When the regulatory sequence is defined
by a mutation, this provides the differential allele for the CRM. The gene acted on
must have two alleles, either coming from different mouse subspecies or one being
an engineered allele. The ultimate demonstration that the characterized altera-
tion in the genome is the cause for the abnormal phenotype will be provided by
recapitulating this phenotype using the altered sequence in a functional test, such
as expression of a reporter in a transgenesis experiment, or phenotypic rescue by
BAC transgenesis, or de novo creation of the suspected mutation by homologous
recombination.
With its very powerful tools (different mutagenesis strategies to generate dif-
ferent types of mutations, screens to identify new dominant and recessive muta-
tions, cis–trans tests, etc.), genetics could play a major role in the identification of
new regulatory modules. However, genetics now has strong competitors over the
whole spectrum: targeted mutations, long-range haplotyping by genome sequenc-
ing strategies, and identification of remote regulatory modules by scanning the
genome via overlapping transgenes. Even before we can directly identify CRMs
using appropriate algorithms, genetic approaches may be outdated by genomic
strategies—which also are considerably less expensive.
5.3.4 Organization of Syntenic Regions

at the Chromosome Level
As we explained in the previous chapters (Chap. 4 in particular), the linear

arrangement of mouse genes along the chromosomes tends to be preserved,
at least to some extent, among the different species of mammals, recalling the
existence of a more or less distantly related common ancestor. This means that
when two genes are found closely linked in the mouse, they have a good chance
of also being linked in the rat and in human genomes, depending on the degree
of linkage. With the ever-increasing resolution of genetic maps and the availabil-
ity of genomic sequences of several different mammalian species, it has become
possible to reconstruct the progressive reshuffling of the chromosomal segments
that occurred across the species in question during evolution. For example, scan-
ning the human, mouse, and rat genomes at high resolution we find that there are
280 orthologous chromosomal segments between human and mouse, 278 between
human and rat and 105 between rat and mouse. Comparisons between dog, cat,
and cow, whose genomes are also completely sequenced, indicate that the number
of chromosome breaks between human and rodents (~280) is consistent with the
number of synteny breaks observed in other species separated by similar evolu-
tionary distances. However, the number of chromosomal rearrangements between
rat and mouse seems to be excessive if the divergence between the two species
really occurred 12–14 Myr ago. Explanations for this discrepancy are lacking.
The existence of these homologies of synteny indicates that, during evolution,
many genomic segments of the different species have been broken and then trans-
located, inverted, or transposed several times. This, however, is difficult to rec-
oncile with the experimental observations presented earlier, indicating that most
alterations in the karyotype structure are in general strongly counterselected by
impeding normal gametogenesis in heterozygotes. Here again, explanations are
awaited to reconcile all these observations, but it is tempting to speculate that this
may be linked to the mechanisms of speciation themselves.
Homologous chromosomal segments display great variations in size across the
different species. Mouse chromosome 11, for example, contains a large homolo-
gous region (almost all) to human chromosome 17q, while some other homolo-
gous chromosomal regions are extremely small-sized, and are sometimes reduced
to a few genes. Human chromosome 21 has homologies with at least three mouse
chromosomes (10, 16, and 17) and this, as we already mentioned, has hampered
the development of mouse models of Down syndrome.
When checked at high resolution, it is sometimes observed that the genes
in one species are not exactly in the same order as in another related species,
although they are within the same syntenic segment. The genes flanking the
OAS cluster on human chromosome 12q are on the same syntenic segment as the
orthologous genes on mouse Chr 5, but are not in the same order, because a short
inversion occurred in one of the lineages (probably in the mouse). Many other
such rearrangements have been observed in other regions of the genome (Fig. 5.6).
Based on observations made in several distantly related eukaryotic species, the
hypothesis has been suggested (Petkov et al. 2007) that the associations or cluster-
ing of genes within short genetic distances might have occurred initially because
the genes in question were cooperating in various cellular and physiological func-
tions (akin to large operons, so to speak). It is then not so surprising that these
associations have remained relatively unchanged during evolution. Some support
for this interesting hypothesis has been provided by the observation of non-allelic
parental associations in recombinant inbred strains. Another stronger line of sup-
port should come from the analysis of the genome sequence of mice from the
Collaborative Cross (see Chaps. 9 and 10).
(a) (b)
MMU5 HSA12q TEL
Mouse Chr 19
- Ung - UNG
- Oasl - CDV1
- p21-ARC
- Rnf10 - ALDH2
- Cox6a1 - PTPN11 HSA11
- Nos1 - OAS1
RNO1
- Oas1 - NOS1
- COX6A1 HSA10
- Ptpn11
- RNF10
- Aldh2 - OASL
- Arc21
- Cdv1 - HPD
- RSN MMU19 MMU19
- 4HD
- Rsn - TJ6
- Tj6
Fig. 5.6 Homologies of synteny. a An example of homology of synteny between mouse Chr 5

and human Chr 12q24 in the region of the Oas/OAS cluster. The genes flanking the OAS cluster
on human Chr 12q are on the same syntenic segment as the orthologous genes on mouse Chr 5,
but not in the same order because a short inversion occurred in the mouse. Many rearrangements
of this kind have been observed in other regions of the genome. b Another example of homol-
ogy of synteny between mouse Chr 19 and human Chrs 10 and 11. The same mouse Chr 19 also
exhibits homology of synteny with a large fragment of rat Chr 1. More than 90 % of mouse and
rat genes are in regions exhibiting homology of synteny with a chromosomal region in humans
(the maps are from MGI database—2013)
5.3.5 Gene Families and Pseudogenes
As we already mentioned, when looking in the mouse genome for a DNA

sequence orthologous to a human or rat gene we generally find them in the
homologous syntenic region, as expected. However, it is not uncommon to find
that the sequence homology between the two species is not always in a 1:1 ratio.
On human chromosome 12q, for example, there is a cluster of three genes encod-
ing 2′,5'-oligoadenylate synthetase (OAS), an enzyme that is induced by interfer-
ons and plays an important role in the inhibition of cellular protein synthesis and
resistance towards viral infections. In this cluster, the human genes are arranged in
the following order: HSA12 cen—… –OAS1–OAS3–OAS2– …– tel.
When looking for the orthologous syntenic region encompassing the OAS
encoding genes in the mouse genome, we find a cluster on chromosome 5 with no
less than ten genes. These genes exhibit a very high degree of sequence similarity
and the linear order: MMU5 Cen–… –Oas2–Oas3–Oas1e–Oas1c–Oas1b–Oas1f–
Oas1h–Oas1g–Oas1a–Oas1d– …– Tel. Thus, the human OAS2 and OAS3 genes
each have, and as expected, a single 1:1 orthologous copy on mouse chromosome
5 while the human OAS1 has no less than eight copies (1:8 orthologs). These Oas1
genes are all transcribed although not always in the same direction, indicating that
they probably result from a series of segmental duplications with subsequent rear-
rangements (inversions). In the rat, the structure and organization of the cluster is
similar to that of the mouse, but with only eight genes; the orthologous copies of
mouse Oas1a and Oas1e are missing (Perelygin et al. 2006). These differences
between the human, rat, and mouse OAS clusters indicate that the genomes of
these three species are in constant evolution. Similar observations have been made
when performing sequence alignments between mice of the same genus Mus but
belonging to different species (Fig. 5.7).
These clusters of genes (the three human genes, ten mouse genes and eight rat
genes), which encode proteins with similar biochemical functions, were presum-
ably formed by recurrent duplications of a single ancestral gene and represent
what geneticists call a gene family. Such gene families are common in mammalian
genomes and include, for example, the genes encoding the globins, the myosins,
the Hox and Sox clusters, etc. Looking at different unrelated vertebrate species,
one observes that the number of repeated copies is highly variable, and the sig-
nificance of these variations in copy number (if any) is not clear. In the case of
the mouse Oas cluster, all ten copies are transcribed but the mouse Oas1b gene
carries a stop codon in its exon 4, resulting in the premature truncation of the
Human OAS gene cluster

telomere centromere
DTX1 RPH3A
30 kb
12q24
OAS2 OAS3 OAS1
Mouse Oas gene cluster
centromere telomere
Dtx1 Rph3a
30 kb
Oas2 Oas3 Oas1e Oas1c Oas1b Oas1f Oas1h Oas1gOas1aOas1d
(Oasl11) (Oasl10) (L1) (L2) (L3)
Fig. 5.7 Gene families. The three genes encoding human 2′, 5' oligoadenylate synthetase (OAS)
are clustered on HSA12, flanked by the same two genes (DTX1 and RPH3A) as in the mouse,
and ordered as indicated in the figure. These three genes are transcribed in the same direction.
The homologous region is on mouse Chr 5 (MMU5) and consists of ten genes with a very high
degree of sequence similarity. The orthologous copies of human OAS2 and OAS3 are well pre-
served, with a 1:1 orthology, while human OAS1 has no less than eight orthologous copies in the
mouse. This cluster of Oas1 genes probably results from a series of segmental duplication with
subsequent rearrangements (inversions). All these genes are transcribed, although not always is
the same direction. Such quantitative differences between the human and mouse OAS clusters
indicate that the genomes of these species are in constant evolution, although with variations in
gene copy numbers (Adapted from Mashimo et al. 2003)
gene product (oligoadenylate synthetase or 2′,5'-OAS), leading itself to its com-

plete inactivation in virtually all mouse laboratory strains. Interestingly, this
mutation does not exist in wild mice and researchers demonstrated that this dif-
ference, which is specific to the Oas1b gene, is responsible for the susceptibility
of practically all laboratory mice to experimental infections with flaviruses. The
function(s) of the proteins encoded by the other genes of the family is (are) not
yet elucidated but, obviously, they do not complement the functional deficiency of
Oas1b in laboratory strains.
The formation of a gene family results from a mechanism that is classical in
evolution. As in the case of the OAS/Oas clusters, a majority of these families
are formed by a succession of tandem duplications of a single ancestral gene and
the different proteins encoded by the genes of the same family (commonly des-
ignated isoforms) generally have similar biochemical functions, but this is not a
rule. Some gene families are easy to identify because the duplicated copies are
closely linked to each other, are arranged in tandem, and have retained similar
sequences. In other instances the situation is more complex because the gene fam-
ily is ancient and has been more or less extensively remodeled during evolution.
This is the case, for example, with the Oas1 gene cluster that we described above
and two other genes with a similar structure (Oas-like1 and Oas-like2—symbols
Oasl1 and Oasl2), located 4 cM away, on the proximal end of the same mouse Chr
5 (Fig. 5.8).
A B C D
A B C D
A B B' C D
A C D
Fig. 5.8 Gene duplications. Some tandem duplications result from unequal crossing-over

between homologous chromatids, as indicated in the illustration. The chromosome with a deleted
gene or segment (in grey) is in general rapidly lost, while the duplicated region (solid black) is
retained. Gene duplications can also result from error (slippage) of the polymerase during DNA
replication, with the enzyme copying the same segment more than once. Gene duplication is an
essential source of evolutionary innovation when the duplicated copies acquire specific functions
(for example, the different isoforms of β-globin, Hox genes etc.)
This is also the case for the genes encoding the globin subunits, which are all
clearly derived from a single ancestral copy that existed some 500 Myr ago, but
are now separated in two different clusters in the mouse genome (α-globin on Chr
11 and β-globin loci on Chr 7). The expansion or contraction of gene families in a
specific lineage can be due to chance, or can be the result of natural selection, and
it is extremely difficult to decide between these two options.
When genes are duplicated in tandem, it is also common to observe that not all
the copies are transcribed in exactly the same way. For example, according to the
strain, laboratory mice have either one or two copies of the gene encoding Renin,
a protein that participates in the regulation of arterial blood pressure (Ren1 and,
sometimes, Ren2-Chr 1). Ren1 encodes the renin mRNAs found in the submaxil-
lary gland while Ren2 encodes the renin mRNAs found in the kidney. This differ-
ence in transcriptional activity can be explained by the promoter regions of these
two genes, where structural differences have been described (Panthier et al. 1984).
Some specific gene families, like those concerned with a reproductive function
(exhibiting, for example, spermatid or oocyte-specific expression), an immunolog-
ical function, or an olfactory function (encoding, for example, the odorant (OR)
or vomeronasal (VR) receptors) originated from relatively recent duplications
(expansions) that occurred in the mouse lineage since the time of its divergence
from the rat, around 12–14 Myr ago. In the initial draft of the C57BL/6 genomic
sequence, for example, scientists were surprised at the identification of some 1,400
OR genes and 332 VR genes. In the human genome the same olfactory or vome-
ronasal receptors are much less numerous. The explanation generally proposed
to explain these considerable differences is that such sequences are preserved
because they are translated into functional proteins that are more or less important
for the host species. Geneticists have coined the expression “genome shaping” to
account for such a situation where the genome structure is influenced by natural
selection triggered itself by environmental factors (Nouvel 1994). Although one
can accept the idea that olfactory receptors are much more important for wild mice
than for human beings, the same argument is less obvious for some other genes
that are members of very large gene families in rodents but are much less repre-
sented in the human genome.
After careful examination and comparison with a consensus (or ancestral)
sequence, it is common to observe that some members of a gene family carry point
mutations (SNPs). These mutations are missense or sometimes nonsense, resulting
in a loss of function for the gene in question. This is the case for the Oas1-like
gene (Oasl1) described above. When this occurs, the mutated gene no longer
encodes a functional protein, even if it is still transcribed. It is then classified as a
pseudogene and its sequence will progressively degenerate, generation after gen-
eration, until it becomes unrecognizable in terms of structure. The pseudogene is
then called a relic, a vestige or a fossil, and the intergenic regions of the genome
have sometimes been described as “cemeteries” for these degenerated genes. The
“death” of a gene is not important for the survival of the species as long as other
copies of the family are present in the genome as potential backups, capable of
taking over the function of the missing copies.
When missense mutations (i.e. leading to an amino acid substitution) occur

in a gene that is a member of a family, this results in the gene encoding a novel
protein, with sometimes new characteristics, a different 3D shape, a different sta-
bility, etc. Evolution will then “decide” whether this novel protein deserves to be
retained or not based on the potential advantages it may confer to the affected indi-
vidual in its current environment (Demuth and Hahn 2009). In this case, one real-
izes that diploid organisms have an advantage since they can put to test, in the
same genome and for a few generations, both the ancient and the new copy (allele)
of a given gene and finally retain the one with the best fit.
An interesting gene family is that of myosins, mostly known for their role in
muscle contraction but also involved in a wide range of motility processes. In fact,
myosins belong to a huge superfamily of genes whose products share the basic
properties of actin binding, ATP hydrolysis (ATPase enzyme activity), and force
transduction. Virtually all eukaryotic cells contain myosin isoforms (alternative
forms). Some isoforms have specialized functions in certain cell types (muscle),
while others are ubiquitous.
A careful analysis of the initial draft of the mouse genome sequence indicated
that, in this species, the rate of nucleotide substitution is approximately twice as
fast as the rate in human, and this explains why, after a few million years, it is
sometimes difficult to establish sequence similarities between some elements of
the human and mouse genomes.
As we discussed, it is clear that the mammalian genome contains a great num-
ber of sequences that look like protein-coding genes but, in fact, are not (or no
longer). The first category of these sequences is the pseudogenes we reported
above, which are duplicated copies of an ancestral (single copy) gene, and have
become non-functional after the accumulation of random mutations (SNPs or
indels). There is, however, another category of pseudogenes that geneticists call
processed pseudogenes. These pseudogenes, unlike the former ones (which are
then called unprocessed pseudogenes), originate from the retrotranscription of
messenger RNAs back into the genomic DNA in more or less random locations.
They have no introns and often exhibit mutations in their sequences (including
frame-shifts and stop codons), indicating that they definitely do not encode pro-
teins. Around 18,000 such pseudogenes have been identified in the mouse genome
assembly (build 38.1), but their identification is often difficult. To discriminate
between a true, bona fide gene (a gene encoding a protein and then submitted to
purifying selection) and a pseudogene (processed or unprocessed), researchers
calculate the so-called Ka/Ks ratio. This ratio compares the number of non-synon-
ymous substitutions (Ka) to the number of synonymous substitutions (Ks) in the
sequence of the two genes. Synonymous mutations, as we will discuss later, do not
modify the amino acid sequence (for example, the GGC codon becomes GGA but
still codes for glycine) and accordingly can occur at the same frequency in genes
and in the pseudogenes, with no consequence. Non-synonymous mutations, on
the contrary, because they generally alter the protein structure, and often its func-
tion, are counterselected and are uncommon in functional genes. Computing the
Ka/Ks ratio is then a reliable assessment of whether a gene is a “true gene” or a
pseudogene. Ka/Ks values approaching 1 are indicative of neutral evolution, sug-

gesting a pseudogene. In addition, most mouse pseudogenes do not have an orthol-
ogous copy in the same syntenic position in the human or rat genomes, whereas
active genes generally do.
As we discussed above, most pseudogenes were considered to be “fossils” or
“relics” of genes that, once transcribed and reintegrated into the genome, became
silent and functionally useless. This view, however, might not be correct or uni-
versally true. In fact, there has been speculation and some evidence has been col-
lected suggesting that pseudogenes, or portions of the latter, may be transcribed
from the opposite strand relative to their functional counterparts, making them
a source of antisense RNA. These RNAs have been proposed to play a role in
the fine regulation of genes of the same family through RNA–RNA interaction
(Balakirev and Ayala 2003). Even more recently, scientists working on the mouse
transcriptome have identified no less than 10,000 full-length cDNAs derived from
expressed pseudogenes—representing approximately 10 % of the known tran-
scriptome—with a good half of them likely participating in various regulatory
mechanisms. As noted by the members of the FANTOM 3 project (see later in this
chapter), we must remain open-minded about the potential function of expressed
pseudogenes. For this reason, pseudogenes have been referred to as “potogenes”
(potential genes) (Balakirev and Ayala 2003; Hayashizaki and Carninci 2006).
5.3.6 Copy Number Variations
In a preceding section (see 5.3.1.1), while discussing the different structural

variations that have been observed at the genome level, we noted that some genes
have been found to be missing (deleted) in some mouse strains and not in others
(for example, Snca on Chr 6), while other genes, in contrast, were duplicated in
some strains and not in others (for example, Ren1 and Ren2 on Chr 1). Variations
of this kind are common in mammals and one can certainly expect many similar
cases to be reported in the future, for example when comparing distantly related
strains or sub-species of the same Mus genus. Many of these duplications are lost
after a few generations, but a few of them may be retained, eventually after a few
changes, either by chance or because they have an adaptive value. We have already
discussed this point.
Copy number variations (CNVs) originate from both coding and non-cod-
ing regions of the genome. The mechanisms leading to these CNVs in a specific
chromosomal region have not yet been completely elucidated, but it makes sense
to consider a priori that CNVs are of three kinds. A substantial proportion prob-
ably results from unequal crossovers, producing both deleted and complemen-
tary duplicated genomic segments. Given that these chromosomal rearrangements
often concern large segments, the duplications have a greater chance to be trans-
mitted to the next generation than the deletions, which are generally unviable and
lost.
Another type of CNV probably results from defects occurring during DNA rep-
lication (for example, defects in replication fork maintenance). This class of CNV
commonly occurs in somatic cell lineages (especially in neoplastic tissues), and,
accordingly, occurs independently of the process of meiotic recombination.
Finally, the observation that some short-length chromosomal duplications have
been found on different chromosomes (cases have been reported in the mouse)
suggests that these duplications are, in fact, transpositions of DNA segments
very similar to those described earlier and classified as transcriptionally active
pseudogenes.
In the mouse, around 100 well-dispersed regions across the 19 autosomes and
the X chromosome have been shown to harbor CNVs. Their greatest preponder-
ance is on chromosomes 7, 12, 14, and X, where some of them appear as large
blocks.
The sequence homology between the different copies is >94 % on the average,
and their size ranges from 62 bp to 8.6 Mb (with an average length of 250 kb). In
total, if we include both the deletions and the duplications, this represents close to
10 % of all polymorphisms (excluding microsatellites), with short deletions being
more frequent than insertions (Cutler and Kassner 2008).
CNVs involving large or very large chromosomal segments, although rare, have
been observed by cytogeneticists using the classical techniques of fluorescence in
situ hybridization. Nowadays, more sensitive techniques, like high-resolution com-
parative genomic hybridization (HR-CGH) or representational oligonucleotide
microarray analysis (ROMA), are adapted to this sort of analysis. Using appropri-
ate DNA arrays, these techniques allow for the detection of structural variations at
a resolution of 200 bp (Egan et al. 2007) (Fig. 5.9).
In the near future, taking advantage of the recent advances in DNA sequenc-
ing technology, it should be possible to identify and quantify many more CNVs at
high resolution in both human and mouse, allowing comparisons to be made at the
individual level.
The occurrence of CNVs at the genome level translates to variations in gene
dosage within the duplicated or deleted regions (0/1–1/1–2/1, etc.), and it makes
sense to think that this may be causative or associated with some pathologies.
A trisomic mouse, for example, can be regarded as carrying a single large CNV,
since the only difference relative to a normal karyotype is an extra chromosome.
This difference can nevertheless result in a severe and often lethal syndrome. A
good example where a CNV has been found to be causative of a pathological
syndrome is Charcot–Marie–Tooth, type A (CMT1A) disease in humans. This
neuropathy was found to segregate with a ~1.4 Mb duplication on human chromo-
some 17p12 among the members of the same family, suggesting a possible causal
relationship. Shortly after this observation, the gene coding for peripheral myelin
protein 22 (PMP22), a component of myelin, was identified within the duplicated
region and mutations in this gene were found to be also responsible for a clini-
cal form of the disease very similar to the form associated with the duplication
(Valentijn et al. 1992a, b). Finally, an almost perfect mouse model of CMT1A was
created by pronuclear injection of a YAC containing a normal, intact copy of the
Strain A Strain B
Fig. 5.9 High-resolution-comparative genomic hybridization (HR-CGH). This technique is use-

ful for comparing two genomes, or two chromosomal regions, for possible quantitative differences
in terms of copy number. The technique consists of two steps. First, a reference DNA sample is
labeled with a fluorophore (for example, Cyanine 3, green) while the DNA from a test sample is
labeled with a different fluorophore (for example, Cyanine 5, red). In the second step, equal quanti-
ties of the two-labeled DNA samples are mixed and co-hybridized to a DNA microarray of sev-
eral thousand evenly spaced cloned DNA fragments previously spotted on the array. Finally, after
hybridization, digital imaging systems are used to capture and quantify the relative fluorescence
intensities of each of the hybridized fluorophores. Obviously, the ratio of the fluorescence intensi-
ties is proportional to the ratio of the copy numbers of DNA sequences in the test and reference
DNA samples. If the intensities of the fluorophores are equal for a given probe, the spot appears
yellow and the region of the genome is interpreted as having an equal quantity of DNA in the
test and reference samples (i.e., no copy number variation (CNV)). If there is an altered Cyanine
3:Cyanine 5 ratio, this indicates a loss or a gain of the test DNA sample at that specific genomic
region. Discovering which regions of the genome have undergone CNVs is achieved by another
test, for example by sequencing followed by fine localization of the sequence. Finely estimating the
CNVs can ultimately help to identify genes that are over- or under-expressed, or even deleted. The
technique can be adapted to the localization of CNVs directly on the chromosomes
human PMP22 gene and a large proportion of its flanking region. The conclusions
of all these observations and experiments are that both point mutations and dupli-
cation of the PMP22 gene can produce the same phenotype of severe demyelina-
tion in the peripheral nervous system.
If the mere duplication of an intact, normal myelin-encoding gene (PMP22-
Pmp22) can induce a pathology in humans and mice, as demonstrated with YAC
transgenics, one can then seriously consider that other CNVs might be at the
origin of (or associated with) some clinical diseases or, at least, influence their
phenotypic expression (penetrance or expressivity, for example) by altering the
transcript level of some essential genes. The presence of some specific CNVs in
the human genome has been found to be associated with susceptibility to autism
(Sebat et al. 2007; Cook and Scherer 2008). A reduction in CNVs involving the
gene Defensin beta 1 (DEFB) has been reported to increase the risk of d eveloping
Crohn disease (Roberts et al. 2012). Other human pathologies are equally sus-
pected to be associated with (or the consequence of) CNVs (e.g., autoimmunity,
susceptibility or resistance to infectious disease).
In the mouse, genes involved in the control of the immune response or environ-
mental sensory perception have also been found to exist in variable copy numbers
in the genomes of the various inbred strains (Watkins-Chow and Pavan 2008). In
these conditions, it should not be so surprising to observe in the future that these
mice exhibit different phenotypes related to these CNVs.
Nowadays, many geneticists consider that the transmission of some complex
traits might be better explained by the transmission of CNVs than by hypothetical
Mendelian characteristics (Canales and Walz 2011). Observations relative to some
infectious diseases in human populations have already provided preliminary clues
to this important question. For example, Gonzalez and colleagues (Gonzalez et al.
2005) reported a strong positive correlation between a high number of copies of
the gene encoding the chemokine CCL3L1 and HIV susceptibility.
5.3.7 Single Nucleotide Polymorphisms
When orthologous sequences from different mice (laboratory mice or wild mice)
are aligned, single nucleotide differences are frequently observed in the DNA
sequence. These differences are base-pair substitutions in most instances, less fre-
quently insertions or deletions of one nucleotide. These sequence differences have
been collectively designated single nucleotide polymorphisms (SNPs, pronounced
“snips”) and are the most common type of genetic variation at the DNA level.
They are found in both coding and non-coding regions and almost all these SNPs
are bi-allelic, i.e., presenting one of two possible nucleotides in an individual (e.g.,
homozygous G/G or T/T or sometimes heterozygous G/T).
SNPs are extremely abundant among the different mouse inbred strains, and
even more so across the different strains recently derived from wild populations.
These SNPs are easy to score and permit the performance of high-density/high-
resolution mapping. They have undoubtedly been an important outcome of the
mouse genome sequencing project, because they represent the ultimate genetic
markers. We described their use and advantages in Chap. 4 (Fig. 5.10).
5.3.8 Tandem Repeated Sequences
Like other mammalian genomes, the mouse genome contains a large num-
ber of repeated (both coding and noncoding) sequences. They are classified as
moderately or highly repeated sequences, and among the latter one must also
FVB/N
Mouse
C57BL
DBA/2
129X1
BALB
CAST
129S1
NZW
NOD
C3H
Strains
A/J
C/T SNP C C C C C C T C C T C
“C” allele “T” allele
Fig. 5.10 Single nucleotide polymorphisms (SNPs). SNPs are single base-pair differences in the
DNA sequence, and are the most common type of genetic variation. As described in Chap. 4,
they are very useful for genetic mapping, they are found in both coding and non-coding regions,
and almost all these SNPs are bi-allelic, i.e., presenting one of two possible nucleotides (e.g.,
G/G, T/T, or G/T genotypes). In the figure, the upper panel represents a C/T SNP that is poly-
morphic between DBA/2 and CAST (homozygous for the T allele) and other strains (homozy-
gous for the C allele). The lower panel presents DNA sequencing electropherograms showing the
SNP (arrow)
distinguish those that are organized as tandem repeats and those that are inter-
spersed. Tandem repeats are those where the nucleotides motifs are repeated
adjacent to each other in a head-to-tail manner. Depending on the number of
nucleotides and on the size of the motif, these tandem repeats are known as satel-
lite DNA (between 120 and 250 nucleotides), minisatellites (between 10 and 60
nucleotides), and microsatellites (between 2 and 6 nucleotides). In these types
of repeats, the polymorphism is a direct consequence of the number of repeats.
The interspersed or dispersed repeats are a totally different category and will be
described below.
5.3.8.1 Satellite DNA
The name “satellite DNA” was coined in reference to a difference in the buoy-
ant density of this category of DNA when compared to the density of bulk DNA.
Satellite DNA constitutes about 5 % of total mouse DNA and is divided into two
major categories: major satellite, which is composed of 234-bp repeats (6 Mb long
altogether—occurring at a few loci on the genome), and minor satellite, which is
composed of 123-bp repeats (from 500 kb to 1.2 Mb in size and located essen-
tially in the centromeric and telomeric regions of chromosomes). Satellite DNA is
the main component of heterochromatin, is not transcribed, and has proved to be
rather difficult to sequence.
5.3.8.2 Minisatellites
Minisatellite loci are also known as variable number of tandem repeats or

VNTRs. They consist of a short series of 10–60 bp repeated in tandem over and
over to reach around 5–10 kb in size. They are extremely abundant and are dis-
tributed at more than 1,000 locations in mammalian genomes. The occasional
slippage occurring during replication is probably at the origin of the minisatel-
lite copy number variations, thereby making each individual unique (Kuznetsova
et al. 2005). These highly polymorphic loci were used as genetic markers in the
late 1980s, particularly in human studies, and became the basis for the famous
DNA fingerprinting method that revolutionized forensic medicine. These “fin-
gerprints” are the individual-specific band patterns resulting from the hybridi-
zation (by use of Southern blotting) of restriction-endonuclease-digested DNA
with probes directed against extremely polymorphic minisatellite (VNTR) loci.
Although it was used in a few mouse linkage studies and also for the genetic
monitoring of inbred strains (isogenic individuals within an inbred strain share
the same band pattern), the use of DNA fingerprinting in the mouse was aban-
doned after the advent of microsatellites as universal molecular markers (Julier
et al. 1990; Silver 1995).
5.3.8.3 Microsatellites
Microsatellites (also known as short tandem repeats (STRs) or simple sequence

length polymorphisms (SSLPs)) are tandem repeats of 1–5-bp elements that are
probably the consequence of polymerase slippages. They are very abundant
(approximately 105 copies per genome), extremely polymorphic, and widely dis-
tributed throughout the genome. Since the early 1990s, microsatellites have been
the genetic marker of choice in mouse genetics because their analysis is extremely
simple, inexpensive, and relatively reliable. For the same reason as for the SNPs
mentioned above, we will review their interest as genetic markers in several chap-
ters of this book and in various contexts (Fig. 5.11).
5.3.8.4 Trinucleotide Repeat Expansions
Some severe human genetic disorders have been found to be the consequence
of the continuous and abnormal expansion of DNA-trinucleotide repeats in cer-
tain genes. The fragile X syndrome is one of these disorders and the first to be
explained at the molecular level. Human geneticists found 230–4,000 CGG tan-
dem repeats in a specific X-linked gene in affected patients compared with the
5–54 repeats in unaffected individuals. Similarly, Huntington disease (HD), which
affects muscle coordination often associated with psychiatric problems, is caused
Fig. 5.11 Microsatellites. Microsatellites (SSLPs) are composed of short DNA sequences,

measuring 1–6 bp, which are repeated in tandem a number of times. They are common in all
mammalian genomes, where they exhibit variations in terms of the number of repeats (size
polymorphism) and for this reason they are sometimes designated as simple sequence length
polymorphisms (SSLPs). Microsatellites can be amplified by PCR with primers designed
from the flanking regions. The number of repeats, which translates into size variations of
the amplification product, can then be used as a reliable genetic marker. Microsatellites have
been extensively used in the mouse for the establishment of high-density/high-resolution
genetic maps and are still used for the acute localization of quantitative traits. As indicated
in the figure, microsatellites are co-dominant markers, allowing for the identification of
heterozygous genotypes. In the figure, we can observe the segregation of the alleles from
a microsatellite locus on a pedigree. The male (square) and the female (circle) from this
breeding pair are both heterozygous (black and white). In the 22 offspring we can clearly
see the segregation of the two alleles, where some mice are homozygous (solid color on the
pedigree) for one allele or the other (one band on the gel), and the rest are heterozygous
(two bands). Note that the percentages are in agreement with Mendelian ratios for this
co-dominant SSLP marker (~54 % heterozygous; ~23 % homozygous larger allele; ~23 %
homozygous smaller allele)
by a CAG repeat expansion in the protein-coding regions of another specific gene

called Huntingtin (HTT—in human Chr 4p16.3). In some other instances, such
repeats are also observed and associated with a severe pathology but they are
located outside of the protein-coding regions of the genes. To date, similar DNA-
trinucleotide repeat expansions have not been reported in the mouse, but trans-
genic mouse models have been created by pronuclear microinjection of DNAs
cloned from affected human patients (Ehrnhoefer et al. 2009).
5.3.9 Interspersed Repeated Sequences: Transposable Elements
Transposable elements (TE), as the name indicates, are small sized DNA
sequences that move within the genome and insert into new chromosomal loca-
tions sometimes leaving behind a copy of their sequence at their original site
(Wessler 2006). These TEs exist in virtually all genomes and have been described
in bacteria, Drosophila, mammals and many other organisms. TEs were identified
and characterized for the first time in plants, more precisely in maize, through the
somatic mutations they induced.4 In the mouse, and more generally in mammals,
these elements are repeated over and over, by thousands of copies, but they are
dispersed in the genome and for this reason they are commonly designated inter-
spersed repeats in opposition to the tandem repeats discussed above. Transposable
elements are generally classified into two categories: (i) the retrotransposons,
which transpose via an RNA intermediate in a “copy and paste” fashion, and (ii)
the transposons, which use a “cut and paste” mechanism to move within the
genome, with no RNA intermediate.
5.3.9.1 The Retrotransposons
Retrotransposons (or class I transposons) are of two kinds based on their size and
structure: the LINEs (Long Interspersed Nuclear Elements) and the SINEs (Short
Interspersed Nuclear Elements). In addition to these two kinds of transposons,
endogenous retroviruses (ERVs) are often considered as equivalent to retrotrans-
posons, as we will explain. Altogether these TEs represent the most abundant com-
ponent of the mammalian genome, estimated at a proportion of greater than 40 %
of genomic DNA.
Long Interspersed Nuclear Elements
The LINE family of retrotransposons, and more precisely the L1 subfamily, is

the most important category of transposable elements in placental mammals,
representing roughly 17–20 % of mouse genomic DNA. The normal, intact L1
sequence measures ~7.5 kb and consists of a promoter at its 5' end, followed by
two non-overlapping open reading frames, ORF1 and ORF2, that encode respec-
tively an RNA-binding protein and a 40-kDa protein with reverse transcriptase and
endonuclease activity, and finally an AT-rich region of variable length at its 3' end.
This basic structure is relatively uniform, but variations resulting from mutations
or deletions, accumulated with time, are common. Thus, only a minority of the
LINE elements (a few thousand) appears intact in the mouse genome. The mRNA
transcribed from these LINEs serves as templates for the reverse transcriptase
II encoded in ORF2, and this explains why this type of transposon is also des-
ignated autonomous transposons. The new cDNA (a new LINE element) is ret-
rotransposed into a different site, at a new position in the genome, with the help
of the endonuclease that nicks the chromosomal DNA and creates the conditions
favorable for integration: in other words a true “copy and paste” mechanism. This
4 Barbara McClintock was awarded the Nobel Prize in 1983 for the discovery of “jumping
genes”.
process of retrotranscription is similar to the one leading to the creation of pro-

cessed pseudogenes, as discussed earlier. Sometimes it fails, and this also explains
why so many LINEs are incomplete and truncated at their 5' end.
As observed after the sequencing of several mammalian genomes and com-
parisons between related species, L1 transposons are active as contributors to the
so-called genome shaping and have been a source of evolutionary novelty by pro-
viding sequence motifs that can be recruited by the host, either for the regulation
of its own genes or among its coding sequences. In contrast to this rather positive
aspect, L1 transposition can also be deleterious for the host, for example when a
transposed copy accidentally inserts within a gene or when it mediates a chromo-
somal rearrangement through ectopic (non-allelic) recombination (Sookdeo et al.
2013). The spastic mutation of the mouse (Glrbspa-Chr 3), which is a model of
human hereditary hyperekplexia (OMIM 149400), is caused by the intronic inser-
tion of a 7.1-kb L1 element resulting in the aberrant splicing of the beta subunit
of the glycine receptor mRNA (Mülhardt et al. 1994). L1 transposition can also
be mutagenic in somatic tissues and was actually discovered through this type of
activity in maize. This finding has potential consequences for the whole organism
which can translate into an increase in cancer occurrences (Belancio et al. 2010).
However, most L1 sequences are silenced by methylation and finally become
inactive.
This mechanism of LINE retrotransposition, as described, would result in a
progressive increase in the size of mammalian genomes unless a compensatory
mechanism operates at some point. Based on recent observations, geneticists
assume that the mechanism in question consists of repeated deletions (sometimes
massive) of some of these constantly burgeoning sequences. Whatever the exact
nature of the regulatory mechanism, the size difference observed between the
human and mouse genome is generally attributed to variations in the number of L1
copies.
Short Interspersed Nuclear Elements
SINEs are a type of non-autonomous retrotransposon whose sequence does not

encode any protein. SINEs have a sequence of around 100–500 bp, which is
closely related to the sequence of some tRNAs or to short RNAs. The most com-
mon category of SINEs in the human genome is the Alu1 sequence, whose equiva-
lents in the mouse genome are the B1 and B2 sequences. SINEs are transcribed
by RNA polymerase III but their retrotranscription, necessary for their mobility
inside the genome, is not completely elucidated and probably depends (at least in
part) upon the LINE machinery—hence their occasional designation as non-auton-
omous retrotransposons.
There are around 1–1.5 × 106 copies of these SINEs in a mouse genome, repre-
senting between 11 and 17 % of the total genomic DNA. Depending on their
sequence, SINEs are classified as lineage-specific (added to the mouse genome
after the divergence from a common ancestor with other rodents) or ancestral
(before the divergence).5 Thus, the sequences of these two categories of SINEs
have great value for research in evolution and systematics.
Using a software program for multiple sequence alignment guided by phylo-
genetic trees, researchers have found a DNA sequence measuring 710 bp in the
close vicinity of the bovine β-globin locus, sandwiched between two SINEs, and
obviously resulting from a transposition (Zelnick et al. 1987). This finding may
be considered circumstantial but it nevertheless indicates that, if such a transposi-
tion of a DNA segment (by “hitch-hiking”, so to speak) can occur in the bovine
genome it may also occur in other species, and this is important in the context of
the constant remodeling of the genome structure.
The existence of a very large number of retrotransposons with nearly identical
sequences, scattered throughout the mouse genome, has some potentially interest-
ing technical applications in the sense that universal (non-specific) primers for PCR
amplification can be designed based on the sequence of these retrotransposons and
used either with another specific primer (for example, for cloning the sequences flank-
ing a transgenic insertion) or with the same primer with the inverted sequence for the
amplification of the host genomic DNA situated between two LINEs or SINEs.
The Endogenous Retroviruses
The endogenous retroviruses (ERVs) are a third kind of element that can affect the
structure and function of the mouse genome. Although uncommon, infections of
mouse germ cells by retroviruses can occur, resulting in the integration of more
or less complete retroviral copies into the mouse genome. These retroviral cop-
ies are easily recognizable at the molecular level because they are flanked by two
classical long terminal repeats (or LTRs) and contain the three classical genes
gag (encoding structural elements of the virus), pol (encoding the reverse tran-
scriptase), and env (encoding the coat protein of the virus). Many ERVs are incom-
plete and no longer move in the mouse genome, and in some cases one LTR is
the only sequence that remains of an ancestral retroviral copy that has been com-
pletely excised or deleted.
Just like the LINEs and SINEs, ERVs occasionally have influence on the
genome’s structure and function. They can be mutagenic, like LINEs, when they
integrate into the host DNA into or around a coding sequence. They can also trigger
various forms of structural rearrangements. A classical example of the role of ERVs
as mutagens is the hairless mutation of the mouse (Hrhr) (Stoye et al. 1988; Cachon-
Gonzalez et al. 1994). This recessive mutation is the result of the retroviral insertion
of murine leukemia proviral sequences into intron 6 of a gene encoding a specific
protein at the Hr locus of chromosome 14, which results in aberrant splicing of the
gene. Many other mutations of this type have also been reported in the mouse. The
viable yellow (Avy) allele, which originated through the retrotransposition of an
5 The ancestral SINEs are sometimes designated MIR3 (for mammalian-wide interspersed repeat
elements).
intracisternal A-particle6 (IAP) upstream to the canonical wild-type transcription

start of the agouti gene (A), is another example.
Some elements of these ERVs can also have functional consequences. This is
the case, for example, when long terminal repeats (LTRs) act as alternative pro-
moters or enhancers leading to the transcription of tissue-specific RNAs. In
humans, diseases have been reported as being caused by TE-generated alleles.
These diseases include, for example, hemophilia A and B, severe combined immu-
nodeficiency, porphyria, predisposition to cancer, and some cases of Duchenne
muscular dystrophy.
Recombination between homologous retroviral sequences has also contributed
to “gene shuffling” and to gene duplications and deletions that largely contribute
to genome plasticity.
Several years ago, the retrotransposons we just described were considered
as examples of the so-called “selfish” or “junk” DNA because, apparently, their
only function was to make more copies of themselves with no apparent benefit
for the host. Nowadays, the perspective has dramatically changed and these DNA
elements are regarded as tools contributing to genome plasticity and “novelty”.
L1 sequences frequently insert into the introns of functional genes, where they
can interfere with the transcription process without permanently harming the gene
product. When the inserted L1 copy is long or very long, the transcription rate is
reduced and this might represent another subtle (and reversible) method of gene
regulation. When inserted into an intron, SINEs or LINEs can also introduce new
splicing sites, allowing the de novo creation of new exons and accordingly of new
protein domains. It is then up to the environment to determine, at no risk, whether
the new protein presents some selective advantage, whether the structural altera-
tion is selectively neutral or, on the contrary, whether it is detrimental and should
be eliminated by returning to the original copy of the gene—which is still in the
genome as a back-up. In other words, thanks to the TEs, evolution can perform
experiments at virtually no cost.
5.3.9.2 The Transposons
Transposons exist in many species including bacteria, plants, insects (for exam-
ple the P elements of Drosophila melanogaster), and mammals. They are rela-
tively short elements, measuring a few kilobases when intact, and they encode an
essential enzyme: a transposase (also called transposonase). The gene encoding
this transposase is flanked by two inverted or palindromic terminal repeats that are
essential for transposition in the genome. These terminal repeats pair with each
other as the transposon folds and forms a loop. This DNA loop is then excised and
released, ready to transpose into another location in the genome, hence the “cut
and paste” mechanism of transposition.
6
IAPs are a class of defective endogenous retroviral sequences measuring ~7 kb. These IAPs are
mostly abundant in the endoplasmic reticulum.
The excision of a transposon from its original location in the host genome often
generates a small gap in the genomic DNA, while its insertion in a new location
disorganizes the neighboring genetic sequences. For these reasons the transposons
are responsible for the occurrence of new mutations in the species where they are
active.
In the mouse genome the vast majority of transposons no longer encode any
functional transposase, and accordingly, they have lost the capacity to transpose:
they are “dormant” or even “dead”. Interestingly, a fish transposon, which had
remained inactive for over 15 million years, could be artificially “resurrected” into
an active one by the transgenic addition of two essential functional components
into the same host genome: (i) the transposon DNA containing the two inverted
terminal repeats, and (ii) the transposase enzyme essential for activation. This
engineered (and resurrected) transposon, named Sleeping Beauty (Izsvák and Ivics
2005), has been shown to transpose efficiently enough in the mouse to be pro-
posed as a tool for the in vivo production of mutations (Carlson and Largaespada
2005). This method of mutagenesis has the advantage that new mutations are cre-
ated simply by breeding mice, and, most importantly, that the transposon DNA
tags the integration site. However, the disadvantage is that the mutation rate
is rather low, especially when compared to other mutagenesis methods. More
recently, SleepingBeauty has also been reported as an interesting tool for cancer
gene discovery and gene therapy (Copeland and Jenkins 2010; Howell 2012),
helping for example to introduce transgenes into host genomes. Other resurrected
transposons (Piggy Bac and Mariner) have also been used for the production of
mutations (by gene trapping) and for transgenesis.
The transposable elements are definitely important elements of the genome,
since they participate actively in its evolution. Together they are often referred to
as elements of the “mobilome,” and it is likely that their role and functions are still
underestimated.
5.4 The Transcriptome: Coding and Non-coding RNAs
In the same issue of the journal Nature announcing the initial draft of the mouse
genome sequence (Nature 420–5 December 2002), another very important report
was published, summarizing the results of the functional and manual annotation of
a large collection (60,770) of full-length mouse cDNA7 collected by the
“FANTOM consortium” (Functional Annotation of the Mouse) of the RIKEN
Genomic Science Center in (Okazaki et al. 2002). This publication, perhaps
because it was released at the same time as the impressive and outstanding
7 Full-lengthcDNA libraries are established from all RNA transcripts (protein-coding and non-
protein-coding). Manual annotation of such libraries is a guarantee of their quality.
5.4 The Transcriptome: Coding and Non-coding RNAs 167
presentation of the mouse genome sequence, did not receive the attention we think
it deserved from the community, at least when published. Ten years later, and
based on the information gathered in the meantime from the analysis of the mouse
and human genomes and transcriptomes, we think that this report should be con-
sidered another breakthrough in our understanding of the ways in which the mam-
malian genome actually works. Not only did it confirm some important
observations that were made independently a few years earlier, for example about
the unjustified overestimation of the number of protein-coding genes in the mouse
genome (which was sometimes estimated to be as high as 120,000) and the con-
comitant underestimation (or mis-appreciation) of some other transcription prod-
ucts (Lander et al. 2001; Kapranov et al. 2002), but it also raised a number of new
ideas that have been confirmed since and widely amplified in successive reports, in
particular those of the same FANTOM consortium as well as in other reviews
devoted to the analysis of the mouse transcriptome (Carninci et al. 2005;
Katayama et al. 2005; Mattick and Makunin 2006; Gustincich et al. 2006; Saxena
and Carninci 2011; ENCODE Project Consortium 2012; Kapranov and St Laurent
2012). The ideas that were developed in these initial reports have radically
changed our views of the transcriptome, in particular the belief which was solidly
anchored in most scientists’ mind that proteins were the most important (if not the
only) bioactive molecules encoded in the genome.
The main conclusions of the reports in question are the following: (i) the pro-
tein-coding RNAs (the mRNAs) and the other RNAs that cooperate with mRNAs
in protein synthesis and processing (rRNAs, tRNAs, snoRNAs, and snRNAs)
represent only a minor (around ~2–3 %) component of the transcriptome; (ii) the
mouse genome is pervasively and extensively transcribed and encodes several thou-
sand non-protein-coding RNAs (ncRNAs), and (iii) sequencing all these RNA mol-
ecules and making in silico alignments with the DNA genomic sequence indicates
that up to 90 % of the euchromatic genome of the mouse is transcribed, sometimes
from both DNA strands, and in both directions (many sense–antisense pairs).
Nowadays, the mammalian genome can no longer be regarded as a mere reposi-
tory of the basic information necessary for the synthesis of thousands of proteins,
but rather as a sophisticated factory releasing a great variety of coding and non-
coding RNAs (ncRNAs) of various sizes and functions. In spite of enormous
progress in the sequencing technology of nucleic acids, the inventory of these mol-
ecules is far from being completed and their annotation may still require several
years. It has been established, for example, that many primary RNA transcripts
are processed into smaller sized molecules, while others are alternatively spliced,
thus tremendously increasing the complexity and diversity of the transcriptome.
For this reason, scientists sometimes refer to this new category of non-coding
RNAs as “the dark matter of the transcriptome”. We will summarize the situa-
tion as it stands presently based on recent reviews on the subject, but it is clear
that this chapter, more than any others in this book, will require regular updat-
ing. Undertaking the exhaustive inventory of the ncRNAs encoded in the mouse
genome and performing their annotation is nothing less than embarking on the
exploration of “a new continent in the RNA world”.
5.4.1 ncRNAs Involved in Protein Synthesis
In addition to the messenger RNAs (mRNAs), which are protein-coding and are
considered as the “noble” RNAs since they represent the message transcribed from
the DNA, four types of ncRNAs have been described as essential components in
the successive steps of protein synthesis and processing: transfer RNAs (tRNAs),
ribosomal RNAs (rRNAs), short non-coding RNAs (snRNAs, sometimes referred
to as U-RNAs) and small nucleolar RNAs (snoRNAs).
5.4.1.1 Transfer RNAs
Transfer RNAs are a relatively homogeneous family of “adaptor” molecules

whose function is to mediate recognition of a specific codon in the processed
mRNA and to provide the corresponding amino acid during the process of protein
synthesis (elongation step). These RNAs are part of a larger transcript, the pre-
tRNAs, in the nucleus, which are subsequently split into smaller molecules (aver-
age 80 nucleotides), with a typical 3D cloverleaf structure that is well adapted to
the function since, as we said, the tRNA binds to the mRNA codon (specific), to
the new incoming amino acid (specific), and to the last amino acid incorporated
into the growing polypeptidic chain. There are around 500 tRNA-encoding genes
in the mouse genome and about the same number of pseudogenes. The tRNA-
encoding genes are dispersed over the whole genetic map, including both the X
and the Y chromosomes. Their computerized prediction is difficult due to their
short size and, mostly, to the existence in the mouse genome of a very high num-
ber of short interspersed sequences (SINEs—see above) that originally derived
from tRNA genes but are now inactive. This is typically a source of confusion and,
for this reason, the experimental verification of the status of a potentially novel
tRNA gene must be part of the annotation process (Coughlin et al. 2009).
5.4.1.2 Ribosomal RNAs
In contrast with the tRNAs, ribosomal RNAs are a relatively heterogeneous family
of molecules with a size between 150 and ~5,000 nucleotides. The family com-
prises four types of RNAs (28S, 5.8S, 5S, and 18S). The 28S RNA (5,070 nt) and
the 5.8S RNA (156 nt) bind to each other and are associated with the 5S RNA
(121 nt) and with at least 45 proteins, to make the ribosomal large unit (60S). The
18S rRNA (comprising 1,869 nt) is associated with around 33 proteins to make the
ribosomal small unit (40S). The two ribosomal subunits, the small and the large,
are tightly associated to make the cytoplasmic ribosomes. The biosynthesis of
mature ribosomes is complex and involves numerous processing events with the
participation of other ncRNAs. When mature, the ribosomes serve as workbenches
for protein synthesis. The mRNA is held sandwiched between the two subunits of
the rRNAs while being “scanned” and then transcoded into proteins. rRNAs are
rapidly degraded in the cytoplasm once they have been used for protein synthe-
sis. The genes encoding ribosomal RNAs are very numerous and spread over the
whole genome (Henderson et al. 1974). They are organized in repeated units that,
in the mouse, are 44 kb long. Each repeat contains three of the genes encoding
rRNA, namely 18S, 5.8S, and 28S, and constitutes a transcription unit produc-
ing polycistronic RNA that is cleaved apart afterwards. These units are tandemly
repeated and constitute the so-called nucleolar organizers (or NORs). These are
distributed over several chromosomes (Chrs 4, 12, 15, 16, 18 and 19) in the case of
Mus m. domesticus, but on all 40 chromosomes except the Y in Mus caroli (Rowe
et al. 1996; Cazaux et al. 2011). At the end of mitosis (telophase) when rDNA
transcription by RNA Polymerase I resumes, the NORs gather in the nucleolus (a
nuclear organelle where rRNAs are produced and assembled with ribosomal pro-
teins to form functional ribosomes). Genes that encode rRNA are expressed in vir-
tually all types of cells and in all species, including prokaryotes. For this reason,
many rRNAs have been sequenced and their sequences are now used as tools for
systematics (ribotyping).
5.4.1.3 Small Nuclear RNAs
Small nuclear RNA molecules are found in the nucleus of eukaryotic cells. As is
the case for many other small-sized RNAs, they are transcribed as larger mole-
cules that are cleaved afterwards. They have an average length of approximately
150 nucleotides and are generally classified into five categories: U1, U2, U4, U5,
and U6. Each of these snRNAs is associated with a large set of specific proteins
(over 150), and the complexes they form with these proteins are referred to as
small nuclear ribonucleoproteins (snRNPs or “snurps”). The snurps are essential
in the splicing process. The splicing of mRNAs is a very complex and extremely
precise process and this is probably why the spliceosome requires so many com-
ponents to make its functioning totally error-proof. Each of the five categories of
snRNAs has specific binding sequences and a specific function on the pre-mRNA
substrate.
5.4.1.4 Small Nucleolar RNAs
The small nucleolar RNAs are small molecules measuring 60–300 nt. They are
involved in the processing of rRNAs and are essential for ribosome maturation.
They can also regulate the splicing of some mRNAs by modifying small nuclear
RNAs (snRNAs) that are the major RNA component of the spliceosome, as we
mentioned. snoRNAs probably have many other functions that have not yet been
described, and the inventory of this family of molecules is difficult because their
computerized prediction and classification is unreliable, yielding many orphan
snoRNAs. snoRNAs encoding genes have been identified at several loci in the
mouse genome (2, 7, 8, 9, 12, 17, and X). The range of functions of these RNAs is
likely to expand with the discovery of new molecules (Gardner et al. 2010).
Some genetic diseases affecting humans (for example spinal muscular atro-
phy and congenital dyskeratosis) have been correlated to abnormal functioning of
the snurps. Prader–Willi syndrome (and the reciprocal Angelman syndrome—see
Chap. 6 for details) is caused by the abnormal imprinting of a cluster of snoRNAs
encoding genes located in the q11-13 region of human chromosome 15 that are
involved in the synthesis of the serotonin-2C receptor mRNA. snRNAs also play
an important role in maintaining the size of the telomeres (see Chap. 3).
5.4.2 The ncRNAs Functioning as Post-transcriptional

Regulators
5.4.2.1 MicroRNAs
MicroRNAs (miRNAs) are small, single-stranded RNAs, measuring 21–24 nt

(average 22 nt), whose function is to negatively regulate specific genes by mRNA
degradation or translational repression. Around 60 % of these miRNAs are
encoded in the intergenic regions and in antisense orientation to certain genes, and
40 % are encoded in the intronic regions of genes encoding proteins. These RNAs
(along with the small interfering RNAs, described later) are the most well-known
family of non-protein-coding RNAs.
The DNA encoding miRNAs is transcribed into precursors called pri-miRNAs.
Each of these pri-miRNAs folds to form a double-stranded structure by base-pair-
ing with itself. This structure looks like a hairpin with a few loops of stranded
RNA. The pri-miRNA is then cleaved into a precursor known as a pre-miRNA,
which is transported into the cytoplasm. Finally, the pre-miRNA is incorporated
into a molecular complex of proteins of the argonaute family called the miRNA-
induced silencing complex or miRISC. The processing of mature miRNAs requires
the participation of an endoribonuclease known as Dicer that cleaves the pre-
miRNA into the mature miRNA. miRISC modulates the activity of the targeted
mRNA by identifying a 2–7-bp complementary sequence, known as the “seed
region”, which is generally located at the 3'-UTR. Both the processing and the
loading of miRNAs into the RISC complex and the function of this machinery are
precisely regulated (Ebert and Sharp 2012).
The fact that these miRNAs exist in several species including invertebrates
and plants, and the way they are transcribed and processed from highly preserved
sequences, with highly sophisticated mechanisms, indicates that they probably
represent an ancestral mechanism of gene regulation (Lewis and Steel 2010).
Because they also have a wide range of spatial and temporal expression patterns,
they probably play important roles at different steps of embryonic development
and in some pathological conditions. Indeed, it is expected that about 60 % of
mammalian protein-coding genes are more or less regulated by miRNAs.
miRNAs are numerous and distributed throughout the genomes of both animals
and plants. In the mouse, as in humans, their number has been estimated in the
range of 1,000. miRNAs are involved in many regulation processes, including cell
proliferation, differentiation, apoptosis, and development. They function via base-
pairing with complementary sequences of mRNA molecules (seed region), leading
either to translational repression or to silencing via target degradation.
miRNA nomenclature consists of the generic or root symbol Mir, followed by
the numbering in the miRBase database (www.mirbase.org), a database that tracks
microRNAs reported for all species. Mouse Mir143 (microRNA 143), for exam-
ple, is represented as mmu-mir-143 in miRBase, with the mmu signifying Mus
musculus (Fig. 5.12).
Demonstration of the involvement of miRNAs in a given developmental or
pathological process is not easy. In the mouse, this can be achieved, for exam-
ple, by performing the complete elimination of all miRNAs in a certain tissue or
cell type and then observing the phenotypic effects. Since the Dicer protein is
essential for the processing of miRNAs, as discussed above, mice with a condi-
tional knockout allele of Dicer targeted in Purkinje cells (see Chap. 8—targeted
knockout) no longer had any miRNAs in these cells, and were found to develop
ataxia with Purkinje cell degeneration. This indicates that at least some miRNAs
are indispensable for the differentiation of these highly specific cells (Schaefer
et al. 2007). Another more specific strategy would be to establish an indisputable
causal and direct relationship between a point mutation in the sequence of a given
miRNA and a particular phenotype. Examples of this type are now accumulating,
(a)
(b)
mmu-mir-143
Fig. 5.12 The microRNAs. MicroRNAs (miRNAs) are short, noncoding, single-stranded RNAs.
These miRNAs are nested within longer non-coding RNA molecules, which are processed in
several successive steps with a double-stranded pre-miRNA (a), and finally a functional single-
stranded RNA molecule measuring 20–22 bp (b). These miRNAs finely regulate the expression
levels of several genes by binding to the 3'-untranslated regions of the corresponding mRNAs.
The seed sequence of miR-143 (represented in bold) matches perfectly with the 3'-UTR of the
mRNA transcribed from the (cytosine-5)-methyltransferase 3A (DNMT3A) gene. mir-143 is
known to be involved in cardiac morphogenesis, it has also been implicated in human colon can-
cer development, and its expression is down-regulated during mouse odontoblast differentiation.
mir–143 is encoded in mouse Chr 18 and is transcribed from the same DNA as another miRNA
(mir-145). miRNAs are highly conserved in vertebrates, and this is suggestive of an important
function. It is expected that about 60 % of mammalian protein-coding genes are more or less
regulated by miRNAs
and one of the first and most well-documented cases is the semidominant mutation
Diminuendo (symbol Mir96Dmdo-Chr 6) (Lewis et al. 2009; Lewis and Steel 2010).
This mutation was observed in the progeny of a male treated with the chemical
mutagen ENU (see Chap. 7) and was presumably induced by this substance. The
phenotype is characterized by progressive deafness, a condition that is quite com-
mon in humans. After positional cloning and careful sequencing of several can-
didate DNA segments in the 4.96-Mb critical interval where Diminuendo was
mapped, the researchers finally found an A →T transversion in the “seed” region
of the miRNA Mir96. This mutation, which was unique to Diminuendo and absent
in all other mice as well as in a large series of vertebrates, was confirmed as the
causative agent of the deafness and was associated with the down-regulation of
several (at least five) proteins, each of them being involved in the function of the
hair cells of the inner ear. These five proteins, which are downstream in the cas-
cade of regulation initiated by Mir96, are all important for the differentiation and
function of the hair cells and were all found to result in deafness when individu-
ally knocked out.
The discovery of the molecular origin of the Diminuendo mutation is an exam-
ple of the role that the myriad of miRNAs may play in the fine regulation of gene
(mRNA) expression in several developmental or pathological processes in verte-
brates. The discovery of a point mutation in the seed region of Mir96 proved that
cell differentiation and organogenesis involve a network of functionally linked
proteins as well as one or several miRNA(s).
Identification of the miRNA targets would certainly represent an enormous step
forward in developmental genetics, and this is therefore a focus in many labora-
tories worldwide. Progress, however, is hampered by the fact that miRNAs are
very small molecules and their sequences are not often totally complementary to
their targets. In addition to this difficulty, many scientists also believe that many
mRNAs, if not all, offer several targets to several miRNAs in their 3'-UTRs, thus
adding even more complexity to the picture.
MicroRNAs definitely have a promising future in medicine because they
are simple molecules but have, at the same time, the power of interfering with
gene regulation. In humans they are intensively studied because their expres-
sion levels have been found increased in certain forms of cancers (for example,
lymphomas or chronic lymphocytic leukemias), in diseases like cardiomyopa-
thies, and in some infectious diseases or autoimmune diseases. These increases
in specific miRNAs can then be used as information for the diagnosis or prog-
nosis of the disease, or as potential treatments. For example, aortic banding
in mice induces cardiac hypertrophy and concomitant up-regulation of many
(over 100) miRNAs including Mir21. When Mir21 was knocked down using
an antisense approach, cardiomyocyte hypertrophy was reduced, suggest-
ing that this particular miRNA plays a key role in the mechanism of cardiac
hypertrophy. This obviously opens perspectives for the development of novel
therapies.
Scientists believe that there are different grades in the process of mRNA reg-
ulation by miRNAs. Some miRNAs regulate specific individual targets, but it
seems that key miRNAs (so-called “super-miRNAs”) can regulate the expres-
sion levels of hundreds of genes simultaneously and cooperatively. These super-
miRNAs are of course actively searched. It has been suggested that miRNAs exert
both absolute and fine-tuned control of gene expression, adjusting levels of tran-
scripts to give either complete repression or simply decreased expression. Such
“fine-tuning” miRNAs will be much harder to identify than those resulting in the
complete “switching off” of a gene, since loss of function of any of these miR-
NAs would presumably have subtle effects, which would be difficult to character-
ize and study.
The discovery over the past ten years of these post-transcriptional regula-
tors has opened up a “new continent of the RNA world”. We just gave a rapid
overview of this continent using the miRNAs as examples, but many other
RNA or RNA-like molecules are just as interesting. We will now consider the
case of siRNAs, another type of ncRNA with post-transcriptional regulatory
functions.
5.4.2.2 Small Interfering RNA
Small interfering RNA, short interfering RNA, or silencing RNAs (all abbrevi-
ated siRNAs) are short double-stranded RNA molecules (20–25 bp) with a 2-bp 3'
overhang and phosphate groups on the 5' end of each strand. These RNAs interfere
with (i.e., reduce or suppress) the expression of specific genes with complemen-
tary nucleotide sequence, and in so doing they obviously have similarities in their
mode of action with the miRNAs discussed above.
The existence of these siRNAs and their remarkable properties were dis-
covered by chance while plant geneticists were performing transgenic experi-
ments with the aim of darkening the color of petunia flowers. The transgene
they were using was that for chalcone synthase, a key enzyme of the
flavonoid/isoflavonoid biosynthesis pathway. The scientists expected that by
increasing the enzyme level with several extra transgenic copies of the gene,
this may influence the pigmentation of the flower (Napoli et al. 1990). In fact,
and to their surprise, instead of obtaining the dark purple flowers they expected
they got light-colored flowers and sometimes flowers with white (unpigmented)
patches, indicating that the chalcone-encoding transgene actually had adverse
effects on the pigmentation process. Other similar experiments revealed that the
observed phenotypes were not exceptional but, on the contrary, the consequence
of an increased rate of mRNA degradation leading to specific gene suppression
or, more precisely, down-regulation. This effect was designated RNA interfer-
ence or RNAi.
In 1998, Fire and colleagues (1998), performing a similar experiment with
the worm Caenorhabditis elegans, concluded that neither the complete mRNA
nor a variety of antisense RNAs had an effect on protein production in experi-
mentally injected worms. However, they found that double-stranded RNAs cor-
responding to a myofilament protein successfully silenced the targeted genes,
once injected under the same conditions. They also demonstrated that only a few
molecules of injected double-stranded RNA were required to induce gene
silencing, thus arguing against stoichiometric interference with endogenous
mRNA and suggesting that there could be a catalytic or amplification compo-
nent in the interference process. This finding had a great impact in biology and
medicine when it was demonstrated that RNAi mechanisms are universal and
active in humans as well as in several model organisms including rats and mice,
offering new tools for gene annotation as well as opening the way to the devel-
opment of novel therapeutic strategies for the treatment of genetic diseases,
including cancers.8
Unlike in many model species, RNA interference cannot be triggered in mam-
malian cells by injecting long double-stranded RNAs, because the cells recognize
these RNAs as viruses and immediately develop a deleterious interferon response
with consequences for cell survival. Short molecules do not trigger this reaction
when injected into the cells.
siRNAs can also be synthesized as single-stranded molecules in the laboratory
and then introduced into cells either by direct injection or by transfection. Direct
chemical synthesis has the great advantage of allowing slight variations in the
sequence, and as a result increasing the efficiency of the siRNAs. Not all native
siRNAs are equally active, and the possibility of synthesizing novel molecules
appears to be a promising strategy (Ramachandran and Ignacimuthu 2013). The
mechanisms by which miRNAs and siRNAs work are similar. However, while
miRNAs cause translational repression or destabilization, the siRNAs cleave their
target RNAs at a particular site.
The use of RNA interference is an interesting and efficient way of altering
the gene function and accordingly of performing gene annotation. However, in
most instances and unlike other strategies described in Chap. 8, RNA interfer-
ence induces down-regulation of gene expression (knockdown) and not knockout
proper. In addition, some of these knockdowns are not specific.
5.4.2.3 Piwi-Interacting RNAs
Piwi-interacting RNAs (piRNAs) are short ncRNAs (26–31 nt long), which are
expressed mainly, not to say specifically, in spermatogenic cells of mammals.
Their function is not yet fully understood, but it is known that they form com-
plexes with the regulatory piwi (or miwi) proteins. These piRNA complexes are
thought to play a role in transposon silencing in male germ line cells, limiting the
expansion of these repeated sequences. They presumably have other functions that
have not yet been characterized.
8 A. Fire and C. Mello were awarded the Nobel Prize in Physiology or Medicine in 2006 for
their discovery of “RNA interference—gene silencing by double-stranded RNA”.

5.4.2.4 Long Non-coding RNAs
Long non-coding RNAs (lncRNAs) have an average size larger than 200 nt and in
many cases, in the range of 2 kb or more. This relatively great size distinguishes
them from all other ncRNAs, but being similar in size to the mRNAs can hamper
their isolation and characterization. Computer algorithms assessing the coding poten-
tial of the two molecules (lncRNAs and mRNAs) have been used to discriminate
between these molecules when necessary, but this criterion has finally proven unreli-
able because some (not all) lncRNAs do have a coding frame or, more precisely, a
nucleotide sequence resembling a coding frame with start and termination codons. So
far, the analysis of the sequences of lncRNAs does not allow sorting them in discrete
families with specific functions. In addition, the sequences of these RNAs are only
poorly conserved across species, even among closely related mammals. Indeed, this
family of ncRNAs is heterogeneous to the point where its very existence has long
been debated. Since lncRNAs are four times more numerous than mRNAs, one can
understand why they have been designated the “dark matter” of the transcriptome.
Aside from this rather confusing situation, some data have recently emerged that
make the situation a little more coherent. First, sequence alignments reveal that lncRNAs
are transcribed from both strands and in both directions overlapping introns, some-
times exons, and intergenic regions: this is never the case with mRNAs. Also, unlike
mRNAs, many of these molecules stay in the nucleus, suggesting that they have a
function at or close to this location. Finally, and as we will discuss further, the den-
sity of lncRNAs seems to be locally associated with some pathologies, suggesting
that they may be involved more or less directly in these processes.
Most of the knowledge we have of the lncRNAs results from the studies of five
important lncRNAs that have been studied in the mouse and whose functions have
now been relatively well characterized: these are the Kcnq1 overlapping transcript 1
(Kcnq1ot1-Chr 7), the antisense IGF2R-RNA (Airn-Chr 17), the HOX transcript
antisense RNA (Hotair-Chr 15), the X-specific transcripts (Xist-Chr X), and the
X (inactive)-specific transcript, antisense (Tsix-Chr X). The function and mode of
action of the lncRNAs involved in the X-chromosome inactivation process will be
analyzed in Chap. 11. Xist is one of the first genes, expressed after fertilization,
leading to silencing of all the genes on the targeted chromosome as a consequence
of histone H3 modifications. The targeting of XIST RNA to only one of the chro-
mosomes is controlled by another lncRNA: TSIX, which is the antisense repressor
of Xist on the active X chromosome.
Antisense repression is also the mode of action of the gene Kcnq1, whose
expression is silenced by the paternally expressed antisense non-coding RNA
KCNQ1OT1.
LncRNAs have extremely variable stability and expression levels. Some have a
half-life of only one hour (for example, KCNQ1OT1), while others are much more
stable. Some are highly expressed, while others are barely detectable.
Indeed, from the many reviews that have been published, one can conclude that
“we have barely begun to scratch the surface of the lncRNA world” (Kung et al.
2013).
5.5 Ultraconserved Elements (UCE) and Long Conserved

Non-coding Sequences
When the mouse genomic sequence is aligned to the genomic sequence of other
vertebrate species, we observe that quite a large number of elements measur-
ing ≥200 bp are conserved, and sometimes highly conserved. These sequence
elements are commonly designated ultraconserved elements (UCEs). UCEs
were first described in the human, rat, and mouse genomes by Bejerano and
coworkers (2004), but were also discovered in many other more distantly related
species (chicken, for example). For the UCEs encoding proteins or functional
RNAs, geneticists have an explanation: they consider that these resemblances
are a consequence of strong selection pressures acting during evolution and that
we mentioned earlier as “genome-shaping forces”. However, the situation is
much less clear for the non-protein-coding UCEs, and in this case explanations
are lacking.
After alignment of the mouse and human genomes, scientists at the RIKEN
Institute identified over 600 such conserved non-coding DNA sequences with
nearly 95 % identity and a size greater than 500 bp, most of them independent
of the previously reported UCEs (Sakuraba et al. 2008). These sequences, which
they provisionally designated long conserved non-coding sequences (LCNS), were
also found scattered throughout the genome of the rat as well as other vertebrate
species (chick, frog, fish) but were not found in non-vertebrate species. Given
that the probability of finding sequence similarities of that kind, just by chance,
is extremely low, two hypotheses were proposed by the researchers to account
for their observations: the first hypothesis was to consider that these LCNS either
have an important although unknown function associated with their structure
(they could have regulatory or structural elements important for the chromosome
structural organization, for example), or that they are transcribed into functional
ncRNAs whose function is not yet established (perhaps a type of lncRNA); in
both cases, this would explain why the sequences in question were selectively
constrained. The second hypothesis is that the LCNS/UCEs have remained intact
for so many years of evolution, simply because they are mutational cold spots
(Katzman et al. 2007). To challenge these hypotheses, the scientists had the clever
idea of performing ENU mutagenesis and measuring, afterwards, the frequencies
of induced mutations in the LCNS and comparing it with other genomic regions.
They did not find any significant differences in the mutation rates after screening
40.7 Mb of conserved sequences (~35 mutations) and concluded that the LCNS
were not mutational cold spots. To date, we do not have any satisfactory expla-
nation to account for the presence of so many of these LCNS/UCEs. The scien-
tists of the ENCODE project consider them to be associated with gene regulation
(ENCODE Project 2012) and their role is probably essential if we consider their
near-universal conservation across extremely divergent species. On the other hand,
it has also been reported that deletions of these UCEs in mice had virtually no
effect on the viability or fertility of the animals (Ahituv et al. 2007). This indicates
5.5 Ultraconserved Elements (UCE) and Long Conserved Non-coding Sequences 177
that extreme sequence constraints do not necessarily correspond to crucial func-

tions. For mouse geneticists, this also indicates that another type of sequence ele-
ment must now be added to the “dark matter of the transcriptome”.
5.6 Mitochondrial DNA
Mitochondria have a genome of their own that is represented by a small, circu-

lar, double-stranded DNA molecule known as mtDNA, sometimes mDNA. In the
mouse (as in humans) there are between two and ten such mtDNA molecules per
mitochondrion, and the number of mitochondria per cell is extremely variable and
depends on the cell type. The oocyte, for example, contains up to 106 mtDNA cop-
ies while a mature sperm cell contains less than 100.
The mtDNA comprises 37 genes encoding 13 mitochondrial enzymes involved
in respiration and oxidative phosphorylation, two ribosomal RNAs (12S and 16S)
and a full set of 22 tRNAs that are essential for the synthesis of these enzymes.
However, this small set of proteins represents only a sampling of the ~1,500 mito-
chondrial proteins, the rest of them being encoded in the nuclear genome. In con-
trast to the mammalian nuclear DNA, mtDNA is a naked DNA molecule (i.e.,
histone-free) with no introns and no sequence repeats (Bayona-Bafaluy et al.
2003). In addition, its two strands are quite different from those in the nuclear
DNA, the heavy strand being very heavy while the light one is much lighter. All
these unique characteristics of the mtDNA molecule are generally correlated with
its presumptive evolutionary origin, which states that the mitochondria are rem-
nants of bacteria that have been incorporated into the primitive eukaryotic cells
and retained as symbiotic organisms due to their selective advantages for cellular
metabolism. This interesting hypothesis, which is also proposed for chloroplasts in
plants, is not formally confirmed but it seems more than likely and fits perfectly
with the molecular data accumulated recently, in particular some fundamental
changes in codon usage9 (Yu et al. 2009).
The consensus sequence of the mouse mtDNA has been established and found
to consist of around 16,300 bp, with point variations (a few SNPs and gaps or
indels) among the most common laboratory inbred strains and the most commonly
used mouse species (Goios et al. 2007, 2008). These sequence polymorphisms
have been cleverly exploited to establish or to confirm the phylogenetic relation-
ships relationships between the different species of the genus Mus and related
genus (see Chap. 1) and the historical phylogeny among the laboratory strains (see
Chap. 9). This has allowed, in particular, the confirmation that a great majority
9 There are a few differences between the vertebrate mtDNA code and the “universal” code. In
the mtDNA, UGA codes for Trp rather than being a stop codon. In the same mtDNA there are
two Met codons (AUA and AUG) rather than only one. Finally, both AGA and AGG are read as
stop codons.
of the most frequently used inbred strains were all derived from the same female
ancestor, as initially established by Yonekawa et al. (1982), and to confirm that
most laboratory strains can be sorted into three groups with independent ances-
tral/geographical origins: the Sino-Japanese mice, the Swiss mice and the “Abbie
Lathrop’s” mice in the United States.
The mtDNA replicates at a much higher rate than the nuclear DNA and does
not possess repair mechanisms as efficient as those of the latter. For this rea-
son, and probably also because the mtDNA is not protected from the mutagenic
action of its environment by a variety of histone proteins, as is the case for mam-
malian DNA, it is more “mutable” and appears to be about 10–20 times more
affected by mutations generating a sequence polymorphism than the nuclear
DNA of the same species. Considering the great differences between male
and female gametes in terms of mitochondria numbers (up to 1/1,000), it is no
surprise to learn that the mtDNA is transmitted by the mother to her offspring
rather than by the father. Although sperm cells do have some mtDNA molecules,
the mtDNA appears to be lost very early during egg development, and in virtu-
ally all species studied so far the only mtDNA molecules found in embryos are
of maternal origin.
When a mutation occurs in a mtDNA molecule of an oocyte (or of a precur-
sor cell), it is generally counter-selected and rapidly eliminated unless it confers
a selective advantage to the mtDNA, for example by increasing its replication
rate (Sharpley et al. 2012). In the latter case, the mutant molecules progressively
overgrow the population of normal mtDNAs and the oocyte (or cell) becomes het-
eroplasmic with two (or more) types of mtDNA. Finally, due to some sort of sam-
pling effect, sometimes referred to as a genetic bottleneck, the mutant form of the
mtDNA may completely replace the pre-existing form and become the standard.
This explains why mtDNA is an attractive molecule to geneticists studying evolu-
tion. It is also interesting to note that mtDNA evolution is completely independent
of nuclear DNA evolution, and accordingly represents another valuable tool for
establishing the systematics of a species. For this reason, it has been extensively
used in many domestic species, including the mouse, and still is.
In the human species, mutations in the mtDNA have been associated with
more or less severe pathologies. Leber hereditary optic neuropathy (LHON), for
example, was the first reported and is one of the most prevalent, with an estimated
frequency of 15 in 100,000 births. This syndrome is the consequence of muta-
tions (several have been described) occurring in the genes encoding the oxida-
tive phosphorylation complex I. Many other mtDNA defects have been reported
in the human species, including a syndrome of maternally inherited diabetes and
deafness (MIDD), Leigh syndrome, a syndrome associating neuropathy, ataxia,
retinitis pigmentosa, and ptosis (NARP), myoneurogenic gastrointestinal encepha-
lopathy (MNGIE), and many other neuromuscular diseases. All these pathologies
are maternally transmitted and exhibit variations in severity presumably associated
with the degree of heteroplasmy. Surprisingly, no such pathologies clearly attribut-
able to an mtDNA defect have ever been reported in the mouse, although mtDNA
mutations have been reported in cell lines transplanted in vitro.
5.6 Mitochondrial DNA 179
Because spontaneous mtDNA defects have never been reported in the mouse,
animal models of human pathologies have been created by introducing defective
human mtDNAs into mouse oocytes.10 In particular, a murine model of LHON
syndrome has been produced by using this strategy. These mice exhibited reduc-
tion in retinal function, indicating that the physiopathology of the syndrome may
result from some oxidative stress (Lin et al. 2012).
Those readers of this chapter who might be interested in the biology and
pathology of mtDNA, in both human and mouse, should refer to the important
contribution of D.C. Wallace (University of Pennsylvania), who wrote several
reviews on the subject (Wallace 2009).
5.7 Conclusions
At the beginning of this chapter we stated that we considered the decision taken
several years ago to completely and systematically sequence the mouse genome
to be a wise one. If we consider the huge amount of information gathered, directly
or indirectly, from this sequencing and the data we can expect to collect in the
near future, our initial feeling is strengthened; indeed, the sequencing of the mouse
genome has had an enormous impact in many areas of genetics and biology.
The knowledge of this sequence has allowed the development of better tools
(for example, SNPs) and allows better experiments to be designed. Nowadays one
can design an experiment of homologous recombination (targeted mutagenesis)
with precision at the base-pair level.
Aside from these technical advances, in silico comparisons of the mouse
sequence with other mammalian (or vertebrate) sequences has allowed the dis-
covery of similarities or differences that have proved a rich source of information
for a better understanding of evolution. Even within individuals of the same spe-
cies, the analysis of copy number variations, for example, has revealed intriguing
differences whose significance and phenotypic expression is not yet completely
clear, even if we suspect that they probably play an important role in quantitative
genetics.
The information gathered concerning the structure of the mouse genome and its
variations across the different inbred strains and different subspecies of the Mus
genus will certainly reveal important clues for understanding the genetic determin-
ism of complex traits, especially when complemented by the constantly increasing
amounts of phenotyping data. The mouse is unique in the sense that one can cross
animals of different subspecies, breed very large progenies, extensively phenotype
all the animals, and sequence the individual genomes when desired.
10 Two inbred strains of mice with the same genomic (nuclear) DNA but different mtDNAs are
said to be conplasmic. The production of such strains can be achieved by normal sexual repro-
duction or by direct cytoplasmic transfer (See Chap. 9).
The sequencing of the genome has also revealed its great plasticity. We now
know that LINEs and ERVs play an important role in gene regulation, and even as
a source of diversity, a point that was totally unexpected.
Finally, a true revolution in our understanding of the transcriptome occurred
during the last ten years. The number of protein-encoding genes has been revised
downward while the number of RNA-encoding genes is constantly being revised
upward. Over the last ten years we have started to realize that a myriad of ncRNAs
(long and short) are transcribed from the genome, exhibiting great although still
incompletely explored functional diversity. From whole-genome analyses using
microarrays and high-throughput transcript sequencing, we estimate that more
than 85 % of the nucleotides in the euchromatic genome are represented in pri-
mary transcripts. Indeed, the proportion of supposedly “junk” DNA shrinks more
every day. We have learnt that the genome is pervasively and bidirectionally tran-
scribed, increasing tremendously the amount of information that can be stored.
The discovery of the role of miRNAs and siRNAs in the fine regulation of gene
activity is another revolution that may have major consequences for the diagnostic
and treatment of some diseases. The long coding RNAs probably play a major role
in gene regulation and imprinting … but we have information about only a handful
of these molecules although we know that there are many.
The role and importance of the ultraconserved elements and long conserved
non-coding sequences remains a mystery. If they are ultraconserved this would
mean that they are under selection pressure. But, alternatively, we know that they
can be experimentally deleted with apparently no consequences. No doubt all
these observations will fuel much research in the years to come and it won’t be
surprising that, at this point, even the concept of gene may be reconsidered11.
Acknowledgements The authors thank Doctor Benoît Robert, Institut Pasteur, for his contribution
to Sect. 5.3.3 of this chapter.
References
Ahituv N, Zhu Y, Visel A, Holt A, Afzal V, Pennacchio LA, Rubin EM (2007) Deletion of ultra-
conserved elements yields viable mice. PLoS Biol 5:e234
Arnold CN, Xia Y, Lin P, Ross C, Schwander M, Smart NG, Müller U, Beutler B (2011) Rapid
identification of a disease allele in mouse through whole genome sequencing and bulk seg-
regation analysis. Genetics 187:633–641
Balakirev ES, Ayala FJ (2003) Pseudogenes: are they “junk” or functional DNA? Annu Rev
Genet 37:123–151
Barash Y, Calarco JA, Gao W, Pan Q, Wang X, Shai O, Blencowe BJ, Frey BJ (2010)
Deciphering the splicing code. Nature 465:53–59
11 Most of the data provided in this chapter concerning the Mouse Genome are from the
Ensembl website: http://www.ensembl.org/Mus_musculus/Info/Annotation#assembly and http://

www.ncbi.nlm.nih.gov/projects/mapview/stats/BuildStats.cgi?taxid=10090&build=38&ver=1.
References 181
Bayona-Bafaluy MP, Acín-Pérez R, Mullikin JC, Park JS, Moreno-Loshuertos R, Hu P, Pérez-

Martos A, Fernández-Silva P, Bai Y, Enríquez JA (2003) Revisiting the mouse mitochondrial
DNA sequence. Nucleic Acids Res 31:5349–5355
Bejerano G, Pheasant M, Makunin I, Stephen S, Kent WJ, Mattick JS, Haussler D (2004)
Ultraconserved elements in the human genome. Science 304:1321–1325
Belancio VP, Roy-Engel AM, Pochampally RR, Deininger P (2010) Somatic expression of
LINE-1 elements in human tissues. Nucleic Acids Res 38:3909–3922
Birnbaum RY, Clowney EJ, Agamy O, Kim MJ, Zhao J, Yamanaka T, Pappalardo Z, Clarke
SL, Wenger AM, Nguyen L, Gurrieri F, Everman DB, Schwartz CE, Birk OS, Bejerano
G, Lomvardas S, Ahituv N (2012) Coding exons function as tissue-specific enhancers of
nearby genes. Genome Res 22:1059–1068
Blanco E, Guigo R (2005) Predictive methods using DNA sequences—analysis at the nucleotide
level. In: Baxevanis AD, Francis Ouellette BF (eds) Bioinformatics: a practical guide to the
analysis of genes and proteins, 3rd edn. Wiley, Hoboken
Buckler AJ, Chang DD, Graw SL, Brook JD, Haber DA, Sharp PA, Housman DE (1991) Exon
amplification: a strategy to isolate mammalian genes based on RNA splicing. Proc Natl
Acad Sci U S A. 88:4005–4009
Burge C, Karlin S (1997) Prediction of complete gene structures in human genomic DNA. J Mol
Biol 268:78–94
Cachon-Gonzalez MB, Fenner S, Coffin JM, Moran C, Best S, Stoye JP (1994) Structure and
expression of the hairless gene of mice. Proc Natl Acad Sci U S A 91:7717–7721
Canales CP, Walz K (2011) Copy number variation and susceptibility to complex traits. EMBO
Mol Med. 3:1–4
Carlson CM, Largaespada DA (2005) Insertional mutagenesis in mice: new perspectives and
tools. Nat Rev Genet 6:568–580
Carninci P, Kasukawa T, Katayama S, Gough J, Frith MC, Maeda N, Oyama R, Ravasi T,
Lenhard B, Wells C, Kodzius R, Shimokawa K et al (2005) The transcriptional landscape of
the mammalian genome. Science 309:1559–1563
Carninci P, Sandelin A, Lenhard B, Katayama S, Shimokawa K, Ponjavic J, Semple CA, Taylor
MS, Engström PG, Frith MC, Forrest AR, Alkema WB, Tan SL, Plessy C, Kodzius R,
Ravasi T, Kasukawa T, Fukuda S, Kanamori-Katayama M, Kitazume Y, Kawaji H, Kai
C, Nakamura M, Konno H, Nakano K, Mottagui-Tabar S, Arner P, Chesi A, Gustincich
S, Persichetti F, Suzuki H, Grimmond SM, Wells CA, Orlando V, Wahlestedt C, Liu ET,
Harbers M, Kawai J, Bajic VB, Hume DA, Hayashizaki Y (2006) Genome-wide analysis of
mammalian promoter architecture and evolution. Nat Genet 38:626–635
Cazaux B, Catalan J, Veyrunes F, Douzery EJ, Britton-Davidian J (2011) Are ribosomal DNA
clusters rearrangement hotspots?: a case study in the genus Mus (Rodentia, Muridae). BMC
Evol Biol 11:124. doi:10.1186/1471-2148-11-124
Cook EH, Scherer SW (2008) Copy-number variations associated with neuropsychiatric condi-
tions. Nature 455:919–923
Copeland NG, Jenkins NA (2010) Harnessing transposons for cancer genes discovery. Nat Rev
Cancer 10:696–706
Coughlin DJ, Babak T, Nihranz C, Hughes TR, Engelke DR (2009) Prediction and verification of
mouse tRNA gene families. RNA Biol 6:195–202
Cutler G, Kassner PD (2008) Copy number variation in the mouse genome: implications for the
mouse as a model organism for human disease. Cytogenet Genome Res 123:297–306
Demuth JP, Hahn MW (2009) The life and death of gene families. BioEssays 31:29–39
Ebert MS, Sharp PA (2012) Roles for microRNAs in conferring robustness to biological pro-
cesses. Cell 149:515–524
Egan CM, Sridhar S, Wigler M, Hall I (2007) Recurrent DNA copy number variation in the labo-
ratory mouse. Nat Genet 39:1384–1389
Ehrnhoefer DE, Butland SL, Pouladi MA, Hayden MR (2009) Mouse models of Huntington dis-
ease: variations on a theme. Dis Model Mech 2:123–129
ENCODE Project Consortium (2011) A user’s guide to the encyclopedia of DNA elements
(ENCODE). PLoS Biol 4:e1001046
ENCODE Project Consortium, Dunham I, Kundaje A, Aldred SF, Collins PJ, Davis CA, Doyle
F, Epstein CB, Frietze S, Harrow J, Kaul R, Khatun J, Lajoie BR et al (2012) An integrated
encyclopedia of DNA elements in the human genome. Nature 489:57–74
Fire A, Xu S, Montgomery M, Kostas S, Driver S, Mello C (1998) Potent and specific genetic
interference by double-stranded RNA in Caenorhabditis elegans. Nature 391:806–811
Gardner P, Bateman A, Poole AM (2010) SnoPatrol: how many snoRNA genes are there? J Biol
9:1–4
Gibbs RA, Weinstock GM, Metzker ML, Muzny DM, Sodergren EJ, Scherer S, Scott G, Steffen
D, Worley KC, Burch PE et al (2004) Genome sequence of the Brown Norway rat yields
insights into mammalian evolution. Nature 428:493–521
Goios A, Pereira L, Bogue M, Macaulay V, Amorim A (2007) mtDNA phylogeny and evolution
of laboratory mouse strains. Genome Res 17:293–298
Goios A, Gusmão L, Rocha AM, Fonseca A, Pereira L, Bogue M, Amorim A (2008)
Identification of mouse inbred strains through mitochondrial DNA single-nucleotide exten-
sion. Electrophoresis 29:4795–4802
Goldberg ML. 1979. PhD Diss. Stanford University, Stanford, CA
Gonzalez E, Kulkarni H, Bolivar H, Mangano A, Sanchez R, Catano G, Nibbs RJ, Freedman BI,
Quinones MP, Bamshad MJ, Murthy KK, Rovin BH, Bradley W, Clark RA, Anderson SA,
O’connell RJ, Agan BK, Ahuja SS, Bologna R, Sen L, Dolan MJ, Ahuja SK (2005) The
influence of CCL3L1 gene-containing segmental duplications on HIV-1/AIDS susceptibil-
ity. Science 307:1434–1440
Gustincich S, Sandelin A, Plessy C, Katayama S, Simone R, Lazarevic D, Hayashizaki Y,
Carninci P (2006) The complexity of the mammalian transcriptome. J Physiol 575:321–332
Hardison RC, Taylor J (2012) Genomic approaches towards finding cis-regulatory modules in
animals. Nat Rev Genet 13:469–483
Harrow J, Nagy A, Reymond A, Alioto T, Patthy L, Antonarakis SE, Guigó R (2009) Identifying
protein-coding genes in genomic sequences. Genome Biol 10:201 Epub
Hatzis P, van der Flier LG, van Driel MA, Guryev V, Nielsen F, Denissov S, Nijman IJ, Koster J,
Santo EE, Welboren W, Versteeg R, Cuppen E, van de Wetering M, Clevers H, Stunnenberg
HG (2008) Genome-wide pattern of TCF7L2/TCF4 chromatin occupancy in colorectal can-
cer cells. Mol Cell Biol 28:2732–2744
Hayashizaki Y, Carninci P (2006) Genome Network and FANTOM3: Assessing the Complexity
of the Transcriptome. PLoS Genet 2(4):e63
Henderson AS, Eicher EM, Yu MT, Atwood KC (1974) The chromosomal location of ribosomal
DNA in the mouse. Chromosoma 49:155–160
Hill RE (2007) How to make a zone of polarizing activity: insights into limb development via the
abnormality preaxial polydactyly. Dev Growth Differ 49:439–448
Hoskins AA, Moore MJ (2012) The spliceosome: a flexible, reversible macromolecular machine.
Trends Biochem Sci 37:179–188
Howell VM (2012) Sleeping beauty—a mouse model for all cancers? Cancer Lett 317:1–8
International Human Genome Sequencing Consortium, Lander E, Linton L, Birren B, Nusbaum
C, Zody MC, Baldwin J, Devon K, Dewar K, Doyle M, FitzHugh W et al (2001) Initial
sequencing and analysis of the human genome. Nature 409:890–921
Izsvák Z, Ivics Z (2005) Sleeping Beauty hits them all: transposon-mediated saturation mutagen-
esis in the mouse germline. Nat Methods 2:735–736
Julier C, de Gouyon B, Georges M, Guénet JL, Nakamura Y, Avner P, Lathrop GM (1990)
Minisatellite linkage maps in the mouse by cross-hybridization with human probes contain-
ing tandem repeats. Proc Natl Acad Sci U S A. 87:4585–4589
Kapranov P, St Laurent G (2012) Dark matter RNA: existence, function, and controversy. Front
Genet. 3:60
References 183
Kapranov P, Cawley SE, Drenkow J, Bekiranov S, Strausberg RL, Fodor SP, Gingeras TR (2002)
Large-scale transcriptional activity in chromosomes 21 and 22. Science 296:916–919
Katayama S, Tomaru Y, Kasukawa T, Waki K, Nakanishi M, Nakamura M, Nishida H, Yap CC,
Suzuki M, Kawai J, Suzuki H, Carninci P, Hayashizaki Y, Wells C, Frith M, Ravasi T, Pang
KC, Hallinan J, Mattick J, Hume DA, Lipovich L, Batalov S, Engström PG, Mizuno Y,
Faghihi MA, Sandelin A, Chalk AM, Mottagui-Tabar S, Liang Z, Lenhard B, Wahlestedt C;
RIKEN Genome Exploration Research Group; Genome Science Group (Genome Network
Project Core Group); FANTOM Consortium (2005) Antisense transcription in the mamma-
lian transcriptome. Science 309:1564–1566
Katzman S, Kern AD, Bejerano G, Fewell G, Fulton L, Wilson RK, Salama SR, Haussler D
(2007) Human genome ultraconserved elements are ultraselected. Science 317:915
Kozak M (1987) At least six nucleotides preceding the AUG initiator codon enhance translation
in mammalian cells. J Mol Biol 196:947–950
Kung JT, Colognori D, Lee JT (2013) Long noncoding RNAs: past, present, and future. Genetics
193:651–669
Kuznetsova IS, Prusov AN, Enukashvily NI, Podgornaya OI (2005) New types of mouse centro-
meric satellite DNAs. Chromosome Res 13:9–25
Lagha M, Bothma JP, Levine M (2012) Mechanisms of transcriptional precision in animal devel-
opment. Trends Genet 28:409–416
Lai F, Orom UA, Cesaroni M, Beringer M, Taatjes DJ, Blobel GA, Shiekhattar R (2013)
Activating RNAs associate with Mediator to enhance chromatin architecture and transcrip-
tion. Nature 494:497–501
Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Baldwin J, Devon K, Dewar K, Doyle
M, FitzHugh W, Funke R, Gage D et al (2001) Initial sequencing and analysis of the human
Lettice LA, Horikoshi T, Heaney SJ, van Baren MJ, van der Linde HC, Breedveld GJ, Joosse M,
Akarsu N, Oostra BA, Endo N, Shibata M, Suzuki M, Takahashi E, Shinka T, Nakahori Y,
Ayusawa D, Nakabayashi K, Scherer SW, Heutink P, Hill RE, Noji S (2002) Disruption of a
long-range cis-acting regulator for Shh causes preaxial polydactyly. Proc Natl Acad Sci U S
A. 99:7548–7553
Lewis MA, Steel KP (2010) microRNAS in mouse development and diseases. Semin Cell
Develop Biol 21:774–780
Lewis MA, Quint E, Glazier AM, Fuchs H, De Angelis MH, Langford C, van Dongen S, Abreu-
Goodger C, Piipari M, Redshaw N, Dalmay T, Moreno-Pelayo MA, Enright AJ, Steel KP
(2009) An ENU-induced mutation of miR-96 associated with progressive hearing loss in
mice. Nat Genet 41:614–618
Li G, Ruan X, Auerbach RK, Sandhu KS, Zheng M, Wang P, Poh HM, Goh Y, Lim J, Zhang J,
Sim HS, Peh SQ, Mulawadi FH, Ong CT, Orlov YL, Hong S, Zhang Z, Landt S, Raha D,
Euskirchen G, Wei CL, Ge W, Wang H, Davis C, Fisher-Aylor KI, Mortazavi A, Gerstein M,
Gingeras T, Wold B, Sun Y, Fullwood MJ, Cheung E, Liu E, Sung WK, Snyder M, Ruan Y
(2012) Extensive promoter-centered chromatin interactions provide a topological basis for
transcription regulation. Cell 148:84–98
Lin CS, Sharpley MS, Fan W, Waymire KG, Sadun AA, Carelli V, Ross-Cisneros FN, Baciu
P, Sung E, McManus MJ, Pan BX, Gil DW, Macgregor GR, Wallace DC (2012) Mouse
mtDNA mutant model of Leber hereditary optic neuropathy. Proc Natl Acad Sci U S A.
109:20065–20070
Mashimo T, Glaser P, Lucas M, Simon-Chazottes D, Ceccaldi PE, Montagutelli X, Desprès P,
Guénet JL (2003) Structural and functional genomics and evolutionary relationships in the
cluster of genes encoding murine 2′,5'-oligoadenylate synthetases. Genomics 82:537–552
Mattick JS, Makunin IV (2006) Non-coding RNA. Hum Mol Genet 15 Spec No 1:R17–29
Modrek B, Lee CJ (2003) Alternative splicing in the human, mouse and rat genomes is associ-
ated with an increased frequency of exon creation and/or loss. Nat Genet 34:177–180
Mouse ENCODE Consortium (2012) An encyclopedia of mouse DNA elements (Mouse

ENCODE). Genome Biol 13:418
Mouse Genome Sequencing Consortium, Waterston RH, Lindblad-Toh K, Birney E, Rogers J,
Abril JF, Agarwal P, Agarwala R, Ainscough R, Alexandersson M, An P, et al (2002) Initial
sequencing and comparative analysis of the mouse genome. Nature 420:520–562
Mülhardt C, Fischer M, Gass P, Simon-Chazottes D, Guénet JL, Kuhse J, Betz H, Becker CM
(1994) The spastic mouse: aberrant splicing of glycine receptor beta subunit mRNA caused
by intronic insertion of L1 element. Neuron 13:1003–1015
Munroe R, Schimenti J (2009) Mutagenesis of mouse embryonic stem cells with ethylmethane-
sulfonate. Methods Mol Biol 530:131–138
Mural RJ, Adams MD, Myers EW, Smith HO, Miklos GL, Wides R, Halpern A, Li PW, Sutton
GG, Nadeau J et al (2002) A comparison of whole-genome shotgun-derived mouse chromo-
some 16 and the human genome. Science 296:1661–1671
Napoli C, Lemieux C, Jorgensen R (1990) Introduction of a chimeric chalcone synthase gene
into petunia results in reversible co-suppression of homologous genes in trans. Plant Cell
2:279–289
Nouvel P (1994) The mammalian genome shaping activity of reverse transcriptase. Genetica
93:191–201
Ohno S (1972) So much “junk” DNA in our genome. In: Smith HH (ed) Evolution of genetic
systems. Gordon and Breach, New York, pp 366–370
Okazaki Y, Furuno M, Kasukawa T, Adachi J, Bono H, Kondo S, Nikaido I, Osato N, Saito R,
Suzuki H, Yamanaka I et al (2002) Analysis of the mouse transcriptome based on functional
annotation of 60,770 full-length cDNAs. Nature 420:563–573
Panthier JJ, Dreyfus M, Roux TL, Rougeon F (1984) Mouse kidney and submaxillary gland renin
genes differ in their 5' putative regulatory sequences. Proc Natl Acad Sci U S A. 81:5489–5493
Perelygin AA, Zharkikh AA, Scherbik SV, Brinton MA (2006) The mammalian 2′-5' oligoad-
enylate synthetase gene family: evidence for concerted evolution of paralogous Oas1 genes
in Rodentia and Artiodactyla. J Mol Evol 63:562–576
Perez CJ, Dumas A, Vallières L, Guénet JL, Benavides F (2013) Several classical mouse inbred
strains, including DBA/2, NOD/Lt, FVB/N, and SJL/J, carry a putative loss-of-function
allele of Gpr84. J Hered 104:565–571
Petkov PM, Graber JH, Churchill GA, DiPetrillo K, King BL, Paigen K (2007) Evidence of a
large-scale functional organization of Mammalian chromosomes. PLoS Biol 5(5):e127
Ramachandran PV, Ignacimuthu S (2013) RNA interference-a silent but an efficient therapeutic
tool. Appl Biochem Biotechnol 169:1774–1789
Roberts RL, Diaz-Gallo LM, Barclay ML, Gómez-García M, Cardeña C, Merriman TR, Gearry
RB, Martin J (2012) Independent replication of an association of CNVR7113.6 with
Crohn’s disease in Caucasians. Inflamm Bowel Dis 18:305–311
Rowe LB, Janaswami PM, Barter ME, Birkenmeier EH (1996) Genetic mapping of 18S riboso-
mal RNA-related loci to mouse chromosomes 5, 6, 9, 12, 17, 18, 19, and X. Mamm Genome
12:886–889
Sakuraba Y, Kimura T, Masuya H, Noguchi H, Sezutsu H, Takahasi KR, Toyoda A, Fukumura
R, Murata T, Sakaki Y, Yamamura M, Wakana S, Noda T, Shiroishi T, Gondo Y (2008)
Identification and characterization of new long conserved noncoding sequences in verte-
brates. Mamm Genome 19:703–712
Savarese F, Grosschedl R (2006) Blurring cis and trans in gene regulation. Cell 126:248–250
Saxena A, Carninci P (2011) Long non-coding RNA modifies chromatin: epigenetic silencing by
long non-coding RNAs. BioEssays 33:830–839
Schaefer A, O’Carroll D, Tan CL, Hillman D, Sugimori M, Llinas R et al (2007) Cerebellar neu-
rodegeneration in the absence of microRNAs. J Exp Med 204:1553–1558
Sebat J, Lakshmi B, Malhotra D, Troge J, Lese-Martin C, Walsh T, Yamrom B, Yoon S, Krasnitz
A, Kendall J, Leotta A, Pai D, Zhang R, Lee YH, Hicks J, Spence SJ, Lee AT, Puura K,
Lehtimäki T, Ledbetter D, Gregersen PK, Bregman J, Sutcliffe JS, Jobanputra V, Chung
References 185
W, Warburton D, King MC, Skuse D, Geschwind DH, Gilliam TC, Ye K, Wigler M (2007)
Strong association of de novo copy number mutations with autism. Science 316:445–449
Sharpley MS, Marciniak C, Eckel-Mahan K, McManus M, Crimi M, Waymire K, Lin CS,
Masubuchi S, Friend N, Koike M, Chalkia D, MacGregor G, Sassone-Corsi P, Wallace DC
(2012) Heteroplasmy of mouse mtDNA is genetically unstable and results in altered behav-
ior and cognition. Cell 151:333–343
Shen Y, Yue F, McCleary DF, Ye Z, Edsall L, Kuan S, Wagner U, Dixon J, Lee L, Lobanenkov
VV, Ren B (2012) A map of the cis-regulatory sequences in the mouse genome. Nature
488:116–120
Silver LM (1995) Mouse genetics—concepts and applications. Oxford University Press, Oxford
Sookdeo A, Hepp CM, McClure MA, Boissinot S (2013) Revisiting the evolution of mouse
LINE-1 in the genomic era. Mob DNA 4:3. doi:10.1186/1759-8753-4-3
Specht CG, Schoepfer R (2001) Deletion of the alpha-synuclein locus in a subpopulation of
C57BL/6 J inbred mice. BMC Neurosci 2:11
Stoye JP, Fenner S, Greenoak GE, Moran C, Coffin JM (1988) Role of endogenous retroviruses
as mutagens: the hairless mutation of mice. Cell 54:383–391
Valentijn LJ, Baas F, Wolterman RA, Hoogendijk JE, van den Bosch NH, Zorn I, Gabreëls-
Festen AW, de Visser M, Bolhuis PA (1992a) Identical point mutations of PMP-22 in
Trembler-J mouse and Charcot-Marie-Tooth disease type 1A. Nat Genet 4:288–291
Valentijn LJ, Bolhuis PA, Zorn I, Hoogendijk JE, van den Bosch N, Hensels GW, Stanton VP Jr,
Housman DE, Fischbeck KH, Ross DA et al (1992b) The peripheral myelin gene PMP-22/
GAS-3 is duplicated in Charcot-Marie-Tooth disease type 1A. Nat Genet 1:166–170
Venter JC, Adams MD, Myers EW, Li PW, Mural RJ, Sutton GG, Smith HO, Yandell M, Evans
CA, Holt RA et al (2001) The sequence of the human genome. Science 291:1304–1351
Wallace DC (2009) The pathophysiology of mitochondrial disease as modeled in the mouse.
Genes Dev 23:1714–1736
Watkins-Chow DE, Pavan WJ (2008) Genomic copy number and expression variation within the
C57BL/6 J inbred mouse strain. Genome Res 18:60–66
Wessler SR (2006) Transposable elements and the evolution of eukaryotic genomes. Proc Natl
Acad Sci U S A. 103:17600–17601
Wong K, Bumpstead S, Van Der Weyden L, Reinholdt LG, Wilming LG, Adams DJ, Keane
TM (2012) Sequencing and characterization of the FVB/NJ mouse genome. Genome Biol
13:R72
Xia Y, Won S, Du X, Lin P, Ross C, La Vine D, Wiltshire S, Leiva G, Vidal SM, Whittle B,
Goodnow CC, Koziol J, Moresco EM, Beutler B (2010) Bulk segregation mapping of muta-
tions in closely related strains of mice. Genetics 186:1139–1146
Yonekawa H, Moriwaki K, Gotoh O, Miyashita N, Migita S, Bonhomme F, Hjorth JP, Petras ML,
Tagashira Y (1982) Origins of laboratory mice deduced from restriction patterns of mito-
chondrial DNA. Differentiation 22:222–226
Yu X, Wester-Rosenlöf L, Gimsa U, Holzhueter SA, Marques A, Jonas L, Hagenow K, Kunz M,
Nizze H, Tiedge M, Holmdahl R, Ibrahim SM (2009) The mtDNA nt7778 G/T polymor-
phism affects autoimmune diseases and reproductive performance in the mouse. Hum Mol
Genet 18:4689–4698
Zelnick CR, Burks DJ, Duncan CH (1987) A composite transposon 3' to the cow fetal globin
gene binds a sequence specific factor. Nucleic Acids Res 15:10437–10453
Chapter 6
Epigenetic Control of Genome Expression
6.1 Introduction
From the standpoint of evolution, diploidy is generally considered advantageous

for two reasons. First, because diploid organisms possess twice as many genes
as haploids and in these conditions twice as many favorable mutations arise per
generation. This of course increases the genetic diversity in the population and,
finally, contributes to the progress of adaptive evolution. Diploidy is also consid-
ered advantageous because, when a recessive mutation occurs in a given gene,
there is always a backup copy of the original allele on the other chromosome,
offering a chance for the population to assess, with no risk, which one of the two
alleles is most advantageous for the future of the species in a given environmental
context. In most cases, the new mutant allele is neutral and has no selective advan-
tage; sometimes it is harmful and is more or less rapidly eliminated. On rare occa-
sions, it is beneficial and can then gradually replace the original allele.
However, mammals are not perfectly diploid since the males have an X and a
Y chromosome while females have two X chromosomes. This difference, which is
associated with sex determination, requires that a mechanism of gene dosage com-
pensation be developed to equilibrate the transcriptional activity of the X-linked
genes between the two sexes. Understanding this mechanism and its determinism
has elicited a great number of investigations over the last fifty years, and scientists
have discovered that, in eutherian mammals, the females are functionally haploid
for the major part of their X chromosome. As we will explain in this chapter, this
functional haploidy is controlled by an epigenetic mechanism leading to the inacti-
vation of one of the two X chromosomes.
In the same mammals, scientists have also discovered that some autosomal
regions, sometimes reduced to one or a few genes, are also functionally haploid,
exclusively expressing the alleles inherited from one of the two parents and not
those inherited from the other, due to the intervention of similar epigenetic mech-
anisms. These discoveries have broken the dogma of the superiority of diploidy
over haploidy and have revealed the existence of a new kind of control of the tran-
scriptional activity of mammalian genes. Although the epigenetic mechanisms

188 6 Epigenetic Control of Genome Expression
controlling the transcriptional activity of the X chromosome are not exactly the
same as those at work for the autosomal regions, they nonetheless have so many
similarities that we will describe them here, in the same chapter.
6.2 X-Chromosome Inactivation in Mammals
In mammals, the XX/XY sex-determination system is common, and only rare excep-
tions have been reported.1 In the mouse, females have two large X chromosomes
while males have an X and a Y on which the sex-determining region (Sry) is the
master regulator of sex determination. In its absence, for example in mice with an
XX or XO karyotype, the embryo develops as a normal, healthy and fertile female.2
The XX/XY system is both simple and robust, since relatively few anomalies in
sex determination (intersexuality or sex ambiguities) have been reported, but the
presence of two X chromosomes in the female versus a single one in males clearly
raises a problem associated with gene dosage imbalance. For this reason, during
their evolution mammals have found an efficient way to compensate (or more pre-
cisely to equilibrate) the transcription of X-linked genes between the two sexes.
6.2.1 In Female Mammals Only One X is Transcriptionally

Active
The XX/XY sex-determination system exists in many diploid organisms, and differ-
ent ways of solving the question of XX/XY dosage compensation have been retained.
In the fruit fly Drosophila melanogaster, for example, the male-specific lethal (MSL)
complex increases transcription of the single X chromosome to equalize expression
of X-linked genes between the two sexes (Larschan et al. 2011). In Cænorhabditis
1 At least two exceptions to the classical XX/XY mechanism of sex determination have been
reported. The first one is found in wood lemmings (Myopus schisticolor), a species of Cricetidae
rodent in which there are two types of X chromosomes (X and X*) and a Y chromosome. XX
genotypes develop as females and XY develop as males, as in other mammals. However, both
X*X and X*Y develop as females because the X* chromosome carries a mutation that inhibits
the male-determining effect of the Y chromosome. The three categories of females (XX, X*X,
and X*Y) are fertile, but X*Y females only produce X* ova. This sex determination system
induces a strong distortion in the sex ratio (3/1 instead of the normal 1/1) and is considered an
adaptation to the extreme seasonal reductions in population size that might otherwise threaten
the survival of the species. Another remarkable exception is the mole vole Ellobius lutescens,
another species of Cricetidae rodent in which both the male and female have the same odd num-
ber of chromosomes with a single X and no Y. In this species the sex-determination process is not
yet completely understood.
2 The development of testes as gonads also depends upon some other genes (Foxl2, NrOb1,
Sox9, etc.).
6.2 X-Chromosome Inactivation in Mammals 189
elegans, dosage compensation is achieved by the female in which transcription

from the two X chromosomes is simply halved (Kelly et al. 2002). In mammals, yet
another solution has evolved with one of the two X chromosomes being inactivated in
the female.
In 1961, the Harwell geneticist Mary F. Lyon suggested that, to ensure correct
gene dosage compensation between male and female mammals, one out of the two
X chromosomes was randomly and permanently inactivated during embryonic and
adult life (Lyon 1961). This hypothesis, as reported by Lyon herself (Lyon 2002),
was based on two main observations: (i) one X chromosome is sufficient for nor-
mal (female) mouse development; and (ii) mice heterozygous for some alleles at
the X-linked coat color loci mottled (Atp7aMo) or dappled (Atp7aMo-dp) show a
variegated effect in heterozygotes, with a pattern of mottling resembling that “seen
in somatic mosaics”. Lyon’s hypothesis has been validated in a great number of
mammalian species, and the mechanism of inactivation has been progressively
elucidated at the molecular level.3
To explain how X-chromosome inactivation works, we have selected two exam-
ples: the first one is common and refers to tortoiseshell and calico female cats,
while the second is historical and refers to glucose-6-phosphate dehydrogenase
deficiency in women. The choice of these two examples may appear paradoxical
in a book dedicated to the mouse, but it has the great advantage of being didactic.
6.2.1.1 Calico Cats and G6PD-Deficient Women
In the cat, the Orange locus is X-linked, it has two alleles: black Ob and orange
Oo (or o), and no homolog on the Y chromosome. In male cats this locus deter-
mines two phenotypes: black (Ob/Y) or orange (Oo/Y), depending on the allele
carried by the X chromosome. In females there are three genotypes: Ob/Ob, Oo/Oo,
and Ob/Oo and also three phenotypes: black (Ob/Ob), orange (Oo/Oo), and a third
phenotype called tortoiseshell, which is observed in the heterozygous Ob/Oo. This
phenotype is clearly different from the uniform phenotype we would expect to
get for a cat heterozygous for an autosomal gene involved in the determinism of
coat color and exhibiting the classical dominant/recessive or semi-dominant allelic
interactions. Here, in contrast, the phenotype suggests that the two alleles, Oo and
Ob, are expressed independently and exclusively, rather than simultaneously in the
pigment-forming cells (the melanocytes). In other words, the fur of each female
cat appears as a mixture of hairs in which the individual melanocytes express
either one or the other of two different alleles at the X-linked Orange locus. This
is a clear-cut and classical example of the functional inactivation of one of the two
X chromosomes in female mammals (Fig. 6.1).
3 On July 22, 2011, at the occasion of the 50 Years of X-Inactivation Conference held in Oxford,
the Lyon hypothesis became the Lyon law.
(a) (b) (c)
Fig. 6.1 Calico cats and dappled mice. a The figure represents a female cat with a typical
“three-color” coat. Cats with such a coat color are called calico and are heterozygous for two
different alleles at the X-linked Orange (O) locus: black Ob and orange Oo. The spots are either
black or orange depending on the active X chromosome in the melanocytes. The white areas rep-
resent the unpigmented background and are due to a recessive autosomal spotting allele, called
piebald. This allele, extremely common in the cat, makes the (orange or black) spots encoded
by the Ob or Oo alleles even more visible (Courtesy of Dr. Abitbol, Alfort Veterinary School,
France). b The diagram represents three contiguous clones of melanocytes, derived from inde-
pendent stem cells in which a different X chromosome is inactivated. Since X inactivation occurs
early in development and is irreversible, many of the observed spots in the adult cat represent a
cluster of cells derived from the same stem cell. c The figure represents a female mouse heterozy-
gous for the Atp7aMo-dp (dappled) allele. Mutations at this X-linked locus are common and affect
copper metabolism (Courtesy of Dr. Eppig, The Jackson Laboratory Bar Harbor, Maine, USA)
Looking at the fur of different female cats with a similar Ob/Oo genetic consti-
tution, one may also note that the X chromosome that is inactivated in the melano-
cytes results from a random process because there is no specific pattern for the
distribution of the orange or black pigment, while the proportion of orange/black
fur remains close to 50 %. Also, it seems clear that once an X chromosome is inac-
tivated, this status persists in the daughter cells, resulting in the appearance of a
mosaic female made up of a mixture of cells, with one or the other X chromosome
actively producing either one of the two alternative gene products at the Orange
locus.4 Since, as we shall discuss later, X inactivation occurs quite early in devel-
opment, patches of cells with a similar pattern of X inactivation can become quite
large and are easily seen on the female’s coat. Some Ob/Oo female cats have an
even more spectacular coat color pattern when, by chance, they also carry an auto-
somal spotting allele (for example piebald), because this allele makes the orange
and black fur patches even more distinct on an otherwise white background.
Female cats with this coat color pattern are called calico.
Another observation that illustrates well the consequences of X inactivation at
the phenotypic level was published in 1962, by Ernst Beutler (Beutler et al. 1962),
4 The term "mosaic" is appropriately used in this context (see Chap. 2) because all the cells in
a female organism derive from the same egg and have the same genetic makeup at the Orange
locus. The difference in gene (or allele) expression depends upon the active/inactive status of one
of the two Xs. This results from an epigenetic mechanism, but not from a difference at the DNA
or chromosome level.
a few months after the publication of Lyon’s theory, and refers to the human
genetic deficiency in glucose-6-phosphate dehydrogenase (G6PD). To explain
their observation concerning the kinetics of dehydrogenation of glutathione (GSH)
by the enzyme G6PD from the erythrocytes of heterozygous human females,
Beutler and colleagues came to the conclusion that two populations of erythro-
cytes co-existed in females heterozygous for the X-linked deficiency (G6PDX
gene) rather than a single one, as would have been the case for enzymes encoded
by autosomal genes. Once more, the situation appeared to be the consequence of
monoallelic and independent expression of G6PD in the individual red cells of the
heterozygous patients.
Many more examples of mosaicism have been reported in female mammals,
including humans, to illustrate this point. The so-called Barr body, which was
observed and reported years ago, even before Lyon’s hypothesis, as a darkly stained
dot in the nucleus of cells prepared from oral swabs, represents a heteropycnotic X
chromosome. Karyotypes with X-chromosome aneuploidy (monosomics or XO, tri-
somics XXX or XXY male patients) display a number of Barr bodies that always
contain one less than the total number of X chromosomes in the karyotype, indicating
that there is a biological mechanism that somehow “counts” the total number of X
chromosomes in mammalian cells in addition to the mechanism inducing inactivation
of all X chromosomes but one.
6.2.1.2 X Inactivation is a Two-Step Process that Occurs Early

During Embryonic Life
The mechanism and precise timing of X-chromosome inactivation (XCI) were

debated for a few years after the initial publications of Lyon’s hypothesis. Nowadays
it is established that, in the mouse, two different forms of X-chromosome inactiva-
tion occur successively during early female embryogenesis. The first is an imprinted
or selective inactivation, which starts at the 4–8-cell (morula) stage and affects only
the paternal X chromosome (Xp).5 By embryonic day 6.5 (i.e., when gastrulation
begins), the Xp is reactivated in the cells that will give rise to the embryo proper,
then the classical Xp or Xm random inactivation ensues that will be retained as such
for the rest of the organism’s life (Morey and Avner 2011; Pollex and Heard 2012).6
In contrast to the tissues of the embryo proper, the imprinted or selective Xp inacti-
vation is maintained in the extra-embryonic tissues (placenta) for the rest of
gestation.
5 This Xp-specific inactivation is consistent with the observation that, at the pachytene step of
male meiosis, the Xp is condensed with the Y chromosome in an inactive XY body while, at the
same pachytene stage of female meiosis, the two X chromosomes are visible and form a normal
bivalent.
6 Unlike in eutherian mammals, the imprinted X inactivation persists in all cells of protherian
p
mammals (marsupials) including in the cells of the embryo proper.
As we already mentioned, the randomness of X-inactivation in the embryonic

and adult tissues is faithfully translated at the phenotypic level. For example, when
looking at the external appearance of adult calico cats one can observe that in most
instances approximately 50 % of their spots are orange (Oo/–) while the other
50 % are black (−/Ob). This randomness was also demonstrated more rigorously
by Davidson and colleagues (Davidson et al. 1963), who derived clones of epithe-
lial cells from female patients heterozygous for two different forms of the enzyme
glucose-6-phosphate dehydrogenase (G6PDA/G6PDB). In their experiment,
the authors found that of 14 clones of cells derived from the same heterozygous
patient, seven showed the A form of the enzyme while the other seven showed the
B form, and none contained both the A and B forms.
6.2.1.3 X-Inactivation Is Complete … or Nearly so
X inactivation is thought to be highly stable in somatic cells and does not revert
in the cells of the developing embryos, after implantation or in the cells of adult
females. However, it has been reported that a few genes on the inactivated X chro-
mosome could reactivate at low levels during aging. This is obviously a conse-
quence of some relapse in the X-inactivation process, but remains marginal and
concerns only a minority of the X-inactivated genes.
The situation is rather different with another limited set of genes, which are on
the mouse X chromosome and escape X inactivation completely. Most of these
“escapees” map to the pseudo-autosomal region (PAR), which means that they
have a homolog on the Y chromosome. The pseudo-autosomal regions on the X
and Y chromosomes pair and recombine during meiosis, (almost) as if they were
autosomal, and it makes sense to believe that this is probably the reason why they
are not inactivated: after all, there is no reason to apply any form of dosage com-
pensation to these genes. Steroid sulfatase (Sts) is the best known example of these
genes mapping to the PAR; mice homozygous for a deficient allele (Sts–/Sts–) have
been reported as a model for a common neurodevelopmental disorder in humans,
attention deficit hyperactivity disorder (ADHD) (Trent et al. 2011). However,
unexpectedly, the same Sts–/Sts– mice are not a model for the human X-linked
recessive disease ichthyosis, although human patients appear to be affected on the
orthologous gene STS.
Besides the genes mapping to the PAR, some other genes mapping to the
X chromosome also escape inactivation and are found to be transcribed from
the inactive X chromosome. Most of these genes have (or had) a homolog on the
Y chromosome or elsewhere in the genome, but this homolog is no longer func-
tional. They are orphan genes and probably do not encode any functional pro-
teins. The reason why these genes escape inactivation is unclear, but this does not
seem to be a problem since XO females appear to be normal though sub-fertile. In
contrast, in humans, where many more genes escape X inactivation, XO females
present a severe phenotype known as Turner syndrome, which is probably due to
these escapee genes, both in the PAR and elsewhere on the X chromosome.
6.2.2 The Mechanisms Controlling X-Chromosome

Inactivation
6.2.2.1 Characterization of an X-Inactivation Center (XIC)
Elucidating the mechanisms leading to X-chromosome inactivation consists of

understanding how two genetically identical and transcriptionally active X chro-
mosomes, that lie within the same nucleus, can be differentially treated in such
a way that one of them remains active while the other is inactivated. The first
important observation in this matter was that all inactivated genes are on the same
chromosome, while all active alleles are on the other. To interpret this observation,
scientists hypothesized that the inactivation was essentially a chromosomal issue
and that a master switch, controlling inactivation, might exist somewhere on the X
chromosome from which inactivation starts and spreads along the rest of the chro-
mosome. The identification of this master switch or inactivation center (XIC) was
achieved in several laboratories in the mid-1990s, using strategies that are com-
mon in genetics, consisting of the demonstration of the physical association of a
short chromosomal segment with its potential to inactivate the flanking regions of
a given chromosome. Studying the consequences of various reciprocal transloca-
tions and deletions involving the X chromosome and mouse autosomes allowed
the demarcation of a region of chromosome X with these properties. Confirmation
of these observations came from experiments of transgenesis with cloned DNAs
of various size followed by analysis of the consequences of the transgene on the
flanking regions.
The extent of the region enclosing the inactivation center has been defined by
studying X-chromosome deletions and by performing transgenesis in embryonic
stem cells. Both experiments have permitted the characterization and delimitation
of a region spanning a few hundred kilobases. Female embryonic stem cells (ES
cells) have been invaluable tools for studying the XIC because these cells have
their two X chromosomes active when undifferentiated, while X inactivation pro-
ceeds, as in embryos, when they start to differentiate in vitro.
A second discovery was that the XIC contains a gene, called Xist (for
X-inactive-specific-transcript) that is transcribed into a non-coding RNA
expressed only from the inactive X chromosome (Brown 1991). The Xist RNA
was found to coat the inactive X chromosome in cis and to correlate with the onset
of X inactivation.
Although the properties and localization of Xist RNA strongly suggest that it
should be a key element in the X-inactivation process, further experimental evi-
dence was required to show that this locus is necessary for inactivation of the X
chromosome. This was shown through the use of various deletions of the Xist gene
that prevent production of full-sized Xist RNA (Penny et al. 1996; Marahrens et al.
1997). In these cases the chromosome bearing the deletion is not inactivated, indi-
cating that a complete, intact Xist gene is required, in cis, for inactivation to take
place.
Further support for a critical role of XIST comes from experiments in which
the Xist-cDNA, under an inducible promoter, was inserted into an autosome in
male ES cells. Induction of Xist-RNA provoked coating of the chromosome in
cis and repression of gene transcription for this autosome. Although other fac-
tors are probably also involved, these experiments demonstrated that Xist-RNA
is a key trigger for chromosome-wide silencing and that it may do so by binding
to the chromosome from which it is expressed. These experiments also demon-
strated, along with previous studies from X-autosome translocations, that specific
X-linked sequences are not required for Xist-RNA to coat a chromosome.
The XIC candidate region harbors four non-protein-coding genes, Xist, Tsix,
Jpx, and Ftx, which are involved in X-inactivation. The XIC also contains binding
sites for both known and unknown regulatory proteins.
The Xist transcript has no significant open reading frame and the product
remains in the nucleus, coating the inactive X chromosome. This suggests that Xist
is among those loci that produce a functional RNA molecule that is never trans-
lated into a protein (a non-coding RNA—see Chap. 5). Xist expression is detected
early in pre-implantation development, often from both X chromosomes, just prior
to X inactivation at the 4–8-cell stage (Okamoto et al. 2004; Patrat et al. 2009).
In the mouse, the paternal X chromosome is initially subject to X inactivation as
a result of an imprint in the gametes that leads to the paternal nonrandom inacti-
vation found in extra-embryonic cells. Later, in the inner cell mass, Xist is acti-
vated from one of the two X chromosomes in cells that will form the epiblast.
Random X inactivation follows and Xist transcription on the active X chromosome
is silenced. Recent evidence suggests that XIST regulation involves a combination
of cis-elements including antisense transcription as well as trans-acting factors
that are tightly integrated with the pluripotent and stem cell proteins (for a recent
review, see Augui et al. 2011).
The inactive X chromosome has been associated with several putative epige-
netic marks (or non-sequence-based heritable changes) including DNA methyla-
tion, histone modifications, and Polycomb group complexes. DNA methylation is
probably the best studied to date. Methylation of the cytosine base occurs enzy-
matically after DNA synthesis, and in mammals is restricted to the dinucleotide
5′-CpG-3′ (CpG). About 7 % of CpGs are present at relatively high density in clus-
ters called CpG islands, which are usually located at the 5′ ends of genes. The
remaining CpGs are dispersed throughout the genome, usually as singlets. Most
CpG islands are unmethylated, but those near inactivated genes on the X chro-
mosome, and those near some imprinted genes on autosomes, are methylated.
Methylated CpG islands repress transcription, and most silent genes on the inactive
X chromosome have such methylated CpG islands in normal cells. It is believed
that DNA methylation acts in a synergistic way with other chromatin modifica-
tions to lock in the inactive state in a highly stable fashion in somatic cells.
The mouse has played a fundamental role in our understanding of the mech-
anisms of gene regulation and expression underlying processes such as X inac-
tivation, as it has rendered observations and experiments possible that were not
possible in any other species up until quite recently.
6.2.2.2 X-Inactivation Skewing
Most women heterozygous for the X-linked mutation DMD (Duchenne muscular
dystrophy—DMD+/DMDmut) remain completely asymptomatic during their life
and are generally unaware that they are carriers until they give birth to an affected
son. This situation is common to many other pathologies where females are het-
erozygous for a deleterious X-linked mutation. The lack of overt phenotype or
only mild phenotype in females is generally explained by considering that around
50% of their cells express the normal allele from the active (transcribed) X chro-
mosome, with the mutated allele being on the silent, inactive X chromosome.
This explains why these carrier females are protected from the clinical effects of
X-linked mutations such as in the case of the DMD gene.
However, such situations of intercellular complementation are far from being
the rule, and after careful analysis of other X-linked human pathologies, it has
been observed that X inactivation may occur randomly at first (i.e., 50 % X+/50 %
Xmut) but, with time, the cells in which the X chromosome carries an allele with
deleterious effects (Xmut) are counter-selected more or less efficiently, depending
on the case, giving the impression of X-inactivation skewing. This is the case in a
form of X-linked mental retardation (XLMR), ATR-X syndrome, which is caused
by mutations in a ubiquitously expressed, chromatin-associated protein and in
which phenotypically normal female carriers have highly skewed X-chromosome
inactivation of the X chromosome that carries the mutant allele. Interestingly, the
homologous disease has been modeled in mice heterozygous for a null Atrx orthol-
ogous allele, and it has been observed that X-chromosome inactivation is bal-
anced early in embryogenesis but becomes skewed over the course of development
because of a strong selection favoring cells expressing the Atrx wild-type allele
(Garrick et al. 2006).
Selection against the cell lineage that carries the mutant allele on the active
X chromosome appears logical, especially if it is the price to pay for surviving
in better conditions, but it is not the rule. For example, unfavorable skewing of
X inactivation has been reported in young females suffering from hemophilia B
where the paternal X chromosome, carrying a normal copy of the FIX gene, was
predominantly the inactive one, leading to the phenotypic expression of hemo-
philia B in these young girls (Espinós et al. 2000).
X-inactivation skewing is sometimes influenced by chromosomal rearrange-
ments. An excellent example of such skewed X inactivation is provided by the
T(X;16)16H (or Searle’s) reciprocal translocation in the mouse. In this translo-
cation, a piece of the telomeric region of chromosome X is attached to the cen-
tromeric part of chromosome 16, and vice versa. As expected, the piece of
X chromosome that carries the X-inactivation center is inactivated, but inactiva-
tion spreads over the breakpoint and concerns all the genes on the piece of chro-
mosome 16 that is attached to the broken X, resulting in a deleterious functional
haploidy. Conversely, all the X-linked genes on the non-inactivated piece of
X chromosome are expressed, where they should not be. In fact, for the female
mice heterozygous for Searle’s translocation, the only way to survive is to
inactivate their normal X chromosome. It is likely that such a situation, which is

extreme in the case of T16H mice, probably exists with other mutations, although
sometimes with a less dramatic effect—leading to a less extreme skewing.
At this point, it is interesting to note that, 40 years ago, Cattanach had already
reported skewed X inactivation in F1 hybrid mice. He observed that, depending
on the cross and strains involved, the percentage of inactivated X chromo-
somes was different for X chromosomes of different genetic origins. Cattanach
quantified his observations by defining four alleles at a locus that he desig-
nated as the X-inactivation controlling element (symbol Xce) with four alleles:
Xcea < Xceb < Xcec < Xced in order of the tendency of the X chromosome to
remain active (Simmler et al. 1993; Thorvaldsen et al. 2012). As of today, the
identity of the Xce locus remains unknown, although its genetic localization has
been much refined, indicating that multiple elements on the X chromosome con-
tribute to the Xce and that some of these may lie within the X-inactivation center
(see below).
6.3 Parental Imprinting of Autosomal Genes
As discussed above, X-chromosome inactivation is an original and sophisticated

method which has evolved to equilibrate the transcriptional activity of the genes
on this chromosome between male and female mammals. Being epigenetic by
nature, X-chromosome inactivation does not alter the DNA sequence and is com-
pletely erased when the primary germ cells enter gametogenesis. However, the
X chromosome is not the only segment of the mammalian genome that can be
modified epigenetically, as we will now discuss.
6.3.1 Evidence of Genomic Imprinting in the Mouse
6.3.1.1 The Unusual Behavior of the Hairpin-Tail Allele at the T-Locus
The T-locus of the mouse (brachyury T-Chr 17) has several mutant alleles; some
are dominant while others, mostly found in wild mice, are recessive. Dominant
alleles have an effect on the notochord derivatives and are characterized, when
heterozygous, by a shortened tail with extensive variation in expressivity. T/T
homozygotes die during embryonic development, at about mid-gestation.
Hairpin-tail (Thp) is unique in the allelic series at the T locus, in the sense that
the phenotype of the heterozygote offspring depends upon the origin of the mutant
allele. When Thp is inherited from a Thp/+ male mated to a wild-type female
(+/+), the offspring are all viable and about 50 % of them have a shortened tail, as
expected. However, when the cross is set up the other way around (i.e., between a
6.3 Parental Imprinting of Autosomal Genes 197
(a) (b) (c)
Fig. 6.2 Inheritance of the hairpin-tail mutation. 17-day-old embryos collected in the uterus of

a Thp*/+ mother mated with a Thp/+ male: +/+; Thp/+; Thp*/+. Embryo (a) is normal. In embryo
(b), the Thp allele is of paternal origin, while it is of maternal origin* in embryo (c). Offspring
with a genotype Thp*/Thp die shortly after implantation and are not represented in the figure. The
tail shortening effect of Thp is not obvious in this figure, especially in embryo (b), probably for
lack of expressivity (from Johnson 1974a). Thp has been characterized as a large deletion on chro-
mosome 17, which includes the (imprinted) gene encoding IGF2R
wild type +/+ male and a Thp/+ mutant female), the progenies are reduced (nearly
halved) and no mutant phenotypes are observed: they all die in utero at a relatively
late stage of gestation.7 This peculiarity of the hairpin-tail allele, which was first
reported in 1974 as “a case of post-reductional gene action in the mouse egg”
(Johnson 1974a, b), is not a simple maternal effect, since Thp/+ × Thp/+ matings
produce two types of Thp/+ heterozygous embryos: one is viable ab utero with a
short tail, while the other is unviable (Fig. 6.2).
Nowadays, we know that the Thp allele is associated with a deletion in the cen-
tromeric region of chromosome 17 (T-associated maternal effect—Tme). We will
later discuss the molecular nature of this structural change and its consequences, but
at this stage and from a historical point of view, it is important to note that the iden-
tification of this allele at the T locus was quite fortunate. If, by chance, the original
Thp mutant allele had occurred in a female germ cell it would have been lost and the
discovery of a “post-reductional gene action” would have been delayed.
6.3.1.2 The Fate of Embryos Resulting from the in Vitro

Re-Association of Pronuclei
For many years, and for technical reasons, it was impossible to grow mouse eggs
in vitro, from the one-cell stage up to the stage of expanded blastocyst. Once this
difficulty was overcome, one of the first experiments undertaken by embryologists
was to try and reconstruct artificially diploid embryos by re-associating pronuclei
7 Some exceptions have been reported, but they are extremely rare and fall well below the
expected 50 %.
from embryos at the one-cell stage in different combinations. The rationale for
undertaking this sort of experiment was to check whether a given haploid genome
could merge with any other haploid genome to result in a viable mouse organism.
Such experiments were completed in the early 1980s, in particular in England and
in the USA, and led to the unambiguous conclusion that the development to term
of reconstructed pseudo-diploid embryos requires the association of a maternally
derived and a paternally derived pronucleus. Any other association (i.e., two male
pronuclei or two female pronuclei) appeared lethal a few days after implantation
(Barton et al. 1984; McGrath and Solter 1984; Surani et al. 1984) (Fig. 6.3).
The result of these experiments suggested that the haploid genome in a pronu-
cleus was marked in a specific manner according to its parental origin, and that
the male and female contributions were not functionally equivalent. This mark has
become known as the parental genomic imprint or simply genomic imprinting.
Other experiments, focusing on the study of the developmental potentialities
of cells derived from either gynogenetic embryos (resulting from the association
of two female pronuclei) or androgenetic embryos (resulting from the association
of two male pronuclei), merged together or associated independently with cells of
a normal embryo in a single chimeric organism, indicated that androgenetic cells
preferentially contribute to the formation of extra-embryonic tissues while gyno-
genetic cells, in contrast, preferentially contribute to the formation of embryonic
2n
Mat.
nm Pat.
Viable
np
Mat.
nm Mat. Pat.
Non-viable
nm Pat.
np np
Non-viable
2n 2n
Fig. 6.3 The fate of reconstructed, pseudo-diploid embryos. The development to term of recon-

structed, pseudo-diploid embryos requires the association of maternally derived and paternally-
derived pronuclei. Reconstructed embryos with either two maternal or two paternal haploid sets
are unviable
tissues. Another conclusion that can be drawn from these experiments is that par-
thenogenetic development is strongly hindered in the mouse although it occurs,
occasionally, in other vertebrate species (it is common in fish and some reptiles,
and has also been reported in birds).
6.3.1.3 The Fate of Embryos Resulting from Uniparental Disomies
The conclusions of the experiments reported above have been confirmed and
refined by another totally different kind of experiments, achieved mostly in
England, in the mid-80 s at the Harwell MRC Laboratory, by B.M. Cattanach,
C.V. Beechey, J. Peters, and A.G. Searle. These experiments made use of two
types of chromosomal rearrangements (Robertsonian translocations and reciprocal
translocations) that were available in the large genetic repository at Harwell.
As described in Chap. 3, devoted to cytogenetics, mice whose genetic con-
stitution consists of a single Robertsonian translocation plus the two acrocentric
chromosomes whose arms are homologous to the arms of the Robertsonian translo-
cation are perfectly normal since they have a balanced karyotype although reduced
by one centromere. Such mice, however, often produce a high percentage of unbal-
anced gametes—i.e., gametes with either one extra (acrocentric) chromosome or,
reciprocally, with one missing (acrocentric) chromosome. As we already discussed,
these unbalanced gametes, resulting from meiotic non-disjunction, yield trisomic
or monosomic embryos when merging with a normal gamete (Fig. 6.4).
In the mouse, most trisomic and all monosomic embryos die in utero at a stage
of development that varies with the chromosome involved.8 However, when by
chance an unbalanced gamete with, for example, one missing acrocentric chromo-
some combines with a gamete with one extra chromosome of the same pair, this
results in an embryo with a [(n – 1) + (n + 1)] = 2n (euploid) chromosome com-
plement, regardless of whether the two chromosomes of the pair in question were
contributed by one and the same parent or not. Such embryos, with the two chro-
mosomes of a given pair originating from the same parent, are said to result from
uniparental disomies (UpDi).9
The observations by Cattanach and colleagues, made on the progenies of mice
with a variety of different chromosomal translocations, were that viable and nor-
mal embryos resulting from complementary double non-disjunctions (UpDis) were
(i) rather rare and (ii) very much dependent on the chromosome pair involved. In
fact, in many instances, dramatic effects on development, including enhanced or
retarded growth and sometimes lethality in utero, could be observed in the prog-
enies (Cattanach and Kirk 1985; Cattanach 1986). Cattanach demonstrated that
only a few chromosomes could be inherited as uniparental disomies, still leading
8 Ts19 is the only trisomy viable ab utero but only a few mice survive after 10 days.
9 Uniparental disomies can be of maternal (MatUpDi) or paternal (PatUpDi) origin.
(a)
(b)
 Fig. 6.4 Double non-disjunction in mice heterozygous for translocations. In mice heterozy-

gous for translocation the meiotic process often results in the production of a high percentage
of aneuploid gametes due to the abnormal segregation of the chromosomes. a Represents the
disjunction of chromosomes 11 or 13 in mice heterozygous for the Robertsonian translocation
Rb(11.13)4Bnr. When a gamete with an extra chromosome arm merges with a normal gamete,
this results in a trisomic embryo (see Chap. 3 for explanations). However, when the same ane-
uploid gamete merges with a complementary unbalanced aneuploid gamete missing the same
chromosome arm, this recreate a normal (2n) karyotype with the exception that, in this case,
the same parent provides the two chromosomes of a given pair and the other parent none of the
gamete of the pair in question. In this case, the embryo is said to result from uniparental disomy
(UpDi). Such embryos are viable only when the two elements of a chromosome pair involved in
the UpDi are not imprinted. In the original experiments by Cattanach and colleagues (see text),
identification of the parental origin of the chromosomes was done by using the phenotypic genetic
markers vestigial tail (vt) for chromosome 13 and dominant, wavy coat (Rewc) for chromosome
11. Nowadays, molecular markers like SNPs or microsatellites would rapidly distinguish the ori-
gin of the different chromosomes in such a cross. In the cross represented here, Cattanach and
colleagues observed that the offspring resulting from MatUpDi11 (maternal uniparental dis-
omy of chromosome 11) were smaller than their normal sibs while the offspring resulting from
PatUpDi11 were bigger. This was a demonstration of the parental imprinting of (at least a seg-
ment of) the chromosome 11. Doted lines show the three different segregations of chromosome
11 including non-disjunctions. (Adapted from Cattanach’s original drawings). b Represents
the disjunction of chromosomes 2 and 8 in mice heterozygous for the reciprocal translocation
T(2;8)26H. These mice produce a variety of gametes (a–a′, b–b′, c–c′) with a variety of chromo-
somal segment association, depending on the type of segregation (see Chap. 3 for explanations).
Some of these gametes carry duplicated segments (for example, b′ and c for Chr 2; b and c′ for
Chr 8) while some others carry segmental deletions (for example, b and c′ for Chr 2; b′ and c for
Chr 8). The gametes with either a segmental deletion or a segmental duplication (b–b′ and c–c′
on the picture) produce unviable offspring when they merge with a normal gamete—only gametes
of the a–a′ type produce embryos with a balanced (viable) karyotype. When the cross is between
two progenitors heterozygous for the same reciprocal translocation, there are rare cases where
two gametes resulting from complementary non-disjunctions fuse together, restoring a balanced
karyotype (i.e., when the duplications complement the deficiencies). These offspring, resulting
from complementary uniparental partial disomies, are rare but they can be identified if genetic
(or molecular) markers segregate in the cross, labeling the various chromosome arms. A major
impediment to this kind of experiment is that many reciprocal translocations, when heterozygous,
are sterile in one sex or the other. The production of neonates resulting from complementary dou-
ble non-disjunction is also laborious because, unlike for Robertsonian translocations, mice het-
erozygous for reciprocal translocations produce small-sized progenies due to embryonic lethality
(semi-sterility; see Chap. 3 for explanations). (Adapted from Cattanach’s original drawings)
to normal healthy offspring. In all other cases, anomalies were observed, generally
associated with difference in body size.
The general conclusions of these experiments are that normal development to
term of a mouse embryo requires that some specific chromosomes be inherited
from the mother or from the father, and sometimes from both the father and the
mother (for example, chromosomes 7 or 11). This again suggested that a parent-
of-origin-specific expression exists, at least for some genes, and for one and/or the
other of the two parental chromosome homologs.
In addition to this series of experiments (made with mice heterozygous for
Robertsonian translocations and concerning intact, complete acrocentric chromo-
somes), scientists at Harwell used another approach to screen the whole mouse
genome for specific imprinted regions. The strategy made use of an assortment
of reciprocal translocations, a very common type of chromosomal rearrange-
ment, resulting from the reciprocal exchange of chromosome arms between two
non-homologous chromosomes. Here again, mice heterozygous for reciprocal
translocations produce a variety of aneuploid gametes and, by inter-crossing such
mice, it is possible to obtain normal, 2n embryos whose genomes result from the
fusion of complementary unbalanced gametes. These experiments were arduous
and required many crosses because, as we explained in Chap. 3, the progenies of
mice heterozygous for a reciprocal translocation are much reduced in number.
After carefully screening hundreds of progenies, the scientists at Harwell could
observe the presence (or suspect the absence) of conceptuses resulting from unipa-
rental duplication/deficiency for a particular chromosomal region and, finally, they
could summarize their observations by drawing a chromosomal map indicating the
maternally or paternally imprinted chromosomal regions (See Fig. 6.5).
Fig. 6.5 The Harwell map of mouse imprinted genes and regions. Some chromosomal seg-
ments (outlined on the map) must be inherited from the male parent or from the female parent
or, sometimes, simultaneously from both the male and the female parents. This is a consequence
of genomic imprinting, which occurs during the process of gamete formation, and results in
the functional inactivation of some specific genes encoding proteins or RNAs. The size of the
imprinted segments has been estimated based on experimental data (see references), and in most
instances it is excessively large compared to the actual size of the cluster of imprinted genes
(1 Mb on average). Most (although not all) imprinted genes in the mouse are also imprinted in
human and rat species. The establishment of this map has required an enormous investment in
terms of crosses, and was possible only in a few laboratories (like MRC Harwell) where a large
repository of translocations of all kinds existed. This map is now being progressively refined by
direct analysis of the transcripts
6.3.2 Characterization of the Imprinted Regions in the

Mouse
6.3.2.1 Imprinted Regions Harbor Genes that are Transcribed

Exclusively from One Allele
The first imprinted region that was (partially) characterized at the molecular level
was precisely the one that was discovered first and is associated with the “hairpin-
tail phenotype”. The characterization of the region in question was achieved by
making a fine genetic map of the chromosome 17 proximal segment and perform-
ing a quantitative assessment of the transcription products of the genes mapping to
that region. Providentially, another allele at the same T/t locus (tLub2) was discov-
ered, which is recessive and associated with similar developmental defects as Thp.
When the chromosome carrying the tLub2 mutation is inherited from the mother,
embryos heterozygous for this mutation are severely affected by edema and death
generally occurs between days 15–17 of gestation, just as for Thp/+ mice born to
a Thp/+ mother (Winking and Silver 1984). Genetic and molecular analyses indi-
cated that Thp and tLub2 were overlapping deletions of chromosome 17, with Thp
spanning a distance of about ~7 Mb and tLub2 only ~0.8 Mb.
The tLub2 haplotype has been characterized in detail, and several genes (Chr
17 cen—Plg, Igf2r, Tcp1, Sod2) have been identified within the deleted region.
Remarkably, among all these genes Igf2r, the gene encoding the insulin-like
growth factor type-2 receptor (IGF2R) appeared to be transcribed exclusively from
the maternal allele, while the other genes were transcribed from both the paternal
and maternal alleles.
Considered together, these observations explain all the observed phenotypes;
in short, since Igf2r is deleted in the Thp and tLub2 chromosomes, and given that
Igf2r is not transcribed from the paternal allele, any embryo with a ThpM/+ P or
tLub2M/+ P constitution has no functional IGF2R and accordingly cannot survive
to birth. Embryos with the reciprocal genotype (i.e., ThpP/+ M or tLub2P/+ M) are
normal since the maternal copy is intact and transcribed, exactly as in normal
embryos. For all other genes, hemizygous embryos survive normally as they gen-
erally do with most other autosomal genes (Barlow et al. 1991).
Igf2r encodes a trans-membrane receptor protein whose function is to transport
mannose-6-phosphate tagged proteins and insulin-like growth factor 2 (IGF2) to
lysosomes; it is an essential protein for the completion of a normal gestation. The
conclusions drawn from these observations have been validated by studying, inde-
pendently, the fate of embryos inheriting a non-functional copy (i.e., a knockout
allele—see Chap. 8) of the Igf2r gene from their mother or from their father.
In a series of experiments performed two years later, i.e., once the detailed
mechanisms generating imprinting were unraveled, scientists created a non-
imprinted allele of Igf2r (designated R2Delta) by deleting an essential element
repressing the paternal allele in mouse ES cells (actually the ICE—see below).
Maternal inheritance of this R2Delta allele had no phenotype, as expected.
However, paternal inheritance resulted in biallelic expression of Igf2r. In this case,
embryos were affected by a 20 % reduction in body weight late in embryonic

development that persisted to adulthood. Paternal inheritance of the functional
R2Delta allele rescued the lethality of a maternally inherited Igf2r null allele and a
maternally inherited Tme (T-associated maternal effect) mutation. These data sug-
gested that one of the biological reasons for imprinting Igf2r is probably to trigger
an increase in body weight at birth. These data confirmed the importance of the
Igf2r gene in the Tme deletion phenotype (Wutz et al. 2001).
The second region that was recognized as imprinted, and characterized at the
molecular level, was the telomeric region of chromosome 7. This region was
identified by studying the progeny of an intercross between mice heterozygous
for the reciprocal translocation T(7;18)50H. Embryos with the maternal duplica-
tion and paternal deficiency of distal Chr 7 (MatDp7/PatDf7) are growth-retarded
and die around day 16 of gestation; the reciprocal maternal deficiency and pater-
nal duplication embryos (MatDf7/PatDp7) die at an unidentified but much earlier
stage. The imprinted region harbors, among others, the gene encoding insulin-like
growth factor 2 gene (Igf2), a gene functionally and physiologically related to the
gene encoding its receptor Igf2r (DeChiara et al. 1991).
IGF2 is a growth-promoting hormone acting during gestation and sharing struc-
tural similarities with insulin. Igf2 is imprinted differently from Igf2r since it is
transcribed exclusively from the paternal allele. The observations relative to the
growth retardation of the embryos resulting from chromosome 7 uniparental diso-
mies have been confirmed by studying the mice carrying a null (knockout) allele
of Igf2. As expected, non-complementation of the Igf – allele by the normal Igf2
allele was observed when the wild-type allele was inherited from the mother.
6.3.2.2 Making the Inventory of Imprinted Genes in the Mouse
Many genes have been progressively discovered in the various imprinted regions
identified by Harwell’s scientists, and a good proportion of these regions have now
been characterized at the molecular level. As indicated on the map (Fig. 6.5), there
are at least 15 and probably up to 25 imprinted regions spread over 16 different
autosomes and these regions are apparently distributed randomly, i.e., with no spe-
cific pattern. They are either telomeric or centromeric and harbor clusters of genes
(from 3 to 11) rather than single independent genes. Some geneticists think that
this clustering of the imprinted genes is probably not by chance, and may reflect
subordination to a common mechanism of inactivation. This conclusion, however,
should be reconsidered when a greater number of imprinted genes or regions are
identified in different mammalian species.
The genes mapping to the same imprinted cluster do not appear to be function-
ally related. Even more surprisingly, some genes in a given cluster are maternally
expressed while others are paternally expressed (for example, Igf2 and H19 on dis-
tal Chr 7). This is in good agreement with the original observation at Harwell that,
at least for some pairs of chromosomes, one element must be inherited from the
father and the other from the mother.
As we mentioned, the function of the genes mapping to the imprinted clusters
is not always fully characterized, and for some of them it may take some time
before we precisely determine all their functions. This is particularly true if we
consider that some of these genes, for example H19, do not encode proteins but
non-coding RNAs instead.
The analysis of the transmission of knockout (null) alleles, produced by in
vitro gene targeting in the mouse, will be of great help for the future identifica-
tion of imprinted regions or genes. It seems, however, that genes of this cat-
egory represent only a minority of the genes because if the wild-type alleles of
the genes that have been knocked out were imprinted, their uniparental trans-
mission to the progeny would be impossible or associated with some pathol-
ogy, and this would almost certainly have already been noticed by researchers.
The analysis of the transmission patterns of knockout alleles is indeed an effi-
cient way to screen for genomic imprinting in the mouse, and the occurrence
of any phenotypic alterations exclusively transmitted by one sex and not by the
other should trigger curiosity and call for further investigation. Similarly, iden-
tification of a new imprinted gene in humans (or any other mammalian species)
should be considered as an indication for a candidate in the homologous region
in the mouse.
As of today, the number of imprinted genes reported in the mouse is around
140. Studies of the total number of imprinted genes are currently being refined
by other methods (Yu et al. 2012). Sequencing the whole transcriptomes of inter-
specific mouse hybrids resulting from crosses in both directions (for example, a
female of a laboratory inbred strain × a Mus m. musculus male or vice versa),
and looking for tissue/cellular distribution of species-specific SNPs is a promis-
ing way of achieving the complete inventory of imprinted genes in the mouse
(Fig. 6.6).
Of the 140 genes that have been reported as being imprinted in the mouse, a
quite large proportion has also been found to be imprinted in humans, but excep-
tions exist. Igf2, for example, has been found to be imprinted in the human, rat,
and mouse species but the gene encoding the receptor for this molecule, Igf2r, is
imprinted in the rat and mouse species but not in humans (Weidman et al. 2006).
In addition to this observation, it is worth noting that, from interspecific com-
parisons that have been made, it seems that the degree of homology in terms of
imprinted genes parallels the phylogenetic distances. This is not so surprising
and, with a better knowledge of the imprinted genes across mammalian species,
it should be possible to learn more about their function. Already, by comparing
the known functions of the imprinted genes in the three above-mentioned species
(human, mouse, and rat), it is obvious that most of these genes code for growth
factors expressed during embryonic life either in the fetal membranes, the placenta
or in the embryo proper.
Fig. 6.6 Molecular identification of imprinted genes using SNPs in the cDNAs. One can eas-
ily check if the two alleles at a given locus are co-expressed in embryonic or adult tissues by
analyzing the SNP pattern of the transcribed RNAs. The figure represents part of the sequence
of the transcripts of the gene encoding β-hemoglobin (HBB) in the bone marrow cells of F1
mice heterozygous for a single, untranslated nucleotide polymorphism (a silent mutation) in
exon 2 of the gene. The figure shows that both alleles are transcribed, since one can recognize
the profile of a C/T SNP (arrow) in the sequence of the corresponding cDNA. If the gene
encoding β-hemoglobin chain (Hbb) was among the genes undergoing genomic imprinting,
one would have found a single transcript (either from the C or from the T allele) depending
on the direction of the cross. Sequencing the whole transcriptome of interspecifc F1 mice is
an efficient way of making the inventory of imprinted genes in a given species or in a given
tissue
6.3.3 What are the Molecular Mechanisms that Control

Genomic Imprinting?
6.3.3.1 DNA Methylation Modifies Transcriptional Activity
Understanding the biological mechanisms involved in the establishment and

maintenance of genomic imprinting has motivated a large number of experiments
carried out mainly in the mouse and using the most sophisticated techniques.
The results obtained have much clarified the situation, even if some aspects
require a closer look. We will summarize the state of knowledge as it stands now.
However, before doing this, it is important to note that the molecular mechanisms
in question had to comply a priori with some basic constraints. First, imprinting
may interact with the transcription process but in no way may it alter the DNA
sequence of the imprinted regions. Imprinting, as we discussed, is strictly epi-
genetic, which means that the information in the DNA sequence is not altered.
Second, the imprinted regions must be transmitted unchanged to the daughter cells
during the development of the embryo and in the adult to ensure the continua-
tion of imprinting, at least for some time, in the different cell lineages. Third, the
epigenetic alteration(s) must be initiated in the paternally or maternally inherited

chromosomes independently, and at a time when they are not in the same nucleus;
that is, during gametogenesis or immediately after fertilization, before the fusion
of the pronuclei. Finally, the parental imprint must be erasable (or reversible) in
order to be set differently when the allele goes into a gamete of the opposite sex
(Ferguson-Smith 2011).
One of the imprinted regions that has been the most extensively studied is,
again, the one that maps to the distal part of mouse chromosome 7. This region,
in fact, contains two contiguous clusters: one with four genes (cen–… H19–Igf2–
Igf2as–Ins2), encompassing around 1 Mb, and another one, more distal, harbor-
ing around 15 genes (around the Kcnq1 locus). Both clusters have a homolog
in human and rat, and the genes in question are equally imprinted in these two
species.
H19 encodes a 2.3-kb ncRNA that is highly preserved across mammalian
species, indicating that it presumably has an important function. Embryos hete-
rozygous for a maternally inherited knockout allele or homozygous for the H19
knockout allele exhibit increased placenta and body weight (Gabory et al. 2009).
Igf2 encodes a hormone that has similarity with insulin and is probably a major
fetal growth factor. Mice heterozygous for an Igf2 knockout allele (Igf2-), trans-
mitted through the male, exhibit pre- and post-natal growth retardation. In con-
trast, when the disrupted (null) allele is transmitted maternally, the heterozygous
offspring are phenotypically normal. Both H19 and Igf2 are widely expressed
during embryonic development, and then they are down-regulated in most adult
tissues.
Shortly after the characterization of the H19–Igf2 cluster and its complete
sequencing, it was demonstrated that imprinting of these two genes is concomi-
tant with the methylation of an imprinting control region (ICR) or differentially
methylated region (DMR), which is 2 kb long and inserted between the two genes.
Proper imprinting of H19 and Igf2 requires the ICR integrity because, when
this region is altered or deleted by genetic engineering, imprinting is abolished.
Similarly, proper imprinting of the H19–Igf2 cluster requires that the ICR be
methylated on the paternal allele and unmethylated on the maternal allele.
As already discussed concerning the mechanisms at work in the case of X inac-
tivation, DNA methylation is a biochemical process that consists of the addition of
a methyl (CH3) group at the C-5 position of cytosine, at specific sites known as
5′-CpG-3′ dinucleotides or CpG islands. When methylation occurs in the 5′ regula-
tory regions of many genes, this generally results in transcriptional silencing of
these genes. Experiments performed in the early 1990s demonstrated that DNA
methylation is probably a crucial step in determining imprinting in mammals,
since deficiency in DNA methyltransferase activity (for example, as a consequence
of a targeted null mutation at the Dnmt1 gene) impedes normal imprinting and the
homozygote mutant embryos die around day E9.5 (Li et al. 1993). Methylation is
stable and can be inherited through mitotic cell division in the differentiated tis-
sues. Methylation alters the spatial conformation of the DNA, making it more
compact and accordingly less accessible to DNA-binding proteins, but it does not
alter the sequence proper. Methylation is also reversible and accordingly complies
with the constraints mentioned above.10
The H19–Igf2 imprinted region of mouse chromosome 7 and its human
homolog on chromosome 11p15.5 have been extensively studied with the aim of
elucidating the mechanisms at work for imprinting establishment and maintenance
in mammals. Most of the results gathered in the mouse have been cross-validated
in humans, and vice versa. As mentioned, these results have revealed the existence
of ICRs and DMRs, as regulatory elements for imprinting of the gene cluster, and
have underlined the role of methylation of the CpG islands as previously observed
in plants. Methylation of these regions results in silencing or activation of the clus-
ter, depending on the initial status of the genes concerned (Ferguson-Smith et al.
1993; Constância et al. 1998; Reik et al. 2001; Reik and Walter 2001).
The DMRs are the main signature of imprinted genes. Some are called primary
or germline DMRs (such as the H19–Igf2 ICR or the Igf2r ICE), because they
acquire their differentially methylated status in the germline, and others are called
secondary or somatic DMRs and acquire their methylation after fertilization. In
the case of the H19–Igf2 locus, the insulator protein, called CTCF, binds only to
the unmethylated ICR and produces a boundary. This results in the interaction
of downstream enhancers with the H19 promoter but not with the Igf2 promoter
on the maternal allele. This was defined as the enhancer competition model and
explains the monoallelic expression of these genes.
6.3.3.2 Other Mechanisms Involved in the Control of Imprinting

The Role of ncRNAs
Analysis of several imprinted regions also revealed that some specific ncRNAs
are probably essential intermediate molecules for the establishment (and main-
tenance) of imprinting. This assumption was validated by observations made on
the imprinted Igf2r cluster on mouse chromosome 17. In this cluster, the ICE
(imprinting control element) acts as a promoter for a long ncRNA named Airn (for
antisense of Igf2r RNA non-coding) from the unmethylated paternal allele. When
Airn expression is abolished, the Igf2r imprint is removed, suggesting a mecha-
nism of transcription interference (Latos et al. 2012). This mechanism, however,
does not exist in humans where Igf2r is not imprinted.
10 Several assays have been designed to assess the methylation status of the genomic DNA.
One of the most popular consists of the initial treatment of DNA with sodium bisulfite, which
converts cytosine residues into uracil (U) or thymidine (T), but leaves 5-methylcytosine resi-
dues unaffected. Once treated with bisulfite the DNA can then be directly sequenced or digested
with restriction enzymes (like BstUI), which only cleave sites that were originally methylated
(CGCG) but not those that were originally unmethylated (TGTG). Combined bisulfite restriction
analysis (or COBRA) is a widespread technique allowing quantification of DNA methylation. It
has been extensively used in cancer research and epigenetics studies.
A recent report indicated that within each cluster all imprinted genes show
concordant parent-of-origin-specific gene expression except for the ncRNAs
that show expression from the opposite parental allele. Such strict reciprocal
parent-specific expression seen between mRNAs and imprinted macro ncRNAs
strongly indicates that ncRNAs regulate imprinting in such clusters (Saxena
and Carninci 2011). This has also been shown for the Kcnq1 locus, in which
the Kcnq1ot1 long ncRNA is required to maintain DNA methylation and tran-
scriptional gene silencing of the adjacent imprinted genes (Mohammad et al.
2012).
The Role of Histones
Histone modifications have also been considered as an important mechanism in
establishing the imprint either directly or indirectly, and in many cases the alleles
that display DNA methylation also carry histone marks associated with inactivity.
Many points still remain to be clarified concerning the mechanisms of establish-
ment and maintenance of imprinting in mammals (Chen and Dent 2014).
6.3.3.3 Marks of Imprinting are (in General) Cleared Between

Generations and Reset During Gametogenesis
The sex-specific marks on DNA, which result in (or lead to) genomic imprinting,
and consequently to functional haploidy of the non-imprinted alleles, persist in
general from conception throughout all embryonic stages and up to the adult state
in most somatic cells. These marks, however, have to be completely erased at a
certain critical period of the life cycle since they are likely to be set differently at
each generation.
Experiments and observations have demonstrated that epigenetic marks
(histone modifications and DNA methylation) on most of the genome start to
become erased in primordial germ cells of both sexes at around day 11.5 of ges-
tation, upon entry of the germ cells into the gonads. Genes then acquire new
sex-specific DNA methylation marks during fetal development in males and
a little later, during the growing oocyte phase, in the early neonatal period in
females. The mechanisms involved during the clearing out of the imprinting
marks (active or passive DNA demethylation) have not been completely unrave-
led (Ferguson-Smith 2011).
More importantly, acquired methylation of the ICRs or DMRs of imprinted
genes needs to be preserved during the massive wave of demethylation that occurs
in the embryo after fertilization. It is now known that imprinted genes display hex-
anucleotide motifs that are methylated and recognized by several proteins (such
as Zfp57, TRIM 28, or Stella). The complex formed between the hexanucleotide
motif and these proteins protects the ICRs from being demethylated at these early
stages of development and is a signature of the imprinted genes. These observa-
tions reveal that both genetic and epigenetic signals are required to establish and
maintain the imprinted status of a gene.
When discussing X-chromosome inactivation we mentioned that the inactive

X chromosome could sometimes reactivate in somatic cells, especially when
the animals age. The situation is similar and even more common with autoso-
mal imprinting, and cases of tissue-specific variations have been reported in the
mouse. For example, in the developing embryos only the paternal allele at the
Igf2 locus is expressed, while the maternal allele is silent. However, in the cho-
roid plexus and leptomeninges the situation is different and both alleles are tran-
scriptionally active (DeChiara et al. 1991). Another example of tissue-specific
imprinting is provided by the Cdh15 gene. The germline DMR of this gene is
protected from erasure of methylation during the first steps of embryogenesis
but becomes methylated after implantation. This led to the proposal of the exist-
ence of both bona fide imprinted germline DMRs and transient germline DMRs
(Proudhon et al. 2012).
Another interesting situation is provided by the viable yellow allele at the
agouti locus (Avy-Chr 2). This mutation is transmitted as a dominant allele; it is
viable when homozygous (unlike the classical yellow allele Ay, which is homozy-
gous lethal), but the coat color of affected mice exhibits variation, ranging from
pure homogeneous yellow, through mottling with dark patches, to an agouti-like
coat (pseudo-agouti) similar to the wild-type allele A. Homozygous (Avy/Avy) and
heterozygous (Avy/a) mice tend to become obese and diabetic, and the degree of
obesity is correlated with the coat color, yellow mice being more affected than
agouti ones (Morgan et al. 1999).
The Avy mutation is the result of the insertion of an intra-cisternal A-particle
(IAP or retrotransposon) into a non-coding exon 5′ of the agouti gene. Functional
analysis revealed that the expression of the mutant allele is controlled by the long
terminal repeat (LTR) of the IAP. When the LTR in question is hypomethylated,
the Avy allele is transcribed, the coat is yellow, and the mouse is bigger than nor-
mal. When the viral LTR is methylated (and accordingly inactivated), the coat is
agouti. Variegation of coat color in Avy/+ mice (which is sometimes also observed
in Ay/+ mice) is very likely the consequence of some mosaicism at the somatic
cell level.
When Avy/+ males are mated with a/a (black non-agouti) females, there is no
significant difference in the proportions of yellow, mottled or pseudo-agouti phe-
notypes in the progenies, and this occurs independently of the coat color (yellow,
mottled or agouti) of the male. The situation is different when the cross is set up
the other way, i.e., between an a/a (non-agouti) male and Avy/+ female. In this
case there is some sort of transgenerational epigenetic inheritance in the sense
that the distribution of phenotypes in the progenies is related to the phenotype of
the dam and not of the sire—for example, yellow mothers produce more yellow
offspring than agouti mothers. Clearly, it appears that imprinting marks are not
erased when transmitted through the female, while they are erased when transmit-
ted through the male. Several laboratories have confirmed these observations and it
has been demonstrated that selection of a certain phenotype (for example, the per-
centage of pseudo-agouti offspring in the progeny) could increase the prevalence
of the trait in successive progenies (Blewitt et al. 2006; Cropley et al. 2012).
The behavior of the Avy allele, which is quite uncommon in mouse genetics, may
appear anecdotal but similar situations might be common if we consider the abun-
dance of IAP in the mammalian genomes (Morgan et al. 1999).
6.3.4 Genomic Imprinting Across Mammalian Species
To date, the differential expression of alleles according to their parental origin has
been reported and documented only in flowering plants (Nowack et al. 2007) and
in mammals. In mammals, it seems to be an exclusive characteristic of the eutheri-
ans and metatherians11 (marsupials), while prototherians (for example the platy-
pus, Ornithorhynchus anatinus) do not exhibit genomic imprinting. In other
words, genomic imprinting seems to correlate with gestation of the embryo inside
the uterus and placentation (viviparity) but not with egg laying (oviparity).
Genomic imprinting has never been reported in fish, amphibians, reptiles or birds
(Dünzinger et al. 2005).
In mammals, the imprinted regions are in general relatively well preserved
across the different species and for each of the imprinted regions in the mouse,
for example, there is in many instances a homologous region in the rat and in
humans—with, however, a few remarkable exceptions. From these phylogenetic
observations one may conclude that genomic imprinting probably appeared con-
comitantly with the viviparous mode of reproduction (i.e., ~180 Myr ago). One
may also observe that the more closely related are any two species, the greater
are the homologies between the different imprinted regions. However, after care-
ful observation it is sometimes discovered that rare but noticeable differences exist
between closely related species, as if the process of genomic imprinting was still
in evolution in that class of vertebrates.
As we discussed in a previous chapter, some morphological differences
between inter-specific hybrids have been reported which depend upon the way the
cross that produced these hybrids was set up. Even in the Mus genus, in which so
many species have been identified including Mus m. musculus and Mus m. domes-
ticus, some morphological and anatomical differences have been noted that could
be attributed to point differences in terms of genomic imprinting. For example,
female mice of the Mus spretus species do not (or very rarely) produce viable off-
spring when crossed with laboratory mouse males, while the reverse is not true.
The placental hypertrophy of some of these rare F1 hybrids or backcross offspring
has been attributed to an X-linked locus (Ihpd for interspecific hybrid placental
dysplasia) with several alleles, but could also be interpreted as differential imprint-
ing due to differential X inactivation.
11 In marsupials, the number of imprinted genes is much lower than in eutherian mammals.
6.3.5 The Origin and Evolution of the Imprinting

Mechanisms in Mammals
The existence of genomic imprinting raises a number of issues that can be sum-
marized in the following question: what advantage can justify, for a mammalian
embryo, having a number of its genes maintained in a functionally haploid status,
while diploidy is generally considered more advantageous with regards to evolu-
tion? The answer to this basic question is not yet definitively known, and several
hypotheses have been developed over the last decade (Wood and Oakey 2006).
One of the first explanations that came to mind was the consideration that
imprinting emerged during evolution as a mechanism to clear the genome of spon-
taneously occurring mutations with lethal or deleterious effects, for the simple
reason that such mutations, when they occur within an imprinted region, are elimi-
nated when the region in question becomes functionally haploid. This hypothesis
unfortunately has several weaknesses, and in particular it does not explain why
such a clever mechanism appeared so late in evolution and has remained an exclu-
sive privilege of mammals.
A more consistent explanation is that genomic imprinting is a very efficient
way of inhibiting parthenogenetic (gynogenetic or androgenetic) development in
mammals. Indeed, and as explained above, the development of a normal mouse
embryo from two female (or two male) pronuclei (i.e., from only one parent or
from two parents of the same sex) is strongly repressed. This is a direct conse-
quence of genomic imprinting at the H19–Igf2 and Dlk1–Gtl2 loci, as demon-
strated by Japanese scientists who succeeded in producing bi-maternal mice after
artificially erasing (i.e., by genetic engineering) the imprinting at these loci (Kono
et al. 2004; Kawahara et al. 2007; Kawahara and Kono 2012). Although more
likely than the previous one, the hypothesis stating that genomic imprinting exists
only to impede parthenogenesis in mammals is not entirely convincing and is
definitely not sufficient. In fact, the possibility that parthenogenetic development
could occur in mammals cannot, a priori, be regarded as a disadvantage, since that
sort of development exists occasionally in some classes of vertebrates as an excep-
tional and alternative way of reproduction, for example to escape a reproductive
dead end. From this point of view, the possibility of the mammals using partheno-
genesis for one or two generations would also appear advantageous.
A third hypothesis on the origin of genomic imprinting is that it has no advan-
tages at all and exists only by chance. According to this hypothesis imprinting is
a mere artifact, a “red herring” so to speak, which results from the uncontrolled
expansion to the neighboring regions of a defense mechanism used by mammals
to control or neutralize the possible invasion of their genome by self-replicating
parasitic DNAs such as retroviruses or retro-transposons (see Chap. 5). Just like
the previous two, this hypothesis has some weaknesses and, in particular, it does
not explain why imprinting exists only in mammals—while birds have to compete
with so many retroviruses and retro-transposons invading their genomes. In the
same way, it does not fit with the fact that imprinting is reversible.
If we summarize the information gathered from the observations made in

humans (see below) and those collected from the many experiments that have been
performed in the mouse species, we can establish correlations and draw some con-
clusions about the essence of imprinting in mammals and finally come to more
coherent hypotheses. An important one is based on the observation that most—
not to say all—of the genes that are imprinted have been found to play a role in
the control of embryonic growth and development, in most instances through the
development of the placenta. Based on this observation, a widely accepted hypoth-
esis to explain the origin and evolution of genomic imprinting is the “parental
conflict hypothesis”, which is also known as the “tug-of-war hypothesis”. The
hypothesis states that the differences between parental genomes due to imprint-
ing are the result of the divergent interests of each parent (or sex) concerning the
evolutionary fitness of their genes (Haig 1997; Sha 2008). Since males can have
a virtually unlimited number of offspring, the father’s genes gain greater fitness
through the vigor of the offspring, eventually at the expense of the mother, and
this explains why paternally expressed genes tend to be growth-promoting for
the embryo. The mother’s interest, on the other hand, is to preserve nutrients and
resources for her own use, to get rid of the offspring that are in her uterus as soon
as possible, and thus be able to produce another litter as rapidly as possible. This
would be in agreement with the observation that maternally expressed genes tend
to be growth-limiting. Indeed, unlike other vertebrate embryos, mammals could
theoretically stay in utero for an unlimited period of time, surviving at the expense
of the mother’s nutrients, unless a mechanism regulating gestation length inter-
venes. Genomic imprinting, indirectly controlling the embryo’s growth, appears a
good way to limit the duration of gestation. This hypothesis has the great advan-
tage of justifying the existence of imprinting and its existence exclusively in mam-
mals and, for this reason, it has been accepted for a good ten years. Nowadays,
however, our understanding of the molecular mechanisms at work in genomic
imprinting has revealed some inconsistencies, and the parental conflict hypoth-
esis would probably need to be revisited. Recent observations have suggested
co-adaptation between the mother and the conceptus at fetal stages (involving pla-
cental exchanges) and at post-natal stages with metabolic and behavior exchanges
(Keverne 2013).
6.3.6 The Pathological Aspects Associated with Genomic

Imprinting
6.3.6.1 Epigenetics and Human Diseases
The same year (1974a) when Johnson reported his observations concerning
the phenotypic differences associated with the parental origin of the hairpin-tail
(Thp) mutant allele in the mouse (see above), Lubinsky and colleagues reported
a similar parental effect in a family transmitting a syndrome now known as
Beckwith–Wiedemann syndrome (BWS) (Lubinsky et al. 1974). In fact, these

two observations independently inaugurated the studies relating to the effect of
genomic imprinting on gene expression in the mouse and human species, respec-
tively. Nearly forty years after these publications, a lot has been learned concern-
ing genomic imprinting and its importance in some human pathologies.
Beckwith–Wiedemann syndrome (OMIM 130650) is a rare disorder with an
incidence of approximately one in 14,000 childbirths. It is characterized by the
association of traits like macroglossia, greater than normal birth weight and size,
neonatal hypoglycemia, and some other visceral defects (of the adrenal gland
in particular). In most cases the BWS is sporadic, but around 15 % of the cases
are familial and in many of these familial cases, mutations or deletions of genes
within a region spanning approximately 1 Mb of human chromosome 11p15.5
have been reported (the mouse homologous region is on distal chromosome 7).
Imprinting defects of genes in the same region have also been described in a very
high proportion of BWS patients having a biallelic (rather than paternal monoal-
lelic) expression of the IGF2 gene. In these cases, the maternal copy of the gene
IGF2 is transcribed where it is normally inactivated in healthy babies. Finally,
some babies affected by the BWS have been found to be the consequence of a
paternal uniparental disomy (PatUpDi) of chromosome 11, and in these rare cases
the two regions of chromosome 11, having escaped maternal imprinting, are both
transcribed. Other patients exhibit loss of imprinting of a gene encoding a long
ncRNA transcript, called KCNQ1OT1, which is also known to be imprinted in the
mouse.
Another rare human syndrome, Russell–Silver syndrome (RSS-OMIM
180860—one in 70,000 childbirths), has also been found to be associated with an
imprinting defect. In a recent survey concerning this disease, 10 % of all the cases
were found to be associated with a maternal uniparental disomy (MatUpDi) of
chromosome 7. In some other cases, the same 11p15.5 region of human chromo-
some 11 harboring the H19 and IGF2 genes appeared to be involved. The defect
in this case is characterized by a suppression of IGF2 growth factor activity that
explains the concomitant growth reduction observed in RSS patients. In these
cases, where the same 11p15.5 region is concerned, the pathological features of
RSS logically appear to be the opposite of those described for BWS (Butler 2009).
Prader–Willi (PWS) and Angelman (AS) syndromes are the two most stud-
ied cases of human diseases commonly related to defective genomic imprinting.
Unlike BWS and RSS, which are often compatible with an almost normal adult
life, PWS and AS are always severe and do not improve with aging. PWS and AS
are caused by mutations, deletions, uniparental disomy or by abnormal imprint-
ing of one or several different members of a gene cluster in the q11-q13 region of
human chromosome 15 (Horsthemke and Wagstaff 2008).
Prader–Willi syndrome (OMIM 176270) occurs in one in 15,000 individuals
and is characterized at a young age by hypotonia, short stature, mental deficiency,
behavioral problems, and feeding difficulties. In a second phase, from the age of
3 years, developmental delay and psychomotor retardation are even more obvious
but obesity becomes a life-threatening issue requiring strict dietary restrictions.
Angelman syndrome (OMIM 105830) is characterized by severe mental

retardation with seizures, ataxia, uncoordinated movements, hypopigmentation,
inappropriate hilarity, lack of speech, etc. In the late 1980s geneticists observed
that PWS and AS were caused by deletions in bands 15q11-q13, and they reported
that the observed phenotypic differences between the two syndromes in fact
depended upon the parental origin of the deletion. Deletions occurring on pater-
nal chromosome 15 generally resulted in PWS, while similar deletions occur-
ring on maternal chromosome 15 resulted in AS. For this reason, PWS and AS
were, and still are, considered as sister syndromes—which fits rather well with the
symptomatology.
Nowadays, the situation has been much clarified, and by and large it is con-
cluded that PWS is a consequence of the lack of the paternal copy of one or a
few genes in the 15q11-q13 region, while AS is a consequence of the lack of a
functional maternal copy of the UBE3A gene encoding ubiquitin protein ligase 3A
(Moncla et al. 1999; Horsthemke and Wagstaff 2008).
In addition to the four syndromes described above, which are relatively well
documented, a few other human diseases and pathological conditions, including
certain forms of cancers, have been described as the very likely consequence of
abnormal imprinting because of a clear effect of the parental inheritance. In most
instances, however, the situation was reported as complex and difficult to analyze
because of the interference of environmental factors and/or epistatic interactions
with elements of the genetic background. It is likely that, with the rapid progress
in sequencing technology and the development of quantitative analysis of RNA
transcription, these diseases or syndromes will be clarified in the near future. This
will definitely allow a better understanding of the role of epigenetic regulation in
gene expression.
6.3.6.2 Epigenetic Manifestations in Some Animal Crosses
At several points in this book we have mentioned that some interspecific mouse
hybrids exhibit a variety of pathological features depending on the direction of the
cross. For example, crosses between male mice of the Mus spretus species and
females of the Mus m. domesticus species produce viable hybrids but the sex ratio
in the offspring progeny is much biased in favor of the female, and the male F1s
are always sterile. This difference is in compliance with the so-called Haldane’s
rule and has been observed in several other cases of interspecific crosses (for
example, between different Drosophila species, between Bos taurus and Bison
bison, and between Chrysolophus pictus and Gallus g. domesticus).12 In the case
of mouse crosses, it has been established that the sterility of hybrids is controlled
by a few genes, some of which have been localized on the genetic map. In con-
trast, the reasons for the shortage of males are still conjectural.
12 Haldane's rule states "when in the offspring of two different animal races one sex is absent,
rare, or sterile, that sex is the heterozygous [heterogametic] sex."

More interesting is the observation that crosses in the other direction (between
Mus m. domesticus males and Mus spretus females) result in stillbirths in most
cases, with a marked enlargement of the placenta.13 A similar situation was
reported for crosses between two other species of rodents of the genus
Peromyscus, with strong parent-of-origin effects involving placental growth.
Female P. maniculatus crossed with male P. polionotus produce neonates smaller
than either parental strain, with placentas half the parental size. In contrast, female
P. polionotus crossed with male P. maniculatus produce dysmorphic overgrown
embryos whose placentas average up to 2.5 times the mass of the parental strains
(Vrana 2007).
Such biases are difficult to explain in terms of Mendelian genetics if we con-
sider that the genetic makeup of the above-mentioned reciprocal F1s are virtually
the same, with one allele of each parental species in both cases. However, a pos-
sible (and likely?) explanation would be to guess that the parental alleles of some
homologous genes are imprinted differently in the two F1s. This would explain all
the observed phenotypes.
A similar observation has been made concerning the offspring of crosses made
in zoological gardens between two species of the Panthera genus: Panthera leo,
the African lion, and Panthera tigris, the Bengal tiger. The liger, a hybrid between
a male lion and a tigress, is an enormous animal, with a total length reaching
3–3.5 m and a weight of up to 380 kg (~800 lb), while the reciprocal hybrid, the
tigon (much less common), is slightly undersized compared to its parents. Here
again, the explanations for these size differences are still somewhat speculative
but, given that the imprinted genes often play a role in issues of hybrid growth, it
is tempting to guess that this applies in the case of these two interspecific hybrids
(Morison et al. 2001, 2005).
Finally, another interesting case is the Callipyge phenotype in sheep (abbr.
CPLG—from the Greek “beautiful buttocks”). This mutation was first discovered
in the USA segregating in a flock in Oklahoma. It causes lambs to develop large
and muscular rumps, and for this important economical value it has been exten-
sively studied by animal geneticists (Georges et al. 2013). It has then been dem-
onstrated that the phenotype is fully expressed only in heterozygous individuals
who receive the CLPG mutant allele from their father. When inherited from the
mother, it is not expressed. This situation is known as polar overdominance and is
another example of phenotypic alteration due to imprinting. The CLPG mutation
is a single nucleotide substitution in what is probably a long-range control ele-
ment (LRCE—see Chap. 5) within the DLK1–GTL2 imprinted domain of several
species of mammals. The mutation also exists in humans and in cattle, and has
been created by genetic engineering in the mouse. It is a very interesting model for
these sorts of phenotypic observations.
13 Only some exceptional viable offspring have been bred from such a cross.
6.4 Conclusions 217
6.4 Conclusions
Initially discovered in the form of anecdotal observations (coat color of calico

cats and the unexpected inheritance of the hairpin-tail mutation), X inactivation
and genomic imprinting appear to be two important ways of regulating genomic
expression. Diploidy, as we said, was generally considered as advantageous with
regards to evolution because, having a backup copy for each and every gene, dip-
loid organisms were more protected against the deleterious effects of mutations.
After the discovery of genomic imprinting, this analysis must be seriously recon-
sidered. Indeed, if a gene mutates, the back-up (normal) copy of this gene may
not be available for replacement if it is in an imprinted region and accordingly
epigenetically inactivated. What then is the evolutionary advantage of imprinting
for mammals? A close association has been established with viviparity, at least
with the development of the embryo in utero, but this association by definition
does not exist in flowering plants where the imprinting phenomenon has also been
described. Nowadays, a theory is emerging suggesting that genomic imprinting
might play an important role as a mechanism of reproductive isolation generating
diversity. Many of these investigations are conducted in mammals (in particular,
laboratory rodents), and it is likely that the evolutionary advantages of genomic
imprinting will be established in the relatively near future.
Unraveling the intimate molecular mechanisms at work in the establishment
and maintenance of imprinting might be laborious, but it is a very important issue
and there is no doubt that, in this matter more than in any other, the mouse will be
an invaluable model.
References
Augui S, Nora EP, Heard E (2011) Regulation of X-chromosome inactivation by the X-inactivation
centre. Nat Rev Genet 12:429–442
Barlow DP, Stöger R, Herrmann BG, Saito K, Schweifer N (1991) The mouse insulin-like
growth factor type-2 receptor is imprinted and closely linked to the Tme locus. Nature
349:84–87
Barton SC, Surani MAH, Norris ML (1984) Role of paternal and maternal genomes in mouse
development. Nature 311:374–376
Beutler E, Yeh M, Fairbanks VF (1962) The normal human female as a mosaic of X-chromosome
activity: studies using the gene for G-6-PD deficiency as a marker. Proc Natl Acad Sci USA
48:9–16
Blewitt ME, Vickaryous NK, Paldi A, Koseki H, Whitelaw E (2006) Dynamic reprogramming of
DNA methylation at an epigenetically sensitive allele in mice. PLoS Genet 2(4):e49
Brown SD (1991) XIST and the mapping of the X chromosome inactivation centre. BioEssays
13:607–612
Butler MG (2009) Genomic imprinting disorders in humans: a mini-review. J Assist Reprod
Genet 26:477–486
Cattanach BM (1986) Parental origin effects in mice. J Embryol Exp Morphol 97(Suppl):137–150
Cattanach BM, Kirk M (1985) Differential activity of maternally and paternally derived chromo-
some regions in mice. Nature 315:496–498
Chen T, Dent SY (2014) Chromatin modifiers and remodellers: regulators of cellular differentiation.
Nat Rev Genet 15:93–106
Constância M, Pickard B, Kelsey G, Reik W (1998) Imprinting mechanisms. Genome Res
8:881–900
Cropley JE, Dang TH, Martin DI, Suter CM (2012) The penetrance of an epigenetic trait in
mice is progressively yet reversibly increased by selection and environment. Proc Biol Sci B
279:2347–2353
Davidson RG, Nitowsky HM, Childs B (1963) Demonstration of two populations of cells in the
human female heterozygous for glucose-6-phosphate dehydrogenase variants. Proc Nat Acad
Sci USA 50:481–485
DeChiara TM, Robertson EJ, Efstratiadis A (1991) Parental imprinting of the mouse insulin-like
growth factor II gene. Cell 64:849–859
Dünzinger U, Nanda I, Schmid M, Haaf T, Zechner U (2005) Chicken orthologues of mam-
malian imprinted genes are clustered on macrochromosomes and replicate asynchronously.
Espinós C, Lorenzo JI, Casaña P, Martínez F, Aznar JA (2000) Haemophilia B in a female caused
by skewed inactivation of the normal X-chromosome. Haematologica 85:1092–1095
Ferguson-Smith AC (2011) Genomic imprinting: the emergence of an epigenetic paradigm. Nat
Rev Genet 12:565–575
Ferguson-Smith AC, Sasaki H, Cattanach BM, Surani MA (1993) Parental-origin-specific epige-
netic modification of the mouse H19 gene. Nature 362:751–755
Gabory A, Ripoche MA, Le Digarcher A, Watrin F, Ziyyat A, Forné T, Jammes H, Ainscough JF,
Surani MA, Journot L, Dandolo L (2009) H19 acts as a trans regulator of the imprinted gene
network controlling growth in mice. Development 136:3413–3421
Garrick D, Sharpe JA, Arkell R, Dobbie L, Smith AJH et al (2006) Loss of Atrx affects trophoblast
development and the pattern of X-inactivation in extraembryonic tissues. PLoS Genet 2(4):e58
Georges M, Charlier C, Cockett N (2013) The callipyge locus: evidence for the trans interaction
of reciprocally imprinted genes. Trends Genetics (in press)
Haig D (1997) Parental antagonism, relatedness asymmetries, and genomic imprinting. Proc Roy
Soc Lond Ser B-Biol Sci 264:1657–1662
Horsthemke B, Wagstaff J (2008) Mechanisms of imprinting of the Prader-Willi/Angelman
region. Am J Med Genet A 146:2041–2052
Johnson DR (1974a) Further observations on the haipin-tail (Thp) mutation in the mouse. Genet
Res 24:207–213
Johnson DR (1974b) Hairpin-tail: a case of post-reductional gene action in the mouse egg.
Kawahara M, Kono T (2012) Roles of genes regulated by two paternally methylated imprinted
regions on chromosomes 7 and 12 in mouse ontogeny. J Reprod Dev 58:175–179
Kawahara M, Wu Q, Takahashi N, Morita S, Yamada K, Ito M, Ferguson-Smith AC, Kono T
(2007) High-frequency generation of viable mice from engineered bi-maternal embryos. Nat
Biotechnol 25:1045–1050
Kelly WG, Schaner CE, Demburg AF, Lee MH, Kim SK, Villeneuve AM, Reinke V (2002)
X-chromosome silencing in the germline of C. elegans. Development 129:479–492
Keverne EB (2013) Importance of the matriline for genomic imprinting, brain development and
behaviour. Philos Trans R Soc Lond B Biol Sci 368:20110327. doi:10.1098/rstb.2011.0327
Kono T, Obata Y, Wu Q, Niwa K, Ono Y, Yamamoto Y, Park ES, Seo JS, Ogawa H (2004) Birth
of parthenogenetic mice that can develop to adulthood. Nature 428:860–864
Larschan E, Bishop EP, Kharchenko PV, Core LJ, Lis JT, Park PJ, Kuroda MI (2011) X chro-
mosome dosage compensation via enhanced transcriptional elongation in Drosophila. Nature
Latos PA, Pauler FM, Koerner MV, Şenergin HB, Hudson QJ, Stocsits RR, Allhoff W, Stricker SH,
471:115–118
Klement RM, Warczok KE, Aumayr K, Pasierbek P, Barlow DP (2012) Airn transcriptional
overlap, but not its lncRNA products, induces imprinted Igf2r silencing. Science 338:1469–1472
References 219
Li E, Beard C, Jaenisch R (1993) Role for DNA methylation in genomic imprinting. Nature
366:362–365
Lubinsky M, Herrmann J, Kosseff A, Opitz JM (1974) Autosomal-dominant sex-dependent trans-
mission of the Beckwith-Wiedemann syndrome. Lancet 1:932
Lyon MF (1961) Gene action in the X-chromosome of the mouse (Mus musculus L.). Nature
190:372–373
Lyon MF (2002) A Personal History of the Mouse Genome. Annu Rev Genomics Hum Genet
3:1–16
Marahrens Y, Panning B, Dausman J, Strauss W, Jaenisch R (1997) Xist-deficient mice are defective
in dosage compensation but not spermatogenesis. Genes Dev 11:156–166
McGrath J, Solter D (1984) Completion of mouse embryogenesis requires both the maternal and
paternal genomes. Cell 37:179–183
Mohammad F, Pandey GK, Mondal T, Enroth S, Redrup L, Gyllensten U, Kanduri C (2012)
Long noncoding RNA-mediated maintenance of DNA methylation and transcriptional gene
silencing. Development 139:2792–2803
Moncla A, Malzac P, Livet MO, Voelckel MA, Mancini J, Delaroziere JC, Philip N, Mattei JF
(1999) Angelman syndrome resulting from UBE3A mutations in 14 patients from eight fami-
lies: clinical manifestations and genetic counseling. J Med Genet 36:554–560
Morey C, Avner P (2011) The demoiselle of X-inactivation: 50 years old and as trendy and mes-
merising as ever. PLoS Genet 7:e1002212
Morgan HD, Sutherland HG, Martin DI, Whitelaw E (1999) Epigenetic inheritance at the agouti
locus in the mouse. Nat Genet 23:314–318
Morison IM, Paton CJ, Cleverley SD (2001) The imprinted gene and parent-of-origin effect data-
base. Nucleic Acids Res 29:275–276
Morison IM, Ramsay JP, Spencer HG (2005) A census of mammalian imprinting. Trends Genet
21:457–465
Nowack MK, Shirzadi R, Dissmeyer N, Dolf A, Endl E, Grini PE, Schnittger A (2007)
By-passing genomic imprinting allows seed development. Nature 447:312–315
Okamoto I, Otte AP, Allis CD, Reinberg D, Heard E (2004) Epigenetic dynamics of imprinted
X inactivation during early mouse development. Science 303:644–649
Patrat C, Okamoto I, Diabangouaya P, Vialon V, Le Baccon P, Chow J, Heard E (2009) Dynamic
changes in paternal X-chromosome activity during imprinted X-chromosome inactivation in
mice. Proc Natl Acad Sci USA 106:5198–5203
Penny GD, Kay GF, Sheardown SA, Rastan S, Brockdorff N (1996) Requirement for Xist in X
chromosome inactivation. Nature 379:131–137
Pollex T, Heard E (2012) Recent advances in X-chromosome inactivation research. Curr Opin
Cell Biol. http://dx.doi.org/10.1016/j.ceb.2012.10.007
Proudhon C, Duffié R, Ajjan S, Cowley M, Iranzo J, Carbajosa G, Saadeh H, Holland ML,
Oakey RJ, Rakyan VK, Schulz R, Bourc’his D (2012) Protection against de novo methylation
is instrumental in maintaining parent-of-origin methylation inherited from the gametes. Mol
Cell 47:909–920
Reik W, Dean W, Walter J (2001) Epigenetic reprogramming in mammalian development.
Science 293:1089–1093
Reik W, Walter J (2001) Evolution of imprinting mechanisms: the battle of the sexes begins in
the zygote. Nat Genet 27:255–256
Saxena A, Carninci P (2011) Whole transcriptome analysis: what are we still missing? Wiley
Interdiscip Rev Syst Biol Med 3:527–543
Sha K (2008) A mechanistic view of genomic imprinting. Annu Rev Genomics Hum Genet
9:197–216
Simmler MC, Cattanach BM, Rasberry C, Rougeulle C, Avner P (1993) Mapping the murine Xce
locus with (CA)n repeats. Mamm Genome 4:523–530
Surani MAH, Barton SC, Norris ML (1984) Development of reconstituted mouse eggs suggests
imprinting of the genome during gametogenesis. Nature 308:548–550
Thorvaldsen JL, Krapp C, Willard HF, Bartolomei MS (2012) Nonrandom X chromosome inactiva-
tion is influenced by multiple regions on the murine x chromosome. Genetics 192:1095–1107
Trent S, Dennehy A, Richardson H, Ojarikre OA, Burgoyne PS, Humby T, Davies W (2011)
Steroid sulfatase-deficient mice exhibit endophenotypes relevant to attention deficit hyperac-
tivity disorder. Psychoneuroendocrinology 37:221–229
Vrana PB (2007) Genomic imprinting as a mechanism of reproductive isolation in mammals. J
Mammal 88:5–23
Weidman JR, Dolinoy DC, Maloney KA, Cheng JF, Jirtle RL (2006) Imprinting of opossum
Igf2r in the absence of differential methylation and Air. Epigenetics 1:49–54
Winking H, Silver LM (1984) Characterization of a recombinant mouse T haplotype that
expresses a dominant lethal maternal effect. Genetics 108:1013–1020
Wood AJ, Oakey RJ (2006) Genomic imprinting in mammals: emerging themes and established
theories. PLoS Genet 2:e147
Wutz A, Theussl HC, Dausman J, Jaenisch R, Barlow DP, Wagner EF (2001) Non-imprinted
Igf2r expression decreases growth and rescues the Tme mutation in mice. Development
128:1881–1887
Yu M, Hon GC, Szulwach KE, Song CX, Zhang L, Kim A, Li X, Dai Q, Shen Y, Park B, Min JH,
Jin P, Ren B, He C (2012) Base-resolution analysis of 5-hydroxymethylcytosine in the mam-
malian genome. Cell 149:1368–1380
Chapter 7
Mutations and Experimental Mutagenesis
7.1 The Importance of Mutations
The word mutation was coined in 1901 by Hugo De Vries to describe “sudden,
spontaneous and drastic alterations in the hereditary material of Oenothera”, the
evening primrose.1 Mutations occur in the genome of all living organisms and
vary in importance, ranging from single base-pair changes to extensive chromo-
somal rearrangements. They can occur either in somatic or germ cells, at all stages
of development, and are transmitted to daughter cells except when they cause
death or a severe selective disadvantage.
When mutations occur in somatic cells with high mitotic activity, such as cells
of the bone marrow, intestinal mucosa, lung or skin, or when the mutations in
question interfere with the mechanisms that regulate the cell cycle or differentia-
tion, then the affected cells may become carcinomatous. When mutations occur in
the cells contributing to the germ line they may be transmitted to the next genera-
tion and, in this case, a proportion of the offspring will be heterozygous for a new
mutant allele. This category of mutations is precisely the one we will focus on in
this chapter.
Germinal mutations, by definition, generate new alleles that enter the gene pool
of the species and contribute to an increase in polymorphism. Most of these new
alleles have no effects or effects that do not influence the fitness of the affected
individuals, and for this reason they are called “neutral mutations”. A small pro-
portion of these new mutations may result in a better adaptation of the animals to
their environment.2 Finally, some mutations have deleterious effects, frequently
leading to pathological conditions. In this case, and if we consider that almost all
genes in the mouse genome have an equivalent in the human genome, it is obvious
that many among the new mutant alleles found in the mouse species represent
potentially interesting models of human genetic diseases.
1 Hugo de Vries also used the word “sport” to define the same sort of sudden genetic changes.
2 Resistance to the rodenticide warfarin is a good example of the mutations that occurred in wild
populations, generating a selective advantage.

222 7 Mutations and Experimental Mutagenesis
Mutations can affect all genomic regions, with a wide range of consequences
at the phenotypic level. They are either dominant, semi-dominant (heterozygotes
have a less severe phenotype than homozygous mutants), co-dominant (both
alleles are equally expressed) or recessive. Detailed study of the phenotype of
these new mutant alleles is part of the process of genome annotation, and is of
great importance for the characterization of gene function(s).
The occurrence of spontaneous mutations in mammalian genomes results from
errors occurring either during meiosis or in the process of DNA replication which are
not mended by the cellular (DNA) repair mechanisms. These repair mechanisms are
very sophisticated, with specific enzymes constantly checking the integrity of cellular
DNA during and after replication, but the system is sometimes defective or saturated
and fails. Taking this into account, one understands that there is no way to prevent
mutations from occurring, and that the spontaneous mutation rate is a basic parameter
that each species must cope with. In addition, many agents such as radiation, some
chemicals, and some viruses and transposons can increase the rate of mutations well
above the spontaneous rate. Some of these agents, as we will discuss in this chapter,
have been used over the last fifty years for performing experimental mutagenesis.
Experimental mutagenesis can be “phenotype-driven”, when unknown genes
are identified based on the phenotypic changes associated with at least one of the
mutant alleles. In this case, the structure of the gene affected by the mutation is
elucidated afterwards, by positional cloning, depending on the potential interest
of the mutant allele. Experimental mutagenesis can also be “genotype- or gene-
driven”, whereby mutations are massively induced and then sought only in pre-
selected genes or DNA regions of unknown function, for example for the purpose
of genome annotation. As we will see, experimental mutagenesis is relatively sim-
ple to achieve, but its efficiency depends upon the mutagenic treatment as well as
on the protocols used for the characterization of the mutant phenotypes.
In this chapter, we will describe in some detail the different types of mutations
that can affect a mammalian genome and their consequences. We will then discuss
the different protocols that can be used for the induction of mutations in the mouse
germline, with special emphasis on chemical mutagenesis, which is highly effi-
cient and accordingly has become widespread.
7.2 The Different Types of Mutations
When considered at the DNA level, mutations are generally classified into two
categories:
• chromosomal mutations, which are detectable by the observation of morpholog-
ical changes at the karyotype level, and
• point mutations, when no alteration in chromosome integrity is detectable.
This classification into chromosomal mutations and point or gene mutations dates
back to a time when the microscope was the only tool available to visualize changes
7.2 The Different Types of Mutations 223
in the hereditary material. Since then, the notion of point mutation has changed and
now covers a group of structurally defined changes occurring in the DNA. We will
describe these changes, from the simplest to the more complex, and in so doing we
will realize that the classification mentioned above, in fact, is not really stringent.
However, it is convenient from a didactic point of view and thus we will adopt it.3
The geneticist H.J. Muller,4 who did pioneering research on experimental
mutagenesis in Drosophila flies using X-rays, proposed a classification of the
mutations into five categories based on the effect of the genetic change on gene
activity. The first category, the amorphic mutations, consisted of those mutations
that completely abolish the activity of the gene and were equivalent to null or loss
of function alleles. Hypomorphic mutations were associated with reduced activity
compared to the wild-type allele, while hypermorphic mutations were the oppo-
site, with an increased activity. Neomorphic mutations were the group of muta-
tions exhibiting a new function, and antimorphic alleles were mutations with
dominant negative effects.
7.2.1 Mutations Resulting from Base-Pair Substitutions

in the Coding Sequences
The information gathered from recent efforts of systematic sequencing of the

genome of various mouse strains have revealed that nucleotide substitutions are
the most frequent type of mutations. We will then take a simple example of this
type of mutation and discuss its consequences. This example will be the DNA
codon 5′-TGT-3′, which is transcribed as UGU and encodes the amino acid
cysteine (Cys, or C when using the one-letter code) and, for the time being, we
will focus on the nucleotide at the third position of this codon (T)5 (Table 7.1).
The first substitution is when the thymine (T) in the DNA strand is replaced by
a cytosine (C) (this substitution of a pyrimidine for another pyrimidine is called a
transition). In this case, the resulting mRNA codon becomes UGC but, like UGU,
it still encodes the same Cys residue. This type of mutation has no effect on the pro-
tein sequence, and for this reason, it is said to be silent or synonymous. In this case,
the base substitution is detected only when it suppresses or creates a restriction site
or, when comparing sequences, as a single nucleotide polymorphism (SNP).
3 Chromosomal rearrangements such as translocations of all types, transpositions, deletions,

duplications, inversions, etc. have been discussed in Chap. 3 (Cytogenetics) and 5 (The Mouse
Genome).
4 Hermann J. Muller was awarded the Nobel Prize in Physiology or Medicine for the “discovery
that mutations can be induced by X-rays”.

5 Here we write the codon sequences as they are read on the sense (5′ to 3′) strand of DNA. In
these conditions, they read the same as the RNA codons (with the exception that T is replaced by
U in mRNAs). However, it must be kept in mind that the mRNA transcripts are synthesized using
the antisense strand of DNA (3′ to 5′) as a template.
Table 7.1 Point mutations in the coding sequences

(a)
(b)
(c)
(d)
This table represents the two strands of a short coding DNA sequence, with three examples of
nucleotide substitution on the third position of the ACA/TGT codon, and their consequences at
the protein level. a The original (non-mutated) DNA strand. b The TGT–TGC substitution (a
transition) has no consequence at the protein level due to the degeneracy of the genetic code. In
this case, the same cysteine residue is incorporated into the protein—it is a synonymous muta-
tion. c The replacement of a TGT by a TGA (a transversion), on the contrary, leads to the termi-
nation of the translation process; this is a nonsense mutation. d Finally, the replacement of a TGT
by a TGG codon (another transversion) results in the incorporation of a different amino acid with
a wide variety of possible consequences; this is a missense mutation
Synonymous mutations are common findings (~23 %) in the sequence databases

of the different mouse inbred strains, especially those recently derived from wild
specimens (Beier 2000; Sakuraba et al. 2005; Frazer et al. 2007; Keane et al. 2011;
Yang et al. 2011; Arnold et al. 2012), and this observed frequency is in keeping with
theoretical computations. Indeed, if we consider that each of the 61 sense codons
(64 − 3) can mutate to one of nine different codons after the substitution of one or
the other of the three nucleotides, we can calculate that out of these 549 (61 × 9)
possible mutations around 25 % are synonymous while most others (75 %) are not
(Graur 2003). If we take a closer look at the distribution of all these mutations we
may notice that the synonymous mutations are much more frequent when the sub-
stitutions occur on the third nucleobase of the codon (70 %) than when they affect
one of the other two positions. This is, of course, because the code is degenerated.
Synonymous mutations occur constantly and regularly, even if at a low rate.
They are also relatively stable and have virtually no impact on the phenotype. For
these reasons they represent an interesting class of polymorphism for evolutionists
and can be considered as a molecular clock useful, for example, for assessing the
time of divergence between any two species or strains (Gilman 1972).6
These synonymous SNPs, when considered with the other flanking SNPs of the
same type on the same chromosome, can also be used for identifying the phyloge-
netic origin of the chromosome (or haplotype) in question. We will come back to
this point when discussing the inheritance of complex or quantitative traits (Keane
et al. 2011) (Chap. 10).
An interesting observation is that, in mammals, some synonymous codons are
found more frequently than others, even when the codons in question encode the
same amino acid. For example, the 5’-AGA-3’ and 5’-AGG-3’ DNA codons both
encode the amino acid arginine (R), but AGA is six times more frequent than AGG
in the transcripts. A similar observation can be made with the codons 5’-ACA-3’
and 5’-ACG-3’, which both encode the amino acid threonine (T), but ACA is five
times more frequent than ACG in the transcripts. The reason(s) for such a bias in
codon use is (are) not yet elucidated: they may be related to the fact that the muta-
tion rate is not the same for the four different nucleotides (discussed later); alterna-
tively, the bias in the codon usage may be related to the fact that the synonymous
codons are not equivalent in terms of efficiency at the translational level; some of
the codons have a selective advantage over the others.
Let us now assume that the third nucleotide of the same 5'-TGT-3' codon, the T,
is replaced by an adenine (A)—this change is designated a transversion (i.e., the
substitution of a pyrimidine for a purine). This mutation results in the incorpora-
tion of the UGA codon instead of UGU in the mRNA transcript, but this is the
signal for the termination of polypeptide synthesis, or stop codon. The resulting
mutations are called nonsense mutations, generating null or non-functional alleles.
Analysis of the sequencing data from positional cloning (in human and mouse) of
mutant alleles with a deleterious effect reveals that mutations of this type represent
around 4–5 % of the overall point mutations found in the coding sequences.
The functional consequences of nonsense mutations depend on the type of
protein encoded by the gene and the potential existence of other genes capable
of achieving the same or similar function(s). If the protein has an important func-
tion in cellular metabolism and if the gene is present as a single copy, the mutation
6 Theaverage spontaneous mutation rate at the DNA level has been estimated to be 2.2 × 10−9
per nucleotide per year in the human species (Kumar and Subramanian 2002).
generally leads to cell and/or embryonic death when in the homozygous state (reces-
sive lethal). If, however, the encoded protein is not essential or if it is expressed only
in a limited number of cells—for example, only the cells that are involved in the
synthesis of melanin pigment (melanocytes)—then only the hair coat and retina of
the animal are affected by the mutation, resulting in albinism (the consequence of
a null allele of the tyrosinase-encoding gene Tyr-Chr 7). All intermediates between
these two extreme cases are possible. Typically, inactive alleles resulting from a
stop codon have no phenotypic expression when heterozygous, except in the case of
haplo-insufficiency or parental imprinting of the normal allele (see Chap. 6).
mRNAs with a premature stop codon are in general rapidly degraded by spe-
cific exonucleases.7 However, in some cases where the stop codon occurs close to
the 3' end of the gene (in the last exon, for example), the transcript often escapes
mRNA decay and the abnormal (truncated) protein may have a dominant negative
effect of variable intensity.
The reciprocal mutations, where one of the three stop codons 5'-TAA-3',
5'-TGA-3', and 5'-TAG-3' reverts to a non stop-codon, are called read-through
mutations. These mutations are exceptional and only a very small number have
been reported (Noveroske et al. 2000). This is understandable if we consider the
relatively small target the three stop codons represent (9 bp altogether) compared
to the rest of the exonic sequences.
The last substitution we must consider is when the third base of the codon 5'-TGT-3'
for thymine (T) is replaced by a guanine (G); this change is another transversion. This
substitution changes the mRNA codon UGU to UGG, and a different amino acid (Trp—
tryptophan or W) is inserted into the polypeptide chain instead of the original cysteine.
These mutations are called non-synonymous or missense, and their effects are almost
unpredictable because they depend upon the site where the substitution occurred and the
type of amino acid replacement. This sort of mutation is by far the most common type
found in sequencing data from positional cloning of mutant alleles with a deleterious
effect. In some cases, the change has extremely limited effects and only some biophysi-
cal characteristics of the protein, such as, for example, its electrical charge, are altered.
In the case of altered electrical charge, the proteins are designated electrophoretic vari-
ants; they are easily identified by electrophoresis in a non-denaturing gel, but the func-
tion of the protein remains generally unchanged (see Chap. 4).
The β-chain of mouse hemoglobin (HBB, encoded by the Hbb gene on Chr 7)
has been extensively studied in wild mice because it represents an interesting sys-
tem for evaluating the functional divergence of duplicated genes during evolution.
In these studies, it has been observed that amino acid changes in the β-globin chain
are very common among the different species that are close relatives of the labora-
tory mice in the genus Mus, but all these “mutant” molecules (called isoforms) are
perfectly functional (Runck et al. 2009).
Another example of a non-synonymous mutation is worth mentioning: the Tyrc-h or
Himalayan allele at the Tyr locus in the mouse. This spontaneous mutation is common
7 This is referred as nonsense-mediated mRNA decay.

in mammals and an orthologous mutant allele also exists in the rat, the Siamese breed
of cats, the rabbit, and several other mammalian species. In the mouse, the mutation
was found to be the consequence of an A → G transition at nucleotide 1,259 of the Tyr
gene, which results in an amino acid change at position 420 from histidine to arginine
(His420Arg—a structurally important change). Because of this mutation, melanin syn-
thesis in Tyrc-h/Tyrc-h homozygous mice becomes temperature-sensitive; the pigment
is synthesized normally in the fur at around 20 °C but not at ~30 °C. As a result, the
mice have a different fur color at their extremities (the tip of their nose, tail, limbs, and
ear are normally pigmented because the temperature is lower at these parts of the body,
while the rest of the mouse is not or weakly pigmented). The Himalayan allele, which
is of ancient origin, has been relatively easy to detect and propagate because it made the
mice quite eye-catching without altering their health. However, if such mutations occur
in genes encoding proteins with an important role in homeostasis of the organism, the
consequences, although unpredictable, might be severe.
So far we have only considered the mutations that are the consequences of sub-
stitutions occurring at the third position of the 5′-TGT-3′ DNA codon. This codon
was selected as an example because it is one of the rare types that can produce the
three classes of mutations (synonymous, nonsense, and missense) with a single
base-pair replacement at the same position. However, as we already mentioned and
because of the degeneracy of the genetic code, mutations at the first and second
nucleotides of mRNA codons are generally more deleterious in terms of conse-
quences than mutations at the third position. Using the same permutation as
explained above, we can calculate, for example, that substitutions at the first or sec-
ond position would generate a missense mutation in 91 and 96 % of the cases,
respectively.8 Even if this theoretical computation must be corrected, taking into
account that the nucleotides are not represented in equal proportions in the mouse
DNA, and accordingly that all 64 codons are not equally frequent, this percentage
of non-synonymous mutations is very close to the data actually collected after posi-
tional cloning of hundreds of mutations and analysis of mouse genome sequences.9
Although predictions concerning the possible deleterious effects associated
with missense mutations are difficult and always depend on the genomic con-
text, a number of observations that have accumulated over time provide some
indications. For example, it has been observed repeatedly that non-synonymous
mutations replacing an aliphatic amino acid with an aromatic one (for example
TCG → TGG) have deleterious consequences in most cases. The same applies to
the mutations replacing one of the two amino acids containing a sulphur (S) atom
(Cys or Met) by another amino acid not containing the S atom. Most amino acid
substitutions occurring in the highly conserved domains of proteins almost always
have deleterious consequences. Finally, missense mutations leading to an impor-
tant structural change at the C-terminus often have severe effects by hampering the
8 Four substitutions of the first nucleobase result in a synonymous codon (lysine or arginine
codons). No substitution of the second nucleobase leads to a synonymous codon.
9 In mouse nuclear DNA, the G + C content is 41.70 %, indicating that codons making use of
these two nucleotides are under-represented in this species.

Fig. 7.1 Missense mutations. The severe mouse neurological syndrome called progressive motor
neuronopathy is the consequence of a missense mutation (Tbcepmn-Chr 13) affecting the gene
encoding the tubulin-specific chaperone E protein (TBCE). This missense mutation leads to the
replacement of the very last amino acid of the protein, a tryptophan residue at position 524, by
a glycine (in short: Trp524Gly). This change, which is unique to the mutant mouse and is not
found in any other species, has consequences for the stability of the protein, and this probably
explains the relatively late onset of the pathology (adapted from Martin et al. 2002)
correct folding of the protein, as is the case for progressive motor neuronopathy
(Tbcepmn) (Fig. 7.1).
Accumulation of new data of this kind contributes to the enrichment of data-
bases, and all of these findings are important for a better understanding of the
molecular mechanisms leading to genetic diseases. In this matter, it must be kept
in mind that the information gathered from observations made in the mouse are
universal and accordingly apply to all mammalian species. In human, around 56 %
of the mutations resulting in a pathology are point mutations of the nonsense or
missense types. Analysis of a large number of nucleotide substitutions associated
with disorders shows that the most common substitutions are T to C, C to T, A
to G, and G to A (Krawczak et al. 1998). In humans, the most common type of
single nucleotide substitution is the CpG dinucleotide that mutates to TpG at a fre-
quency which is about five times higher than mutations in all other dinucleotides
(Youssoufian et al. 1988; Antonarakis et al. 1995; Krawczak et al. 1998). There is
no reason to think that this frequency might be different in the mouse.
7.2.2 Base-Pair Substitutions in the Non-coding Regions
Base-pair substitutions in non-coding regions of the genome are innumerable, and

the data gathered from mouse, rat, and human sequencing efforts provide many
examples of such substitutions that, in most instances, have been recorded as mere
SNPs with no detectable phenotypes. Exceptions are when the changes occur in
splicing sites or in regulatory regions. These two kinds of mutations represent,
respectively, 9.3 % and 1.9 % of the mutations associated with a pathological syn-
drome in humans, and it is likely that the proportion is similar in the mouse.
Mutations that interfere with the splicing process result in exon skipping or in
the reciprocal defect known as intron retention. In some other instances, a cryptic
splicing site is activated after a single base-pair substitution, and this results in the
incorporation of a DNA segment of intronic origin into the transcript and possibly
into the encoded protein (Figs. 7.2 and 7.3).
(a)
(a’)
(b)
(c)
(d)
(e)
Fig. 7.2 Examples of splicing defects generated by nucleotide substitutions. a Schematic repre-

sentation of a normal gene. Exons are shown as grey boxes and introns as lines between exons. a′
represents the mature mRNA transcript after splicing of all introns. b A nucleotide substitution in
a 3′ splicing site results in the skipping of an exon. c A nucleotide substitution leads to the activa-
tion of a cryptic splicing site and results in the incorporation of some intronic sequence into the
mRNA transcript. d A nucleotide substitution deactivates the normal splicing site, while a cryptic
one is used a few base-pairs downstream in the intronic sequence. e The substitution leads to the
skipping of the last exon. All of these situations have been observed after positional cloning of
mouse mutations
All types of splicing defects that are theoretically possible have been actu-
ally identified in the mouse, altering more or less significantly the function of the
encoded protein. A situation that is quite common and has severe consequences
is when a 3′ splicing site (3′ss) is altered, leading to the attachment of a stretch
of intronic DNA at the 3′ end of the mRNA molecule. In this case a number of
amino acid residues are added to the C-terminus of the protein until, by chance, a
stop codon occurs to terminate the aberrant transcription. In this case the protein
is almost always abnormally folded and accordingly non-functional. Sometimes it
also happens that cryptic 3′ or 5′ splice sites are activated after a single point muta-
tion. In this case the consequences are unpredictable although, in general, severe.
Unlike for the splicing sites, mutations affecting DNA binding sites or regula-
tory regions are not common. This is either because these sites do not represent an
important target in which mutations can occur or, alternatively, because mutations
occurring at these sites have consequences that are not critical and accordingly are
more or less tolerated or compensated for.
Most of the spontaneous mutations which have been found in the mouse, and
which have been characterized at the molecular level after positional cloning, have
(a) (b)
(c) (d)
Fig. 7.3 Mutations resulting in abnormal splicing. Lrp4mdig and Lrp4dan are two independent
recessive mutations affecting the gene encoding the mouse lipoprotein receptor 4 (Lrp4-Chr 2).
a Schematic representation of exons 14–17 of the Lrp4 gene indicating skipping of exon 15 in
Lrp4mdig/Lrp4mdig mice. b RT-PCR amplifications performed on total cDNAs with specific prim-
ers (green arrows) allow the detection of an amplification product of the expected size in wild
type (+/+) whereas only a faint band is observed with Lrp4dan/Lrp4dan cDNA. This is because
a retroviral insertion in intron 2 of the Lrp4dan allele hampers the transcription of a messenger
RNA. However, the retroviral insertion does not suppress the transcription entirely since a faint
band can be observed with cDNAs from homozygous Lrp4dan/Lrp4dan. PCR amplification with
the same primers yields a product shorter than expected in homozygous Lrp4mdig/Lrp4mdig mice.
Here again, skipping of exon 15 is probably not absolute since a faint band is still observable. c
and d Genomic sequence in Lrp4+/Lrp4+ and Lrp4mdig/Lrp4mdig co-isogenic mice. An A → T
transversion alters the splicing donor site 3′ of exon 15 (from Simon-Chazottes et al. 2006)
been explained by the observation of a non-ambiguous structural defect. Among

the few exceptions, one may cite the case of the Agtpbp1pcd allele at the ATP/GTP
binding protein 1 locus (formerly known as Purkinje cell degeneration—pcd; Chr
13). At this locus six spontaneous alleles and five chemically induced alleles have
been reported, which all belong to the same complementation group (i.e., they fail
to complement each other in a complementation test). For all the mutant alleles,
obvious changes have been described in the coding region or splicing sites except
for the original Agtpbp1pcd allele. For this allele, Northern blot analysis failed
to detect a transcript in all tissues of homozygotes except for the testis, where
reduced levels were noted. In this case, the researchers suggested that the struc-
tural defect for this mutation should likely be in a regulatory region. However, as
of today, the question is still open (Fernandez-Gonzalez et al. 2002).
With the rapid development of DNA sequencing techniques and the concomi-
tant reduction in costs, it is likely that many regions of the mammalian genomes
suspected of having particular importance in the regulation of gene expression
will be easily compared between different strains or subspecies. In so doing, many
point mutations of potential interest are likely to be discovered outside of splicing
sites and regulatory regions. The discovery of a point mutation in the seed region
of miRNA96, which is responsible for or associated with the semi-dominant deaf-
ness phenotype of Diminuendo mice (Mir96Dmdo), is a good example and might be
the first in a long series of such findings (Lewis et al. 2009).
7.2.3 Insertions, Deletions, and Duplications
Insertions are mutations resulting from the intercalation of a DNA sequence of

variable size into the genome. The reciprocal alterations, those that are character-
ized by a missing sequence or portion of DNA, are called deletions. Insertions/
deletions can be as small in size as a single nucleotide or, on the other hand, they
can expand over several kilobases of DNA, affecting a variable number of genes
on a chromosome and sometimes making their analysis difficult.
When aligning DNA sequences in the non-coding regions it is not always easy
to select the appropriate designation between insertion and deletion. Sometimes it
is noted that a single nucleotide makes a difference, but it is impossible to deter-
mine whether the mutation represents an insertion in one of the sequences or a
deletion in the other. The situation is even more complex when this single nucle-
otide difference is frequent and co-localized across different strains. For these
cases, geneticists have coined the word indel (from insertion/deletion), indicating
their ignorance concerning the historical sequence of the molecular change and
the co-existence of the two forms as alleles. In short, an indel is a gain or loss in
nucleotides, at a specific site, that is polymorphic in a given species.
Microindels are indels that result in a net gain or loss of 1–50 nucleotides.
Insertions, deletions, and indels are potentially innumerable since nucleotides can
be either deleted or inserted almost anywhere in a DNA strand as a consequence
of aberrant replication, unequal crossing-over or transposition. Interestingly, how-
ever, deletions are more commonly observed in practice than insertions in both the
mouse and human genomes (17 % versus 6.4 %, respectively).
Deletions or insertions of nucleotides have consequences when they occur in an
open reading frame (ORF), in the close vicinity of splicing sites or at DNA binding
sites. When they occur in an ORF and have a size greater or less than three nucle-
otides (i.e., a number not divisible by three), they result in frameshift mutations,
whose effects are similar to those of the mutations occurring in the splicing sites
and are transcribed, in general, into aberrant mRNA molecules (Perez et al. 2013).
When indels have a size of three or a multiple of three nucleotides, they result in
the incorporation of additional amino acids into the protein chain, and their effects
are difficult to predict. One such example has been described for another allele at
the same Agtpbp1 locus (already mentioned above), the Agtpbp1pcd-5J allele of
spontaneous origin. Positional cloning of this mutation demonstrated that, in this
case, a GAC triplet was inserted at position 775, adding an additional aspartic acid
(Asp) to the protein. Northern blotting demonstrated comparable expression to that
of wild-type mice, indicating normal RNA expression. However, Western blot anal-
ysis showed that the protein level is dramatically reduced (Chakrabarti et al. 2006).
Many mouse mutations of spontaneous origin, or discovered via studies of
the effects of radiation on the germline, are the consequence of deletions encom-
passing several contiguous genes. Although common, this type of mutation is of
limited interest for modeling human defects or even for annotating the mouse
genome, because it is in general difficult to establish a direct link between a par-
ticular phenotypic trait and the genotypic defect. The mouse mutation oligotriche
(olt-Chr 9) is an example of such a deletion. This mutation has been found to be
a 234-kb deletion affecting no less than six contiguous genes: Vill, Plcd1, Dlec1,
Acaa1b, and parts of Ctdspl and Slc22a14, but the gross phenotypic expression is
relatively modest: some hair loss on the hind legs and male sterility due to severe
sperm defects (Runkel et al. 2012).
Duplications are another type of mutation whose effects and consequences are
similar to insertions. The gene encoding the leptin receptor (Lepr-Chr 4), with all
its many alleles, is a good example illustrating both indels and duplications (see
Fig. 7.4).
7.2.4 Triplet Expansions
Triplet expansion or trinucleotide expansion is a defect in DNA replication that

is responsible for a dozen severe human diseases (i.e., Huntington disease,
Friedreich ataxia, X-fragile syndrome, Kennedy syndrome, Steinert myotonic
dystrophy, to mention just a few). These diseases are characterized at the DNA
level by an increase in the number of tandemly repeated specific trinucleotides,
for example CGG, CTG, CAA or CAG, occurring in specific genes and caused
by slippage during DNA replication. Huntington disease (HD), for example, is
caused by the expansion of CAG repeats in the gene encoding huntingtin (HTT).
The number of CAG repeats increases with age in some patients and encodes an
expanded glutamine (Gln) tract within the huntingtin protein. When the number of
repeats passes a critical number (actually 36 for HTT), then the enlarged polyglu-
tamine fragment in the protein leads to the formation of the huntingtin aggregates
that are observed in the brain as well as in some other tissues, leading to severe
pathologies.
Although the mechanism leading to triplet expansion is only poorly understood,
geneticists have established that the number of repeats is frequently variable from
ex1 ex2 ex3 ex4 ex5 ex6 ex7
(a)
ex1 ex2 ex3 ex4 ex5 ex6’ex4 ex5 ex6 ex7
Lepr db-Pas1
(b) ex1 ex2 ex3 ex4 ex11 ex12 ex13 ex14

Lepr db-Pas2

..TGGAGAAAAAT..
..TGGAG^AAAAT..
Lepr db-Pas2/Lepr db-Pas2
Fig. 7.4 Mutations resulting from duplications and deletions. In the mouse, over 15 spontaneous
mutations have been reported at the locus of the gene encoding the leptin receptor (Lepr-Chr 4).
This gene normally consists of 18 exons and has multiple splice variants, comprising at least five
isoforms. a Among these mutant alleles, Leprdb-Pas1 is the consequence of a partial duplication
that spans the entirety of exons 4 and 5, plus 21 bp of coding exon 6 (as well as the two introns
between exons 4 and 6). This produces a null allele that is unable to encode a functional receptor
(from Liu et al. 1998). b Another spontaneous allele, Leprdb-Pas2, is the consequence of a 1-bp
deletion producing a frameshift in exon 12, altering another domain of the protein. The mutant
allele is inactive and the mouse becomes obese and diabetic
tissue to tissue in the same patient, suggesting that distinct expansion processes
can occur in different tissues. Human geneticists have also established strong cor-
relations between the length of the triplet repeats and the severity of the disease.
Such spontaneous cases of trinucleotide expansions have not been reported in
the mouse but mouse models of HD, displaying phenotypes relevant to the human
disease, have been created by transgenesis (Menalled and Chesselet 2002). These
models will aid the understanding of the fundamental mechanisms underlying
unstable triplet expansion in humans, and hopefully will also provide useful tar-
gets for inhibiting disease development.
7.2.5 Mutations Resulting from the Insertion

of Mobile Elements
As we discussed in Chap. 6, many mobile elements (retrotransposons, retroviruses,

LINEs, SINEs, etc.) are well-known and quantitatively important components of the
mouse genome. These elements move within the genome by duplication and retro-
transposition and, depending on their integration site, they may have a mutagenic
action. Many such mutations have been identified in the mouse. For example, the
dilute (Myo5ad-Chr 9) mutation, a very ancient mutation of the mouse with sev-
eral alleles, is the result of the integration of the ecotropic murine leukemia virus
Emv-3 into the myosin VA (Myo5a) gene. The a (non-agouti-Chr 2) mutation is
also the consequence of the insertion of a 5.5-kb virus-like element (VL30) into
the first intron of the agouti gene, which interferes with the transcription process.
At the same Agouti locus, we previously reported the case of the dominant muta-
tion Avy (viable yellow), which is the consequence of the insertion of an intra-cis-
ternal A-particle (IAP or retrotransposon) into a non-coding exon at the 5′ end of
the agouti gene. Similarly, the spontaneous mutation spastic (Glrbspa-Chr 3) results
from the insertion of a 7.1-kb LINE-1 element within intron 6 of the gene encod-
ing the glycine receptor, beta subunit (Mülhardt et al. 1994). Finally, the hairless
(Hrhr-Chr 14) mutation in mice was caused by the insertion of a murine leukemia
virus into intron 6 that results in aberrant splicing of the Hr gene (Stoye et al. 1988).
Some strategies have been designed to make use of the capacity of transposons to
move in the mammalian genome, for the induction of new mutations in the mouse
and mostly in the rat. We will come back to this point later in this chapter (Sect. 7.6).
7.2.6 Mutations Due to Non-homologous Recombination

or Non-homologous End Joining
Non-homologous DNA recombination or non-homologous end joining (NHEJ)

occur in mammalian genomes when double-strand DNA breaks are imprecisely
repaired, leading to loss (or duplication) of a segment of nucleotides. Spontaneous
mutations that are the consequence of NHEJ have been reported in humans (for
example, a β-thalassemia leading to hemoglobin Lepore syndrome). To date, no
such mutations have been reported in the mouse, although they may theoretically
occur spontaneously. However, the NHEJ DNA repair mechanism, along with
homologous recombination, is the molecular basis of new genome editing tech-
nologies with engineered nucleases (this point will be discussed in Chap. 8).
7.2.7 Copy Number Variations
As already discussed in Chap. 6, structural changes that result in copy number var-
iations (CNVs) in a specific chromosomal region are common in all genomes. In
the mouse, approximately 100 genomic regions across the 19 autosomes have been
shown to harbor CNVs, ranging in size from 20 kb to 2 Mb, with more than 90 %
sequence conservation. These CNVs may be considered to be mutations of a new
class: the “multi-duplications”. They certainly affect gene expression by altering
the transcript dosage and, accordingly, the phenotypic variability in genetic dis-
eases by affecting the penetrance of the trait (Cutler and Kassner 2008). CNVs
probably play an important role in quantitative genetics.
7.3 Spontaneous Mutation Rates 235
7.3 Spontaneous Mutation Rates
Spontaneous mutation rates are difficult to assess accurately in any mammalian

species for a number of reasons. First, it is clear that only a fraction of the muta-
tions are detectable at the phenotypic level, and this fraction fluctuates from one
locus to the next. For example, dominant alleles resulting in lethality in utero
or shortly after birth and those impairing the reproductive capacities of animals
are often not even identified as heritable traits. Another major difficulty in the
detection of mutations is that some of them frequently exhibit wide variations
in expressivity or have a very subtle phenotypic expression and accordingly are
underevaluated. Recessive mutations are easier to detect because they are in gen-
eral observed recurrently, especially when they occur in an inbred strain but, even
in this case, many mutations have not been identified simply because they have a
late onset or because they are expressed only in some particular conditions. For
example, most inbred mouse strains are susceptible to experimental infections
with flaviviruses (yellow fever or dengue, for example) while a few others are
resistant. This susceptibility is caused by a recessive mutation (Oas1b locus-Chr 5)
and was discovered incidentally during an experiment, but the mice of both strains
(resistant and susceptible) look perfectly “normal” for all other characteristics and,
for this reason, the mutation remained undetected for many years. Similarly, some
strains are susceptible to the antiparasitic drug ivermectin while most others are
resistant but, here again, the mutation is cryptic, conditional, and can be detected
only when the drug is administered. In conclusion, one can say that the identifica-
tion of a spontaneous mutant phenotype depends upon the quality and accuracy of
the phenotyping and, as a consequence of this, one must bear in mind that muta-
tion rates are, in general, underestimated unless they are computed at a specific
locus.
The first estimation concerning the mutation rate towards a recessive allele was
published in 1966 by two scientists at The Jackson Laboratory (Schlager and
Dickie 1966, 1967). Their estimations were established from the observation of
1,349,725 interstrain F1 progeny at five specific coat-color loci (non-agouti, a;
brown, formerly b and now Tyrp1b; albino, formerly c and now Tyrc; dilute, for-
merly d and now Myo5ad; and leaden, formerly ln and now Mlphln).10 The authors
reported 12 mutations from the wild-type allele towards a recessive allele (forward
mutations) and calculated an average mutation rate of 8.9 × 10−6 per locus per
gamete (with 4.6–15.5 × 10−6 as 95 % confidence limits).
Another estimation based on similar crosses, although between different
strains, was published a few years later by (Russell and Russell 1996). A total of
1,485,036 F1 progeny were scored at seven loci (the same five as above except
10 Computations of the mutation rates were made on several interstrain F1 hybrids expected to
be all heterozygous for one or several of the recessive coat color alleles and the corresponding
wild-type allele. In such an F1 population, the mice with a non-wild-type phenotype are potential
carriers of a new mutant allele. This was confirmed by setting up separate crosses.
PT stock C57BL/6
a/a; b/b; c ch-p/ cch-p; d-se/d- se; s/s a/a; +/+; +-+/+-+; +-+/+-+; +/+
F1 population
a/a; +/b; +-+/cch-p; +-+/d-se; +/s
Fig. 7.5 Assessing the mutation rate at specific loci. Mice of the PT stock are homozygous for seven
recessive mutant alleles involved in the determinism of coat color. When crossed with mice of the
C57BL/6 inbred strain (which are non-agouti a/a and homozygous for the wild type allele at the
other six loci), all F1 are expected to have a non-agouti (a/a = solid black) coat color phenotype.
Phenodeviants, with a coat color different from the expected one (boxed), are potentially heterozygous
for a new recessive allele at one of the six loci of the PT stock, and their status must be characterized
by additional crosses. This historical PT stock, developed at Oak Ridge by W. Russell and colleagues,
has been extensively used for assessing the mutagenic activity of radiation or chemical compounds.
Another similar stock, the HT stock, with different alleles has been developed at MRC Harwell
leaden Mlphln, plus pink-eyed dilution Oca2p, piebald Ednrbs, and short ear
Bmp5se) and the authors calculated a rate of 6.6 × 10−6 mutations per locus per
generation.11 In addition to the “complete” mutations, the same authors also found
several “mosaic” mutations at five loci, which led them to calculate a corrected
mutation rate of 11 × 10−6 per locus per generation (Fig. 7.5).
These mutation rates, calculated independently, are relatively close to each
other and definitely represent a good estimation for the loci described above.
However, this rate (~10 × 10−6 mutation per locus per generation) is certainly not
representative of the “average” mouse locus because the same scientists at The
Jackson Laboratory reported a total of only 28 recessive mutations at 26 different
loci from a total of 83,368,463 mice examined, yielding an overall spontaneous
11 These observations were made on the F1 progeny of a cross between a tester stock, known as
PT stock, homozygous for seven fully penetrant recessive alleles, and mice homozygous for the
wild-type alleles at the same seven loci.
7.3 Spontaneous Mutation Rates 237
recessive mutation rate of 6.7 × 10−7 per locus per gamete (95 % confidence lim-
its: 5.1–8.7 × 10−7). This rate, which is only 1/13th of the rate calculated for the
forward mutations at the five/seven specific coat-color loci, is probably a better
estimate of the overall spontaneous mutation rate towards a recessive allele in the
mouse. This was confirmed by scientists working at Harwell using an independ-
ent tester stock, the so-called HT stock, homozygous for six recessive alleles with
only one recessive allele (non-agouti a) in common with the PT stock.
Schlager and Dickie also recorded the number of mutations towards a dominant
allele. They collected this information by observing breeding colonies during a
3-year period (36 mutations were collected from a total of 67,161,745 mice), yield-
ing an estimated spontaneous mutation rate of 0.54 × 10−6 per locus per gamete,
with 95 % confidence limits of 0.38–0.74 × 10−6 (Schlager and Dickie 1967).
A careful analysis of the mutations (both recessive and dominant) collected by
the scientists at The Jackson Laboratory indicated that there are great differences
in the mutation rates at the different loci. As we already mentioned, this is cer-
tainly a consequence of the fact that many mutant alleles escape detection either
because of their unobtrusive (or very severe!) phenotype or late onset phenotype.
This may also be explained by differences in the size of the different loci at the
DNA level or the splitting of the coding regions into many exons, offering a wider
target to the mutagenic events. However, these two explanations are clearly not
sufficient to explain some of the observed differences, and it is now well estab-
lished that some genes have an unexpectedly higher mutation rate than average.
This is the case, for example, with the gene encoding the Kit receptor tyrosine
kinase (Kit-Chr 5), in which 18 spontaneous mutant alleles were recorded in a
population of mice analyzed by Schlager and Dickie during their survey.12 This is
also the case with a locus on chromosome 4, where no less than seven independent
mutations were found in a single experiment (Kiernan et al. 2002). Other exam-
ples are the non-agouti locus (a-Chr 2) with 58 spontaneous alleles, and the dilute
locus (Myo5a-Chr 9 with 53 alleles. Regardless of the loci and observed variations
in the mutation rates, these rates remain very low. This explains why mammalian
geneticists, like other geneticists, have invested in the development of strategies to
increase the rates of mutation.
7.4 Mutagenesis in the Mouse
Over the last century, mice have been extensively used by geneticists as “living test
tubes” for assessing the genetic hazards associated with the domestic use of nuclear
energy. Mice have also been used by toxicologists for assessing the mutagenic activity
of potentially hazardous chemical compounds in the human environment (drugs, food
additives, pollutants, pesticides, etc.), and hundreds of mutations of all types have been
12 On a total of 36 dominant mutations.

produced as “by-products” of these activities. These mutations, in addition to the spon-

taneous mutations that were previously collected at low frequency in breeding facili-
ties, have been instrumental for the development of mouse genetic maps because, at
that time, they were the only available genetic markers. They also provided geneticists
with many potentially interesting models of human diseases. However, experimental
mutagenesis sensu stricto, which means the treatment of animals with known mutagenic
agents to purposefully increase the mutation rate, is only recent.
7.4.1 Gametogenesis and Experimental Mutagenesis
Experimental mutagenesis consists of exposing progenitors of either sex to a

mutagenic agent, with the aim of increasing the occurrence of novel mutations in
the progenies of the treated animals. For practical reasons, only male progenitors
are exposed to the mutagens because spermatogenesis is a continuous process,
starting at puberty and lasting several months or even years. In females, on the
contrary, gametogenesis is a cyclic process and the number of cells that are poten-
tial targets for mutagenesis is much reduced in adult mice (Fig. 7.6).
Spermatogonia are the stem cells of the male germline. When they divide they
produce two daughter cells: a spermatogonium type A0, which stays in the pool of
Stage XII
Me SC
Step 14 ESp
Stage II
Step 11 ESp
Step 2 RSp
Stage XI
PS
ZS
Sg
SC
Fig. 7.6 Histological appearance of seminiferous epithelium. Sections from an adult mouse tes-

tis indicating several stages of the spermatogenetic process. SC = Sertoli cell, Sg = spermatogo-
nia, ZS = zygotene spermatocytes, PS = pachytene spermatocytes, Me = meiotically dividing
spermatocytes, step 2 RSp = step 2 round spermatids, step 11 Esp = step 11 elongating sperma-
tids, step 15 Esp = step 14 elongating spermatids, stage II, stage XI, and stage XII = tubules in
different stages of the spermatogenic cycle (Figure courtesy of Dr. Dianne Creasy, Huntingdon
Life Sciences, East Millstone, NJ, USA)
7.4 Mutagenesis in the Mouse 239
stem cells, and a spermatogonium type A1 that undergoes several mitotic rounds,
producing A2, A3, A4, and Intermediate types, and finally type B spermatogonia.
The type B spermatogonia divide and form pre-leptotene spermatocytes, which
are almost identical to type B spermatogonia in appearance, but they become
much larger as they duplicate their chromosomes to form tetraploid cells and pro-
ceed through meiotic prophase (zygotene, pachytene, diplotene, and diakinesis).
The first meiotic division produces two short-lived diploid secondary spermato-
cytes, which rapidly divide again (second meiotic division) to produce four round
haploid spermatids. These round spermatids then undergo a complex morphologi-
cal transformation into spermatozoa, developing condensed heads covered by an
acrosome and attached to a motile tail, which are then shed into the tubular lumen
(spermiation). In theory, a single A1 spermatogonium would give rise to 256 sperm
cells in 5 weeks, but there is some attrition of cells during spermatogenesis so that
the actual number of sperm is smaller than the theoretical maximum. A few of
these mature sperm cells will fertilize ova, and most others are eliminated while a
new cycle of spermatogenesis follows. The duration of the spermatogenetic cycle
is much shorter in the mouse than in most other species; spermatogonia become
mature spermatids that are released into the lumen in only 5 weeks (Russell et al.
1990). By comparison, the spermatogenic cycle is 8 weeks in the rat and 10 weeks
in humans. It then takes another 1–2 weeks for the released sperm to reach the tail
of the epididymis, where they are stored prior to ejaculation.
Mutagenic agents (physical or chemical) exert their effects as soon as they are
in contact with the genetic material of the treated mice and this effect terminates, in
general, immediately or shortly after treatment ends. The cells that have been muta-
genized repair most of the damage resulting from the treatment, but, depending on
the severity of this damage, some cells may recover and pass genetic alterations to
their daughter cells while others die and are eliminated. The success of a mutagenic
treatment is reflected in the percentage of cells that survive and carry a mutation,
and the higher this percentage the better. As we will discuss, this depends upon the
mutagenic treatment, the type of cells exposed to the mutagen, the dose and dura-
tion of the treatment, and the dose rate and the possible splitting of the dose.
Because spermatogenesis is a continuous and precisely timed process, we can
calculate the precise stage of development of a specific germ cell, at the time of
exposure to a mutagen, depending on the time elapsed between the treatment and
the fertile mating. For example, if male mice are exposed to a mutagen and mated
3–4 weeks later, the embryos that result from the mating will have originated
from germ cells that were mature spermatids (post-meiotic stage) or spermatozoa
entering the epididymis at the time of treatment. In contrast, if the mating takes
place more than 7 weeks after the treatment, the embryos result from cells that
were exposed as spermatogonia. When the stem cells of spermatogenesis (i.e., the
spermatogonia A0) are successfully mutagenized, the male becomes a permanent
provider of mutations. On the other hand, when the targeted cells are post-meiotic
(spermatids or spermatozoa), the mutagenesis is transient.
An important point to mention is that a very efficient selection process oper-
ates during gametogenesis to eliminate the mutations that may have occurred either
spontaneously or after the mutagenic treatment. This process is much more effi-
cient during the early (diploid) phases of gametogenesis, where the cells divide and
have an active metabolism with efficient DNA repair mechanisms, than during the
haploid phase, when the cells differentiate but no longer undergo mitosis. In the
same way, meiosis occurring at the spermatocyte stage is an efficient filter to elimi-
nate the chromosomal rearrangements that interfere with the normal distribution
of chromosomes in the daughter cells. Reciprocal translocations or inversions, for
example, are strongly counter-selected when they occur in spermatogonia, whereas
many of them are transmitted to the offspring when induced in early spermatids.
When males receive a mutagenic treatment, the number of affected stem cells
depends on the dose. If the dose is elevated, most spermatogonia are killed and
the male becomes permanently sterile. Conversely, if the dose is too low, the lethal
effect is limited but the mutation rate is low and the experiment might not be suc-
cessful. Selecting the best dose is very important and may require preliminary
experiments.
7.4.2 The Induction of Mutations by Radiation
Hermann Muller (1927) was among the first to report that X-rays can cause muta-
tions and chromosomal damage in Drosophila flies. However, most of the knowl-
edge geneticists have gathered concerning the mutagenic effects of radiations in
the mouse results from research conducted at MRC Harwell in England and at
Oak Ridge National Laboratory in the United States. An excellent review of these
fundamental studies, which may still be useful, can be accessed online in the book
“Biology of the Laboratory Mouse” in a chapter by Green and Roderick (1966).
In short, we can say that all types of radiation are mutagenic, provided they have
sufficient energy to come into contact with the genetic material. Cosmic radia-
tion, a mixture of photons and high-energy protons originating from outer space,
constantly showers on all living organisms and is probably responsible of many
“spontaneous” mutations. In contrast, UV radiation, consisting of photons with a
wavelength between 100 and 400 nm, is mutagenic (and carcinogenic!) only for the
cells of the epidermis. Their energy is insufficient to reach the gonads, and accord-
ingly their impact on the genetic material of mammalian species is virtually nil.
Countless experiments have been performed to understand the mutagenic
effects of electromagnetic (X- and γ-rays) and corpuscular (protons and
β-particles) radiation. These types of radiation are mutagenic because they have a
direct effect on the chromosomes and DNA strands; they produce breakages or
deletions that are more or less efficiently repaired, depending on the extent of the
damage and the efficiency of the repair mechanisms. They are also mutagenic
because they produce ionization as they dissipate their energy into living matter,
producing a very large number of hydroxyl and hydroperoxyl free radicals that are
highly reactive and diffusible elements. From the experiments conducted by health
physicists between 1950 and 1970, it was concluded that the mutagenic activity
and the type of mutations produced by exposure to radiation depend on a physical

parameter known as linear energy transfer (LET), and, of course, on the dose dis-
tributed, the duration of exposure, and whether the dose is fractionated. Heavy
particles like protons or α- particles have a very high LET and dissipate their
energy over a short distance while passing through living matter. Accordingly, they
exhibit high mutagenic activity and produce extensive chromosomal breakage. On
the other hand, photons such as X- and γ-rays have a much lower LET and are
much less mutagenic, producing mostly point mutations or small-sized deletions.
For X- and γ-rays, the rate of induced mutations varies linearly with the dose from
0 to 7 grays (abbreviated Gy).13 Beyond 7 Gy repair mechanisms are saturated,
and many cells are affected by several mutations and die.
When a dose of X- or γ-rays is distributed over a short period of time (at high
dose-rate), the mutagenic effect of the radiation is more intense compared to the
same dose distributed over a longer period of time. Similarly, a single dose of radi-
ation is more damaging to the genetic material than the same dose split into sev-
eral sessions. This is a consequence of the fact that the DNA repair mechanisms
are saturated when the dose is delivered over a short period of time. Males, whose
germ cells are constantly in mitotic activity (from puberty until death), are more
susceptible to the mutagenic effects of radiation than females, whose germ cells
are resting at the time of birth.
All germ cells are sensitive to radiation, but haploid cells (post-meiotic stages,
with n chromosomes) are more sensitive than spermatogonia (2n) because the
DNA repair mechanisms are almost inactive in these highly differentiated cells.
As we already mentioned, mutations induced in spermatogonia may be transmitted
to the offspring throughout the life of the mutagenized animal, whereas mutations
affecting the post-meiotic haploid cells are transmitted only during the short lifes-
pan of these cells (3 weeks), provided that they fertilize an oocyte.
In mice, the mutation rate after exposure of spermatogonia to 10 Gy of X-rays
at 0.9 Gy/min, split into two doses of 5 Gy distributed 24 h apart, was reported
to be ~ 500 × 10−6 per locus per gamete, compared to the spontaneous rate of
~10 × 10−6 as mentioned above (Russell 1962, 1963). This mutation rate seems
to be the highest possible for X-and γ-rays (~50× the spontaneous rate). This
increase in mutation rate due to the splitting of the dose suggests that the first
irradiation imposes some sort of synchronization and enhances mutability of the
cells, while the second dose yields more mutations than otherwise expected. One
could, in theory, obtain a higher frequency of point mutations by using neutrons.
However, most mutations produced by this type of radiation are deletions that are
frequently incompatible with the survival of heterozygotes. Deletions are also dif-
ficult to analyze in the molecular context, in particular when they encompass more
than one gene, as is often the case.
13 Since 1970, the gray (Gy) has replaced the rad as a unit of absorbed radiation in terms of
energy per unit of mass. One gray corresponds to one joule of energy absorbed per kilogram of
living matter. One Gy is equal to 100 rads.
7.4.3 The Induction of Mutations by Chemicals
Studies of the mutagenic activities of chemicals were initiated by C. Auerbach

(Auerbach and Robson 1946; Auerbach 1962), who first reported that
1,1′-thiobis[2-chloroethane], a chemical warfare agent known as “mustard gas”
and used during World War I, could cause mutations in Drosophila flies. Since
these initial studies, toxicologists have identified a large number of chemicals with
mutagenic activity. Identification of such molecules has been rationalized by the
introduction of laboratory tests using bacteria (for example, the Ames test devel-
oped in the 1970s; Ames et al. 1973). More recently, transgenic mice with sev-
eral copies of bacterial genes integrated into their genome have been developed
as tools for mutation assays; for example, the lacI model, commercially avail-
able as the Stratagene Big Blue® mouse, and the lacZ model, available as the
Muta™Mouse (Wahnschaffe et al. 2005a, b). These tests, which are very sensitive,
relatively inexpensive, and simple to use, allowed the establishment of a very long
(and ever-increasing) list of substances with demonstrated mutagenic activity in
mammals.
Chemical mutagens have been classified into four categories based on the type
of interaction they have with DNA (Vogel and Rohrborn 1970). The first category
includes molecules known as base analogs. Molecules of this category (6-amino-
purine, for example) are mistakenly used by bacteria during DNA synthesis, lead-
ing to the production of transitions or transversions after replication. However,
these substances have not been found to be mutagenic in mammals, probably
because the metabolic pathways leading to the synthesis of nucleotides are not
exactly the same in mammals and in bacteria.
Intercalating agents represent another important group of mutagens. Examples
include acridine orange, ethidium bromide, and proflavine. These molecules insert
into the DNA helix and bind covalently to the bases of the two strands, leading to
deletions occurring during the next round of replication. As with the base analogs,
these substances have little effect on pre-meiotic germ cells of mammals and have
not been used frequently for the purpose of experimental mutagenesis. Some of
these agents, however, are active on post-meiotic germ cells and induce transloca-
tions and deletions at a low rate.
The third class of mutagens includes the deaminating agents that are best rep-
resented by nitrous acid and sodium bisulfite. Because deamination of guanine
(G) or adenine (A) occurs spontaneously in most eukaryotic cells, it has been sug-
gested that deaminating agents might be mutagenic by increasing the basic, natural
level of deamination. This, however, has never been clearly demonstrated in mam-
mals, and these agents are not currently used for mutagenesis.
Alkylating agents are, by far, the most potent mutagens in mammals (mus-
tard gas belongs to this category). Molecules of this type transfer alkyl radicals
(methyl, CH3, or ethyl, C2H5) onto DNA bases, particularly on adenine but also
on guanine. If this alkylation is not repaired promptly by the DNA repair mecha-
nisms, transitions or transversions ensue during the next step of replication.
Alkylating agents are mutagenic in the mouse, but most of them are only active
on post-meiotic germ cells (type-2 spermatocytes or spermatids). Among these
substances, we must mention the anticarcinogenic drugs TEPA and Thiotepa™,
ethyl methane sulfonate (EMS), methyl methane sulfonate (MMS), triethylen-
emelamine (TEM), procarbazine, and chlorambucil, all of which have been used
during the last thirty years as chemical mutagens in the mouse.
In 1979, William Russell from Oak Ridge National Laboratory reported that
a simple alkylating agent, N-ethyl-N-nitroso-urea (ENU), has considerable muta-
genic power, and even more remarkably, that this substance is active on both pre-
and post-meiotic germ cells (Russell et al. 1979). These observations had a major
impact on genetic research and must be considered as an important milestone in
the history of mouse genetics (Fig. 7.7).
ENU is generally sold in the form of a light yellow powder in dark glass bottles,
sealed with a rubber stopper. This packaging makes the chemical relatively easy to
handle safely. The molecule is light-, heat-, and pH-sensitive and does not dissolve
easily in water, but adding a few drops of ethanol avoids this drawback. The muta-
genic activity of ENU results from its capacity to transfer an ethyl group to oxy-
gen or nitrogen radicals in the DNA molecule, inducing mis-pairing and ultimately
leading to base-pair substitutions or deletions (Van Zeeland et al. 1989; Vogel and
Natarajan 1995). In fact, the mutagenic activity of ENU results from two mecha-
nisms acting in opposite directions: the alkylation of the DNA molecule resulting in
Fig. 7.7 Alkylating agents with

mutagenic activity. a Methyl- (a)
CH3 – O – SO2 –CH3
methane-sulfonate (MMS).
b Ethyl-methane-sulfonate
(EMS). c N-Ethyl-N-nitrosourea (b)
(ENU). d N, N′, N″- CH3 – CH2 – O – SO2 – CH3
Triethylenethiophosphoramide
(ThioTEPATM). All four
molecules are potent mutagens
but ENU is, by far, the most NH2
potent and the only one active CH3 – CH2 – N – C
(c)
on pre-meiotic germ cells O
N
O
CH2 CH2
(d) N
CH2 CH2
N–P–N
CH2 CH2
S
the creation of adducts on the one hand, and the efficiency of the enzymatic DNA
repair mechanisms on the other. In spermatogonia, the ENU-alkylated nitrogen
atoms are efficiently repaired, while ENU-alkylated oxygen atoms are repaired with
a much lower efficiency.
Many ENU-induced germline mutations have been studied at the molecular level
after positional cloning and it has been found that, in the great majority of cases,
adenine (A) is the main target of ENU activity with the primary genetic alteration
being either AT to TA transversions or AT to GC transitions (Justice et al. 1999).
The mutagenic activity of ENU has been evaluated using several tests (Russell
et al. 1979; Favor 1986; Lewis 1991; Lewis et al. 1991, 1992; Favor 1994; Ashby
et al. 1997; Schmezer and Eckert 1999). In his initial paper, (Russell et al. 1979)
found 35 confirmed mutations at the seven specific loci mentioned above (those
homozygous in the PT stock) among 7,584 offspring in the treated group (one
injection of 250 mg/kg of body weight), compared to 28 mutations among 531,500
mice in the control group. This indicated a mutation rate 90 times higher than the
spontaneous rate and five times higher than for 6 Gy of γ-rays.
Plotting the mutation rates calculated with the same “multiple loci” assay to the
doses of ENU injected in male mice, (Favor et al. 1990) observed that the muta-
tion rate for ENU increased roughly linearly with dose, from the threshold dose
of ~34 mg/kg of body weight up to 300 mg/kg, a dose that seems to be the high-
est tolerable by an adult mouse. If the dose remains low, say less than 30 mg/kg
of body weight, the mutation rates are not significantly different from the rate of
spontaneous mutations in the same assay. Favor’s calculations can be summarized
in the following two formulae:
MR × 10−5 = (1.2 ± 0.3) for D < 33.9 mg/kg
MR × 10−5 = (1.2 ± 0.3) + (0.4 ± 0.05) × (D – (33.9 ± 5.0))
for D ≥ 33.9 mg/kg
where MR = mutation rate and D = dose in mg/kg of body weight.
The threshold effect observed by Favor and colleagues is probably explained
by the fact that, when the number of alkylated sites remains low, the repair mecha-
nisms can cope, but when it becomes high or very high, these mechanisms become
saturated and mis-pairing increases in proportion to the dose of mutagen.
W. Russell and colleagues reported a few years after their initial publica-
tion that three or four injections of 100 mg/kg of body weight, each delivered
at weekly intervals, enhanced the mutation rates by a factor 1.8 and 2.2, respec-
tively, compared with a single dose of 250 mg/kg of body weight, while allow-
ing greater survival and fertility of the treated mice (Russell et al. 1982a, b;
Hitotsumachi et al. 1985). With such a treatment, the maximum mutation rate
of 125–152 × 10−5 per locus could be obtained that roughly corresponds to 150
times the spontaneous mutation rate. It is probably difficult, if not impossible, to
increase this mutation rate further because the risk of inducing dominant lethal
damage would then be maximized (Fig. 7.8).
This linear dose relationship for induced mutation rates at these seven spe-
cific loci demonstrates the extraordinary power of ENU as a mutagen, but cannot
adequately predict the absolute rate of induced mutation at an “average” locus in
MR x10–5= (1.2±0.3) +(0.4±0.05) x (D – (33.93±5.08)
300mg/Kg -4 x 100 mg/Kg

100 -
Mutation rate x 10-5
75 -
50 -
25 -
-
-
50 100 150 200 250 300 ENU – mg/Kg
Fig. 7.8 A dose–response analysis of ethylnitrosourea (ENU)-induced recessive locus-specific

mutations in treated spermatogonia. Predicted locus-specific mutation rates (MR × 10−5) fol-
lowing ENU treatment of spermatogonia in the mouse. This diagram represents the dose–effect
linear relationships for the mutagen, between ~34 mg/kg of body weight (the threshold dose
under which there is no detectable effect) and 300 mg/kg of body weight, which seems to be the
highest dose tolerated by the mouse. This linear model was computed based on extensive data
from Neuherberg (Germany) and Oak Ridge (USA) (adapted from Favor et al. 1990)
the mouse genome. Lewis and co-workers, for example, calculated the number of
electrophoretic variants induced at 32 loci after treatment with increasing doses of
ENU (from 0 to 250 mg/kg of body weight) in DBA/2 and C57BL/6 male mice
(Lewis 1991). In these experiments, the mutation rates again appeared to increase
linearly with dose but were on average 2.6 times lower than for the “multiple loci”
test performed by Russell and colleagues. This latter observation, which has been
reported by many other scientists with different tests, indicates that the sensitiv-
ity of a locus to the mutagenic activity of ENU probably depends on a variety of
parameters such as its “molecular” size, the gene structure (density in A-T, num-
ber of introns, etc.), and presumably several other unknown parameters. It is likely
that some regions of DNA are more susceptible than others to the mutagenic activ-
ity of ENU, validating the idea that hot spots of mutagenesis exist in the mouse
genome (Kiernan et al. 2002; Arnold et al. 2012).
The mutation frequency, established by Russell and co-workers for seven spe-
cific loci, was later refined by Bode (1984) in another experimental context. Bode
considered that, from an optimally mutagenized male, one can expect to obtain,
on average, one mutation at a given locus per 1,500 of its gametes. It must, how-
ever, be kept in mind that a given male can produce only a limited number of
mutations, and this number is dependent on the number of targets that have been
hit by the mutagen. From his experimental data, Bode concluded that this number
is close to 500 with a dose of 250 mg/kg of mouse body weight. This important
observation means that, when the ultimate goal of an experiment is to produce

a great variety of mutations, as is often the case, it is strongly recommended to
inject a quite large batch of males rather than to breed many offspring from only a
few males.
The mutagenic activity of ENU has been assessed directly at the DNA level in
several laboratories, by performing a careful characterization of the number and
type of nucleotide substitutions induced, then by matching the nature of these sub-
stitutions to the phenotype of the affected mice—if any (Justice et al. 1999;
Noveroske et al. 2000; Concepcion et al. 2004; Quwailid et al. 2004; Keays et al.
2006; Takahasi et al. 2007; Arnold et al. 2012).14 The results of these analyses
indicate that ENU induces mutations at a frequency of one for every 0.7–1.9 Mbp
of genomic DNA, depending upon the strain and dose. Analysis of the mutations
confirms that AT-to-TA transversions occur in about 44 % of the cases while
AT-to-GC transitions occur in about 38 % of the cases. When they fall within the
coding regions these substitutions cause missense mutations (64 %), splicing
defects (26 %) or nonsense mutations (10 %). Another interesting observation,
which is a direct consequence of the observed AT-to-TA and AT-to-GC bias men-
tioned above, is that some amino acid changes are more likely to occur after ENU
treatment than others. As such, it must be kept in mind that ENU mutagenesis does
not merely increases the spontaneous mutation rate but its action is biased towards
certain amino acid changes.
ENU has also been used as a mutagen in the rat and has proved efficient. In this
species, however, the dose must be reduced to 90 mg/kg of body weight (Mashimo
et al. 2010) and, here again, splitting of the doses has proved more efficient than a
single dose.
ENU is a powerful, easy-to-use, inexpensive, and remarkably efficient muta-
gen. Its effectiveness varies with the strain of mouse treated, and this is why it is
absolutely essential to calibrate the experimental parameters as precisely as possi-
ble before embarking on a new mutagenesis project. Non-optimal use of the muta-
gen (i.e., using a dose that is either too high or too low) will inevitably lead to a
waste of both time and animal lives.
7.5 Protocols of Experimental Mutagenesis
When a male mouse is treated with a mutagen, for example by performing a sin-
gle injection of 250 mg ENU per kilogram of body weight, it stays fertile for a
few days after the treatment and then becomes sterile for a period spanning
10–18 weeks (Oakberg and Crosthwait 1983). This sterility period is a con-
sequence of spermatogonial cell killing and it is, in large part, strain- and dose-
dependent. BTBR, BALB/c, C3H/He, C57BL/6, and DBA/2 strains have been used
14 The publication by Arnold et al. (2012) is a rich source of information calculated on a very
large sample.
7.5 Protocols of Experimental Mutagenesis 247
for many years, in particular for the large ENU mutagenesis programs conducted
in Germany, England, and the USA (Hrabe de Angelis et al. 2000; Nolan et al.
2000; Arnold et al. 2012). These strains appeared to be relatively resistant to ENU,
although a relatively higher percentage of C57BL/6 males did not recover fertility
after the ENU treatment (Lewis et al. 1991, 1992). Strain FVB, which has several
advantages over the other strains for the production of embryos for transgenesis,
appeared quite susceptible to ENU, and, accordingly, is not a good choice for
experimental mutagenesis (Justice et al. 2000).
While information concerning the toxicity of ENU for the different strains of
mice is available, information about the differences in mutation rates is scarce. In
an experiment aimed at the production of electrophoretic mutant proteins, Lewis
and colleagues (Lewis et al. 1991) made use of C57BL/6 and DBA/2 males, mated
to DBA/2 and C57BL/6 females respectively, and did not observe any statistically
significant differences in mutation rate between the two strains. Considering the
many experiments that have been performed with the classical laboratory inbred
strains and the mutagen ENU, one would conclude that, if inter-strain differences in
mutation rate were important, this would have been noticed, but this is not the case.
After the sterility period, the spermatogonia that survive ENU treatment pro-
gressively repopulate the testis, the sperm concentration rises progressively and
the males regain fertility and produce spermatozoa derived from the several dif-
ferent clones of mutagenized spermatogonia. In the sperm population (and later
in the embryos), all types of mutations are present but, while dominant mutations
can be observed directly in the F1 (or G1) progeny, recessive mutations must be
homozygous to express a phenotype. This requires two more generations and the
establishment of so-called individual or micro-pedigrees.
The production of mutations in laboratory rodents can be achieved either
genome-wide (i.e., at any locus), or in more or less precisely targeted regions,
depending on the aim of the experiment and the protocol used. These protocols do
not depend upon the mutagen and can apply to radiation as well as to chemicals.
We will review the most commonly used mutagenesis strategies.
7.5.1 Phenotype-Driven, Genome-Wide Mutagenesis
A phenotype-driven, genome-wide mutagenesis program consists of four succes-

sive steps. First, males (G0) are treated with the mutagen and then mated with
females of the same strain, or of another inbred strain, once they have recovered
from the sterility period.15 Unusual phenotypes (often called phenodeviants) that
could result from dominant mutations are then looked for by careful examination
of the offspring of this first cross (G1 population).
15 The choice of the strain must be considered with care depending on the future use of the
mutant potentially discovered. If mutations are induced, it will definitely be important to identify
the background strain in which the mutation occurred.
In the second step, G1 males, which are all potential carriers of recessive muta-
tions at a number of unknown loci, are gathered for the establishment of individual
micro-pedigrees. For this, each G1 male is mated to a few females, either of the
same or from a different strain, and a sample of six G2 females, offspring of this
cross, is selected and crossed (backcrossed in this case) to their father to produce
a G3 population. This G3 generation is then carefully examined for the detection
of possible recessive mutations. The rigorous and systematic examination of the
G3 progeny is part of the phenotyping process and requires much care. Indeed,
the higher the number of parameters screened, the higher the number of mutations
detected (Fig. 7.9).
Because their deleterious effects are compensated for by the presence of a nor-
mal allele in heterozygotes, the recessive mutations induced in the G0 males recur
in G3 of the same micro-pedigree and, accordingly, they are easier to detect and
preserve than the dominant mutations, which, in most instances, appears only once
in the G1 population.
In these micro-pedigrees, when six heterozygous (+/mut?) G2 females are
backcrossed to the individual G1 males and a minimum of ten G3 offspring are
phenotyped per G2 female, the probability of not detecting, just by chance, a
ENU
X
G0
G1 1 2
G2 3
G3 4
Fig. 7.9 Phenotype-driven genome-wide mutagenesis. Phenotype-driven mutagenesis consists of

four successive steps. In the first step, males are treated with the powerful mutagen ENU (see
text for doses) and mated with 2–3 females after recovery from a 10 to 13-week sterility period
(this the G0 generation). The entire G1 progeny is then carefully scrutinized, looking for possible
dominant mutations (arrow). In the second step, males of the G1 generation (which are potential
heterozygous carriers of recessive mutations of all kinds) are selected for the establishment of
micro-pedigrees. First, they are mated with females of either the same or a different strain, and
4–6 female offspring (G2) are backcrossed to their G1 father. Finally, the progenies of the G1
male × G2 female offspring (G3) are subjected to careful phenotypic examination (for exam-
ple, in a “Mouse Clinic”). Micro-pedigrees producing mutant phenotypes are then isolated for in-
depth analysis. The number of G2 females and their G3 offspring are established after statistical
computation to optimize the possibility of detection of new phenotypes
recessive mutation with a visible phenotype that would have been heterozygous in
the +/mut? G1 males is less than 2 % at the 95 % confidence level.
Bode et al. (1988), followed by McDonald et al. (1994), were among the first to
use a whole-genome, phenotype-driven ENU mutagenesis program to produce rel-
evant animal models of phenylketonuria (PKU-OMIM 261640. G0 males were
treated with ENU, the G1 male offspring were mated to females of the same strain
to produce the G2 progeny, and finally the G1 males and their G2 female offspring
were intercrossed to produce the G3 progeny. Blood samples from G1, G2, and G3
mice were analyzed by using the popular Guthrie test, a biochemical test that was
used some years ago for detecting elevated levels of phenylalanine in the blood of
human newborns.16 In these experiments, three independent mutant alleles were
identified in the G3 populations (hph1, hph2, and Pahhph5). In addition, it is inter-
esting to note that, using such a phenotype-driven genome-wide strategy, the bio-
chemical pathways at work in the catabolism of the amino acid phenylalanine
were literally “dissected” out, with one mutation identified at each biochemical
step. This was done in exactly the same manner in which the bacterial geneticists
of the early days disentangled the metabolic pathways in bacteria (McDonald
1995).
Nowadays, after much progress in genotyping and phenotyping, several
projects have been undertaken by which the G1 and G3 progenies of ENU-
mutagenized males have been systematically and extensively phenotyped using
a number of criteria by a team of specialists in so-called “mouse clinics”. Many
interesting mutations have been discovered in these projects that would probably
not have been noticed in other laboratories (Hoebe and Beutler 2005; Massironi
et al. 2006; Arnold et al. 2012). Among the many interesting mutations identified
are Clock, which modifies the circadian rhythm of affected mice (Wilsbacher et al.
2000), and Ticam1Lps2, which results in impaired defense mechanisms against
viral and bacterial diseases (Beutler et al. 2007). In a European project compris-
ing six different laboratories and focusing on deafness syndromes, no less than
thirteen new independent genes involved in inner ear differentiation and pathology
were identified by ENU mutagenesis (Quint and Steel 2003).
The genome-wide production of recessive mutations is a tedious enterprise that
requires both intensive animal care and large breeding programs. The advantage
of this approach is that no a priori assumptions are made about the genes involved
in any pathway. Phenotype-driven mutagenesis is thus an effective method for the
identification of novel genes. Numerous projects are now in progress in several
laboratories worldwide, where groups of novel mutations, once identified, are
roughly phenotyped, mapped to a chromosome, and finally made available to the
scientific community for further study. There is no doubt that genome annotation
will benefit from all these programs, even if a significant amount of work remains
to be achieved after a gene is identified in the form of a mutant allele.
16 The Guthrie test (a bacterial assay) was routinely used for the neonatal diagnostic of phe-
nylketonuria. It is now replaced either by an immunoassay or by a tandem mass spectrometry

assay that measures the amino acid proportions.
7.5.2 The Induction of New Mutant Alleles at Specific Loci
As we remarked above, the main advantage of the genome-wide, phenotype-driven

mutagenesis approach is that most of the mutations collected are new alleles
appearing at loci where no mutations had ever been isolated before. However, in
some cases, it may be desired to induce new alleles at a given locus, for exam-
ple to explore the possibility that the severity of a given phenotype might be
allele-dependent.
The induction of new alleles at a specific locus is well illustrated by the so-
called “multiple loci test”, which was used for the assessment of spontaneous
mutation rates, and which we introduced earlier in this chapter. Using this test sev-
eral new alleles (in particular at the Tyr—albino locus) have been induced after
treatment with mutagens in the gametes of wild-type male partners, and were
then observed directly in the F1 progeny of these males after mating with females
homozygous for a set of recessive viable alleles (Rinchik and Carpenter 1999). A
similar strategy can be applied to any situation where the production of a series of
new alleles at a given locus might be informative, and is ideal when at least one
viable recessive allele is available.
Bode (1984) used ENU to produce new alleles at the Brachyury (T), quaking
(qk) and tufted (tf) loci by mutagenizing +++/+++ (wild-type) male mice and
crossing them to females with the genetic constitution T qk tf/++ tf. In the F1
progeny of these mice the researcher found three tufted phenotypes [tf], one quak-
ing [qk], and one with a short tail (t-interacting or tint) out of 5,172 offspring. In
similar experiments, Justice and Bode (1986) and Bode et al. (1988) produced sev-
eral new alleles at the same three loci, with some of the new alleles at the quak-
ing locus exhibiting interesting and unexpected properties (Justice and Bode 1990;
Cox et al. 1999).
Chapman and colleagues performed a similar experiment (Chapman et al.
1989) with the aim of understanding why mice affected by the X-linked Dmdmdx
mutation, homologous to the human mutation producing Duchenne muscular dys-
trophy, were not clinically affected. Chapman hypothesized that this striking phe-
notypic difference might be because the original Dmdmdx mutation hits a domain
of the gene encoding dystrophin that is not functionally essential, and supposed
that alleles affecting one of the other four domains of the protein might have more
severe effects. To test this hypothesis he created four new alleles at the Dmd locus
(Dmdcv2, Dmdcv3, Dmdcv4, and Dmdcv5) by ENU mutagenesis. These new mutant
alleles were detected by checking for an increase in the creatine phosphokinase
(CPK) plasmatic levels in the female progeny of ENU treated +/Y males crossed
with Dmdmdx/Dmdmdx homozygous females.17 The result of these experiments was
that all five alleles, the four ENU-induced and the original one, had a very similar
17 An increase in the plasma level of creatine phosphokinase (CPK) in these F1 mice reveals
some damage to the muscular tissue, and is often an indication of the likely occurrence of a new
mdx allele.
phenotype with no obvious muscular pathology, although the four mutations were
found to affect totally different domains. Later, it was demonstrated that mice
homozygous for the original Dmdmdx allele and one of the ENU-induced series
(Dmdcv5) had a weaker effect than the other three alleles on the electro-retinogram
(ERG) phenotype of the mutant mice (Figs. 7.10 and 7.11).
This observation indicated that the position of the mutation in the dystrophin-
encoding gene, although it had no effect on the muscular phenotype, nonetheless
had some direct consequences on the ERG phenotype (Pillers et al. 1999). This
contributed to the fine annotation of the different domains of the Dmd gene, but
Mutagen
+ m
x
+ m
m' +
Genotype
G1 m m
[m] [+] Phenotype
New mutant allele
Fig. 7.10 Targeted chemical mutagenesis. Male mice are mutagenized and then mated to females
homozygous for a recessive allele (m) at a specific locus. The G1 offspring of this type of cross
are expected to be all wild type. Any deviation from this phenotype must be considered a possible
new mutant allele at the m locus, especially if some similarities exist between the new pheno-
type and the phenotype of the female (m). For example, this strategy allowed the generation of an
allelic series at the dystrophin gene (Dmd) (Chapman et al. 1989)
Mutagen
G1 M F 1 2 3 5
Electrophoretic
bands
Fig. 7.11 Targeted chemical mutagenesis. A male homozygous for a polymorphic protein is

treated and then crossed with a female homozygous for another electrophoretic variant, and the
G1 progeny are analyzed with the same technique. Mice nos. 1 and 5, as expected, are heterozy-
gous for the two parental forms. Mouse no. 2 is heterozygous for an inactive allele (dotted line)
inherited from its (mutagenized) father. Mouse no. 3 is heterozygous for the maternal form and a
new functional electrophoretic variant derived from its father
did not explain the phenotypic differences between the human pathology and the
mouse model.
A variation of the above-mentioned strategy is to analyze the electrophoretic
pattern of enzymatic proteins in an interstrain F1 hybrid where one parent (usually
the male) has been mutagenized. Such an “electrophoretic multiple loci test” has
been successfully used to identify new mutations at loci encoding for enzymatic
proteins (Johnson and Lewis 1981; Marshall et al. 1983; Lewis et al. 1991, 1992).
ENU mutagenesis has also been used to induce mutant alleles in the genes
encoding the β-chain of hemoglobin (Peters et al. 1986) as well as to produce sev-
eral null or functionally different alleles (Charles and Pretsch 1987; Pretsch et al.
1994).
The production of new alleles is also interesting in that it allows the produc-
tion of slightly different animal models. An excellent example of this situation is
provided by the existing animal models of human citrullinemia type I (Perez et al.
2010), where it was demonstrated that some alleles, because they hit a different
domain of the protein, appeared to be much better animal models of the human
syndrome of citrullinemia (OMIM 215700).
The condition set above—that at least one recessive and viable mutant allele for
the locus of interest is available to allow the production of other mutant alleles—
is not an absolute prerequisite, and alternative strategies are possible. Let us sup-
pose, for example, that other alleles are desired at the Mut locus, which to date has
only been characterized by the unviable (or sterile) mutation mut1. In this case,
several F1 (or G1) males, potentially heterozygous for many new ENU-induced
mutations (among which is a potentially new mut2 allele?) are produced and then
crossed to +/mut1 females. If, by chance, a mouse with an abnormal [mut] phe-
notype is detected in the progeny of one of these females, this suggests that a new
mut2 allele at the Mut locus has very likely been induced by the treatment. The
new allele can then be recovered from the G1 progeny.
7.5.3 The Induction of Mutations in Specific Regions

of the Genome
Many strategies have been used to induce and identify the mutations in a specific
chromosomal region. Here, we describe three of these strategies that may be of
interest in the future: the first makes use of deletions, the second uses consomic or
congenic strains, and the last strategy requires a set of overlapping inversions.
Using deletions to detect recessive mutations can only be applied to regions
where haploidy is compatible with life. The basic principle is that, when a muta-
tion is induced in the chromosomal segment in front of a deletion, a new phenotype
(often lethal) is observed when the chromosome carrying the induced mutation
and the deleted chromosome are associated in the same genome. In these condi-
tions, the breeding protocol requires more than one generation, since the induced
mutation must be kept in the heterozygous state while it is revealed by the deletion.
The deletion strategy has been used many times (Justice et al. 1997; Rinchik and
Carpenter 1999) and has been included in modern mutagenesis programs (Nolan
et al. 2000) to identify potential models of human diseases (Fig. 7.12).
The use of consomic strains is an interesting strategy to safely collect the muta-
tions induced in a particular chromosome. Consomic strains (see Chap. 9) are
strains in which an entire chromosome has been backcrossed from a donor strain
into a different recipient or background strain. Such strains are completely iden-
tical for all chromosome pairs but one. These are not common, but at least one
set exists (Nadeau et al. 2000), and this is sufficient for the strategy to be appli-
cable. The strategy, presented in Fig. 7.13, is an interesting approach to studying
the mutations that have a weak effect or that require sophisticated tests for their
detection, because it is possible to establish a co-isogenic strain where the newly
induced mutations are safely stored before being studied. This is a great advan-
tage when populations (not only individuals) are to be compared at the phenotypic
level; for example, histocompatibility, susceptibility to infectious diseases, and
QTL analysis. The same co-isogenic strain that is homozygous for the targeted
chromosome can be used several times in successive rounds of mutagenesis exper-
iments, resulting in the progressive accumulation of several new alleles in the tar-
geted chromosome (Fig. 7.13).
Mutagen
+ + a +
x
+ + ∆
+ + + +
Genotype
∆ a +
[+] [+] Phenotype
G1
+m +m
Genotype
∆ a +
[m] [+] Phenotype
Fig. 7.12 Using deletions to detect recessive mutations in a specific region. A male mouse is

mutagenized and then mated to females heterozygous for a recessive marker a, and for a via-
ble deletion (Δ) (its phenotype is [a]). Most offspring of this cross have a wild-type phenotype,
except when a recessive viable mutation m is induced by the mutagen in the chromosome seg-
ment encompassed by the deletion. In this case, a new phenotype is observed (m). In the cases
where the induced mutation is lethal, the progenies are reduced in size and the mice heterozy-
gous for a are also heterozygous for m, the new mutation (except for a few recombinants)
x
 Fig. 7.13 Accumulating mutations in a specific chromosome. Male mice are mutagenized and

then mated to female mice consomic for a specific (targeted) chromosome (solid black). F1 male
offspring of this cross are mutagenized again and crossed to female mice of the same strain, con-
somic for the same chromosome. The same cycle of mutagenesis—cross with a consomic partner
is perpetuated a number of times, and in so doing mutations accumulate only on the targeted
chromosome at each generation. Finally a few female offspring of this series of backcross are
selected by microsatellite genotyping and backcrossed to their consomic father. This allows re-
establishment of a fully co-isogenic strain with many independent mutations accumulated in the
same chromosome pair. These mice are then carefully phenotyped. If one of the induced muta-
tions is lethal, the experiment cannot be completed, but the induced mutations can be kept and
studied in another context, for example after outcrossing. This strategy requires very little work
and a very limited number of animals to be used. This can easily be coupled with a gene-driven
mutagenesis experiment, reducing the time spent on genotyping
The use of a set of overlapping inversions is similar in principle to the use of

consomic strains and is reminiscent of the balancer chromosome developed in the
past by Muller and colleagues for the collection of X-ray induced mutations in
Drosophila melanogaster. An example of this strategy was the use of a geneti-
cally engineered inversion in chromosome 11, which has been described in detail
(Zheng et al. 1999; Kile et al. 2003).
Many other strategies have been used to generate and keep mutations in spe-
cific areas of the mouse genome that cannot be described in detail here. We
will just briefly mention that Shedlovsky et al. (1986, 1988), using a specially
designed strategy, were able to induce and study a dozen new lethal alleles within
a region spanning two centiMorgans on each side of the T/t region on mouse
chromosome 17.
7.5.4 A Gene-Driven Strategy for the Production

of Mutations at Specific Loci
With the expansion of advanced techniques for the structural analysis of DNA,
approaches have been developed that are based on the direct, in vitro detection
of DNA alterations, either at specific loci or in specific regions of the genome.
These techniques, when applied to the offspring of mutagenized males, allow the
production of new mutations in specific regions, ultimately into a preselected (or
targeted) gene.
The strategy generally consists of four steps. First, adult males of an inbred
strain are treated with an appropriate dose of mutagen (ENU in most instances)
and then mated with females of the same inbred strain for the production of a large
G1 population.18 In the second step, sperm samples are collected from adult G1
18 In this type of experiment it is necessary to exclusively use mice of an inbred strain to enable
the non-ambiguous characterization of the mutations potentially induced in the progeny by the
mutagen.
ENU
X
G0
G1
DNAsamples
Sperm cells
(cryopreserved)
G3 G2
Fig. 7.14 Genotype-driven mutagenesis. Male mice are treated with ENU and mated to females
(preferably of the same inbred strain) once they have recovered from the sterile period (G0). A
large number of G1 males, which all are heterozygous carriers of a great number of independent
point mutations (mostly base-pair changes), are then bred. Sperm samples from each G1 mouse
are collected and preserved deep-frozen, while DNA samples from the same mice are processed
and stored with the same reference. Identification of the mutations generated by the ENU treat-
ment in a specific target (a gene or any other specific sequence) is carried out by molecular tech-
niques to identify DNA mismatches, or directly by sequencing. Once the base-pair changes are
identified and considered potentially interesting (stop codons, missense, etc.), the corresponding
sperm cells are thawed and heterozygous mice are produced by in vitro fertilization with oocytes
of the same background strain. A major advantage of this method is that it produces all types of
point mutations, not only knockouts. A drawback is the difficulty of and time required for iden-
tifying the mutations in the targeted region. With the rapid expansion of new sequencing tech-
niques, the identification step should be somewhat easier
offspring of this initial cross and stored deep-frozen for performing future in vitro
fertilization. Simultaneously, DNA samples from the same G1 males are prepared,
cross-referenced with the sperm samples, and stored (Fig. 7.14).
The third step consists of the analysis of the DNA sequence of all G1 mice,
looking for any structural changes that may have occurred in a selected and well-
delimited region of the genome. This can be achieved by using a sensitive, high-
throughput, physical technique, detecting all single nucleotide mismatches after
pooling of the DNA samples. This can also be achieved by direct sequencing or
SNP genotyping.
When a mutation is found and registered as potentially interesting (i.e., exclud-
ing synonymous base-pair changes but retaining nonsense or missense mutations
with predicted severe effects), the fourth and last step is performed: the sample of
sperm cells corresponding to the potentially interesting mutant mouse is thawed,
oocytes of the same strain are fertilized in vitro and implanted in pseudo-preg-
nant mothers, and, once born, the potentially heterozygous offspring are bred
and crossed in order to produce homozygous offspring whose phenotype is then
observed. In this micro-pedigree, the molecular characterization of the offspring is
fundamental.
This gene-driven protocol allows the production of all types of mutations (and
not only knockouts) in all regions of the genome (coding and non-coding). A
drawback is the difficulty and time required for identifying the mutations in the
targeted regions. However, with the rapid expansion of modern sequencing tech-
niques, the identification step should be somewhat simplified and shortened in the
near future.
The gene-driven or targeted mutagenesis approach has several advantages. It
is fast and relatively inexpensive compared to other gene-driven strategies (for
example, the engineering of knockouts in ES cells—see Chap. 8). Once identi-
fied in a batch of frozen sperm cells, a mutation can be retrieved and made avail-
able as heterozygous adult mice in 4–5 months’ time. Another interesting point is
that a repository comprising a very large number of (non-characterized) mutant
alleles can be established by progressively accumulating and storing samples of
deep-frozen sperm cells from ENU-treated mice. As we already mentioned, and as
observed by direct sequencing of samples prepared from ENU-treated mice, one
expects ~0.7–1.9 nucleotide change(s) to be induced per Mbp of mouse DNA after
the injection of a single dose of 250 mg/kg. If we consider that the mouse genome
consists of 2.7 × 109 bp, one can then expect between ~2,000 and 5,000 de novo
substitutions in each G1 progeny from an ENU-treated male mouse. If these
nucleotide changes are randomly distributed, one can then expect between ~30 and
75 of the latter to be in the coding DNA or the splicing sites, of which ~25–60 will
generate a missense, a nonsense or a splicing defect (77 %).
In addition to these theoretical considerations (but based on actual sequenc-
ing data!), one can also calculate that a repository with frozen sperm samples
from 20,000 individual G1 animals will be a resource with the potential presence
of six independent mutations at any gene of the mouse genome (at the 5 % risk
level).
The identification of specific gene alterations can be achieved using pooled
DNA samples and run concurrently in several different laboratories to increase
the efficiency and ultimately lower the cost of mutagenesis. The final advantage is
that, in a species such as mouse where sperm cells can be frozen for long periods
and thawed for fertilization, there is no time limit for the identification of muta-
tions. Several laboratories have already published interesting results in this man-
ner (Coghill et al. 2002; Augustin et al. 2005; Michaud et al. 2005; Gondo 2008;
Gondo et al. 2010), demonstrating that this gene-driven strategy for the induction
of mutations in the mouse might be very promising. This is even truer if we con-
sider that the technique in question can also be applied to the annotation of DNA
sequences that are highly conserved across different species; for example, those
that are transcribed into non-coding RNAs or not transcribed at all, and whose
function is still under scrutiny.
7.6 Other Techniques for the Production of Mutations in

the Mouse
In addition to those described earlier in this chapter, a few other strategies have
been proposed in the past for the induction of novel mutant alleles in the mouse
genome. Most of these techniques have not proved to be significantly more advan-
tageous than the techniques currently in use (ENU mutagenesis in particular) and,
for this reason, they have been abandoned. However, exceptions must be made for
two strategies that have demonstrated some real advantages. The first consists of
treating embryonic stem cells (ES cells) with chemical mutagens (ENU or EMS):
this approach will be discussed in the next chapter. The second strategy consists
of using transposable elements as insertional mutagens in the mouse, just as the
P elements were used in Drosophila melanogaster, i.e. with the assumption that,
when by chance the random insertion of a transposon occurs into a gene, it gener-
ally hinders the transcription of a normal mRNA at or near the insertion site and
causes a loss-of-function mutation. This technique is known as transposon-based
insertional mutagenesis or TIM. We will describe it briefly.
As discussed in Chap. 5, transposable elements (TEs or transposons) are short
DNA sequences that move (transpose) within the genome of a great variety of
organisms, including bacteria, plants, insects, and vertebrates, by using a cut-and-
paste mechanism (i.e., with no RNA intermediate). This mechanism of transposi-
tion requires a specific structure of the transposon, with inverted repeats at both
ends, and a specific enzyme (a transposase or transposonase), which is synthe-
sized either by the TE itself (in the case of autonomous transposons) or “in trans”
by an independent gene (in the case of non-autonomous transposons). Transposons
are very active in the genome of plants and bacteria, as well as in some other spe-
cies, and play an important role in evolution.19 In mammalian genomes, on the
other hand, transposons are inactive and the transposase-encoding genes are
degenerated and no longer functional.
Starting from these observations, geneticists had the clever idea to “synthesize”
a transposon by genetic engineering using an active transposase in the context of a
mammalian genome. To do this, they selected the sequence of a transposon of the
Tc1/mariner family active in fish (salmon) and, taking into account some phylo-
genetic data, they could “resurrect” a functional transposon system that they judi-
ciously named Sleeping Beauty (SB10) in memory of its historical origins. SB10
was confirmed active in the mouse and rat genomes, inducing mutations by trans-
position as expected (Ivics et al. 1997).
In experiments making use of the SB10 transposon system, two transgenic
strains are prepared independently, one carrying the transposon proper (some-
times modified to carry a marker cassette that helps track the animal carriers of a
novel mutant allele) and the other expressing the indispensable transposase. When
19 The transposons were discovered and studied in maize by Nobel laureate B. McClintock, pre-
cisely because of their mutagenic activity.

7.6 Other Techniques for the Production of Mutations in the Mouse 259
desired, the two strains are crossed to generate F1s in which transposition can occur.
In the mouse, the frequency of SB transposition was estimated to be in the range
of 0.2–2.0 events per spermatid (Copeland and Jenkins 2010). Although the rate of
production of transposon knockout mutations (TKOs) is less than the rate of muta-
tions resulting from ENU treatment, the TKOs are, in most instances, easier to
identify and to clone. By outcrossing the animals carrying the TKO mutations of
interest, one can separate the transgene-encoding transposase from the other compo-
nents of the SB system (the mutator element) and transposition immediately stops.
To illustrate the use of transposons as mutagens and the great versatility of
this strategy, we recommend a set of interesting publications (Carlson et al. 2003;
Lu et al. 2007; Takeda et al. 2007, 2008; Largaespada 2009; Ivics et al. 2011;
Furushima et al. 2012). Finally, a review paper by Copeland and Jenkins (2010) is
a beautiful illustration of the contribution of the SB10 system to the analysis of the
determinism of cancer and the discovery of cancer genes.
In the mouse, and as we will explain in the next chapter, the transposon
Sleeping Beauty as well as another one called piggyBac have been used extensively
both for the transfection and the production of mutations in ES cell lines in vitro.
If transposon-based insertional mutagenesis has some obvious advantages for
the production of mutations, it is also interesting for the transfer of genes with sta-
ble expression in mouse ES cells. Finally, it may also have applications enabling
the persistent expression of therapeutic genes in patients.
7.7 Conclusions
Spontaneous mutations, which are generally identified through the observation

of an abnormal phenotype, present several advantages. The first and probably
the most important is that they are produced at virtually no cost and are in gen-
eral freely available. Another advantage is that they have, in general, an obvious
phenotype given that they are identified based on observation. Also, spontaneous
mutations represent a great variety of molecular events, such as deletions, inser-
tions, and point mutations, generating not only loss-of-function alleles but also
hypomorphic and hypermorphic alleles. The problem is that not all mutant genes
have an obvious phenotype or, conversely, the phenotype of some mutant alleles is
sometimes so severe that affected offspring die in utero.
When mutant allelic forms of a gene are not readily available, the only possible
approach for gene annotation is to generate de novo mutations. Thus, the discovery
of the extraordinary virtues of ENU as a mutagen can certainly be regarded as a
milestone in the history of mouse genetics. With this substance at our disposition,
it is now possible to produce and store a great number of new mutant alleles for
each protein-coding gene, and all these mutations are a valuable tool for genome
annotation. Another advantage is that it is now also possible to induce mutations
in those regions of the genome that are highly conserved but whose function is not
yet elucidated. The only drawback that should be considered is that mutagens act
randomly, forcing us to make a sometimes lengthy and costly selection among the
collected mutations. In this regard, and as we will discuss in the next chapter, the
widespread availability of a variety of genetic engineering technologies, including
new genome editing tools, has opened the field to the creation of subtle modifica-
tions in the mouse genome at will. Even though the identification of genes account-
able for single-gene phenotypes is very important, in particular in the context of
gene annotation, most of the pathologies that affect human patients are not “mono-
genic” but are influenced by multiple genes with additive or synergistic effects. As
such, our present challenge is to advance the genetic analysis of complex traits.
References
Ames BN, Lee FD, Durston WE (1973) An improved bacterial test system for the detection and
classification of mutagens and carcinogens. Proc Ntl Acad Sc USA 70:782–786
Antonarakis SE, Kazazian HH, Gitschier J, Hutter P, de Moerloose P, Morris MA (1995)
Molecular etiology of factor VIII deficiency in hemophilia A. Adv Exp Med Biol 386:19–34
Arnold CN, Barnes MJ, Berger M, Blasius AL, Brandl K, Croker B, Crozat K, Du X,
Eidenschenk C, Georgel P, Hoebe K, Huang H, Jiang Z, Krebs P, La Vine D, Li X, Lyon S,
Moresco EM, Murray AR, Popkin DL, Rutschmann S, Siggs OM, Smart NG, Sun L, Tabeta
K, Webster V, Tomisato W, Won S, Xia Y, Xiao N, Beutler B (2012) ENU-induced pheno-
variance in mice: inferences from 587 mutations. BMC Res 5:577
Ashby J, Gorelick NJ, Shelby MD (1997) Mutation assays in male germ cells from transgenic
mice: overview of study and conclusions. Mutat Res 388:111–122
Auerbach C (1962) Mutation: an introduction to research on mutagenesis. Part I: methods. Oliver
and Boyd, Edinburgh
Auerbach C, Robson JM (1946) Chemical production of mutations. Nature 157:202
Augustin M, Sedlmeier R, Peters T, Huffstadt U, Kochmann E, Simon D, Schöniger M, Garke-
Mayerthaler S, Laufs J, Mayhaus M, Franke S, Klose M, Graupner A, Kurzmann M, Zinser
C, Wolf A, Voelkel M, Kellner M, Kilian M, Seelig S, Koppius A, Teubner A, Korthaus D,
Nehls M, Wattler S (2005) Efficient and fast targeted production of murine models based on
ENU mutagenesis. Mamm Genome 16:405–413
Beier D (2000) Sequence-based analysis of mutagenized mice. Mamm Genome 11:594–597
Beutler B, Du X, Xia Y (2007) Precis on forward genetics in mice. Nat Immunol 8:659–664
Bode VC (1984) Ethylnitrosourea mutagenesis and the isolation of mutant alleles for specific
genes located in the T region of mouse chromosome 17. Genetics 108:457–470
Bode VC, McDonald JD, Guénet JL, Simon D (1988) hph-1: a mouse mutant with hereditary
hyperphenylalaninemia induced by ethylnitrosourea mutagenesis. Genetics 118:299–305
Carlson CM, Dupuy AJ, Fritz S, Roberg-Perez KJ, Fletcher CF, Largaespada DA (2003)
Transposon mutagenesis of the mouse germline. Genetics 165:243–256
Chakrabarti L, Neal JT, Miles M, Martinez RA, Smith AC, Sopher BL, La Spada AR (2006)
The Purkinje cell degeneration 5 J mutation is a single amino acid insertion that destabilizes
Nna1 protein. Mamm Genome 17:103–110
Chapman VM, Miller DR, Armstrong D, Caskey CT (1989) Recovery of induced mutations for
X chromosome-linked muscular dystrophy in mice. Proc Natl Acd Sc USA 86:1292–1296
Charles DJ, Pretsch W (1987) Linear dose-response relationship of erythrocyte enzyme-activity
mutations in offspring of ethylnitrosourea-treated mice. Mutat Res 176:81–91
Coghill EL, Hugill A, Parkinson N, Davison C, Glenister P, Clements S, Hunter J, Cox RD,
Brown SD (2002) A gene-driven approach to the identification of ENU mutants in the
mouse. Nat Genet 30:255–256
References 261
Concepcion D, Seburn KL, Wen G, Frankel WN, Hamilton BA (2004) Mutation rate and pre-
dicted phenotypic target sizes in ethylnitrosourea-treated mice. Genetics 168:953–959
Copeland NG, Jenkins NA (2010) Harnessing transposons for cancer genes discovery. Nat Rev
Cancer 10:696–706
Cox RD, Hugill A, Shedlovsky A, Noveroske JK, Best S, Justice MJ, Lehrach H, Dove WF
(1999) Contrasting effects of ENU induced embryonic lethal mutations of the quaking gene.
Genomics 57:333–341
Favor J (1986) The frequency of dominant cataract and recessive specific-locus mutations in
mice derived from 80 or 160 mg ethylnitrosourea per kg body weight treated spermatogo-
nia. Mutat Res 162:69–80
Favor J (1994) Specific-locus mutations tests in germ cells of the mouse: an assessment of the
screening procedures and the mutational events detected. In: Mattison DR, Olsham AF (eds)
Male-mediated developmental toxicity. Plenum Press, New York, pp 23–36
Favor J, Sund M, Neuhauser-Klaus A, Ehling UH (1990) A dose-response analysis of ethylnitro-
sourea-induced recessive specific-locus mutations in treated spermatogonia of the mouse.
Mutat Res 231:47–54
Fernandez-Gonzalez A, La Spada AR, Treadaway J, Higdon JC, Harris BS, Sidman RL, Morgan
JI, Zuo J (2002) Purkinje cell degeneration (pcd) phenotypes caused by mutations in the
axotomy-induced gene, Nna1. Science 295:1904–1906
Frazer KA, Eskin E, Kang HM, Bogue MA, Hinds DA, Beilharz EJ, Gupta RV, Montgomery
J, Morenzoni MM, Nilsen GB, Pethiyagoda CL, Stuve LL, Johnson FM, Daly MJ, Wade
CM, Cox DR (2007) A sequence-based variation map of 8.27 million SNPs in inbred mouse
strains. Nature 448:1050–1053
Furushima K, Jang CW, Chen DW, Xiao N, Overbeek PA, Behringer RR (2012) Insertional
mutagenesis by a hybrid piggyBac and sleeping beauty transposon in the rat. Genetics
192:1235–1248
Gilman JG (1972) Hemoglobin beta chain structural variation in mice: evolutionary and func-
tional implications. Science 178:873–874
Gondo Y (2008) Trends in large-scale mouse mutagenesis: from genetics to functional genomics.
Nat Rev Genet 9:803–810
Gondo Y, Fukumura R, Murata T, Makino S (2010) ENU-based gene-driven mutagenesis in the
mouse: a next-generation gene-targeting system. Exp Anim 59:537–548
Graur D (2003) Single-base mutation—in nature encyclopedia of the Human Genome Macmillan
Publishers Ltd
Green EL, Roderick TH (1966) Radiation genetics. In: Green EL (ed) Biology of the laboratory
mouse. Dover Publications, New York, pp 165–185
Hitotsumachi S, Carpenter DA, Russell WL (1985) Dose-repetition increases the mutagenic
effectiveness of N-ethyl-N-nitrosourea in mouse spermatogonia. Proc Ntl Acd Sc USA
82:6619–6621
Hoebe K, Beutler B (2005) Unraveling innate immunity using large scale N-ethyl-N-nitrosourea
mutagenesis. Tissue Antigens 65:395–401
Hrabe de Angelis MH, Flaswinkel H, Fuchs H, Rathkolb B, Soewarto D, Marschall S, Heffner S,
Pargent W, Wuensch K, Jung M, Reis A, Richter T, Alessandrini F, Jakob T, Fuchs E, Kolb
H, Kremmer E, Schaeble K, Rollinski B, Roscher A et al (2000) Genome-wide, large-scale
production of mutant mice by ENU mutagenesis. Nat Genet 25:444–447
Ivics Z, Hackett PB, Plasterk RH, Izsvák Z (1997) Molecular reconstruction of sleeping beauty, a
Tc1-like transposon from fish, and its transposition in human cells. Cell 91:501–510
Ivics Z, Izsvák Z, Chapman KM, Hamra FK (2011) Sleeping beauty transposon mutagenesis of
the rat genome in spermatogonial stem cells. Methods 53:356–365
Johnson FM, Lewis SE (1981) Mutation-rate determinations based on electrophoretic analysis of
laboratory mice. Mutat Res 82:125–135
Justice MJ, Bode VC (1986) Induction of new mutations in a mouse t-haplotype using ethylnitro-
sourea mutagenesis. Genet Res 47:187–192
Justice MJ, Bode VC (1990) ENU-induced allele of brachyury (Tkt1) exhibits a developmental
lethal phenotype similar to the original brachyury (T) mutation. J Exp Zool 254:286–295
Justice MJ, Zheng B, Woychik RP, Bradley A (1997) Using targeted large deletions and high-
efficiency N-ethyl-N-nitrosourea mutagenesis for functional analyses of the mammalian
genome. Methods 13:423–436
Justice MJ, Noveroske JK, Weber JS, Zheng B, Bradley A (1999) Mouse ENU mutagenesis.
Hum Mol Genet 8:1955–1963
Justice MJ, Carpenter DA, Favor J, Neuhauser-Klaus A, Hrabé de Angelis M, Soewarto D, Moser
A, Cordes S, Miller D, Chapman V, Weber JS, Rinchik EM, Hunsicker PR, Russell WL,
Bode VC (2000) Effects of ENU dosage on mouse strains. Mamm Genome 11:484–488
Keane TM, Goodstadt L, Danecek P, White MA, Wong K, Yalcin B, Heger A, Agam A, Slater G,
Goodson M, Furlotte NA, Eskin E, Nellåker C, Whitley H, Cleak J, Janowitz D, Hernandez-
Pliego P, Edwards A, Belgard TG, Oliver PL, McIntyre RE, Bhomra A, Nicod J, Gan X,
Yuan W, van der Weyden L, Steward CA, Bala S, Stalker J, Mott R, Durbin R, Jackson IJ,
Czechanski A, Guerra-Assunção JA, Donahue LR, Reinholdt LG, Payseur BA, Ponting CP,
Birney E, Flint J, Adams DJ (2011) Mouse genomic variation and its effect on phenotypes
and gene regulation. Nature 477:289–294
Keays DA, Clark TG, Flint J (2006) Estimating the number of coding mutations in genotypic-
and phenotypic-driven N-ethyl-N-nitrosourea (ENU) screens. Mamm Genome 17:230–238
Kiernan AE, Erven A, Voegeling S, Peters J, Nolan P, Hunter J, Bacon Y, Steel KP, Brown SDM,
Guénet JL (2002) ENU mutagenesis reveals a highly mutable locus on mouse chromosome
4 that affects ear morphogenesis. Mamm Genome 13:142–148
Kile BT, Hentges KE, Clark AT, Nakamura H, Salinger AP, Liu B, Box N, Stockton DW, Johnson
RL, Behringer RR, Bradley A, Justice MJ (2003) Functional genetic analysis of mouse
chromosome 11. Nature 425:81–86
Krawczak M, Ball EV, Cooper DN (1998) Neighboring-nucleotide effects on the rates of germ-
line single-base-pair substitution in human genes. Am J Hum Genet 63:474–488
Kumar S, Subramanian S (2002) Mutation rates in mammalian genomes. Proc Natl Acad Sv
USA 99:803–808
Largaespada DA (2009) Transposon mutagenesis in mice. Meth Mol Biol 530:379–390
Lewis MA, Quint E, Glazier AM, Fuchs H, De Angelis MH, Langford C, van Dongen S, Abreu-
Goodger C, Piipari M, Redshaw N, Dalmay T, Moreno-Pelayo MA, Enright AJ, Steel KP
(2009) An ENU-induced mutation of miR-96 associated with progressive hearing loss in
mice. Nat Genet 41:614–618
Lewis SE (1991) The biochemical specific-locus test and a new multiple-endpoint muta-
tion detection system: considerations for genetic risk assessment. Environ Mol Mut
18:303–306
Lewis SE, Barnett LB, Sadler BM, Shelby MD (1991) ENU mutagenesis in the mouse electro-
phoretic specific-locus test, 1. Dose-response relationship of electrophoretically-detected
mutations arising from mouse spermatogonia treated with ethylnitrosourea. Mutat Res
249:311–315
Lewis SE, Barnett LB, Shelby MD (1992) ENU mutagenesis in the mouse electrophoretic spe-
cific locus test. 2. Mutational studies of mature oocytes. Mutat Res 296:129–133
Liu SM, Leibel RL, Chua SC Jr (1998) Partial duplication in the Leprdb-Pas mutation is a result
of unequal crossing over. Mamm Genome 9:780–781
Lu B, Geurts AM, Poirier C, Petit DC, Harrison W, Overbeek PA, Bishop CE (2007) Generation
of rat mutants using a coat color-tagged sleeping beauty transposon system. Mamm Genome
8:338–346
Marshall RR, Raj AS, Grant FJ, Heddle JA (1983) The use of two-dimensional electrophoresis to
detect mutations induced in mouse spermatogonia by ethylnitrosourea. Can J Genet Cytol
25:457–466
References 263
Martin N, Jaubert J, Gounon P, Salido E, Haase G, Szatanik M, Guénet JL (2002) A mis-

sense mutation in Tbce causes progressive motor neuronopathy in mice. Nat Genet
32:443–447
Mashimo T, Ohmori I, Ouchida M, Ohno Y, Tsurumi T, Miki T, Wakamori M, Ishihara S, Yoshida
T, Takizawa A, Kato M, Hirabayashi M, Sasa M, Mori Y, Serikawa T (2010) A missense
mutation of the gene encoding voltage-dependent sodium channel (Nav1.1) confers suscep-
tibility to febrile seizures in rats. J Neurosci 30:5744–5753
Massironi SM, Reis BL, Carneiro JG, Barbosa LB, Ariza CB, Santos GC, Guénet JL, Godard AL
(2006) Inducing mutations in the mouse genome with the chemical mutagen ethylnitrosou-
rea. Braz J Med Biol Res 39:1217–1226
McDonald JD (1995) Using high-efficiency mouse germline mutagenesis to investigate complex
biological phenomena: genetic diseases, behavior, and development. Proc Soc Exp Biol
Med 209:303–308
McDonald JD, Trischler M, Stoorvogel W, Ullrich O (1994) The PKU mouse project: its history,
potential and implications. Acta Paediatr Suppl 407:122–123
Menalled LB, Chesselet MF (2002) Mouse models of Huntington’s disease. Trends Pharmacol
Sci 23:32–39
Michaud EJ, Culiat CT, Klebig ML, Barker PE, Cain KT, Carpenter DJ, Easter LL, Foster
CM, Gardner AW, Guo ZY, Houser KJ, Hughes LA, Kerley MK, Liu Z, Olszewski RE,
Pinn I, Shaw GD, Shinpock SG, Wymore AM, Rinchik EM, Johnson DK (2005) Efficient
gene-driven germ-line point mutagenesis of C57BL/6 J mice. BMC Genom 6:164
Mülhardt C, Fischer M, Gass P, Simon-Chazottes D, Guénet JL, Kuhse J, Betz H, Becker CM
(1994) The spastic mouse: aberrant splicing of glycine receptor beta subunit mRNA caused
by intronic insertion of L1 element. Neuron 13:1003–1015
Muller HJ (1927) Artificial transmutation of the gene. Science 66:84–87
Nadeau JH, Singer JB, Matin A, Lander ES (2000) Analysing complex genetic traits with chro-
mosome substitution strains. Nat Genet 24:221–225
Nolan P, Peters J, Strivens M, Rogers D, Hagan J, Spurr N, Gray IC, Vizor L, Brooker D,
Whitehill E, Washbourne R, Hough T, Greenaway S, Hewitt M, Liu X, McCormack S,
Pickford K, Selley R, Wells C, Tymowska-Lalanne Z et al (2000) A systematic genome-
wide, phenotype-driven mutagenesis programme for gene function studies in the mouse. Nat
Genet 25:440–443
Noveroske JK, Weber JS, Justice MJ (2000) The mutagenic action of N-ethyl-N-nitrosourea in
the mouse. Mamm Genome 11:478–483
Oakberg EF, Crosthwait CD (1983) The effect of ethyl-, methyl- and hydroxyethyl-nitrosourea
on the mouse testis. Mutat Res 108:337–344
Perez CJ, Jaubert J, Guénet JL, Barnhart KF, Ross-Inta CM, Quintanilla VC, Aubin I,
Brandon JL, Otto NW, DiGiovanni J, Gimenez-Conti I, Giulivi C, Kusewitt DF,
Conti CJ, Benavides F (2010) Two hypomorphic alleles of mouse Ass1 as a new ani-
mal model of citrullinemia type I and other hyperammonemic syndromes. Am J Pathol
177:1958–1968
Perez CJ, Dumas A, Vallières L, Guénet JL, Benavides F (2013) Several classical mouse inbred
strains, including DBA/2, NOD/Lt, FVB/N, and SJL/J, carry a putative loss-of-function
allele of Gpr84. J Hered 104:565–571
Peters J, Ball ST, Andrews SJ (1986) The detection of gene mutations by electrophoresis, and
their analysis. Prog Clin Biol Res 209B:367–374
Pillers DA, Weleber RG, Green DG, Rash SM, Dally GY, Howard PL, Powers MR, Hood DC,
Chapman VM, Ray PN, Woodward WR (1999) Effects of dystrophin isoforms on signal
transduction through neural retina: genotype-phenotype analysis of Duchenne muscular
dystrophy mouse mutants. Mol Genet Metab 66:100–110
Pretsch W, Favor J, Lehmacher W, Neuhauser-Klaus A (1994) Estimates of the radiation-induced
mutation frequencies to recessive visible, dominant cataract and enzyme-activity alleles in
germ cells of AKR, BALB/c, DBA/2 and (102xC3H)F1 mice. Mutagenesis 9:289–294
Quint E, Steel KP (2003) Use of mouse genetics for studying inner ear development. Curr Top
Dev Biol 57:45–83
Quwailid MM, Hugill A, Dear N, Vizor L, Wells S, Horner E, Fuller S, Weedon J, McMath H,
Woodman P, Edwards D, Campbell D, Rodger S, Carey J, Roberts A, Glenister P, Lalanne
Z, Parkinson N, Coghill EL, McKeone R, Cox S, Willan J, Greenfield A, Keays D, Brady
S, Spurr N, Gray I, Hunter J, Brown SDM, Cox RD (2004). A gene-driven ENU-based
approach to generating an allelic series in any gene. Mamm Genome 15:585–591
Rinchik EM, Carpenter DA (1999) N-ethyl-N-nitrosourea mutagenesis of a 6- to 11-cM subre-
gion of the Fah-Hbb interval of mouse chromosome 7: Completed testing of 4557 gametes
and deletion mapping and complementation analysis of 31 mutations. Genetics 152:373–383
Runck AM, Moriyama H, Storz JF (2009) Evolution of duplicated β-globin genes and the struc-
tural basis of hemoglobin isoform differentiation in Mus. Mol Biol Evol 11:2521–2532
Runkel F, Hintze M, Griesing S, Michels M, Blanck B, Fukami K, Guénet JL, Franz T (2012)
Alopecia in a viable phospholipase C delta 1 and phospholipase C delta 3 double mutant.
PLoS ONE 7(6):e39203
Russell LB, Russell WL (1996) Spontaneous mutations recovered as mosaics in the mouse spe-
cific-locus test. Proc Natl Acad Sci USA 93:13072–13077
Russell LD, Ettlin RA, SinhaHikim AP, Clegg ED (1990) Histological and histopathological
evaluation of the testis. Cache River Press, Clearwater
Russell WL (1962) An augmenting effect of dose fractionation on radiation-induced mutation
rate in mice. Proc. National Acad. Sc. USA 48:1724–1728
Russell WL (1963) The effect of radiation dose rate and fractionation on mutation in mice. In:
Sobels F (ed) Repair from genetic radiation damage, vol 4. Pergamon Press, New York, p
205–217
Russell WL, Kelly EM, Hunsicker PR, Bangham JW, Maddux SC, Phipps EL (1979) Specific
locus test shows ethylnitrosourea to be the most potent mutagen in the mouse. Proc Ntl
Acad Sc USA 76:5818–5819
Russell WL, Hunsicker PR, Carpenter DA, Cornett CV, Guinn GM (1982a) Effect of dose frac-
tionation on the ethylnitrosourea induction of specific-locus mutations in mouse spermato-
gonia. Proc Ntal acad Sc USA 79:3592–3593
Russell WL, Hunsicker PR, Raymer GD, Steele MH, Stelzner KF, Thompson HM (1982b)
Dose—response curve for ethylnitrosourea-induced specific-locus mutations in mouse sper-
matogonia. Proc Natl Acad Sc USA 79:3589–3591
Sakuraba Y, Sezutsu H, Takahasi KR, Tsuchihashi K, Ichikawa R, Fujimoto N, Kaneko S, Nakai
Y, Uchiyama M, Goda N, Motoi R, Ikeda A, Karashima Y, Inoue M, Kaneda H, Masuya
H, Minowa O, Noguchi H, Toyoda A, Sakaki Y, Wakana S, Noda T, Shiroishi T, Gondo Y
(2005) Molecular characterization of ENU mouse mutagenesis and archives. Biochem
Biophys Res Commun 336:609–616
Schlager G, Dickie MM (1966) Spontaneous mutation rates at five coat-color loci in mice.
Science 151:205–206
Schlager G, Dickie MM (1967) Spontaneous mutation and mutation rates in the house mouse.
Schmezer P, Eckert C (1999) Induction of mutations in transgenic animal models: BigBlue and
Muta Mouse. Int Agency Res Cancer-Res Publ 146:367–394
Shedlovsky A, Guénet JL, Johnson LL, Dove WF (1986) Induction of recessive lethal mutations
in the T/t-H-2 region of the mouse genome by a point mutagen. Genet Res 47:135–142
Shedlovsky A, King TR, Dove WF (1988) Saturation germline mutagenesis of the murine t
region including a lethal allele at the quaking locus. Proc Ntl Acad Sc USA 85:180–184
Simon-Chazottes D, Tutois S, Kuehn M, Evans M, Bourgade F, Cook S, Davisson MT, Guénet
JL (2006) Mutations in the gene encoding the low-density lipoprotein receptor LRP4 cause
abnormal limb development in the mouse. Genomics 87:673–677
Stoye JP, Fenner S, Greenoak GE, Moran C, Coffin JM (1988) Role of endogenous retroviruses
as mutagens: the hairless mutation of mice. Cell 54:383–391
References 265
Takahasi KR, Sakuraba Y, Gondo Y (2007) Mutational pattern and frequency of induced nucleo-
tide changes in mouse ENU mutagenesis. BMC Mol Biol 8:52
Takeda J, Keng VW, Horie K (2007) Germline mutagenesis mediated by Sleeping Beauty trans-
poson system in mice. Genome Biol 8(Suppl 1):S14
Takeda J, Izsvák Z, Ivics Z (2008) Insertional mutagenesis of the mouse germline with sleeping
beauty transposition. Meth Mol Biol 435:109–125
Van Zeeland AA, Mohn GR, Mullenders LH, Natarajan AT, Nivard M, Simons JW, Venema J,
Vogel EW, Vrieling H, Zdzienicka MZ et al (1989) Relationship between DNA-adduct for-
mation, DNA repair, mutation frequency and mutation spectra. Annali dell’Instituto superi-
ore di sanita (Ann 1st Super Sanita) Istituto Superiore di Sanita (ISDIS) 2003 25:223–228
Vogel EW, Natarajan AT (1995) DNA damage and repair in somatic and germ cells in vivo.
Mutat Res 330:183–208
Vogel F, Rohrborn G (1970) Chemical mutagenesis in mammals and man. Springer, New York,
p 519
Wahnschaffe U, Bitsch A, Kielhorn J, Mangelsdorf I (2005a) Mutagenicity testing with trans-
genic mice. Part I: Comparison with the mouse bone marrow micronucleus test. J Carcinog
4:3
Wahnschaffe U, Bitsch A, Kielhorn J, Mangelsdorf I (2005b) Mutagenicity testing with trans-
genic mice. Part II: comparison with the mouse spot test. J Carcinog 4:4
Wilsbacher LD, Sangoram AM, Antoch MP, Takahashi JS (2000) The mouse clock locus:
sequence and comparative analysis of 204 kb from mouse chromosome 5. Genome Res
10:1928–1940
Yang H, Wang JR, Didion JP, Buus RJ, Bell TA, Welsh CE, Bonhomme F, Yu AH, Nachman NW,
Pialek J et al (2011) Subspecific origin and haplotype diversity in the laboratory mouse. Nat
Genet 43:648–655
Youssoufian H, Antonarakis SE, Bell W, Griffin AM, Kazazian HH Jr (1988) Nonsense and mis-
sense mutations in hemophilia A: estimate of the relative mutation rate at CG dinucleotides.
Am J Hum Genet 42:718–25
Zheng B, Sage M, Cai WW, Thompson DM, Tavsanli BC, Cheah YC, Bradley A (1999)
Engineering a mouse balancer chromosome. Nat Genet 22:375–378
Chapter 8
Transgenesis and Genome Manipulations
8.1 Introduction
In the early 1980s, the expression of transgenic animals was proposed to define
animals having foreign DNA sequences stably and deliberately inserted into their
genome through human intermediaries. With time and the advent of new tech-
niques, this concept has progressively evolved, and nowadays, it is probably more
appropriate to consider that transgenic animals are animals whose genetic char-
acteristics have been altered using one of the techniques of genetic engineering.
Whatever the definition, transgenic animals belong to the category of genetically
modified or genetically engineered organisms (GMOs).
Transgenic mice can be created by using a variety of experimental procedures
depending upon the aim of the experiment. Among these procedures, the micro-
injection of foreign DNA fragments directly into one of the pronuclei of embryos
at the one-cell stage has been, and still is, widely used. Another popular technol-
ogy, which was developed almost concomitantly, makes use of pluripotent stem
cell lines derived from mouse embryos [embryonic stem (ES) cells], which can
be cultivated and manipulated in vitro just like somatic cells and subsequently
inserted into a blastocyst to participate in the formation of the germline of a chi-
meric organism. Transgenic animals have also been created by lentiviral infection
of early embryos, by transposable elements, and by a few other techniques such as
those recently reported that make use of specially designed site-specific nucleases.
Transgenic mice are produced routinely in an ever-increasing number of labo-
ratories. They are also made to order by several private companies. All these trans-
genic animals have been invaluable for answering biological questions related to
gene function and regulation. They are instrumental in the analysis of tissue dif-
ferentiation and ontogeny, for example, by allowing the tracking of cell lineages.
Finally, they allow the development of refined animal models of human genetic
diseases.
In the previous chapter we concluded that the discovery of the mutagen
ethyl-nitrosourea (ENU) could be considered a milestone in the history of
mouse genetics, essentially because it made possible the creation of a virtually

268 8 Transgenesis and Genome Manipulations
unlimited number of new mutant alleles. Similarly, the advent of transgenic

technology has been a true revolution, eliciting unprecedented changes in mam-
malian genetics and related fields. This chapter focuses on the production and
use of transgenic mice.
8.2 Transgenesis Resulting from Pronuclear Injection of

Cloned DNAs
The stable insertion of foreign DNAs into the germ line through microinjec-
tion into the pronuclei of fertilized mouse eggs was reported in the early 1980s
in simultaneously several laboratories using the same technique but with differ-
ent DNA molecules (Brinster et al. 1981; Costantini and Lacy 1981; Gordon and
Ruddle 1981; Harbers et al. 1981; Wagner et al. 1981a, b). It was not until 1982
that the first transgenic mouse with a clear phenotype was developed by Palmiter,
Brinster, and colleagues: a “giant” mouse carrying (and overexpressing) a rat
growth hormone gene (Palmiter et al. 1982). Since these first descriptions, the
technique has been improved and a variety of protocols for the efficient genera-
tion of transgenic mice has been published. Among the most popular “cookbooks”
dealing with the subject, we recommend those by Hogan et al. (1994), and more
recently by Hammes and Schedl (2000), Jackson and Abbott (2000), Houdebine
(2003), Nagy et al. (2003), and Koentgen et al. (2010). We also recommend visit-
ing the webpage of the International Society for Transgenic Technologies (ISTT)
at http://www.transtechsociety.org/.
8.2.1 The Basic Experimental Protocol
The production of transgenic mice is achieved by injection, generally with a

sharpened glass micropipette, of a few picoliters of a DNA solution (concentra-
tion ~2 ng/μl) directly into one of the pronuclei, while the egg proper (the
zygote) is held, by suction, to another glass micropipette (the holding pipette).
In most instances, the foreign DNA is injected into the male pronucleus because
it is a little bigger and closer to the egg membrane than the female pronucleus.1
In skilled hands, around 10–20 % of the microinjected eggs develop to term
into a transgenic animal. Identification of the transgenic status is achieved by
PCR amplification of DNA samples prepared from the presumptive transgenic
animals with specific primers, and confirmation is obtained by using Southern
blotting (Fig. 8.1).
1 For this reason, the technique is sometimes designated “pronuclear transgenesis.”

8.2 Transgenesis Resulting from Pronuclear Injection of Cloned DNAs 269
x x
Day 4:
Day 1: Day 3: Mating of female mice
PMSG injection HSG injection & mating with vasectomized males
Day 4: Day 5:
Isolation of oocytes and Oviduct transfer to pseudo-
microinjection of DNA pregnant foster mothers
Day 25:
Birth of offspring
Day 46:
Genotype (DNA)
analysis of offspring
Fig. 8.1 Producing transgenic mice by pronuclear injection. The chart represents the different
steps for the production of transgenic mice by pronuclear injection. Eggs are flushed out of the
oviduct immediately after fertilization and then the transgene is microinjected in vitro with a
glass micropipette. Once injected, the eggs are kept in vitro for a few hours and then transplanted
into pseudo-pregnant females. Genotyping of the G0 (presumptive) transgenic mice can be
achieved at any time from birth onwards. Every pup genotyped as positive by PCR (i.e., hemizy-
gous Tg/0 carrier) should be considered a “founder,” and independent lines should be developed
from each founder
The DNA that is injected into the pronucleus can be either an unmodified or a
natural copy of a gene cloned in its native genomic configuration, with its natural
promoter, all its introns and other 5′or 3′ regulatory sequences, plus a few tenths
of kb upstream and downstream of the sequences of interest. In most instances,
however, the DNA that is used for transgenesis (the “transgene” proper) is artifi-
cial and designed in the laboratory according to the purpose of the experiment. It
generally consists of several elements gathered in vitro, one piece at a time, then
assembled using the most appropriate recombinant DNA technology. Finally, the
transgene is cloned into a plasmid for amplification, mass production, and storage.
When constructing such a fusion or chimeric gene for expression in transgenic
mice, it is often easier to use a cDNA clone incorporating the coding sequences
rather than the genomic DNA. This is especially true when the coding sequences
in question stretch over a very long DNA segment or when they comprise many
exons. Unfortunately, the levels of gene expression obtained with cDNA-based
constructs are often lower than those obtained when genomic sequences are used.
Among the many explanations that can account for this observation, the existence
of enhancers in the introns is the most likely (see Chap. 5).
Once selected, the relevant cDNA is placed under the control of a promoter,
whose choice depends upon where and when it is desired that the transgene be
expressed. When using cDNA (rather than genomic DNA) as a source of coding
sequences, it is important to make sure that there is a translational start codon
(AUG) within an upstream Kozak sequence (A/GCCPuCCAUGG), which lies
within the short 5′ untranslated region and directs translation of mRNA, and that
there is an in-frame stop codon (UGA, UAG, UAA) for translational termination.
Finally, it is also recommended to add an intron at the 5′ or 3′ end of the transgene
because this allows the production of a more stable mRNA transcript and, finally,
better transgenic expression (Brinster et al. 1988).
Experience teaches that the integration of the foreign DNA into the chromo-
some of the host probably occurs at random. In most instances, DNA integra-
tion occurs at the one-cell stage and at a single site but this is not a rule, and in
10–20 % of cases, the integration is delayed and occurs later during development.
The mechanism of stable integration into the host genome is not precisely known,
but it likely requires a double break (a nick) in the host (or recipient) DNA that
is promptly repaired. Some scientists have suggested that this break might be the
consequence of a trauma caused by the glass micropipette or by the injection of
the DNA suspension. Even if this suggestion makes sense, it is probably not the
only way for a transgene to integrate into a genome since delayed integrations,
which are observed occasionally, are obviously not trauma dependent. When the
foreign DNA does not integrate and stays isolated (as an episome, for example)
in the nucleus for a few hours and integrates only at a later stage of development
(2-cell; 4-cell), the organism develops as a mosaic. In this case, the detection of
the transgene is more difficult and its transmission is unpredictable. In the case
where the foreign DNA is present in all cells of the founder transgenic animal
(noted F0, sometimes G0), it is then transmitted generation after generation as a
new dominant “Mendelian” character.
The generic symbolic designation for a transgenic insertion is Tg. When the
structure of the transgene is known, which is generally the case, a more precise
designation applies. In this regard, we encourage the readers to refer to the guide-
lines for the standardized genetic nomenclature of transgenes in mice and rats at:
http://www.informatics.jax.org/mgihome/nomen/gene.shtml#transg.
In contrast to gene and allele symbols, transgene symbols must not be italicized
when they result from insertions of foreign DNA because they are not part of the
native mouse genome.
The founder transgenic animals are hemizygous for the DNA segment (the
symbol should be Tg/0, not Tg/–), and accordingly, the establishment of a “trans-
genic strain,” in which the transgene is propagated by sexual reproduction,
requires genotyping at each generation to avoid losing the transgenic DNA, unless
the carriers have an obvious phenotype.
A method of safely maintaining a transgene in a mouse strain is to put it in
the homozygous state, but this is difficult to achieve in practice. One reliable way
Fig. 8.2 Fluorescence in situ hybridization. Fluorescence in situ hybridization (FISH) with a

transgene-specific probe indicates the localization of the transgene (green dots) in the karyotype
(duplicated metaphase chromosomes). In this case, the transgenic insertion is homozygous (two
copies). Using a chromosome or gene-specific probe with a different fluorescent staining allows
for localization of the transgene on a specific chromosome (see Chap. 3)
of sorting out homozygous (Tg/Tg) from hemizygous (Tg/0) mice relies on the
statistical analysis of their progeny when mated with a wild-type (WT or non-
transgenic) partner (i.e., a progeny testing). A male mouse, identified as a carrier
of the transgenic insertion based on a DNA test, producing only Tg/0 transgenic
offspring in a progeny of 10 pups, when crossed with a non-transgenic partner
has a greater than 90 % chance of being homozygous for the transgene (Tg/Tg).
When the progeny size increases to 15, with only Tg/0 offspring, the probability
increases to 99 %. Other possible means of identifying homozygous Tg/Tg mice
are by quantitative real-time PCR (qRT-PCR) to determine zygosity and to distin-
guish hemizygous from homozygous transgenic mice (Ballester et al. 2004), or by
cloning a segment of the DNA, flanking the transgene by inverse PCR and using
it as a chromosomal marker for transgene localization. The transgenic insertion
can also be visualized by in situ hybridization with a fluorescent dye (FISH) and
accordingly located on a specific chromosome (see Chap. 3) (Fig. 8.2).
8.2.2 Factors Influencing Transgenic Expression
The number of copies of the transgene that integrates into the host genome is not
controlled and ranges from one to several tens or even hundreds. Because sticky
ends are generated when the foreign DNA is processed for injection, the cloned
DNA copies are generally arranged in head-to-tail arrays in the transgenic inser-
tion with frequent, and sometimes extensive, rearrangements generated in the
flanking regions. Using quantitative PCR technology, it is possible to roughly esti-

mate the number of copies of the transgenic DNA; however, this is neither accu-
rate nor reliable and, as we shall see, it sometimes changes with time.
As we already mentioned, investigators have no way of choosing the loca-
tion where the foreign DNA will integrate. However, the integration site can seri-
ously influence the transcription, and accordingly the expression, of the transgene.
This is the case, for example, when the insertion site is in a heterochromatic or
untranscribed (hypermethylated) region of the genome or when it is strongly influ-
enced by a silencer sequence operating in its close vicinity. In these two cases, the
transgene is weakly expressed or not expressed at all. Conversely, the sequences
surrounding the transgene may contain regulatory elements acting on its promoter
as enhancers of transcription. These enhancer sequences sometimes lead to an
ectopic expression of the transgene; in other words, to an expression pattern that
does not match with the spatial or temporal expression normally expected from the
transcriptional regulatory elements the transgene contains.
These unexpected and somewhat erratic variations in expression are the
consequences of a phenomenon known as the position effect and represent one
of the main weaknesses of pronuclear transgenesis. The position effect and
variations in copy numbers are two serious drawbacks, because both can affect
transgenic expression. For this reason, in all cases, it is absolutely essential to
make sure that the transgene is indeed fully expressed by checking whether all
of the expected transcription products are present. One should also verify, as
thoroughly as possible, that the structure of the transgene has not been affected
by the mechanical handling of the DNA during the process of injection. This
recommendation is especially important for large transgenes such as those
made from yeast artificial chromosomes (YACs) or bacterial artificial chromo-
somes (BACs).
Since it is impossible to predict the effects of the integration site (i.e., the
genomic environment) and of the number of copies on transgenic expression, it is
highly recommended, when developing a transgenic strain for experimental pur-
poses, to compare the offspring of several different founder transgenic mice and
to consider only the features common to at least two independent strains as reli-
ably attributable to the transgenic DNA. For this reason, it is not recommended to
intercross mice originating from different founders but, on the contrary, to develop
independent Tg lines from each founder.
Another classical observation when breeding transgenic animals is that 7–10 %
of the transgenic insertions appear to be lethal when homozygous, presum-
ably because a recessive lethal mutation (most probably a gene disruption) was
mechanically generated at the time of integration in the recipient genome (inser-
tional mutagenesis).
Finally, one must keep in mind that transgenic insertions are not always sta-
ble over time, and many investigators have reported the spontaneous and unex-
pected loss of the transgene from their favorite transgenic line. When a transgenic
line is considered optimal and reliable, it is wise to preserve it as frozen sperm or
embryos.
8.2.3 Using Transgenic Mice for Studying Gene Function

and Regulation
A virtually unlimited number of transgenes can be engineered in vitro by the asso-

ciation of any coding sequence—normal or mutant—taken from any gene of any
species, including plants and bacteria, and controlled by any regulatory elements.
The use of transgenic mice is then a very convenient and efficient way to assess
the function of genes. We will consider a few cases that have been selected as
informative examples.
8.2.3.1 The Use of Transgenic Mice to Define the Function of Genes
Examples of this approach are provided by the homeogenes and the oncogenes,
both of which are important actors in mammalian development. Homeobox-
containing genes, the homeogenes, are transcriptional regulators with a remote
ancestral origin, which are present in mammalian genomes and arranged in four
paralogous clusters (Hoxa, Hoxb, Hoxc, and Hoxd). Because their structures are
very similar, it was impossible to decide a priori whether each of these genes had
a specific function, whether they had an effect because of the copy number (addi-
tive effect) or whether some of the copies were simple “backup” copies, preserved
by evolution for unknown purposes. Transgenic mice were then made for some of
these homeogenes with an intact coding sequence driven by a regulatory sequence
different from the native one (driving ubiquitous expression, for example). In most
instances, the embryos born with such extra transgenic insertions exhibited severe
“homeotic” transformations indicating that indeed, most of the homeogenes in the
Hox clusters had a specific function in the developmental patterning of the mouse
embryo, a patterning reminiscent of their function in Drosophila, where they were
initially discovered (Duboule 1998).
Transgenic mice have also been created with the coding sequence of (intact or
mutated) oncogenes, or the sequence of genes whose function were not completely
understood, downstream of a variety of regulatory sequences. Among these genes
are the oncogenes Abl1, Jun, Mos, Nras, and Myc, as well as the tumor suppres-
sor genes Trp53 and Rb. Transgenic mice overexpressing oncogenes develop neo-
plasias in different tissues, depending on the promoter selected for the construct.
For example, mice overexpressing the oncogene Myc driven by immunoglobu-
lin enhancers develop lymphoid malignancies (Adams et al. 1985). The famous
OncoMouse™ (the name is a trademark) is another example, but in this case, it
carries the activated oncogene v-Ha-ras under the control of the MMTV promoter
and, hence, produces mammary tumors (Hanahan et al. 2007). The subsequent
analysis of these transgenic animals has provided an enormous amount of infor-
mation concerning the role of these oncogenes in the regulation of several basic
cellular functions and during the process of malignant transformation. The unique
advantage of transgenesis in the case of homeogenes, oncogenes, and tumor
suppressor genes is to make the analysis of gene function(s) possible at the level
of the whole organism.
8.2.3.2 Using Transgenic Mice to Identify and Characterize the

Regulatory Sequences of Genes
While many mammalian genes are constantly and ubiquitously expressed, others
are expressed in a tissue-specific manner, or only during embryonic life or only in
the adult organism. Such variations in expression patterns occur because the genes
are controlled by regulatory sequences that are in many cases, although not
always, located in cis and upstream of the coding regions.2 A good example of
such tissue-specific regulation was reported for the gene encoding the cytokine
leptin, which is expressed almost exclusively in adipocytes. After positional clon-
ing of the mouse mutant gene obese (Lepob-Chr 6) (Zhang et al. 1994), it was
demonstrated that the obese phenotype was a consequence of a nonsense mutation
in codon 105 of the gene encoding the 16 kDa leptin protein. Researchers also
learned that the highly tissue-specific expression of the Lep gene is controlled by a
cis-acting regulatory sequence 161 bp long located upstream of exon 1 (He et al.
1995). For many genes, unfortunately, the regulatory sequences are not yet charac-
terized and geneticists must design experiments to identify them accurately (see
Chap. 5). This is important for a better understanding of gene regulation, of
course, but it is also important if we consider that accumulating such data will cer-
tainly help in the future in silico identification of the regulatory elements based on
sequence analogies.3
Transgenic mice are helpful for the identification of these regulatory sequences
because experience teaches us that genes cloned in their native genomic config-
uration and introduced into the mouse germ line by transgenesis retain, in most
instances, their tissue-specific and stage-specific patterns of expression, despite
their integration at random sites. A popular strategy is to design in the laboratory
a series of transgenes whose coding sequence encodes an easy-to-detect product
which is not normally encoded in a mammalian genome (such a sequence is called
a reporter gene), and to associate it by genetic engineering with a variety of regu-
latory DNA sequences, either upstream of the coding region, at the 5′ end or, less
frequently, downstream of the 3′ end.
The gene encoding chloramphenicol acetyltransferase (CAT), from a transposon
of Escherichia coli, has been extensively used to characterize the specific expression
2 The genetic elements regulating gene expression are sometimes numerous and not always
located in the close vicinity of structural genes. This explains (at least in part) why cloned struc-
tural genes, when used as transgenes, are sometimes regulated differently from the same genes in
their natural, native environment (see Chap. 5). This point is inherent to transgenesis by in ovo
injection and must always be kept in mind.
3 In situ hybridization with labeled cDNAs is another way of analyzing the expression profile of
a given gene.
associated with regulatory sequences because CAT activity can be assayed thanks to
a very sensitive enzymatic test that has no background in eukaryotic cells (Overbeek
et al. 1985). CAT has been progressively replaced by the gene encoding luciferase in
the firefly (Photinus pyralis), largely because the assay to measure it is easier (Lira
et al. 1990). lacZ, the historical gene encoding β-galactosidase of Escherichia coli
(Goring et al. 1987), has been the cellular marker of choice to track cells in embryos
and adults because of the ease of its detection and high cellular resolution in fixed
embryos and tissues. The lacZ gene appeared to be particularly useful for studies
of tissue- or position-specific gene expression. However, a major limitation is that
lacZ cannot be used to mark cells in living tissues because the protocol to detect
its expression requires tissue fixation. Fluorescent proteins offer advantages over
enzyme-based reporters (e.g., lacZ, CAT) in the sense that their visualization does
not require tissue fixation and is both quantitative and noninvasive. Indeed, fluores-
cent proteins make it possible to mark specific cells in living organisms, and also to
follow such cells using fluorescence-imaging techniques (Fig. 8.3).
A classical reporter gene has been developed that consists of the sequence
of the green fluorescent protein (GFP) of the jellyfish Aequora victoria (Misteli
and Spector 1997). The product of this gene emits a green fluorescence elicited
by direct illumination with blue light, and the analysis of the expression pattern
requires neither fixation of the tissue nor cofactor or specific substrate, only UV
light. Several variants of the wild-type GFP have been produced that emit in the
blue (BFP), cyan (CFP), and yellow (YFP) regions. A series of variants derived
from the red fluorescent protein (RFP) of the sea anemone Discosoma sp. are
increasingly used because they emit a range of wavelengths in the red region,
from the dark red of cherry to the yellow of banana. Interestingly, these differ-
ent reporter genes can be combined allowing multiplexing and co-visualization
Fig. 8.3 Analysis of gene expression with a reporter gene. Left expression of the structural gene
encoding LacZ with regulation by the Desmin promoter. Observation of this embryo allows for
detection of the tissues in which Desmin, a type III intermediate filament, is expressed (Courtesy
C. Babinet). Right the embryo (recovered 13 days post-fertilization) is heterozygous for a knock-
in allele in which the H2B-GFP coding sequence has been inserted in-frame into the gene encod-
ing the platelet-derived growth factor receptor, alpha polypeptide (Pdgfra+/H2B-GFP) (Courtesy J.
Artus)
of several fluorescent proteins expressed in different tissues of a mouse

(Passamaneck et al. 2006). Transgenic mice with reporter genes have been, and
still are, extensively used by developmental geneticists (Lichtman et al. 2008).
They have also greatly contributed to the annotation of the noncoding sequences.
8.2.4 The Use of Transgenic Technology to Generate

Tissue- or Cell-Specific Ablations
Transgenic animals have been designed using tissue-specific regulatory sequences

associated with sequences encoding cytotoxic proteins, with the aim of program-
ming the genetic ablation of specific cell types either in the developing embryo
or in the adult (Breitman et al. 1989). The most common strategy makes use of
sequences encoding toxic proteins such as the A chains of the diphtheria toxin
(DT-A) or of ricin (R-A), both of which block protein synthesis. In this case, the
cytotoxic effect takes place as soon as the transgene is expressed. These studies
indicate that programmed ablation of specific cell types can be stably transmitted
generation after generation through the germ line (Breitman et al. 1987).
Another strategy has been developed that relies on the induced intracellu-
lar expression of the enzyme thymidine kinase (tk) of the herpes simplex virus
(HSV). This enzyme is not directly toxic to animal cells but, unlike the mamma-
lian thymidine kinase, it can phosphorylate certain nucleoside analogs such as acy-
clovir or ganciclovir, converting them into drugs that are toxic to dividing cells.
In this particular case, the cell-killing effect becomes conditional since it depends
both on the expression of the gene coding for viral thymidine kinase and on the
administration of nucleoside analogs.
These methods of genetic ablation can be used to confirm the tissue specificity
of a promoter; from this point of view, they appear complementary to the methods
described above. Unfortunately, these methods, particularly the one using the highly
toxic DT-A or R-A toxins as cell-killing agents, have a major drawback—the con-
sequence of the extreme sensitivity of eukaryotic cells to these toxins. If the regula-
tory elements used in transgenic construction are not specific enough, a background
expression of the transgene in cells that are not targeted results in misleading patho-
logical conditions. This is mostly why, nowadays, this strategy of cell- or tissue-spe-
cific ablation has been abandoned for more specific approaches (see further).
8.2.5 Transgenic Complementation of a Mutant Allele

Identified by Positional Cloning
As we already mentioned in the previous chapters, positional cloning of mouse

mutations is an efficient approach for assessing the function of genes because the
strategy directly associates a mutant phenotype with a specific gene. For example,
cloning a gene that is responsible for a leukodystrophy, once mutated, will point by
definition to a gene involved in the development and organization of the white mat-
ter of the nervous system. However, when the candidate gene has only two alleles—
one normal and one mutant—with the mutant being, for example, the consequence
of a missense mutation (which occurs in about 75 % of cases), it is risky to conclude
that the mutant allele is indeed responsible for the phenotype because there is always
a chance, even if small, that the two observations (the phenotype and the muta-
tion) are independent. In this case, it is generally necessary to prove that the mis-
sense allele is indeed causative of the pathology, and this can be achieved either by
generating other alleles by mutagenesis (see Chap. 7 and later in this chapter) or by
attempting to rescue the mutant phenotype by transgenic complementation. In this
case, an appropriate breeding protocol is used to obtain genotypes that are certainly
homozygous for the recessive mutation in question (mut/mut), normally leading to
the deleterious phenotype, plus an additional (normal), functional transgenic copy of
the candidate gene. The observation of a normal or nearly normal phenotype for this
genotype validates the candidacy of the gene cloned by a positional approach. An
example of transgenic rescue was reported endorsing the suspicion that a missense
mutation in the gene encoding tubulin-specific chaperone E (Tbcepmn-Chr 13) was
indeed responsible for the deleterious phenotype of the mouse mutation progressive
motor neuronopathy (Martin et al. 2002).
8.2.6 Using Transgenic Mice for Modeling Human Diseases
Different types of transgenic mice have been designed either to allow scientists to
conduct experiments that were not possible with normal mice or to model a patho-
logical condition that exists only in humans. We will provide a few examples to
demonstrate the versatility of this transgenic technology.
8.2.6.1 Making Transgenic Mice Susceptible to Human Infectious

Diseases
Poliovirus, the causative agent of poliomyelitis, infects primates but cannot spon-
taneously infect mice except for some type 2 virulent strains. Transgenic animals
susceptible to all three poliovirus serotypes have been produced by pronuclear
injection of the cloned human gene encoding the cellular receptor for the virus
(Koike et al. 1991). These transgenic mice, when inoculated with poliovirus,
mimic some of the clinical symptoms observed in humans and monkeys and are
good models for studying the molecular mechanisms of pathogenesis of the virus
as well as for testing vaccines against poliovirus infections.
Another example is the bacteria Listeria monocytogenes. These bacteria, once
ingested by humans, can produce severe and sometimes fatal infections. The
mechanisms by which the bacteria passes through the human intestinal barrier
is well known: It requires the intervention of a surface protein called interna-

lin, which interacts with a host receptor, E-cadherin, to promote entry into intes-
tinal epithelial cells. Murine E-cadherin, in contrast to human or guinea pig
E-cadherins, does not interact with internalin, excluding the mouse as a model
for experimental oral infection with L. monocytogenes. In contrast, in transgenic
mice expressing human E-cadherin, internalin was found to mediate invasion of
enterocytes and crossing of the intestinal barrier (Lecuit et al. 2001). These results
illustrate well the value of transgenesis for understanding the physiopathology of
human infections.
Models of the kind we have just described are, of course, of greatest interest for the
study of infectious pathology, in particular for the development of efficient therapies
and vaccines. Unfortunately, they illustrate rather exceptional situations and, in many
cases where transgenesis was used to make animals susceptible to human pathogens,
the situation has been discouraging. The determinism of susceptibility to infectious
agents is sometimes complex and is rarely determined by the presence or absence of a
single, species-specific cellular receptor. Progress in this area certainly awaits the dis-
covery of genes whose products facilitate viral integration into the cell and full devel-
opment of the replicative cycle. For example, scientists at the Rockefeller University
and at the Scripps Research Institute demonstrated that the genes encoding CD81 and
occludin were required for Hepatitis C virus (HCV) to enter human cells, and they
demonstrated that making mice transgenic for these human genes made it possible to
infect these transgenic animals with HCV (Dorner et al. 2011).
Even if perfect and faithful models cannot be made available simply by the
mere addition of a few DNA segments, transgenic technology remains an interest-
ing strategy to make progress in some aspects of infectious pathology. Transgenic
technology, for example, allowed scientists to clarify the role of the complex clus-
ter of genes encoding oligo-adenylate synthetase 1 (Oas1) in mouse susceptibility
to flaviviruses (Scherbik et al. 2007; Simon-Chazottes et al. 2011) and has already
provided insights into the pathogenesis of HIV-1.
8.2.6.2 Transgenic Models of Human Genetic Diseases
A mouse model of the human disease osteogenesis imperfecta type II (OMIM

166210) has been produced by injecting in ovo an abnormal mouse pro-α1 (I)
collagen gene (Col1a1), orthologous to the abnormal human gene (Stacey et al.
1988; Pereira et al. 1993). The animals carrying such a transgene appeared very
sick soon after birth, because of the modification of the extracellular matrix by
the abnormal collagen fibers. In this case, the transgene had a dominant deleteri-
ous effect, the affected animals were almost impossible to breed, and the model
proved to be of limited value. Nowadays, much better models can be generated
using advanced techniques of transgenesis, as we will describe later.
A transgenic mouse strain has been created by pronuclear injection of both the
normal human α-globin and the abnormal βs-globin gene characteristic of sickle-
cell anemia (Ryan et al. 1990). These animals were bred to β-thalassemic mice
to reduce endogenous mouse globin levels. When erythrocytes from these mice
were deoxygenated, greater than 90 % of the cells displayed the same characteris-
tic sickle shapes as erythrocytes from humans with sickle-cell disease. Compared
to controls, the mice had decreased hematocrits, elevated reticulocyte counts,
reduced hemoglobin concentrations, and splenomegaly, which are all indications
of human sickle-cell disease. Such models are also of great help in the understand-
ing of the pathophysiology of this debilitating disease as well as in the develop-
ment of new drugs and therapies.
8.2.7 Transgenic Animals with Large DNA Inserts
Several techniques have been used to create mice transgenic for large DNA frag-
ments. Among these techniques, the direct pronuclear microinjection of purified
YACs or BACs has been the most popular (Jakobovits et al. 1993; Schedl et al.
1993; Lee and Jaenisch 1996; Van Keuren et al. 2009; Rossant et al. 2011). Such
transgenic mice, when available, are very helpful for understanding the mecha-
nisms operating when, for example, the genetic defect results from an unknown
alteration occurring in a relatively large genetic region, or simply when the molec-
ular origin of the defect is not completely clear. Several examples documenting the
ability of wild-type alleles carried in YACs to complement mutations have been
reported. The first one was the simple, complete rescue of the classical mouse
albino mutation after injection into the germ line of albino (Tyrc/Tyrc) mice of a
250 kb YAC encompassing the wild-type mouse tyrosinase (Tyr) gene with all its
introns and 155 kb of the 5′ flanking region (Schedl et al. 1992).
Original animal models of human genetic diseases have also been created using
YAC transgenes. Among these, we must cite a model for Charcot–Marie–Tooth
disease type 1A (Huxley et al. 1996) and a model for Huntington disease in which
large intergenerational trinucleotide repeat expansions could be recreated, endors-
ing the use of these transgenic mouse models to refine the understanding of triplet
repeat expansion and the resulting pathogenesis (Gomes-Pereira et al. 2011).
The possibility of inserting large-sized DNA fragments into the mouse genome
will certainly be very useful for a better understanding of the phenotypic impact
of the variations in genomic copy number (CNVs) (discussed in Chap. 5), as well
as for the production of better models of Down syndrome (discussed in Chap.
3). Many fragments cloned from human chromosome 21 have been added to the
mouse genome by in ovo transgenesis, producing phenotypes more or less remi-
niscent of those of human trisomy 21 (Smith et al. 1995; O’Doherty et al. 2005; Yu
et al. 2010; Herault et al. 2012; Rueda et al. 2013). None of these models is per-
fect because of the complexity of the phenotype when several genes on different
mouse chromosomes are used, but good progress is being made and transgenesis
appears to be a technique of choice in this matter.
Many transgenic models of Alzheimer disease have been developed over
the past several years. Most of these models replicate some of the pathological
features of the disease, such as plaque-like amyloid accumulations and astrocytic

inflammation, but not all phenotypic aspects. In particular, the behavioral deficits
are not faithfully modeled (Lithner et al. 2011).
Transgenesis with BACs or other large chromosomal segments is bound to
become a very popular technology, with the foreseeable development of quantita-
tive genetics in the years to come. The reason is that, unlike in the case of single
Mendelian mutations, the genomic regions that have a quantitative effect on the
phenotype are mostly unknown and, in this case, BACs containing the DNA seg-
ment where a quantitative trait locus (QTL) has been localized can be transferred
into zygotes and the resulting mice tested for the quantitative trait in question.
However, for this system to be applicable, BAC libraries must be available that
contain the appropriate alternative alleles at the QTL in question (Heintz 2001;
Abiola et al. 2003). Such libraries are now being prepared for different mouse spe-
cies and strains.
8.2.8 Transgenic Knockdowns
In Chap. 5, when describing the different sorts of RNAs that are encoded in the
mouse genome, we discussed the case of siRNAs and their possible use for gene
silencing. Experiments of that kind have been undertaken several years ago by
Katsuki et al. (1988) to assess the possibility of controlling gene expression by
inducing the production of antisense RNAs in the genome. For their experiment,
the Japanese scientists constructed a plasmid containing the promoter of the gene
encoding the mouse myelin basic protein (MBP), followed by a portion of the rab-
bit β-globin gene associated with the mouse MBP-cDNA in the antisense orien-
tation and a polyadenylation site. They observed that several transgenic mice for
this transgenic construction had a phenotype similar to that of the mutant mouse
shiverer (Mbpshi-Chr 18). Antisense MBP messenger RNA was transcribed at
high level in these mice, while the endogenous messenger RNA was reduced. The
researchers concluded that the mice with an abnormal phenotype were constitutive
knockdowns and that the transgene expression in vivo resulted in RNA interfer-
ence (RNAi).
Since this first (successful) experiment, several other attempts at production
of knockdown have been undertaken; some have been successful but most have
failed. The reason is that, unlike in plants or invertebrates, double-stranded RNAs
(dsRNAs) elicit an interferon response in mammals, resulting in global inhibi-
tion of protein synthesis and non-specific mRNA degradation. For this reason,
short synthetic dsRNAs, whose length is below 30 bp, have been used to trigger
the specific knockdown of mRNAs in mammalian cells without interferon induc-
tion. In the best experimental conditions, the efficiency of target knockdown can
be as high as 90 % or greater, with permanent gene silencing in transgenic organ-
isms indicating that the production of transgenic antisense RNA is an interesting
approach to assessing gene function in vivo (Hitz et al. 2009).
8.2.9 Assessing the Mutagenic Activity of Chemicals with

Transgenic Mice
As briefly mentioned in Chap. 7, at least two independent mice transgenic strains

have been developed to assess in vivo the mutagenic activity of chemical sub-
stances of the environment: These strains are commercially available under the
names of Big Blue® and Muta™Mouse (Wahnschaffe et al. 2005a, b). Transgenic
mice of the Big Blue® strain have ~30–40 copies of the lambda LIZα shuttle
phage vector integrated into their genome and the target for mutagenesis is the lacI
gene. Muta™Mouse mice have ~80 copies of the lambda-gt10-lacZ shuttle vector
integrated into their genome and the target is the entire lacZ gene.
Whatever the transgenic strain, the chemical compound to be tested is adminis-
tered to the transgenic mice under several forms and at different doses. After a few
days, DNA samples are then extracted from several tissues of the tested mice. The
targeted genes are excised and packaged into lambda phage heads by using specially
designed molecular kits, and the phages are transfected into bacteria. Finally, the
transfected bacteria are plated on indicator plates containing selected chromogenic
substances. Under these conditions, the phage-transfected bacteria with mutations in
the targeted genes form plaques of a different color from those of bacteria with a
non-mutated target gene, and the ratio of colored plaques to colorless plaques is a
reliable measurement of the mutagenicity of the compound tested.4 These transgenic
strains have been extensively used and have provided fast and reliable estimations.
8.2.10 Mutations Induced by Pronuclear Transgenesis
As mentioned earlier, approximately 8–10 % of transgenic insertions result in reces-

sive lethal mutations, and a much lower percentage in recessive viable. Good exam-
ples of the situation are two independent alleles at the Formin locus5 (Fmn1-Chr 2)
and a mutation described as cryptorchidism with white spotting (crsp-Chr 5)
(Woychik et al. 1985; Messing et al. 1990; Overbeek et al. 2001). Such mutations by
insertions would appear a priori to be interesting situations, considering that the
inserted DNA (whose sequence is known) could be used as a tag for the identifica-
tion of the mutated gene and accordingly for facilitating its positional cloning (for
reviews, see (Jaenisch 1988; Gridley et al. 1990; Meisler 1992). In practice, how-
ever, the cloning of DNA flanking the insertion loci has often proved difficult as a
consequence of the structural changes generated by the insertion.
4 The phage-transfected bacteria with mutations in the lacI gene form blue plaques, whereas
bacteria with a non-mutated lacI form colorless plaques in tests with the Big Blue® strain. With
the Muta™Mouse strain, the basic principle is similar but the color of the plaques depends upon
the experimental conditions.
5 The first of the two alleles resulting from a transgenic insertion at the Formin locus (Fmn-Chr
5) has been known for a long time under the name of limb deformity (ld).
8.3 Generating Alterations in the Mouse Genome Using

Embryonic Stem Cells
The technique of transgenesis by pronuclear injection of exogenous DNA

sequences has been a true revolution in mammalian genetics. It has enabled hun-
dreds of experiments that have provided answers to fundamental questions regard-
ing the organization and functioning of the mammalian genome and has permitted
the creation of many useful and original animal models for biomedical research.
Unfortunately, the technique has some important limitations. One is that it allows
the addition but not the deletion or substitution of genomic material meaning
that, except in some rare situations, it is not possible to produce alterations with
a recessive phenotypic expression. Another limitation is that the injected DNA
inserts randomly in the genome of the host, and for this reason, the expression of
the transgene often varies from one founder transgenic mouse to the next due to
unique interactions with other genomic sequences in the background and discon-
nection of the transgene from its natural regulatory elements (see Chap. 5). Such
limitations do not apply to the genetic alterations produced by using the tech-
niques of genetic engineering in embryo-derived stem cells (ES cells). These tech-
niques have been developed over the last 30 years and are still extensively used in
the mouse for the production of a variety of targeted alterations. We will review
the most commonly used.
8.3.1 Embryonic Stem Cells and their Advantages
ES cells were developed in the early 1980s (Evans and Kaufman 1981; Martin
et al. 1981). They were derived from cells dissected from the inner cell mass
(ICM) of blastocysts that were cultured in vitro, generally on feeder lay-
ers of fibroblasts, in tissue culture media supplemented with a few percent
of fetal calf serum, with a high concentration of glucose, with glutamine and
β-mercaptoethanol. To prevent these cells from differentiating in vitro, low con-
centrations of leukemia inhibitory factor (LIF) were added to the medium and the
cells were re-plated at a relatively rapid pace.
ES cells represent a material of choice for geneticists because they can be
manipulated (almost) like ordinary somatic cells, as long as they are maintained in
vitro, while retaining all their developmental potentialities, in particular their
capacity to differentiate into derivatives of all three embryonic germ layers (pluri-
potency). In addition, and most importantly, when merged with the cells of the
ICM of a recipient blastocyst, many ES cells are capable of participating in the
formation of chimeric embryos, and provided that these ES cells are euploid (i.e.,
with 2n chromosomes, a normal XY or XX complement, and no deletions or other
types of chromosomal rearrangements), they are often capable of participating in
the formation of the germ-cell lineage of the embryos in question. It is then
8.3 Generating Alterations in the Mouse Genome Using Embryonic Stem Cells 283
possible to apply to ES cells the classical techniques used in somatic cell genetics
while they are in vitro (e.g., selection based on resistance or susceptibility to a
specific drug), to isolate clones of cells with a pre-defined genetic characteristic, to
“shuttle” them back into the germ line of a chimeric mouse, and finally to breed a
strain of mice that have integrated into their genome an alteration engineered in
vitro. The first experiments on genetic engineering with this type of cells were car-
ried out by Gossler et al. (1986) and by Robertson et al. (1986). They were real
breakthroughs,6, 7 when these experiments were performed, most of the ES cell
lines available for the purpose of scientific research were derived either from
embryos of the 129/SvPas inbred strain (new nomenclature 129S2) or from the
129/J strain (new nomenclature 129P3/J). Nowadays, taking advantage of techno-
logical progress, especially in terms of culture conditions, many other ES cell lines
have been derived from a variety of strains and most of them are stable and relia-
ble, producing a high percentage of chimeric animals and a good germ line trans-
mission ratio. The ES cell lines derived from strain C57BL/6N have become
popular and have been selected in many transnational projects. This was a wise
choice given that the reference sequence of the mouse genome is also from the
C57BL/6 inbred strain.8 ES cell lines derived from NOD, BALB/c, and some
immunodeficient strains (such as NSG) are also available or under development.
On the other hand, in the laboratory rat, the development of germ line-competent
ES cells was only possible very recently (Ping et al. 2008).
Chimeras resulting from the fusion of an engineered ES cell with cells of the
ICM of a recipient embryo can be identified, a few days after birth, for example,
on the basis of their dappled coat color. This is very obvious when, for example,
6 Well before the development of ES cells, another kind of cell, the embryonal carcinoma or
EC cells, was used by oncologists and geneticists for investigating the genetics of cell–tissue
differentiation. These cells were derived from spontaneous or experimentally induced testicular
or ovarian teratocarcinomas (Stevens 1960). They were cultured in vitro, in the form of stable
undifferentiated cell lines and then transplanted into mice of the same strain (syngeneic trans-
plantation). Most of these cell lines, once engrafted, were able to differentiate into a variety of
tissue (nervous tissue, bone, fat tissue, muscle, etc.), and some even proved able to participate
in the formation of a chimeric organism (Papaioannou et al. 1975). They had, however, major
drawbacks for the study of tissue differentiation: They were malignant and became rapidly ane-
uploid, and accordingly, they could not be used for the production of chimeric mice with germ
line transmission.
7 Induced pluripotent stem cells (iPSCs) are pluripotent cells derived from adult somatic cells
after forced re-expression of some specific genes that are normally inactive. Such cells have been
established in many species including human and mice. These iPSCs have many characteristics
in common with ES cells and are being used in many experiments (for example, in the area of
regenerative medicine). However, they have no obvious advantages over the long-established ES
cells for the production of transgenic mice, and accordingly, they will not be considered in this
chapter.
8 The two strains C57BL/6N (ES cells) and C57BL/6J (genome sequence) are not completely
identical, and recent estimates indicate a difference of ~1–2 % (SNPs) at the genome level (see
Chap. 9).
ES cells ES cells ES cells
ES cells in vitro Genetically engineered ES cells

injected into a blastocyst
Heterozygous and wild type mice Chimeric mice resulting from

resulting from engineered ES cells engineered ES cells and normal cells
Fig. 8.4 Targeted mutagenesis in the mouse using engineered ES cells. The chart represents the
different steps for the production of transgenic mice from genetically modified ES cells. ES cells
can be cultured in vitro for several generations, remaining in an undifferentiated status. While
in vitro, the ES cells can be manipulated like ordinary somatic cell lines and, in particular, can
then be selected on the basis of specific criteria. ES cells can also be placed inside full-grown
blastocysts where they spontaneously merge with the inner cell mass. Provided that the ES cells
are still pluripotent and euploid, fertile chimeric mice can result from these reconstructed blas-
tocysts. Mice with a dappled coat color in the figure are chimeras derived from blastocysts of
(albino) hybrid mice (CSJF1) into which ES cells derived from a pigmented strain (129/Sv) were
injected after several generations of in vitro culture. The size of the spots may vary according to
the experimental conditions, but this does not faithfully reflect the percentage of chimerism in
the germline. All of the other pigmented offspring of the chimeric mice are heterozygous for the
genetic alteration(s) that may have been engineered in the ES cells. Two more generations are
then necessary to observe the alteration in the homozygous state, and selection of the progenitors
requires DNA genotyping
the ES cells are derived from the C57BL/6N inbred strain (which is non-agouti
a/a—i.e., solid black) and the recipient blastocyst from either a wild-type (agouti
A/A) or albino (Tyrc/Tyrc) strain. In these conditions, the chimeras exhibit a mix-
ture of black and agouti (or albino) spots (Fig. 8.4).
Using coat color as a reference, one can estimate the percentage of chimerism,
but a high level of chimerism does not necessarily correspond to a high rate of
germ line transmission. Although chimeras can be from either sex, males are gen-
erally the only sex with germ line transmission because the majority of ES cell
lines are XY. When grown in vitro for several generations, many (male) ES cells
have a tendency to lose their Y chromosome and become XO.
8.3.2 Targeted Mutagenesis in ES Cells
The basic principle that characterizes targeted mutagenesis consists of applying a

selection pressure on ES cells cultured in vitro that confers an advantage to the
cells that may have lost (or acquired), spontaneously or after experimental manip-
ulation, a characteristic encoded by a specific gene. The loss of a specific char-
acteristic may result from a mutation, a deletion or any other kind of alteration,
impairing the function of a given gene. The acquisition of a new heritable char-
acteristic generally results from the transfection of foreign DNA molecules into
the ES cells, followed by selection of the transfected cells based on a selective
advantage conferred by the exogenous DNA. Once selected, the mutant or geneti-
cally modified ES cells are used for the production of chimeras, with the hope that
a substantial proportion of the modified ES cells will still participate in the forma-
tion of the gametes. This will allow the production of transgenic mice with a tar-
geted alteration in their genome.
8.3.2.1 The In Vitro Production of Mouse Models

of Lesch–Nyhan Syndrome
Lesch–Nyhan syndrome (OMIM 308000) is a rare and severe X-linked metabolic

disease in humans. The defect is characterized by the absence or inactivity of the
enzyme hypoxanthine phosphoribosyl transferase (HPRT), an essential enzyme
for the catabolism of purines. No animal model for this disease was available up
to the mid-1980s, when two independent teams published the isolation of Hprt-
clones of ES cells resulting from mutations in the Hprt gene. This was achieved
after in vitro selection of Hprt-/Y mutant cells that occurred spontaneously or after
mutagenic treatment and became resistant to the toxic effect of the purine analog
6-thioguanine (6TG) when added to the culture medium.
Hooper et al. (1987) isolated a few Hprt- clones that occurred spontaneously
and were selected with 6TG, injected them into blastocysts, bred chimeric mice
and finally succeeded in establishing an Hprt- mutant strain. The isolation of
clones of mutant ES cells by the mere in vitro selection on a particular phenotype
(and genotype) was proved successful, although with a very low yield.
A technical improvement came from the use of mouse retroviruses as muta-
genic agents and was a consequence of the early observations by Jaenisch and col-
leagues (1981). Two major conclusions of these pioneering experiments were that
(i) retroviral vectors can be efficiently used as mutagenic agents for mammalian
embryonic cells; and (ii) these vectors insert into the genome without generating
extensive chromosomal rearrangements. Based on these observations, ES cells
infected with the Moloney murine leukemia virus (M-MuLV) and mutant (null)
alleles were recovered after selection with 6TG, at the same Hprt locus (Hprt-),
but this time, at a higher frequency (Kuehn et al. 1987).
The experiments reported above by Hooper and colleagues and Kuehn and col-
leagues were published simultaneously. They were the first experiments reporting
the generation of a mutant strain in vitro, in ES cells, after selection of a particular
phenotype. Surprisingly, however, the mutant mice, supposed to be a model of
Lesch–Nyhan syndrome, did not exhibit any symptoms reminiscent of the human
syndrome.9, 10 From the genetic point of view, the result was somewhat disap-
pointing but was nevertheless a great technical achievement, opening the way to
many other technical refinements.
Considering the relatively high efficiency of the technique in terms of pro-
viral integration numbers, massive infections of ES cells have been achieved
from which embryos heterozygous for random insertions have been bred. These
mutations by insertion have been put into the homozygous state using the clas-
sical two-generation micro-pedigrees (cross, backcross), and mutant phenotypes
have been observed on some rare occasions. An interesting example is the reces-
sive lethal mutation Nodaltm1.1Mku (Chr 10), with a block at the gastrula stage,
which was found to be the consequence of a proviral insertion causing the loss
of function of Nodal, a TGFβ-related gene (Lowe et al. 2001). Another mutation
of the same kind (Lrp4dan-Chr 2) was found to cause a syndrome of polysyndac-
tyly as a consequence of the insertion of the proviral copy into the gene encoding
MEGF7/LRP4, a member of the low-density lipoprotein receptor family (Simon-
Chazottes et al. 2006) (Fig. 8.5).
The strategy that consists of infecting ES cells with M-MuLV, or any other
kind of retrovirus, followed by the breeding of mice derived from the infected ES
cells, allowed the identification of a few genes with effects on development. The
retroviruses are mutagenic when they integrate into an exon or when they insert
into an intron and disorganize the splicing process of the transcript encoded in
the neighboring exons. An advantage in this case is that the retroviral insertion
can also be used as a tag to identify DNA clones containing the mutated gene.
Unfortunately, the yield of the strategy is low because, in most instances, retrovi-
ral insertions occur in noncoding regions and accordingly they have no direct or
mechanical mutagenic effects. Another major drawback is that, for most autoso-
mal genes in the mammalian genome, there is no efficient way to select in vitro
the cells heterozygous for a recessive allele. In these conditions, it is necessary to
breed mice homozygous for each proviral insertion and to unambiguously associ-
ate homozygosity for the proviral insertion with a specific phenotype, in general
by the observation of tight linkage. This, however, is a tedious, risky and time-
consuming enterprise.
9 Mutations at the mouse Hprt locus probably occurred spontaneously in the past but were not
recorded due to the complete absence of symptoms in the affected mice. We will never know for
sure.
10 The observation of differences (sometimes dramatic) in the symptomatology associated with a
human syndrome and those observed in mice affected by mutations in the same orthologous gene
is common. This, however, does not affect the value of the model.
Fig. 8.5 Proviral insertional mutagenesis. After experimental infection of ES cells with a defective

Moloney retrovirus, a proviral copy was inserted, by chance, into the first intron of the gene encod-
ing the low-density lipoprotein receptor related protein 4 (Lrp4-Chr 2). This insertional mutation
disorganized the splicing process of the gene in question, making it virtually inactive. This resulted
in the production of a fully penetrant, autosomal recessive mutation characterized by severe poly-
syndactyly (allele Lrp4dan). The images on the left show normal paws from wild-type mice. On the
right, the images depict paws from homozygous mutant mice with malformed digits and syndactyly
8.3.2.2 Another Model of Lesch–Nyhan Syndrome Resulting from

Gene Targeting
In addition to the drawbacks mentioned above, one must also remember that one
cannot target the integration of retroviruses at a specific site in the genome. In
these conditions, the mutations generated are random and unpredictable. From this
point of view, homologous recombination of extrinsic DNA molecules in ES cells
resulting in the replacement of an endogenous gene by a different allele, in most
cases non-functional, has been another breakthrough due to its potential applica-
tions. This technique is generally referred to as gene targeting.
The principle for the production of targeted mutations by homologous recom-
bination is based on the observation that DNA fragments, once introduced into
ES cells by an appropriate experimental procedure (e.g., electroporation or trans-
fection), can recombine with the DNA of the host cells to become part of their
genome. In most instances, the recombination occurs at non-homologous (or ille-
gitimate) sites, but in some rare instances, it occurs at the homologous site. As a
consequence, and provided that the transfected DNA molecules have been previ-
ously adequately modified by genetic engineering in vitro, a homologous recom-
bination event can result in the replacement of an active and functional gene by an
inactive one.
The idea that homologous recombination could occur in mammalian cells, and
in particular in ES cells, originated from observations made in other eukaryotic
organisms, in particular in the yeast Saccharomyces cerevisiae, where similar
experiments had been successfully achieved. The detailed molecular mechanisms
at work in the recombination process are not yet fully understood. It is likely that
the mechanisms of homologous recombination overlap with those of illegitimate
recombination, but a number of experiments indicate that they are not completely
identical (for review, see Hooper 1992). Homologous recombination, of course,
occurs at a much lower frequency than random integration (Smithies et al. 1985;
Wong and Capecchi 1986). At this point, it should be noted that the idea of devel-
oping such a strategy was quite audacious if one compares the relatively small size
of a cloned DNA that can be handled experimentally, to the gigantic dimensions of
a mammalian genome!
To increase the yield of homologous recombination events, experience teaches
us that the DNA molecule transfected into the ES cells must be linear, as large
as technically possible, for instance up to 10 kb and more if possible, and should
have the greatest possible length of sequence homology with the targeted DNA in
the ES cell.
The first endogenous mouse gene that was modified by homologous recombi-
nation in ES cells was again the one encoding hypoxanthine-guanine phospho-
ribosyl transferase (Hprt-Chr X) (Thomas and Capecchi 1987). The experiment
consisted of three steps. In the first step, a DNA molecule cloned from the Hprt
targeted region and containing a few exons, the intervening introns and some
flanking DNA sequences was cloned. In the second step, one exon in the cloned
Hprt-DNA molecule was replaced by a piece of DNA of roughly the same size but
with a different origin. Finally, the engineered cloned DNA was transfected into
normal ES cells by electroporation. The idea underlying this manipulation was
that, in the event of successful homologous recombination, the substitution of an
exon by a segment of exogenous DNA would make the modified Hprt gene unable
to transcribe a functional mRNA, thus generating a null allele.
While designing these “faked” or “counterfeit” DNA constructs to replace the
targeted gene, scientists, instead of using segments of noncoding DNA as a for-
eign sequence, had the clever idea to use a minigene of bacterial origin encoding
the enzyme neomycin phosphotransferase (neor) and capable of conferring to the
transfected cells the capacity to resist to the toxic effect of neomycin. In these con-
ditions, when plated in a culture medium with the antibiotic neomycin or, more
precisely, with one of its amino glycoside analogs, G418, the normal ES cells were
all killed while the cells synthesizing neomycin phosphotransferase (neor) resisted
the cytotoxic effect of the drug. In other words, only those ES cells having stably
integrated an engineered DNA molecule into their chromosomes, either at the tar-
geted locus site or anywhere else in the genome, could survive. The rare ES cells
clones where a strictly homologous recombination occurred would likely have
reciprocally exchanged a functional copy of the Hprt gene for a non-functional
one, and at the same time, they would also have acquired the property to resist the
toxic effects of 6TG just like the Hprt- mutant cells reported above.
The advantages of this technique are twofold. The first is that, after selection
with G418 (eliminating all cells with no stable DNA integration) and selection
with 6TG (eliminating all cells with a functional Hprt gene), the only ES cells
that would still grow in vitro are those where a homologous recombination event
occurred. In other words, only the cells where the gene actually targeted has been
effectively inactivated, or “knocked-out,” would survive. The second advantage is
that the mutation frequency by homologous recombination is higher than with any
other technique. In the case reported above, for example, one stably transfected ES
cell clone out of 150 was found to be a knockout (Capecchi 1989). This frequency
of recombination events was considered high enough to adapt the technique to all
cases where it was suitable for generating a null allele, even though the sorting out
of the homologous recombinant ES cells from the non-homologous recombinant
cells could not be achieved by the same, in vitro selection as in the case, we just
reported for Hprt- cells.
Since these early experiments, thousands of genes have been inactivated using
the gene-targeting strategy.11 Genes inactivated by homologous recombination in
ES cells are now collectively designated by the name of “knockout” or “knock-
out” (KO). The in vitro engineered DNA molecule used for targeting the homolo-
gous native counterpart in the chromosome of the ES cells is designated the
“recombination vector” Nowadays, in all experiments of this kind, confirmation
that the expected event of homologous recombination actually occurred in the
manipulated ES cells is sought by PCR amplification of critical DNA fragments
with an appropriate set of primers followed by sequencing and confirmation by
Southern blotting. The ES cells in question are then placed into a recipient blasto-
cyst for the production of a chimera. The genetically engineered ES cells, once
confirmed “reliable” and capable of participating in the germ line of the chimeric
mouse, are stored deep-frozen for future use or distribution to the community.
8.3.2.3 Generating a Variety of Knockout Alleles by Homologous

Recombination
Many of the knockout mutations that have been generated in mouse ES cells
over the past several years have resulted from the use of replacement vectors as
described above. In this case, after homologous recombination, the targeted gene
is deleted by one of its specific coding sequences, which is replaced by a heter-
ologous DNA that is, in many cases, a selection cassette. As a consequence of
this substitution, the gene is inactivated and, at the same time, the manipulated
ES cells acquire a selective advantage over a drug and can be positively selected.
Several variations on this basic scenario have been used, and it is impossible to
describe them all in this chapter. However, we can say that most of these strategies
11 For their discoveries of the principles for introducing specific gene modifications in mice by
the use of embryonic stem cells Drs. Mario Capecchi, Martin Evans, and Oliver Smithies were
awarded the Nobel Prize in Medicine or Physiology in 2007.
consist of using a variety of selection cassettes, making use of bacterial genes

encoding either resistance to hygromycin B or puromycin as alternatives to the
neor cassette.
The design of the selection cassettes in the replacement vectors for homologous
recombination depends on the nature of the targeted gene. If the gene in question
is transcriptionally active in the ES cells, then the selection cassette is transcribed
and positive selection with a drug can operate. However, if the gene in question is
not expressed in the ES cells or if its expression pattern is unknown, it is then nec-
essary to design a vector that incorporates a promoter active in ES cells, allowing
the gene to be “switched on” when requested.
Replacement vectors allowing positive/negative selection have also been
designed by inserting a neor mini-gene between two regions of homology, and
inserting a gene encoding herpes simplex virus thymidine kinase (HSVtk) out-
side of the regions of homology. ES cells that are transfected in vitro with such
a replacement vector are then subjected to a double selection: (i) the first one
with G418, inducing the destruction of all ES cells that had not integrated at least
one copy of the vector (and accordingly the neor mini-gene); and (ii) the second
selection with the guanosine analog ganciclovir (GANC, sometimes spelled gan-
cyclovir), killing the cells containing a functional thymidine kinase (tk) gene.
This second level of selection eliminates the ES cells, in which non-homologous
recombination occurred because in this case the HSVtk component of the vector
is generally retained, while it is deleted after homologous recombination. In these
conditions, only the very few cells where homologous recombination occurred can
survive (Figs. 8.6 and 8.7).
The techniques for gene inactivation which we just mentioned have been
described in detail in several review papers and book chapters (Hooper 1992;
Hasty et al. 2000; Babinet and Cohen-Tannoudji 2001; DeChiara 2001; Koentgen
et al. 2010). The use of replacement vectors with a selectable marker has been
and still is very popular for the generation of null alleles because it is relatively
Promoter 1 2 3 4 5
A
2 neo r 4
B
Promoter 1 2 neo r 4 5
C
Fig. 8.6 Gene targeting with a replacement vector 1. Recombination events occurring in the

regions flanking the neor cassette result in the deletion of exon 3 and its replacement by the neor
cassette. The neor cassette confers a selective advantage on the recombined ES cells. The longer
the sequence homology between the replacement vector and the host DNA, the better
Promoter 1 2 3 4 5
A
2 neo r HSVtk
B
Promoter 1 2 neo r 4 5
C
Fig. 8.7 Gene targeting with a replacement vector 2. Gene targeting with a replacement vec-
tor engineered with a positive/negative selection cassette. After homologous recombination, the
HSVtk cassette is deleted while the neor cassette replaces exon 3. This recombination confers to the
recombinant ES cells a selective advantage to G418 and a selective disadvantage to Ganciclovir
straightforward and produces stable, permanent alterations. Unfortunately, cases

have been reported where the selection cassette alters to some extent the expres-
sion of neighboring genes.
8.3.2.4 Generating Point Mutations by Homologous Recombination in

ES Cells
The strategies described above, which make use of replacement vectors, require
the introduction of extrinsic DNA sequences of various sizes into the genome of
ES cells. Although mostly unknown, the consequences of this manipulation may
have some possible adverse effects. This is why scientists have developed an alter-
native strategy, in two steps, leading to the creation of specific base-pair changes
(missense or nonsense) in a specific DNA sequence, allowing the generation of so-
called knock-in (KI) animals.12
The strategy in question is based on two successive steps of homologous recom-
bination, with positive and negative selection, and makes use of mutant Hprt- ES
cells similar to those resulting from the experiments reported above (Hooper et al.
1987; Kuehn et al. 1987) and two replacement vectors. The first replacement vector
is designed to replace an exon of the targeted gene in HPRT-deficient (Hprt-) cells
with a functional Hprt minigene after the first homologous recombination
(Selfridge et al. 1992).13 After this first replacement, the recombinant ES cells are
no longer resistant to the toxic effect of 6-thioguanine (6TG) and can grow
12 The definition of knock-in also applies to the targeted insertion (and substitution) of any cod-
ing sequence at a particular locus of an organism. In these conditions, and in most instances, the
inserted coding sequence is controlled by the regulatory regions of the targeted gene.
13 The HPRT mini-gene is a selection cassette that is unique, since selection may be applied for
its presence or absence.

Promoter 1 2 3 4
A
HPRT
Minigene HSVtk
B
HPRT
Promoter 1 2 Minigene
4
C
2 3’ HSVtk
D
Promoter 1 2 3’ 4
E
Fig. 8.8 Induction of point mutations. Induction of point mutations with two replacement vec-
tors in Hprt- mutant ES cells. The first replacement vector substitutes an Hprt (functional) mini-
gene for exon 3 and confers resistance to HAT (hypoxanthine, aminopterine, thymidine). The
second recombination replaces the Hprt mini-gene by a mutated exon 3 (exon 3′) engineered in
vitro. The ES cell then becomes sensitive to HAT but insensitive to 6-thioguanine. This homolo-
gous recombination is a knock-in (KI) because the original gene is replaced by a modified ver-
sion, even if the gene is merely a mutant allele with only a point mutation
normally in so-called Littlefield’s hypoxanthine, aminopterin, and thymidine

(HAT) culture medium because they have a functional HPRT.14 In these recombi-
nant Hprt+ cells, the targeted gene is deleted by one exon after replacement by the
Hprt mini-gene. A second replacement vector is then designed that corresponds
perfectly to the sequence of the original targeted gene, with the exception of a sin-
gle base pair difference (an SNP) in the targeted exon.15 This vector is synthesized
in vitro, using a PCR technique of directed mutagenesis that is now routine in most
laboratories. After this second replacement, homologously recombined ES cells are
killed in HAT medium but survive selection by 6-thioguanine (6TG), as in the case
of the original Hprt- deficient ES cells (Stacey et al. 1994) (Fig. 8.8).
Finally, and although it has no deleterious effects in the mouse, if the Hprt-
mutation is considered undesirable, it can be easily eliminated by two rounds of
sexual reproduction once the “offspring” of the mutated ES cells are born.
14 Hprt- cells cannot grow in HAT medium because aminopterin blocks the endogenous synthe-
sis of both purines and pyrimidines.
15 Mice of this type are not transgenic animals sensu stricto because they do not have any exog-
enous DNA sequences “stably inserted into their genome.” However, they are still GMOs.
This sophisticated technique of double replacement has potential applications,

among which is the generation of a series of co-isogenic strains of mice (see Chap.
9). This can be achieved by using the same Hprt- ES cells, which are derived from
a highly inbred strain, then by targeting these cells with a variety of second-set
vectors from different inbred strains (replacing the Hprt mini-gene). One can then
generate a variety of different point mutations in the same genetic background (the
same inbred strains).
Alternative techniques using an insertion vector instead of a replacement vector,
and known as the hit-and-run or in-and-out techniques, have been used for the gen-
eration of point mutations in targeted genes (Hasty et al. 1991; Valancius and Smithies
1991). These techniques required two rare intra-chromosomal recombination events to
occur, and accordingly, they appeared to be less efficient than the technique making
use of two replacement vectors. For this reason, they have been abandoned.
8.3.2.5 Knock-Ins Are Sometimes Sophisticated Knockouts
An interesting variation of the technique used for obtaining knockout mutations

has been designed by introducing, via the replacement vector, the coding sequence
of reporter genes in-frame with the promoter of the targeted gene. To give an
example of the high degree of sophistication of this method, we refer to an experi-
ment designed to assess the function of the genes encoding connexins (Filippov
et al. 2003). Connexins are expressed in the various cell types of the central nerv-
ous system and are thought to regulate some of the functional properties exhibited
by immature and mature cells. Understanding the specific role of each connexin in
these processes required an unambiguous characterization of their spatial and tem-
poral pattern of expression. To achieve this aim with connexin 26 (CX26) (gene
symbol Gjb2, for gap junction membrane channel protein beta 2), scientists gen-
erated a reporter allele (Gjb2lacZ) in which the pattern of expression of the gene
encoding β-galactosidase was controlled by the endogenous Gjb2 promoter. Then,
by observing +/Gjb2lacZ heterozygous mice, the researchers could easily identify
the tissues expressing CX26 (i.e., liver, kidney, skin, cochlea, small intestine, pla-
centa, and thyroid gland) and demonstrated that the expression of CX26/Gjb2 is
restricted to the meninges both in embryonic and adult brain. The same research-
ers also noted that homozygous Gjb2lacZ/Gjb2lacZ knockout embryos died early in
utero, indicating that at least one intact copy of the Gjb2 gene is necessary for nor-
mal embryonic development.
Such a mutation, where a gene is inactivated by the insertion of a foreign cod-
ing sequence driven by the same promoter, is also designated as a knock-in.16 The
knock-in strategy is universal and can be applied to any gene to inactivate it and, at
16 In short, the main difference between a knock-out and a knock-in allele is that, in the case of
a knock-in, the gene product is different from the normal allele but still has a function, even if
the function in question is totally unrelated to the function of the original allele. In the case of a
knock-out, the gene has simply been made inoperative.
the same time, visualize its expression pattern in the developing embryo or in the
adult. The knocked-in genes are in general more faithfully expressed than the
transgenes produced by pronuclear injection.
8.3.2.6 Engineering Conditional Knockout Mutations—The Cre-loxP

Strategy
When produced by using one of the techniques described above, knockout muta-
tions affect all the cells of the developing embryo in which the gene is normally
expressed, starting from the early stages of development. For this reason, the
mutations in question are often designated constitutive knockouts. Since most of
the knockout alleles behave as recessives, the situation is in general well toler-
ated as long as the allele stays heterozygous. However, when the knockout allele
is homozygous, the gene is permanently switched off in all cells, and the situa-
tion may become problematic. This is the case, for example, when the knockout
allele results in early embryonic lethality because this hinders the analysis of the
gene function(s) in later developmental stages or in the adult. It is also a drawback
when the inactivation of the targeted gene results in the deregulation or misregula-
tion of the expression of other genes.
To bypass these drawbacks, gene-targeting strategies have been developed that
allow the (knockout) mutations to be made conditional (conditional knockout or
cko mice). With conditional mutations, both the timing of gene inactivation and
the cells or tissues in which the gene is to be “switched off” can be controlled.
The discovery and development of these techniques has been another fundamental
achievement in transgenesis.
The strategies used for the production of conditional knockouts make use of
two transgenic strains: one in which the targeted gene is modified in a way that
ensures its future inactivation and the other where the time- or tissue-specific
expression of the mutation is programmed. Each of the two strains is normal and
fully viable, but when intercrossed, all the ingredients necessary for inactivation
are merged into the genome of their offspring.
The most popular strategy is known as the Cre-loxP strategy and makes use of
Cre recombinase (from cyclization recombinase), a 38 kDa enzyme derived from
the bacteriophage P1 (Utomo et al. 1999; Nagy 2000). Cre recombinase cuts and
recombines the DNA strand at specific sites called loxP sites (short for locus of
X-ing over P1) (Sauer 1993). These loxP sites consist of two 13 bp inverted (pal-
indromic) repeats separated by an 8-bp asymmetric spacer region that defines the
orientation of the site. Such sites do not exist in the mammalian genome (Fig. 8.9).
When the loxP sites are in the same orientation and on the same strand (or chromo-
some), the intervening stretch of DNA is excised as a circular loop. When two loxP
sites are in opposite orientations and on the same chromosome, the intervening
DNA segment is inverted. Finally, when the loxP sites are on two different chromo-
somes, the recombinase generates a reciprocal translocation. When there are more
than two loxP sites in the same genome, a variety of recombinations can occur.
Fig. 8.9 loxP and Frt sites. A loxP site (top) consists of two 13-bp palindromic sequences
(arrowed) flanking an 8-bp spacer region (boxed). These 8-bp define the directionality of the loxP
site. When two loxP sites are placed on the same strand and in the same orientation, the Cre
recombinase deletes the intervening sequence plus one loxP site. When the sites are in oppo-
site orientations, Cre generates an inversion of the intervening sequence and both loxP sites are
retained. When the loxP sites are on different chromosomes, the Cre-recombinase generates a
reciprocal translocation. Nucleotide sequence of the 34-bp-long FRT site (below). The palindro-
mic sequences bind the recombinase, whereas the spacer is the site of DNA break, exchange, and
ligation
To illustrate the basic principle of the method, we will take a historical exam-
ple: the case of T-lymphocyte-specific inactivation of the gene encoding the DNA-
directed βpolymerase (Polb-Chr 8)(Gu et al. 1994). In this experiment, a strain of
mice (strain A) had its Polb gene specifically modified by targeted homologous
recombination with a replacement vector. The replacement vector was designed in
such a way that an essential sequence of the Polb gene, actually the promoter and
the first exon, became flanked by two loxP sites. The replacement vector was also
designed in such a way that it contained two selection cassettes: a neor cassette and
a thymidine kinase (HSVtk) cassette, themselves flanked by a third loxP site as indi-
cated in Fig. 8.10. After homologous recombination, the targeted gene, Polb, ended
up with three loxP sites inserted in the same orientation: the first one upstream of the
promoter and exon 1, a second one in intron 1 upstream of the selection cassettes,
and a third site downstream of the cassettes but upstream of exon 2. As geneticists
say, the gene was then floxed (flanked by loxP sites) but, at this point, it was still
functional and normally transcribed, and the mutation was only cryptic, or “premed-
itated”, so to speak. The neor and HSVtk cassettes were useful for positive/negative
selection with the classical drugs G418 or ganciclovir, should it be necessary.
Concurrently, another strain of mice (strain B) transgenic for a gene encoding
Cre-recombinase was produced by classical pronuclear microinjection. The Cre-
encoding transgene in this case was driven by a lymphocyte creatine kinase (lck)
promoter, which is specific for T cells. When strains A and B were intercrossed,
generating double transgenic (bigenic) mice, the product of the Lck-Cre transgene
triggered deletion of the floxed segment in one or both chromosomes according to
the genetic constitution (heterozygous or homozygous) of strain A, but in T cells
exclusively. The consequences of the mutation (symbolized Polb-)17 on T cells
could then be analyzed because mutant mice were viable, whereas they would
have died if the mutation had been expressed ubiquitously during development.
17 According to the official nomenclature rules, the symbol for this mutation should be
Polbtm1.1Rsky. This was the first targeted mutagenesis at this locus in Rajewsky’s laboratory.
Promoter 1 2
A
Promoter 1 HSVtk neo r

B
Promoter 1 HSVtk neo r 2

C
HSVtk neo r 2
D
Promoter 1 2
D′
2
D′′
Fig. 8.10 Inducing gene-targeted deletions with the Cre-loxP system. In this experiment, the replace-
ment vector (B) was designed in such a way that the Polb targeted region ended up with three loxP sites
inserted in the same orientation: the first one upstream of the promoter and exon 1, a second one in
intron 1, upstream of the selection cassettes, and a third one downstream of the cassettes but upstream
of exon 2 (C). When Cre is synthesized, the segments flanked by two loxP sites (the floxed regions)
are deleted, producing three different types of ES cells (D, D′, D″). The ES cells in which the targeted
gene is deleted (and permanently inactivated—D & D″) are the most interesting. The neor and HSVtk
cassettes were useful for positive/negative selection with the classical drugs G418 and Ganciclovir
Hundreds of experiments of the type described above, leading to tissue- or cell-

specific gene inactivation, have been performed in recent years using either the
Cre-loxP system or a similar system known as FLP-Frt (FLP for Flippase recom-
bination enzyme—Frt for Flippase Recognition target). The FLP-Frt system is
very similar to the Cre-loxP system but makes use of a yeast recombinase with
another specific restriction site.
With these systems, an unlimited number of mutations may be designed, keeping
in mind that Cre (or FLP) deletes any DNA segment once the latter is flanked by two
loxP (or Frt) sites, provided these sites are oriented in the same direction. When there
are more than two loxP sites in the same cell, as is the case in Fig. 8.10, Cre cuts at
each site and, under specific conditions, generates a variety of deletions or transloca-
tions. Selection can then be applied to retain one cell type and not the others if selec-
tion cassettes have been judiciously inserted in critical regions (Gu et al. 1994).18
18 This explains why, with such molecular tools, any kind of chromosomal rearrangement can be
engineered in vitro. In the past, these chromosomal rearrangements were occasionally collected
in the progenies of mice after irradiation in the post-meiotic stages (see Chap. 3).
Promoter STOP Cassette 1 2 3 4

(a)
Promoter
1 2 3 4
(b)
Fig. 8.11 Cre-loxP regulation of transcription. a A floxed “stop” cassette hampers transcrip-

tion of the gene downstream. b When the “stop” sequence is deleted by the action of the Cre-
recombinase, transcription resumes
A similar strategy has been employed using the same strain A (with floxed
Polb) and another strain (strain C) with the interferon-inducible promoter of the
gene Mx1 to regulate Cre expression. After crossing strain A with strain C, Polb
inactivation was induced in adult animals after interferon treatment. In this case,
inactivation was complete in liver, spleen, and bone marrow while it was incom-
plete in other tissues (Kuhn et al. 1995). These experimental results demonstrated
that Cre-mediated recombination could also be effectively induced in nondividing
cells. The expression of the Cre transgene can be made inducible, adding more
sophistication to the system. The tamoxifen-inducible CreERT2, which can be acti-
vated by administration of tamoxifen to the transgenic mice, is very popular (Feil
et al. 2009). Nowadays, many Cre-expressing lines are being produced as knock-in
mice that incorporate the Cre sequence into the gene of interest (instead of creat-
ing transgenic lines using pronuclear microinjection).
The Cre-loxP strategy can also be used to regulate the expression of a spe-
cific protein in a tissue- or cell-specific way using a strategy that is schematically
outlined in Fig. 8.11. In this example, the lacZ gene is a reporter gene driven
by a ubiquitous promoter (e.g., Rosa 26) with a floxed “stop” sequence inserted
between the promoter and the lacZ coding sequence. The “stop” sequence is a
short segment of DNA with several terminator codons that impede translation of
the protein. When the floxed “stop” sequence is deleted by the action of Cre in
some specific cells or tissues, then the lacZ gene is transcribed following the same
pattern of cell/tissue specificity (Lakso et al. 1992; Pichel et al. 1993) (Fig. 8.12).
To add versatility to the method, it must be kept in mind that both the Cre
and FLP recombinases can be used, simultaneously or successively, in the same
experiment.
Since experiments on conditional targeting all entail the use of mouse strains
that synthesize Cre (these strains are designated Cre-deleters), either ubiquitously
or in specific tissues or cell types (strain B or C, in the case of Polb, reported
above), geneticists have agreed to establish a specific database listing all the Cre
strains available (The Cre-X-Mice database at http://nagy.mshri.on.ca/cre_new/
search/Search.php and The Jackson Laboratory Cre Resources at http://www.crep
ortal.org/). These strains are, in general, freely available on the basis of a material
Promoter 1 2 3
A
Promoter 1 2 3
B
+ FLP
Promoter 1 2 3
C + Cre
Promoter 1 3
D + Cre
+ FLP
Promoter 1 3
E
Fig. 8.12 Inducing targeted deletions with the Cre-loxP and FLP-Frt systems. The Cre and FLP
recombinases can be used successively in the same experiment. In the case presented here, when
FLP is used first, the selection cassette (shaded box) is deleted (B → C). Alternatively, if Cre is
used first, exon 2 is deleted (B → D). Finally, when Cre and Frt are used successively, the selec-
tion cassette and exon 2 are both deleted (B → E)
transfer agreement (MTA). This attitude, which is more and more common in the
community of mouse geneticists, has saved and still saves a lot of research money.
It has been made simpler every day with the use of the internet.
8.3.2.7 Gene Trapping and Targeted Trapping in ES Cells
In an earlier section of this chapter (Sect. 8.3.2.1), we reported experiments

in which retroviruses were successfully used for producing insertional muta-
tions in ES cells. These experiments revealed that, unlike the cloned DNA mol-
ecules injected into the pronucleus, retroviruses integrate into the genome of ES
cells without generating extensive chromosomal rearrangements. In these con-
ditions, when mutations were induced, the proviral copy could be used as a tag
for cloning the flanking sequences and finally for identifying the mutated genes.
Unfortunately, other than these advantages, the retroviruses have two major draw-
backs: first, they insert randomly in the genome and infrequently in exons; second,
using a proviral insertion for “harpooning” the flanking sequences is sometimes
misleading, especially when there are many proviral insertions in the same ES
cells.
In order to improve the efficiency of recovering mutations that are likely to
have a phenotypic expression, an original strategy known as gene trapping was
developed in several laboratories (Gossler et al. 1989; Friedrich and Soriano
1991; Skarnes et al. 1992; Evans et al. 1997; Cecconi and Meyer 2000; Stanford
Promoter 1 2 3 4
A
A′ Normal protein
Promoter 1 βgeo 2 3 4
C
C′ Fusion protein (inactive)
Splicing acceptor site

Promoterless βgeo
Polyadenylation signal
Fig. 8.13 Gene trapping. When a promoterless synthetic reporter gene, such as βgeo, sand-
wiched between a splice acceptor site and a polyadenylation signal (B) inserts, by chance, into
one of the introns of an expressed gene (A → C), the reporter gene is transcribed as if it were
an exon of the gene. This generates a fusion mRNA, which is (sometimes) translated into a
non-functional fusion protein C′ (the trapped gene is inactivated). (This figure is redrawn from
Skarnes et al. 1992)
et al. 2001; Hansen et al. 2003; Stryke et al. 2003). The principle of this strat-
egy consisted of transfecting ES cells with a promoterless reporter gene and/or a
selectable genetic marker flanked, upstream, by a 3′ splice acceptor (SA) site, and
downstream by a polyadenylation signal (pA) (Fig. 8.13).
In early experiments, a popular promoterless gene was engineered by fusion
of a β-galactosidase moiety (acting as a reporter) with a neomycin-resistant moi-
ety (acting as a selectable marker) and was designated βgeo (contraction of β-gal
with neo). When such a cassette was inserted in an intron, the gene was said to be
“trapped.” Nowadays, a variety of promoterless artificial genes have been designed
with different reporter sequences, making the method more efficient and more
versatile.
Transcription of the trapped genes, controlled by the endogenous promoter,
resulted in a fusion (or hybrid) RNA molecule, which in turn, was translated into a
non-functional protein with some sequence of the endogenous trapped gene beside
some others from the sequence of the reporter/selectable marker.19 Since the
encoded fusion protein was non-functional, the trapped genes were equivalent to
19 Trapping cassettes have also been designed with a marker gene or a selectable gene coupled to
a suitable promoter but lacking a downstream polyadenylation signal. In this case, the transcript
was also a hybrid molecule, utilizing the 3′ sequences of the host gene to acquire a poly (A) tail.
knockout (or loss-of-function) alleles and the sequence of the cassette could then
be used as a tag for gene identification.
Although the strategy of gene-trapping works exclusively with those genes that
are transcribed in ES cells, it is nevertheless a high-throughput approach for the
identification of genes. It has been (and still is) widely used. Several laboratories,
working in an International Gene-Trap Consortium (IGTC), have undertaken the
establishment of large libraries of ES cells harboring gene-trap insertions. From
recent estimates, over 126,500 ES cell lines, each with a trapped gene, are offered
to the community on a non-collaborative basis.20 This represents ~13,300 trapped
genes (i.e., around 50 % of all the known genes in the mouse).
In the laboratories performing this type of experiment, the trapped genes are
systematically identified unambiguously by using a PCR-based strategy such as
5’RACE (rapid amplification of cDNA ends), to generate a sequence tag unique
for each insertion. By the way, this is greatly facilitated by the availability of the
mouse genome sequence. Researchers who are interested can search and browse
the IGTC database (www.genetrap.org) looking for the ES cell lines they are inter-
ested in, using accession numbers or IDs, keywords, sequence data, tissue expres-
sion profiles, or biological pathways.
As we already mentioned, newer gene-trap vectors have been developed, offer-
ing a variety of possibilities for post-insertional modification and the generation of
a wide spectrum of alleles.
The trapped-gene libraries that exist nowadays have become an indispensa-
ble source of ready-made mutations in mice. For those readers who would like to
know more about these libraries, the way they were established and their potential
interest we recommend three general publications co-authored by scientists who
were deeply involved in their development (Guan et al. 2010; Skarnes et al. 2011;
Bradley et al. 2012). The Web site of the International Knockout mouse consor-
tium http://www.knockoutmouse.org/ is also an important source of information,
which is user-friendly and explains all the technical steps in the gene-trapping
strategy.
As we explained above, gene trapping depends on the random insertion of a
reporter cassette in an intron, but the cassette in question can also be inserted in
a predefined position by homologous recombination. This strategy is known by
the generic name of targeted trapping (Friedel et al. 2005). In this case, the vec-
tor (basically the same as the one used for gene trapping) is flanked by genomic
sequences of the host, completely excluding the promoter. Targeted trapping in
mouse ES cells is a simple though powerful tool for analysis of mammalian gene
function. Provided the promoterless construct is carefully designed, most random
insertions are eliminated by drug selection and the targeting frequencies can reach
50 % or even more (Fig. 8.14).
20 With, however, some handling fees.

Promoter 1 2 3 4
A
A′ Normal protein
Promoter 1 βgeo 2 3 4
B
B′ Fusion protein (inactive)
Splicing acceptor site

Promoterless βgeo
Polyadenylation signal
Fig. 8.14 Targeted trapping. In this case, insertion of the promoterless reporter gene βgeo is not
random, as in the case of gene trapping, but instead results from homologous recombination with
a selected region of the targeted gene (A → B). As in the case of gene trapping, the promoterless
gene in the cassette is activated and possibly translated into a fusion protein (B′). In this experi-
ment, it is important that the targeted region does not contain the promoter of the gene. After
characterization, the targeted or trapped ES cell clones can be deep-frozen and stored for further
use. (This figure is redrawn from Skarnes et al. 1992)
8.3.3 Induction of Mutations in ES Cells with Chemical

Mutagens
In Chap. 7, we explained that the induction of mutations in the mouse germ line
with radiation or chemical mutagens was an efficient method for the annotation of
mammalian genes because it produced all kinds of mutations (nonsense, missense,
etc.) and all kinds of alleles (recessive and dominant etc.)—unlike most tech-
niques of ES cell engineering, which produce mostly knockouts (i.e., null alleles).
However, a major drawback of chemical mutagenesis is the cost of breeding and/
or the time necessary to identify and characterize the new mutations. In addition,
all these induced mutations are scattered throughout the whole genome, they are
a mixture of different kinds, and they do not necessarily match the interest of the
scientist. The genotype-based screens, which consisted of the identification, after
analysis performed at the DNA level, of mice heterozygous for a mutation induced
by ENU in a specific gene (as described in Chap. 7—Sect. 7.5.4), were consid-
ered more advantageous, especially when a deep-frozen sperm bank was available.
Unfortunately, here again, this may still be insufficient if a series of alleles at a
given locus is desired.
A genotype-based screen for ENU-induced mutations has been adapted with

success for the identification of mutations induced in ES cells, in specific genes of
interest (Chen et al. 2000; Munroe et al. 2000). In a series of experiments focused
on two loci of importance for mouse early development, Smad2 and Smad4,
Vivian and colleagues (Vivian et al. 2002) mutagenized 2,060 ES cell clones by
incubating the cells for 2 h in a culture medium with 0.2 mg/ml of ENU. They
found a total of 29 mutations, out of which 20 were non-silent (yielding a non-
silent mutation rate of 1 per 673 kb of screened DNA). This indicates that chemi-
cal mutagenesis in mouse ES cells, associated with high-throughput mutation
detection, is another interesting method for the identification of mutations in non-
selectable genes.
Other experiments on chemical mutagenesis of mouse ES cells have also
been suggested as an alternative approach to the chemical mutagenesis of sper-
matogonia (Becker et al. 2006; Munroe and Schimenti 2009). This strategy has
(at least) two major advantages: first, it enables the use of a variety of chemicals
with different mutational spectra (different from ENU); second, it allows (at least
in theory) the induction of a higher number of mutations in the mouse genome
as a consequence of the possibility of performing several successive rounds of
mutagenesis in vitro. In addition to these advantages in terms of efficiency, the
chemical mutagenesis of ES cells has the same advantages as the gene-driven
strategy described in Chap. 7 that it requires only two generations of breeding to
reveal the phenotype of the induced mutations (breeding G1s, then intercrossing
the G1). In addition, just like for the sperm cells in the case of gene-driven strat-
egy, samples of successfully treated ES cells can be stored deep-frozen as long as
necessary for the further detection of induced mutations. This method has not been
used very much, probably because the techniques of genetic engineering were
developed concurrently, but their advantages, as outlined above, are unique and
should be kept in mind.
8.4 Inducible Transgenesis: The Tet-off and Tet-on

Expression Systems
The Cre-loxP and the FLP-Frt strategies allow the induction of conditional gene
knockout. With these strategies, researchers can inactivate virtually any gene,
in any specific tissue or cell lineage, and when desired. However, once the Cre-
recombinase has excised a floxed DNA segment, the situation is irreversible: the
gene is permanently inactivated (or activated) in all daughter cells. Obviously, this
may represent a drawback in experiments where only a transient inactivation (or
activation) would be desired. It also may be desirable, in some experiments with
transgenic mice, to have a transgene expressed only during a certain period but
switched off the rest of the time. Unfortunately, this is not possible with the tech-
niques described above.
The Tet-off and Tet-on inducible expression systems overcome these problems,
placing the transcription of a given transgene under the control of the researcher.
8.4 Inducible Transgenesis: The Tet-off and Tet-on Expression Systems 303
In this system, the expression of a transgene is dependent on a tetracycline-con-

trolled transactivator protein and can be regulated, both reversibly and quantita-
tively, by exposing the transgenic mice to the antibiotic tetracycline (Tc) or to one
of its derivatives such as doxycycline (Dox). The technology was developed by
Bujard and colleagues at the University of Heidelberg (Gossen and Bujard 1992;
Baron and Bujard 2000).
The Tet-off system requires two critical ingredients. The first is the tetracycline-
controlled transactivator protein (in short tTA). tTA is an artificial protein created
(a) Tet-off system

Promoter TetR VP16
z
tTA + Dox
tTA
tet07 CMV tet07 CMV

Target transgene
(b)
Tet-on system
Promoter rTetR VP16
z
rtTA
rtTA + Dox
tet07 CMV tet07 CMV

Target transgene
Fig. 8.15 The “Tet-off” and “Tet-on” Expression Systems. The Tet-off and Tet-on inducible
expression systems enable transgene expression to be dependent on a tetracycline-controlled
transactivator protein (tTA). Under these conditions, transgenic expression can be regulated. a
The Tet-off system requires two ingredients. The first is the tTA, which is a fusion protein created
with the TetR (tetracycline repressor), found in Escherichia coli transposon Tn10 and encoding
resistance to the antibiotic tetracyclin, and a strong trans-activating domain of an herpes simplex
virus protein called VP16. The second ingredient is the tetracycline-responsive promoter element
(TRE) that is composed of a concatemer of seven tet operators (tetO7) fused to the minimal pro-
moter sequences of the human cytomegalovirus immediate early gene 1 (hCMVIE1) promoter/
enhancer. In the absence of tetracyclin (Tc) or doxycyclin (Dox), tTA binds to TRE and acti-
vates expression of the targeted gene. This induction returns to basal levels or is suppressed upon
administration of Tc or Dox. The Tet-on system works in exactly the opposite manner. This sys-
tem is based on a reverse tetracycline-controlled trans-activator (rtTA), which is also a fusion
protein composed of the TetR and the VP16 transactivation domain. However, a four amino acid
change in the TetR DNA-binding moiety alters rtTA’s activity binding characteristics in such a
way that it can recognize the tetO sequences in the TRE of the target transgene only in the pres-
ence of the Dox effector (delivered in the water or the food). Thus, in the Tet-on system, tran-
scription of the TRE-regulated target is stimulated by rtTA only in the presence of Dox. b As
explained in the text, both systems require the generation of double transgenic (or bigenic) mice
carrying, in the same genome, the TRE-regulated target transgene and the tetracycline-controlled
transactivator (tTA or rtTA)
by fusion of the TetR (tetracycline repressor), found in Escherichia coli transposon

Tn10 and encoding resistance to Tc, with a strong transactivating domain of the
herpes simplex virus protein called VP16. The second critical ingredient required
for the Tet-off system to operate is the Tc-responsive promoter element (TRE).
This promoter is composed of a concatamer of seven tet operators (tetO7) fused to
the minimal promoter sequences of the human cytomegalovirus immediate early
gene 1 (hCMVIE1) promoter/enhancer. In the absence of Tc or Dox, tTA binds to
TRE and activates expression of the target gene. This induction returns to basal
levels or is suppressed upon administration of Tc or Dox (Fig. 8.15).
The Tet-on system works in exactly the opposite manner. It is based on a reverse
tetracycline-controlled transactivator (rtTA), which is also a fusion protein com-
posed of TetR and the VP16 transactivation domain; however, a four-amino-acid
change in the TetR DNA-binding moiety alters rtTA’s activity binding characteris-
tics such that it can recognize the tetO sequences in the TRE of the target transgene
only in the presence of the Dox effector. Thus, in the Tet-on system, transcription
of the TRE-regulated target is stimulated by rtTA only in the presence of Dox (i.e.,
when the drug is delivered either in the drinking water or with the food). A good
example of the value of this system for cancer research is a model where an acti-
vated Kras oncogene is inducibly expressed in an epithelial compartment using
keratin 5 (K5)-rtTA: tet-Kras bigenic mice (Vitale-Cross et al. 2004).
These Tet-off and Tet-on systems can be used, for example, to design dominant
gain-of-function experiments in which temporal control of transgene expression is
required (Gossen and Bujard 1992; Furth et al. 1994; Kistner et al. 1996; Schonig
and Bujard 2003). The Tet-off expression system is more popular than the Tet-on
system because it does not require the constant administration of a drug whose
effects might be deleterious in the longterm.
8.5 Other Techniques for the Production of Transgenic

Mice
Considering the efficiency of ENU mutagenesis and the potentialities of genetic

engineering applied to ES cells, it is clear that mouse geneticists have at their dis-
position an unmatched arsenal of strategies allowing them to generate virtually
any type of alteration in the genome of their favorite species. This is unfortunately
not the case with other species of mammals, especially the rat, which is yet
another important source of model for human diseases21. However, techniques
have been developed to generate genomic alterations in these species and some
have proved very promising. Most of these techniques have been efficiently and
successfully transposed to the mouse species. We will describe the most important.
21 Some domestic species (the rat in particular), present phenotypes that have not yet been docu-
mented in the mouse; this is why it would be important that the genetic arsenal that has been
developed for the mouse be replicated in these other species.
8.5 Other Techniques for the Production of Transgenic Mice 305
8.5.1 Transgenesis by Retroviral Infection of Early Embryos
The integration of exogenous DNA into the germ line through experimental infec-
tion of mouse embryos with retroviruses was successfully achieved a long time
ago (Jaenisch 1976). Newborns and preimplantation embryos (4–8 cell stage) were
infected with the Moloney murine leukemia virus (M-MuLV), and it was observed
that infection of preimplantation embryos, in contrast to infection of newborns,
could lead to stable integration of proviral copies into the germline. These initial
experiments have yielded several mouse strains with stable germ line integra-
tions of retroviral DNA at distinct chromosomal loci (for example, the Mov loci;
Jaenisch 1976). One of these integrations was in the gene encoding procollagen,
type I, alpha 1 (Col1a1Mov13) (Stacey et al. 1988).
Experimental infections of preimplantation embryos have the advantage that
the viral integrations are in general stable and do not generate the sort of chro-
mosomal rearrangements that often occur with the classical pronuclear techniques.
Since these integrations occur almost at random, they sometimes hit a gene (as in
the case of Col1a1) and produce a visible mutant phenotype. Here again, the DNA
of the retrovirus can be used as a “hook” to clone the DNA sequences flanking the
insertion site, and this helps in the characterization of the mutant allele.
Viral infection can also be used to introduce foreign DNA into embryos or eukar-
yotic cells in culture, and the advantages of using mouse retroviruses as shuttles for
transgenesis have been explained in detail in a review by Nicolas and Rubenstein
(1988). Two of these advantages are noteworthy in the context of this chapter:
• All the sequences of the viral genome required for its replication, transcription,
and integration are grouped in or adjacent to the long terminal repeat (LTR).
• All the necessary proteins for infection, reverse transcription, and integration of
the viral genome can be removed from the “shuttle” virus and provided in trans
by a “helper” virus, leaving space for foreign DNA inserts of up to 8–10 kb.
For transgenesis in rodents (mostly in rat), the lentiviruses derived from human
HIV have been the most widely employed (Wiznerowicz and Trono 2005). The
reason for this choice is that lentiviruses, unlike most other retroviruses, have the
capacity to infect nondividing cells. Shuttle viruses are produced by transfection
of the construct into packaging cell lines, which are engineered to provide the
essential viral proteins for assembly of infectious particles. The viruses are har-
vested from the cell culture medium and used for microinjection into the perivi-
telline space of single-cell embryos (Koentgen et al. 2010). Infected embryos
reverse-transcribed the lentiviral RNA into DNA (provirus) that inserts back into
the genome. However, because they are defective, the viruses are capable of com-
pleting only a single infectious cycle but cannot replicate further.
Lentiviral integrations, in addition to being relatively stable and because they are
less invasive than pronuclear injections, sometimes yield survival rates approach-
ing 90 %. Another advantage is that lentiviruses integrate as single copies and are
expressed more reliably than the transgenes obtained by pronuclear injections; in
particular, they are less prone to epigenetic silencing (Koentgen et al. 2010). The
major weakness of this technique is the limit of 8–10 kb for the transgene size.
8.5.2 In Vivo Genome Editing: The Production of Targeted

Alterations Using Engineered Nucleases
Over the last 10 years, a totally new kind of technique has been developed for the
production of gene- (or locus-) targeted mutations that make use of engineered
hybrid molecules which associate sequence-specific DNA-binding domains with
a non-specific DNA cleavage domain. These techniques have demonstrated sig-
nificant advantages for the production of a variety of mutations at targeted sites
in several species commonly used by geneticists, including Arabidopsis thaliana,
Caenorhabditis elegans, the sea urchin Echinus melo, Drosophila melanogater,
and Danio rerio, to cite just a few. Recently, the techniques in question have been
successfully adapted to the production of targeted mutations (knockout and knock-
in) in mammals, mainly in the rat (Geurts et al. 2009), the mouse (Carbery et al.
2010), and other domestic species (reviewed in Rémy et al. 2010; Gaj et al. 2013;
Kim and Kim 2014; and Mashimo 2014). We will describe some of these tech-
niques and discuss their possible applications for genome editing.
8.5.2.1 Zinc-Finger Nucleases and Transcription Activator-like

Effector Nucleases
The molecules used in initial experiments associated zinc-finger DNA-binding

motifs with the restriction endonuclease FokI, and for this reason, they were called
zinc-finger nucleases (ZFNs). For the production of mutations, two complemen-
tary ZFNs must be designed, each of them recognizing a specific DNA sequence
spanning 9–18 bp on either side of a 5–6-bp sequence defining the targeted region.
When injected into a cell or a pronucleus, the ZFNs assemble tightly on both sides
of the targeted site, one on each strand, and FokI performs double-strand breaks
(DSBs).22 Once cleaved by the endonuclease, the cellular mechanisms controlling
DNA integrity are immediately triggered to repair the damage. These mechanisms
are of two types. The first is known as the homology-dependent repair (HDR)
mechanism, which requires a homologous (template) sequence to guide the repair:
it is precise and accurate and re-establishes ad integrum the original sequence of
the cleaved DNA strand. The second mechanism, known as non-homologous end
joining (NHEJ), is more common but is more rapidly activated. NHE is much less
precise and only approximately restores the damaged strands, leaving behind dele-
tions of nucleotides and accordingly frame shift mutations that are in most
instances loss-of-function mutations (Fig. 8.16).
22 A specific ZFN binds with 3 bp at the DNA level. Since there is a great variety of such motifs,
a judicious selection of 3–6 of them allows the targeting of a 9–18-bp DNA domain, which is
highly specific. Libraries of ready-made ZFNs are also available which allow the targeting of vir-
tually any sequence in the mouse genome.
(a) (b) (c)

Zinc finger nucleases (ZFNs) TALE nucleases (TALENs) CRISPR/Cas9
20-bp target sequence
Zinc finger binding motifs Transcription Activator-Like Effector FokI nuclease PAM
G TTTCCAGGATTATGTAATAG T
G
FokI nuclease TGTCA
T G
GTCCCT
ACAGT G UUUCCAGGAUUACGUAAUAG AGGGA
A G
5’-TAGCCATCAGGTTTTATGTGATGGAACACCTGAGGGACCACTATTACGTA-3’ C
C AAAGGTCCTAATACATTATCA
G
G UAAGGCUAGUCCGUUAUCAACUUG
3’-ATCGGTAGTCCAAAATACACTACCTTGTGGACTCCCTGGTGATAATGCAT-5’ U A
U A A
A
Cas9 U A
U A
A U
UUUU GUGAA A
FokI nuclease G UG
A AGuide
C G
G C
U A
G CA G C
Zinc finger binding motifs FokI nuclease Transcription Activator-Like Effector C G
U A G C
C G
A U
GAAA RNA UGA
Double strand break (DSB)
non-homologous end joining homology directed repair

(NHEJ) (HDR)
(e) (d)
Knockout (KO) Knockin (KI)
Fig. 8.16 The production of targeted genome alterations using site-specific engineered nucle-

ases or the CRISPR/Cas9-based RNA-guided DNA endonuclease. The figure schematically rep-
resents three strategies used for genome editing. Zinc fingers (a) and TALEN modules (b) both
bind to adjacent DNA sequences in opposite directions, leaving a small gap in between for the
FokI endonuclease to perform a double-stranded break (DSB). c With the CRISPR strategy, Cas9
unwinds the DNA duplex and performs a DSB after recognition of a specific (~20 bp) target by
the gRNA, provided that the correct protospacer adjacent motif (PAM) is present. Whatever their
origins, DSBs are ligated through non-homologous end joining (NHEJ) (d) or repaired through
homology-directed repair (HDR) (e). For HDR to occur, a DNA molecule or a single-stranded
synthetic DNA must be added as a template. If the sequence of the template differs from the
endogenous sequence by the addition or substitution of some nucleotides, this results in a knock-
in. These methods for producing mutations at specifically targeted sites are very efficient. The
CRISPR/Cas9-based RNA-guided strategy permits the production of several independent point
mutations in the same genome (Courtesy T. Mashimo)
The technique is simple in its practical aspects. Messenger RNAs transcribed in

vitro from engineered ZFN plasmids are injected into the (male) pronuclei of
mouse zygotes, exactly as in the case of pronuclear (in ovo) transgenesis, then the
embryos are transferred into the oviduct of pseudo-pregnant females. With this
technique a homozygous knockout mutation can be obtained in 4–5 months, which
is much faster than with the traditional knockout strategies using ES cells.23
Another important advantage of the technique is that it is applicable to all strains
of mice, allowing for the production of a series of mutations at the same locus in
different inbred backgrounds (co-isogenic strains) (Carbery et al. 2010). Finally,
the technique can produce a variety of mutations, mainly deletions ranging from
1 bp to more than 1 kb, and more rarely, insertions of a few bp, but also sequence-
specific mutations—which are all potential tools for the analysis of the targeted
23 This comment concerning the time necessary to produce a knockout mutation in the mouse
genome by using the ZFN strategy, although reduced, must nevertheless be compared with the
time necessary to purchase, when available, an ES cell line harboring the same ready-made
knockout, when the latter is available in a repository such as KOMP (https://www.komp.org/).
gene’s function. Knock-in mice and rats carrying sequence-specific modifications

have already been produced using ZFN technology (Cui et al. 2011).
The expression of artificial nucleases in embryonic cells at early stages of
development does not seem to be toxic or to have any breakage activity outside
of the targeted DNA sequence (little or no off-target events). The only drawback
of the technique, which is not a major one, is that alterations may still occur at the
targeted site several days after the injection, making some founder animals behave
like mosaics. Another potential drawback is that, for technical reasons, the tech-
nique is probably not applicable to gene families, since the design of sequence-
specific domains of the ZFNs would be difficult or even impossible in this case.
This basic technique of genome modification making use of ZFNs has under-
gone several improvements and developments. The first one is based on the obser-
vation, already mentioned above, that when DSBs are induced in cells by any
means (for example, as a consequence of irradiation or of nuclease activity—and
regardless of the nuclease) the homology-dependent repair mechanism (HDR) is
activated. These mechanisms increase the potentialities of insertion of exogenous
DNA that have sequence homologies at their ends with the sequence flanking
the DSB. For example, adding the cloned RNA of the lacZ reporter gene to the
mRNAs injected into the pronucleus allowed the production of knock-in trans-
genic mice with lacZ integrated between the boundaries of the DSB.
Another improvement has been the replacement of the DNA-binding com-
ponents of the ZFNs by molecules from the plant bacterium Xanthomonas with
similar DNA-binding properties. These molecules are a family of transcrip-
tion activator-like effectors (TALEs) and the DNA-binding hybrid molecules are
known as TALENs. TALEs have binding capacities greater than ZFNs and can
match with virtually any sequence, further increasing the efficiency of the tech-
nique (Tesson et al. 2011). Over recent years, several groups have used TALENs to
modify endogenous genes in a wide variety of species including insects, amphib-
ians, fish (zebrafish), and mammals (rat, mouse, pig, and cow) (Joung and Sander
2013; Sung et al. 2013). The advantages of TALENs over ZFNs include their ease
of design and assembly, their specificity, and their lower cost. Injection of the exo-
nuclease Exo1 in substitution for endonuclease FokI in the TALEN technique has
been another improvement in the production of knockouts in rats (Mashimo et al.
2013).
8.5.2.2 The CRISPR/Cas9 System
The strategies that we described in the section above consisted of the production
of double-strand breaks (DSBs) by the protein-guided DNA cleavage activity of
engineered ZFNs or TALENs. Recently, another technique has been developed
that depends on small RNAs for the production of sequence-specific cleavages
(RNA-guided DNA cleavage). This strategy was developed after the identification
and characterization of a defense mechanism, known as the CRISPR/Cas system,
which operates in bacteria and archaea and allows these organisms to fight infec-
tions by viruses, plasmids, or phages (Pennisi 2013).24
A CRISPR locus consists of a series of short direct repeats (average size 32 bp)
of identical sequences, interspersed with intervening regions called spacers, which
consist of small but variable sequences. Analysis of the sequence of these spacers
indicates great similarities with the sequences of some phages and plasmids, pro-
viding a possible interpretation for the mechanism of recognition of the genome of
the invaders by the CRISPR.
The CRISPR loci are transcribed into short CRISPR RNAs (crRNA). These
crRNAs anneal to transactivating crRNAs (tracrRNAs) and direct sequence-
specific cleavage of DNA by Cas proteins. Target recognition by the CRISPR-
associated nuclease (Cas9) protein requires a seed sequence within the crRNA and
a conserved dinucleotide-containing protospacer adjacent motif (PAM) sequence
upstream of the crRNA-binding region (Fig. 8.16).
Engineered modifications of the CRISPR, as well as the Cas9 part, have led to
an efficient way of producing DSBs at will. The CRISPR component is usually
referred to as a guide RNA (gRNA). Cas9 utilizes gRNA that binds to specific
DNA sequences to produce the DSBs.
The Cas9 protein consists of three more or less independent domains: one DNA-
binding domain and two catalytic domains that independently cut one DNA strand.
The two domains with nuclease activity can be inactivated separately by simple
point mutations, and these modified versions of Cas9, with one cutting domain disa-
bled, introduce single-strand breaks or DNA nicks. Even though DNA nicking is less
efficient for genome editing, it dramatically reduces the chance of so-called off-tar-
get effects, since unwanted nicks are faithfully reconstructed by homology-directed
repair (HDR). DSBs can be achieved at the targeted site by a pair of DNA-binding
gRNAs, with sites close to each other but on opposite strands.
The RNA-guided endonucleases can be engineered to cleave virtually any DNA
sequence by appropriately designing the crRNA; for example, to generate knock-
in animals carrying conditional or reporter alleles (Yang et al. 2013). This tech-
nique exhibits several advantages over the methods using ZFNs or TALENs. One
can, for example, generate mice carrying mutations in multiple genes across the
genome in a single step by simultaneously injecting various gRNAs (Horii et al.
2014). This technique is known as multiplex gene editing and has been applied
successfully not only to cells cultured in vitro but also to mouse and rat embryos
(Wang et al. 2013; Wei et al. 2013). It saves a lot of breeding time when an experi-
mental project requires the presence of several mutations in the same genome.
The genomic alterations that can be produced by using the CRISPR/Cas9 technol-
ogy are not limited to the production of indels but can also consist of knock-ins. If we
consider that the strategy is relatively easy to apply and somewhat faster than the other
strategies using engineered nucleases, we see that CRISPR/Cas9 may well-revolution-
ize genomic engineering in the near future (Mashimo 2014; Zhang et al. 2014).
24 CRISPR is an acronym for clusters of regularly interspaced short palindromic repeats.
8.6 Conclusion
Contemplating all the many possibilities for creating transgenic mice, one can see
that geneticists now have all the tools in hand to answer virtually any questions
that may arise in their analysis of gene functions. They also have at their disposi-
tion a very large collection of ready-made mutations of all kinds, waiting to be
used, for example, as models of human diseases.25 All these tools and models will
be important for performing genome annotation.
References
Abiola O, Angel JM, Avner P, Bachmanov AA, Belknap JK, Bennett B, Blankenhorn EP, Blizard
DA, Bolivar V, Brockmann GA, Buck KJ, Bureau JF, Casley WL, Chesler EJ, Cheverud
JM, Churchill GA, Cook M, Crabbe JC, Crusio WE, Darvasi A, de Haan G, Dermant P,
Doerge RW, Elliot RW, Farber CR, Flaherty L, Flint J, Gershenfeld H, Gibson JP, Gu J,
Gu W, Himmelbauer H, Hitzemann R, Hsu HC, Hunter K, Iraqi FF, Jansen RC, Johnson
TE, Jones BC, Kempermann G, Lammert F, Lu L, Manly KF, Matthews DB, Medrano JF,
Mehrabian M, Mittlemann G, Mock BA, Mogil JS, Montagutelli X, Morahan G, Mountz
JD, Nagase H, Nowakowski RS, O’Hara BF, Osadchuk AV, Paigen B, Palmer AA, Peirce
JL, Pomp D, Rosemann M, Rosen GD, Schalkwyk LC, Seltzer Z, Settle S, Shimomura K,
Shou S, Sikela JM, Siracusa LD, Spearow JL, Teuscher C, Threadgill DW, Toth LA, Toye
AA, Vadasz C, Van Zant G, Wakeland E, Williams RW, Zhang HG, Zou F; Complex Trait
Consortium. (2003) The nature and identification of quantitative trait loci: a community’s
view. Nature Review Genetics 4:911–916
Adams JM, Harris AW, Pinkert CA, Corcoran LM, Alexander WS, Cory S, Palmiter RD, Brinster
RL (1985) The c-myc oncogene driven by immunoglobulin enhancers induces lymphoid
malignancy in transgenic mice. Nature 318:533–538
Babinet C, Cohen-Tannoudji M (2001) Genome engineering via homologous recombination in
mouse embryonic stem (ES) cells: an amazingly versatile tool for the study of mammalian
biology. Anais da Academia Brasileira de Ciencias 73:365–383
Ballester M, Castelló Anna, Ibáñez E, Sánchez A, Folch JM (2004) Real-time quantitative PCR-
based system for determining transgene copy number in transgenic animals. BioTechniques
37:610–613
Baron U, Bujard H (2000) Tet repressor-based system for regulated gene expression in eukary-
otic cells: principles and advances. Methods Enzymol 327:401–421
Becker S, de Angelis MH, Beckers J (2006) Use of chemical mutagenesis in mouse embryonic
stem cells. Methods Mol Biol 329:397–407
Bradley A, Anastassiadis K, Ayadi A, Battey JF, Bell C, Birling MC, Bottomley J, Brown SD,
Bürger A, Bult CJ, Bushell W, Collins FS, Desaintes C, Doe B, Economides A, Eppig JT,
Finnell RH, Fletcher C, Fray M, Frendewey D, Friedel RH, Grosveld FG, Hansen J, Hérault
Y, Hicks G, Hörlein A, Houghton R, Hrabé de Angelis M, Huylebroeck D, Iyer V, de Jong
PJ, Kadin JA, Kaloff C, Kennedy K, Koutsourakis M, Lloyd KC, Marschall S, Mason J,
McKerlie C, McLeod MP, von Melchner H, Moore M, Mujica AO, Nagy A, Nefedov M,
Nutter LM, Pavlovic G, Peterson JL, Pollock J, Ramirez-Solis R, Rancourt DE, Raspa M,
25 To paraphrase the title of an interesting review on the subject one could say that, nowadays,
geneticists have at their disposition “a mouse for all reasons” (International Mouse Knockout
Consortium 2007).
References 311
Remacle JE, Ringwald M, Rosen B, Rosenthal N, Rossant J, Ruiz Noppinger P, Ryder E,

Schick JZ, Schnütgen F, Schofield P, Seisenberger C, Selloum M, Simpson EM, Skarnes
WC, Smedley D, Stanford WL, Stewart AF, Stone K, Swan K, Tadepally H, Teboul L,
Tocchini-Valentini GP, Valenzuela D, West AP, Yamamura K, Yoshinaga Y, Wurst W (2012)
The mammalian gene function resource: the International Knockout Mouse Consortium.
Mamm Genome 23:580–586
Breitman ML, Clapoff S, Rossant J, Tsui LC, Glode LM, Maxwell IH, Bernstein A (1987)
Genetic ablation: targeted expression of a toxin gene causes microphthalmia in transgenic
mice. Science 238:1563–1565
Breitman ML, Bryce DM, Giddens E, Clapoff S, Goring D, Tsui LC, Klintworth GK, Bernstein
A (1989) Analysis of lens cell fate and eye morphogenesis in transgenic mice ablated for
cells of the lens lineage. Development 106:457–463
Brinster RL, Chen HY, Trumbauer M, Senear AW, Warren R, Palmiter RD (1981) Somatic
expression of herpes thymidine kinase in mice following injection of a fusion gene into
eggs. Cell 27:223–231
Brinster RL, Allen JM, Behringer RR, Gelinas RE, Palmiter RD (1988) Introns increase tran-
scriptional efficiency in transgenic mice. Proc Natl Acad Sci USA 85:836–840
Carbery ID, Ji D, Harrington A, Brown V, Weinstein EJ, Liaw L, Cui X (2010) Targeted genome
modification in mice using zinc-finger nucleases. Genetics 186:451–459
Capecchi MR (1989) Altering the genome by homologous recombination. Science
244:1288–1292
Cecconi F, Meyer BI (2000) Gene trap: a way to identify novel genes and unravel their biological
function. FEBS Lett 480:63–71
Chen Y, Yee D, Dains K, Chatterjee A, Cavalcoli J, Schneider E, Om J, Woychik RP, Magnuson T
(2000) Genotype-based screen for ENU-induced mutations in mouse embryonic stem cells.
Nat Genet 24:314–317
Costantini F, Lacy E (1981) Introduction of a rabbit beta-globin gene into the mouse germline.
Nature 294:92–94
Cui X, Ji D, Fisher DA, Wu Y, Briner DM, Weinstein EJ (2011) Targeted integration in rat and
mouse embryos with zinc-finger nucleases. Nat Biotechnol 29:64–67
DeChiara TM (2001) Gene targeting in ES cells. Methods Mol Biol 158:19–45
Dorner M, Horwitz JA, Robbins JB, Barry WT, Feng Q, Mu K, Jones CT, Schoggins JW,
Catanese MT, Burton DR, Law M, Rice CM, Ploss A (2011) A genetically humanized
mouse model for hepatitis C virus infection. Nature 474:208–211
Duboule D (1998) Vertebrate Hox gene regulation: clustering and/or colinearity? Curr Opin Gene
Develop 8:514–518
Evans MJ, Kaufman MH (1981) Establishment in culture of pluripotential cells from mouse
embryos. Nature 292:154–156
Evans MJ, Carlton MB, Russ AP (1997) Gene trapping and functional genomics. Trends Genet
13:370–374
Feil S, Valtcheva N, Feil N (2009) Inducible Cre mice methods. Mol Biol 530:343–363
Filippov MA, Hormuzdi SG, Fuchs EC, Monyer H, Robertson E, Bradley A, Kuehn M, Evans M
(2003) A reporter allele for investigating connexin 26 gene expression in the mouse brain.
Eur J Neurosci 18:3183–3192
Friedel RH, Plump A, Lu X, Spilker K, Jolicoeur C, Wong K, Venkatesh TR, Yaron A, Hynes M,
Chen B, Okada A, McConnell SK, Rayburn H, Tessier-Lavigne M (2005) Gene targeting
using a promoterless gene trap vector (“targeted trapping”) is an efficient method to mutate
a large fraction of genes. Proc Natl Acad Sci U S A 102:13188–13193
Friedrich G, Soriano P (1991) Promoter traps in embryonic stem cells: a genetic screen to iden-
tify and mutate developmental genes in mice. Genes Dev 5:1513–1523
Furth PA, St Onge L, Böger H, Gruss P, Gossen M, Kistner A, Bujard H, Hennighausen L (1994)
Temporal control of gene expression in transgenic mice by a tetracycline-responsive pro-
moter. Proc Natl Acad Sci USA 91:9302–9306
Gaj T, Gersbach CA, Barbas CF III (2013) ZFN, TALEN, and CRISPR/Cas-based methods for
genome engineering. Trends Biotechnol 31:397–405
Geurts AM, Cost GJ, Freyvert Y, Zeitler B, Miller JC, Choi VM, Jenkins SS, Wood A, Cui X,
Meng X, Vincent A, Lam S, Michalkiewicz M, Schilling R, Foeckler J, Kalloway S, Weiler
H, Ménoret S, Anegon I, Davis GD, Zhang L, Rebar EJ, Gregory PD, Urnov FD, Jacob HJ,
Buelow R (2009) Knockout rats via embryo microinjection of zinc-finger nucleases. Science
325:433
Gomes-Pereira M, Cooper TA, Gourdon G (2011) Myotonic dystrophy mouse models: towards
rational therapy development. Trends Mol Med 17:506–517
Gordon JW, Ruddle FH (1981) Integration and stable germline transmission of genes injected
into mouse pronuclei. Science 214:1244–1246
Goring DR, Rossant J. Clapoff S, Breitman ML, Tsui LC (1987) In situ detection of beta-
galactosidase in lenses of transgenic mice with a gamma-crystallin/lacZ gene. Science
235:456–458
Gossen M, Bujard H (1992) Tight control of gene expression in mammalian cells by tetracycline-
responsive promoters. Proc Natl Acad Sci USA 89:5547–5551
Gossler A, Doetschman T, Korn R, Serfling E, Kemler R (1986) Transgenesis by means of blas-
tocystderivedembryonic stem cell lines. Proc Natl Acad Sci USA 83:9065–9069
Gossler A, Joyner AL, Rossant J, Skarnes WC (1989) Mouse embryonic stem cells and reporter
constructs to detect developmentally regulated genes. Science 244:463–465
Gridley T, Gray DA, Orr-Weaver T, Soriano P, Barton DE, Francke U, Jaenisch R (1990)
Molecular analysis of the Mov 34 mutation: transcript disrupted by proviral integration in
mice is conserved in Drosophila. Development 109:235–242
Gu H, Marth JD, Orban PC, Mossmann H, Rajewsky K (1994) Deletion of a DNA polymerase
betagene segment in T cells using cell type-specific gene targeting. Science 265:103–106
Guan C, Ye C, Yang X, Gao J (2010) A review of current large-scale mouse knockout efforts.
Genesis 48:73–85
Hammes A, Schedl A (2000) Generation of transgenic mice from plasmids, BACs and YACs. In:
Jackson IJ, Abbott CM (eds) Mouse genetics and transgenesis: a practical approach. Oxford
University Press, New York, pp 217–245
Hanahan D, Wagner EF, Palmiter RD (2007) The origins of oncomice: a history of the first trans-
genic mice genetically engineered to develop cancer. Genes Dev 21:2258–2270
Hansen J, Floss T, Van Sloun P, Fuchtbauer EM, Vauti F, Arnold HH, Schnutgen F, Wurst W, von
Melchner H, Ruiz P (2003) A large-scale, gene-driven mutagenesis approach for the func-
tional analysis of the mouse genome. Proc Natl Acad Sc USA 100:9918–9922
Harbers K, Jahner D, Jaenisch R (1981) Microinjection of cloned retroviral genomes into mouse
zygotes: integration and expression in the animal. Nature 293:540–542
Hasty P, Ramirez-Solis R, Krumlauf R, Bradley A (1991) Introduction of a subtle mutation into
theHox-2.6 locus in embryonic stem cells. Nature 350:243–246
Hasty P, Abuin A, Bradley A (2000) Gene targeting, principles, and practice in mammalian cells.
In: Joyner AL (ed) Gene targeting: a practical approach. Oxford University Press, New
York, pp 1–36
He Y, Chen H, Quon MJ, Reitman M (1995) The mouse obese gene. Genomic organization,
promoter activity, and activation by CCAAT/enhancer-binding protein alpha. J Biol Chem
270:28887–28891
Heintz N (2001) BAC to the future: the use of BAC transgenic mice for neuroscience research.
Nat Rev Neurosci 2:861–870
Herault Y, Duchon A, Velot E, Maréchal D, Brault V (2012) The in vivo Down syndrome
genomic library in mouse. Prog Brain Res 197:169–197
Hitz C, Steuber-Buchberger P, Delic S, Wurst W, Kühn R (2009) Generation of shRNA trans-
genic mice. Methods Mol Biol 530:101–129
Hogan B, Beddington R, Costantini F, Lacy E (1994) Manipulating the mouse embryo: a labora-
tory manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor
References 313
Hooper ML (1992) Embryonal stem cells. Harwood Academic, Chur 147 pp

Hooper M, Hardy K, Handyside A, Hunter S, Monk M (1987) HPRT-deficient (Lesch-Nyhan)
mouse embryos derived from germline colonization by cultured cells. Nature 326:292–295
Horii T, Arai Y, Yamazaki M, Morita S, Kimura M, Itoh M, Abe Y, Hatada I (2014) Validation
of microinjection methods for generating knockout mice by CRISPR/Cas-mediated genome
engineering. Sci Rep 4:4513. doi:10.1038/srep04513
Houdebine LM (2003) Animal transgenesis and cloning. Wiley, New York
Huxley C, Passage E, Manson A, Putzu G, Figarella-Branger D, Pellissier JF, Fontes M (1996)
Construction of a mouse model of Charcot-Marie-Tooth disease type 1A by pronuclear
injection of human YAC DNA. Hum Mol Genet 5:563–569
International Mouse Knockout Consortium, Collins FS, Rossant J, Wurst W (2007) A mouse for
all reasons. Cell 128:9–13
Jackson IJ, Abbott CM (2000) Mouse genetics and transgenesis: a practical approach. Oxford
University Press, Oxford
Jaenisch R (1976) Germ line integration and Mendelian transmission of the exogenous Moloney
leukemia virus. Proc Natl Acad Sci USA 73:1260–1264
Jaenisch R (1988) Transgenic animals. Science 240:1468–1474
Jaenisch R, Jahner D, Nobis P, Simon I, Lohler J, Harbers K, Grotkopp D (1981) Chromosomal
position and activation of retroviral genomes inserted into the germline of mice. Cell
24:519–529
Jakobovits A, Moore AL, Green LL, Vergara GJ, Maynard-Currie CE, Austin HA, Klapholz S
(1993) Germline transmission and expression of a human-derived yeast artificial chromo-
some. Nature 362:255–258
Joung JK, Sander JD (2013) TALENs: a widely applicable technology for targeted genome edit-
ing. Nat Rev Mol Cell Biol 14:49–55
Katsuki M, Sato M, Kimura M, Yokoyama M, Kobayashi K, Nomura T (1988) Conversion of
normal behavior to shiverer by myelin basic protein antisense cDNA in transgenic mice.
Science 241:593–595
Kim H, Kim JS (2014) A guide to genome engineering with programmable nucleases. Nat Rev
Genet 15:321–334
Kistner A, Gossen M, Zimmermann F, Jerecic J, Ullmer C, Lübbert H, Bujard H (1996)
Doxycycline-mediated quantitative and tissue-specific control of gene expression in trans-
genic mice. Proc Natl Acad Sci USA 93:10933–10938
Koentgen F, Suess G, Naf D (2010) Engineering the mouse genome to model human disease for
drug discovery. Methods Mol Biol 602:55–77
Koike S, Taya C, Kurata T, Abe S, Ise I, Yonekawa H, Nomoto A (1991) Transgenic mice suscep-
tible to poliovirus. Proc Natl Acad Sci USA 88:951–955
Kuehn MR, Bradley A, Robertson EJ, Evans MJ (1987) A potential animal model for Lesch–
Nyhan syndrome through introduction of HPRT mutations into mice. Nature 326:295–298
Kuhn R, Schwenk F, Aguet M, Rajewsky K (1995) Inducible gene targeting in mice. Science
269:1427–1429
Lakso M, Sauer B, Mosinger B Jr, Lee EJ, Manning RW, Yu SH, Mulder KL, Westphal H (1992)
Targeted oncogene activation by site-specific recombination in transgenic mice. Proc Natl
Acad Sci USA 89:6232–6236
Lecuit M, Vandormael-Pournin S, Lefort J, Huerre M, Gounon P, Dupuy C, Babinet C, Cossart P
(2001) A transgenic model for listeriosis: role of internalin in crossing the intestinal barrier.
Science 292:1722–1725
Lee JT, Jaenisch R (1996) A method for high efficiency YAC lipofection into murine embryonic
stem cells. Nucleic Acids Res 24:5054–5055
Lichtman JW, Livet J, Sanes JR (2008) A technicolour approach to the connectome. Nat Rev
Neurosci 9:417–422
Lira SA, Kinloch RA, Mortillo S, Wassarman PM (1990) An upstream region of the mouse ZP3
gene directs expression of firefly luciferase specifically to growing oocytes in transgenic
mice. Proc Natl Acad Sci USA 87:7215–7219
Lithner CU, Hedberg MM, Nordberg A (2011) Transgenic mice as a model for Alzheimer’s dis-
ease. Curr Alzheimer Res 8:818–831
Lowe LA, Yamada S, Kuehn MR (2001) Genetic dissection of nodal function in patterning the
mouse embryo. Development 128:1831–1843
Martin GR, Stevens ME, Bissada N, Nasir J, Kanazawa I, Disteche CM, Rubin EM, Hayden MR
(1981) Isolation of a pluripotent cell line from early mouse embryos cultured in medium
conditioned by teratocarcinoma stem cells. Proc Natl Acad Sci USA 78:7634–7638
Martin N, Jaubert J, Gounon P, Salido E, Haase G, Szatanik M, Guénet JL (2002) A missense
mutation in Tbce causes progressive motor neuronopathy in mice. Nat Genet 32:443–447
Mashimo T (2014) Gene targeting technologies in rats: zinc finger nucleases, transcription acti-
vator-like effector nucleases, and clustered regularly interspaced short palindromic repeats.
Dev Growth Differ 56:46–52
Mashimo T, Kaneko T, Sakuma T, Kobayashi J, Kunihiro Y, Voigt B, Yamamoto T, Serikawa T
(2013) Efficient gene targeting by TAL effector nucleases coinjected with exonucleases in
zygotes. Sci Rep 3:1253. doi:10.1038/srep01253
Meisler MH (1992) Insertional mutation of ‘classical’ and novel genes in transgenic mice. Trends
Genet 8:341–344
Messing A, Behringer RR, Slapak JR, Lemke G, Palmiter RD, Brinster RL (1990) Insertional
mutation at the ld locus (again!) in a line of transgenic mice. Mouse Genome 87:107
Misteli T, Spector D (1997) Applications of the green fluorescent protein in cell biology and bio-
technology. Nat Biotechnol 15:961–964
Munroe RJ, Bergstrom RA, Zheng QY, Libby B, Smith R, John SW, Schimenti KJ, Browning
VL, Schimenti JC (2000) Mouse mutants from chemically mutagenized embryonic stem
cells. Nat Genet 24:318–321
Munroe RJ, Schimenti JC (2009) Mutagenesis of mouse embryonic stem cells with ethylmeth-
anesulfonate. Methods Mol Biol 530:131–138
Nagy A (2000) Cre recombinase: the universal reagent for genome tailoring. Genesis 26:99–109
Nagy A, Gertsenstein M, Vintersten K, Behringer R (2003) Manipulating the mouse embryo, a
laboratory manual, 3rd edn. Cold Spring Harbor Press, New York
Nicolas JF, Rubenstein JL (1988) Retroviral vectors. Biotechnology 10:493–513
O’Doherty A, Ruf S, Mulligan C, Hildreth V, Errington ML, Cooke S, Sesay A, Modino S, Vanes
L, Hernandez D, Linehan JM, Sharpe PT, Brandner S, Bliss TV, Henderson DJ, Nizetic D,
Tybulewicz VL, Fisher EM (2005) An aneuploid mouse strain carrying human chromosome
21 with Down syndrome phenotypes. Science 309:2033–2037
Overbeek PA, Chepelinsky AB, Khillan JS, Piatigorsky J, Westphal H (1985) Lens-specific
expression and developmental regulation of the bacterial chloramphenicol acetyltransferase
gene driven by the murine alpha A-crystallin promoter in transgenic mice. Proc Natl Acad
Sci USA 82:7815–7819
Overbeek PA, Gorlov IP, Sutherland RW, Houston JB, Harrison WR, Boettger-Tong HL, Bishop
CE, Agoulnik AI (2001) A transgenic insertion causing cryptorchidism in mice. Genesis
30:26–35
Palmiter RD, Brinster RL, Hammer RE, Trumbauer ME, Rosenfeld MG, Birnberg NC, Evans
RM (1982) Dramatic growth of mice that develop from eggs microinjected with metal-
lothionein–growth hormone fusion genes. Nature 300:611–615
Papaioannou VE, McBurney M, Gardner RL, Evans MJ (1975) The fate of teratocarcinoma cells
injected into early mouse embryos. Nature 258:70–73
Passamaneck YJ, Di Gregorio A, Papaioannou VE, Hadjantonakis AK (2006) Live imaging
of fluorescent proteins in chordate embryos: from ascidians to mice. Microsc Res Tech
69:160–167
Pennisi E (2013) The CRISPR craze. Science 341:833–836
Pereira R, Khillan JS, Helminen HJ, Hume EL, Prockop DJ (1993) Transgenic mice expressing
a partially deleted gene for type I procollagen (COL1A1). A breeding line with a phenotype
of spontaneous fractures and decreased bone collagen and mineral. J Clin Invest 91:709–716
References 315
Pichel JG, Lakso M, Westphal H (1993) Timing of SV40 oncogene activation by site-specific
recombination determines subsequent tumor progression during murine lens development.
Oncogene 8:3333–3342
Li, P, Tong, C, Mehrian-Shai R, Jia L, Wu N, Yan Y, Maxson RE, Schulze EN, Song H, Hsieh
C-L, Pera MF, Ying Q-L (2008) Germline competent embryonic stem cells derived from rat
blastocysts. Cell 135:1299–1310
Rémy S, Tesson L, Ménoret S, Usal C, Scharenberg AM, Anegon I (2010) Zinc-finger nucleases:
a powerful tool for genetic engineering of animals. Transgenic Res 19:363–371
Robertson E, Bradley A, Kuehn M, Evans M (1986) Germline transmission of genes introduced
into cultured pluripotential cells by retroviral vector. Nature 323:445–448
Rossant J, Nutter LM, Gertsenstein M (2011) Engineering the embryo. Proc Natl Acad Sci USA
108:7659–7660
Rueda N, Flórez J, Martínez-Cué C (2013) Apoptosis in Down’s syndrome: lessons from studies
of human and mouse models. Apoptosis 18:121–134
Ryan TM, Townes TM, Reilly MP, Asakura T, Palmiter RD, Brinster RL, Behringer RR (1990)
Human sickle hemoglobin in transgenic mice. Science 247:566–568
Sauer B (1993) Manipulation of transgenes by site-specific recombination: use of Cre recombi-
nase. Methods Enzymol 225:890–900
Schedl A, Beermann F, Thies E, Montoliu L, Kelsey G, Schutz G (1992) Transgenic mice
generated by pronuclear injection of a yeast artificial chromosome. Nucleic Acids Res
20:3073–3077
Schedl A, Larin Z, Montoliu L, Thies E, Kelsey G, Lehrach H, Schutz G (1993) A method for
the generation of YAC transgenic mice by pronuclear microinjection. Nucleic Acids Res
21:4783–4787
Schonig K, Bujard H (2003) Generating conditional mouse mutants via tetracycline-controlled
gene expression. In: Hofker M, van Deursen J (eds) Transgenic mouse methods and proto-
cols. Humana Press, Totowa, pp 69–104
Selfridge J, Pow AM, McWhir J, Magin TM, Melton DW (1992) Gene targeting using a mouse
HPRT minigene/HPRT-deficient embryonic stem cell system: inactivation of the mouse
ERCC-1 gene. Somat Cell Mol Genet 18:325–336
Simon-Chazottes D, Tutois S, Kuehn M, Evans M, Bourgade F, Cook S, Davisson MT, Guénet
JL (2006) Mutations in the gene encoding the low-density lipoprotein receptor LRP4 cause
abnormal limb development in the mouse. Genomics 87:673–677
Simon-Chazottes D, Frenkiel MP, Montagutelli X, Guénet JL, Desprès P, Panthier JJ (2011)
Transgenic expression of full-length 2′, 5′-oligoadenylate synthetase 1b confers to BALB/c
mice resistance against West Nile virus-induced encephalitis. Virology 417:147–153
Scherbik SV, Kluetzman K, Perelygin AA, Brinton MA (2007) Knock-in of the Oas1b(r) allele
into a flavivirus-induced disease susceptible mouse generates the resistant phenotype.
Virology 368:232–237
Skarnes WC, Auerbach BA, Joyner AL (1992) A gene trap approach in mouse embryonic stem
cells: the lacZ reported is activated by splicing, reflects endogenous gene expression, and is
mutagenic in mice. Genes Dev 6:903–918
Skarnes WC, Rosen B, West AP, Koutsourakis M, Bushell W, Iyer V, Mujica AO, Thomas M,
Harrow J, Cox T, Jackson D, Severin J, Biggs P, Fu J, Nefedov M, de Jong PJ, Stewart
AF, Bradley A (2011) A conditional knockout resource for the genome-wide study of mouse
gene function. Nature 474:337–342
Smith DJ, Zhu Y, Zhang J, Cheng JF, Rubin EM (1995) Construction of a panel of transgenic
mice containing a contiguous 2-Mb set of YAC/P1 clones from human chromosome
21q22.2. Genomics 27:425–434
Smithies O, Gregg RG, Boggs SS, Koralewski MA, Kucherlapati RS (1985) Insertion of DNA
sequences into the human chromosomal betaglobin locus by homologous recombination.
Nature 317:230–234
Stacey A, Bateman J, Choi T, Mascara T, Cole W, Jaenisch R (1988) Perinatal lethal osteogenesis
imperfecta in transgenic mice bearing an engineered mutant pro-alpha-1(I) collagen gene.
Nature 332:131–136
Stacey A, Schnieke A, McWhir J, Cooper J, Colman A, Melton DW (1994) Use of double-
replacement gene targeting to replace the murine alpha-lactalbumin gene with its human
counterpart in embryonic stem cells and mice. Mol Cell Biol 14:1009–1016
Stanford WL, Cohn JB, Cordes SP (2001) Gene-trap mutagenesis: past, present and beyond. Nat
Rev Gene 2:756–768
Stevens LC (1960) Embryonic potency of embryoid bodies derived from a transplantable testicu-
lar teratoma of the mouse. Dev Biol 2:285–297
Stryke D, Kawamoto M, Huang CC, Johns SJ, King LA, Harper CA, Meng EC, Lee RE,
Yee A, L’Italien L, Chuang PT, Young SG, Skarnes WC, Babbitt PC, Ferrin TE (2003)
BayGenomics: a resource of insertional mutations in mouse embryonic stem cells. Nucleic
Acids Res 31:278–281
Sung YH, Baek IJ, Kim DH, Jeon J, Lee J, Lee K, Jeong D, Kim JS, Lee HW (2013) Knockout
mice created by TALEN-mediated gene targeting. Nat Biotechnol 31:23–24
Tesson L, Usal C, Ménoret S, Leung E, Niles BJ, Remy S, Santiago Y, Vincent AI, Meng X,
Zhang L, Gregory PD, Anegon I, Cost GJ (2011) Knockout rats generated by embryo
microinjection of TALENs. Nat Biotechnol 29:695–696
Thomas KR, Capecchi MR (1987) Site-directed mutagenesis by gene targeting in mouse embryo
derived stem cells. Cell 51:503–512
Utomo AR, Nikitin AY, Lee WH (1999) Temporal, spatial, and cell type-specific control of Cre-
mediated DNA recombination in transgenic mice. Nat Biotechnol 17:1091–1096
Valancius V, Smithies O (1991) Testing an ‘in-out’ targeting procedure for making subtle
genomic modifications in mouse embryonic stem cells. Mol Cell Biol 11:1402–1408
Van Keuren ML, Gavrilina GB, Filipiak WE, Zeidler MG, Saunders TL (2009) Generating trans-
genic mice from bacterial artificial chromosomes: transgenesis efficiency, integration and
expression outcomes. Transgenic Res 18:769–785
Vitale-Cross L, Amornphimoltham P, Fisher G, Molinolo AA, Gutkind JS (2004) Conditional
expression of K-ras in an epithelial compartment that includes the stem cells is sufficient to
promote squamous cell carcinogenesis. Cancer Res 64:8804–8807
Vivian JL, Chen Y, Yee D, Schneider E, Magnuson T (2002) An allelic series of mutations in
Smad2 and Smad4 identified in a genotype-based screen of N-ethyl-N-nitrosourea-
mutagenized mouse embryonic stem cells. Proc Natl Acad Sci USA 99:15542–15547
Wagner EF, Stewart TA, Mintz B (1981a) The human beta-globin gene and a functional viral thy-
midine kinase gene in developing mice. Proc Natl Acad Sci USA 78:5016–5020
Wagner TE, Hoppe PC, Jollick JD, Scholl DR, Hodinka RL, Gault JB (1981b) Microinjection of
a rabbit beta-globin gene into zygotes and its subsequent expression in adult mice and their
offspring. Proc Natl Acad Sci USA 78:6376–6380
Wahnschaffe U, Bitsch A, Kielhorn J, Mangelsdorf I (2005a) Mutagenicity testing with trans-
genic mice. Part I: Comparison with the mouse bone marrow micronucleus test. J Carcinog
4:3
Wahnschaffe U, Bitsch A, Kielhorn J, Mangelsdorf I (2005b) Mutagenicity testing with trans-
genic mice. Part II: Comparison with the mouse spot test. J Carcinog 4:4
Wang H, Yang H, Shivalila CS, Dawlaty MM, Cheng AW, Zhang F, Jaenisch R (2013) One-step
generation of mice carrying mutations in multiple genes by CRISPR/Cas-mediated genome
engineering. Cell 153:910–918
Wei C, Liu J, Yu Z, Zhang B, Gao G, Jiao R (2013) TALEN or Cas9—rapid, efficient and spe-
cific choices for genome modifications. J Genet Genomics 40:281–289
Wiznerowicz M, Trono D (2005) Harnessing HIV for therapy, basic research and biotechnology.
Trends Biotechnol 23:42–47
Wong EA, Capecchi MR (1986) Analysis of homologous recombination in cultured mamma-
lian cells in transient expression and stable transformation assays. Somatic Cell Mol Gene
12:63–72
References 317
Woychik RP, Stewart TA, Davis LG, D’Eustachio P, Leder P (1985) An inherited limb deformity
created by insertional mutagenesis in a transgenic mouse. Nature 318:36–40
Yang H, Wang H, Shivalila CS, Cheng AW, Shi L, Jaenisch R (2013) One-step generation of
mice carrying reporter and conditional alleles by CRISPR/Cas-mediated genome engineer-
ing. Cell 154:1370–1379
Yu T, Li Z, Jia Z, Clapcote SJ, Liu C, Li S, Asrar S, Pao A, Chen R, Fan N, Carattini-Rivera S,
Bechard AR, Spring S, Henkelman RM, Stoica G, Matsui S, Nowak NJ, Roder JC, Chen C,
Bradley A, Yu YE (2010) A mouse model of Down syndrome trisomic for all human chro-
mosome 21 syntenic regions. Hum Mol Genet 19:2780–2791
Zhang F, Wen Y, Guo X (2014) CRISPR/Cas9 for genome editing: progress, implications and
challenges. Hum Mol Genet. Apr 7. [Epub ahead of print]
Zhang Y, Proença R, Maffei M, Barone M, Leopold L, Friedman JM (1994) Positional cloning of
the mouse obese gene and its human homologue. Nature 372:425–432
Chapter 9
The Different Categories of Genetically
Standardized Populations of Laboratory
Mice
9.1 Introduction
At the beginning of the twentieth century when genetics began to emerge as an

experimental science, laboratory mouse resources were extremely limited. Apart
from a few coat color mutants, which were bred by fanciers as pets, the only ani-
mals available for experiments were “albino” mice. These mice were bred with
no specific mating protocol and were, in most cases, genetically heterogeneous.
At that time, and based on the experience of dog and horse breeders, inbreeding
was a practice to be avoided by all possible means because it was thought to lead
to a decline in vigor and ultimately to the extinction of the colony. In fact, it is
no exaggeration to say that the only qualities that were required of these “albino”
mice were prolificacy, robustness, and tameness.
Regardless of the conditions under which they were bred, these mice (as well
as the albino rats) were suited for most of the experiments that were undertaken
at that time. For example, they could be used for performing experimental infec-
tions with a variety of pathogens, for the evaluation of the biological effects of
drugs or for experiments in physiology, nutrition, etc. However, the transplantation
of tumor cells, which was undertaken to study the process of cancer origin and
progression, resulted in a high percentage of rejection, yielding unreliable results.
Miss Maude Slye, for example, working at the University of Chicago, completed
an extensive study concerning cancer inheritance in mice, but her contribution was
nearly forgotten because she could not repeat her observations with “albino mice”
from another supplier (Strong 1978).
At the same time, researchers began to develop a number of colonies with
unique characteristics. At the origin of these colonies were a handful of mice
selected for some interesting phenotypic traits. They were bred as closed popu-
lations, with no contribution from external breeders, to avoid losing the charac-
teristics in question. Almost simultaneously, and from an increasing number of
experiments involving tumor transplantations, it became progressively clear that
the rate of success was higher when grafts were performed among mice of the
same “family” than when grafts were performed among mice of unrelated origins.

320 9 The Different Categories of Genetically Standardized …
In other words, instead of being disadvantageous, consanguinity appeared to be

advantageous in the case of tissue/tumor transplantations. From this observation,
as well as from a few others, it was then considered worth developing true inbred
strains by systematically mating brothers to their sisters, generation after genera-
tion; the concept of inbred strains was born.
C.C. Little was the first to attempt the development of “pure” mouse lines by
inbreeding (simultaneously, Helen D. King undertook the same sort of experiment
with rats at the Wistar Institute). The first mouse inbred strain, dba, was started
in 1909 by inbreeding mice homozygous for three recessive coat color alleles of
independent origin (d: dilute, now Myo5ad-Chr 9; b: brown, now Tyrp1b-Chr 4;
and a: non-agouti-Chr 2). Similarly, strain C57BL/6 was established in 1921, also
by C. C. Little, from a cross between two “black” mice, female 57 and male 52,
obtained from Miss Abbie Lathrop, a mouse supplier from Massachusetts. A few
other strains were developed at about the same time by other scientists, in particu-
lar L.C. Strong in Cold Spring Harbor and N. Dobrovolskaia Zavadskaia in Paris.
Readers who are interested in the history of mouse genetics are invited to read
the excellent book edited by Morse III (1978), which contains several chapters
written by mouse geneticists of the early days, including L.C. Strong himself. In
his chapter, Strong reveals the recipes he used for the successful development of
his lines, and it is interesting to note that, a century later, most of these recom-
mendations are still relevant. Some points from Strong’s contribution are amusing
anecdotes; for example, when he explains that he captured wild mice in a pigeon
coop close to his lab, for “sorting out their hereditary traits” and was obliged,
for practical reasons, to breed these mice “under his bed in his honeymoon resi-
dence!” These wild mice were also used by Strong to introduce some “vigor” in
one of his stocks after an outbreak of “paratyphoid” (salmonellosis). Finally, and
even though this is not clearly stated, it is more than likely that the wild-type allele
at the agouti (A) locus, which is now homozygous in strains CBA and C3H (two
of Strong’s strains), is inherited from the wild mice trapped in the pigeon coop at
Cold Spring Harbor on Long Island! Another interesting book on the early years
of genetically standardized mice is Making Mice by Karen Rader (2004).
In addition to the contribution of North American researchers, which has
clearly been fundamental for the development of most inbred strains of mice, one
must also mention the contribution of Japanese scientists who established a num-
ber of colonies from fancy mice. Details of this contribution are reported in a book
by Moriwaki et al. (1994).
Nowadays, a great variety of mouse strains are used routinely and many experi-
ments involving laboratory mice would not be possible if these lines had not been
patiently and carefully developed one century ago. In this matter, one must note
that the mouse is, by far, the mammalian species with the largest variety of geneti-
cally defined strains, while the rat comes second, but far behind. In this chapter we
will describe the genetic structure of the different types of genetically standardized
strains and their optimal use in experimental genetics.
9.2 Inbred Strains 321
9.2 Inbred Strains
According to Hans Grüneberg, the “introduction of inbred strains into biology is

probably comparable in importance with that of the analytical balance into chem-
istry” (Grüneberg 1952). This statement may have been considered somewhat per-
emptory 60 years ago but it does not appear exaggerated in the present context,
as it has been validated by numerous examples. Indeed, and as we will discuss,
inbred strains can be regarded as a “basic ingredient” in most experimental pro-
jects in mammalian genetics.
9.2.1 Inbred Mice are Isogenic and Homozygous

at All Loci
According to the definition by the International Committee on Standardized

Nomenclature for mice, “Strains can be termed inbred if they have been mated
brother × sister for 20 or more consecutive generations, and individuals of the
strain can be traced to a single ancestral pair at the 20th or subsequent genera-
tion”. At this point the individuals’ genomes will, on average, have only 1 % resid-
ual heterozygosity and can be regarded for most purposes as genetically identical
(Fig. 9.1).
In practice, most of the mouse strains that are commonly used in research labo-
ratories nowadays have undergone several tens of generations of brother × sister
matings (indicated with an “F”, for filial), and some of the most ancient (DBA/2
for example) may have passed 200 generations.
The definition of an inbred strain calls for a few comments. As stated, mice
of the same inbred strain are genetically identical except, of course, for the sex-
linked characteristics. One must note in particular that, because of strict inbreed-
ing, all the mice of a given strain have become homozygous at all loci that were
segregating in the founder ancestors (the original or ancestral breeding pair) and
they all have become homozygous for the same allele (meaning that the mater-
nal and paternal chromosomes are identical). This is also known as autozygosity
because the two alleles are copies of the same ancestral allele. To describe this
important characteristic, geneticists say that the mice in question are isogenic; in
other words: they are genetically identical.
The process leading to homozygosity by progressive allele loss (or fixation) is
easy to understand if we consider that, when by chance an allele that was present at
generation Fn is not transmitted to at least one member of the breeding pair at gen-
eration Fn+1, it is then permanently lost. In other words, as inbreeding progresses,
alleles are constantly lost but none are ever introduced, with the exception of rare
de novo mutations. This, obviously, leads to both homozygosity and isogenicity!
F M
F1
F2
F3
F4
F 15…
Fig. 9.1 Inbred Strains. This drawing represents schematically the breeding protocol commonly
used to produce an inbred strain: mating a male and a female from the same litter (brother × sis-
ter) in successive generations. Theoretical calculation would indicate that parent × offspring
exceptional matings (F4 in the example) would not affect the progression toward homozygosity
provided that the parent selected for mating is the youngest of the pair. The uppercase letter F
followed by a number represents the number of inbreeding generations. When this number is
not known, a question mark is used: F? + 27, for example, would indicate that the number of
brother × sister matings was not known when the strain was imported, but 27 generations of unre-
laxed inbreeding have been added since this time. F13 + F28 indicates that 13 generations of strict
inbreeding have been achieved in a breeding laboratory and an additional 28 in another laboratory
The categorization of the alleles that are lost or retained at each generation
depends on chance for a large part, and if the inbreeding protocol could be reset
with the same founder animals, it would lead to a strain with a different genetic
constitution after the same 20 generations. This means that an inbred strain repre-
sents a unique and fortuitous assortment of alleles.
100
80
% Heterozygosity
60
40
20
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Generations
Fig. 9.2 Effects of inbreeding. The curve was drawn based on the ratios 1/1, 2/2, 3/4, 5/8, 8/16,
13/32, 21/64, and so on. In these ratios, the denominator doubles at each generation while the
numerator is given by the Fibonacci sequence; each number being the sum of the two preceding
numbers. This recursion relationship represents relatively accurately the decreasing percentage
of genes that are still in the heterozygous state as inbreeding progresses. From generation F5
onwards, this percentage corresponds to ~19.6 % at each generation
To get a fairly accurate idea of what the genetic makeup of the individuals of an
inbred strain actually looks like, one could imagine a totally virtual and theoretical
experiment where the male pronucleus is removed immediately after fertilization,
before it merges with the female pronucleus, while the remaining female pronu-
cleus is duplicated, for example, after a short treatment with the alkaloid cytocha-
lasin D, to become the diploid nucleus of a one-cell stage embryo. This totally
artificial embryo would be a female, with the two chromosomes of each pair abso-
lutely identical: this is precisely what the genome of all members of an inbred
strain look like with the exception, of course, of the sex chromosomes.1
During the process of inbreeding, the progression toward homozygosity is relatively
fast during the first few generations, where a great number of genes become homozy-
gous, then it slows down and after 20 generations no more than 1–2 % of the loci that
were heterozygous in the ancestors are still segregating. A mathematical series, based
on Fibonacci’s numbers, is traditionally used to model the decrease in heterozygosity
as the number of sib-matings increases. Even though this curve is only an approxima-
tion, it represents fairly accurately the evolution of heterozygosity over time (Fig. 9.2).
When explaining the progression toward full homozygosity during inbreeding,
we often consider the genome as a little bag full of genes, themselves considered
1 In Chap. 6 we explained that, as a consequence of epigenetic modifications at the genome

level, such a uniparental mouse could not exist in practice.
as independent entities. In reality, one must keep in mind that the genes are linked
and arranged linearly on the chromosomes, and the evolution towards homozy-
gosity involves blocks or “chunks” of chromosomes of variable sizes rather than
individual genes. This explains why independent inbred strains carrying the same
allele at a given locus have a great chance of sharing the same short segment of
neighboring DNA (haplotype) on both sides of the allele in question, and this for
historical reasons. For example, four strains homozygous for the albino (Tyrc)
allele (A, AKR, BALB/c, and SJL) are probably homozygous for the same short
segment of chromosome 7 flanking the albino mutation (Tyrc) because the muta-
tion shared by these strains results from the same mutational event that occurred
well before the creation of these strains (i.e., identical by descent or IBD). This
peculiarity must be kept in mind because it applies to many other situations and
may be advantageous (or unfavorable?) in the design of an experimental protocol.
We will come back to this point in the section concerning congenic strains.
In most mammalian species, inbreeding of a natural population often has delete-
rious effects of variable intensity and phenotypic expression. In some (rare) cases,
stillbirths are observed or newborns exhibit growth retardation and finally die. In
other instances, there is a decrease in fitness or/and fertility, which is sometimes
severe to the point that it leads to the extinction of the strain. All these adverse man-
ifestations are commonly referred to as inbreeding depression. The basis of inbreed-
ing depression has been debated over the last century, including by Darwin himself.
Modern genetic studies suggest that inbreeding depression is predominantly caused
by the presence of recessive deleterious mutations in natural populations that are
progressively fixed in the homozygous state as inbreeding progresses (Charlesworth
and Willis 2009). Alternative explanations, such as epistatic interactions, are also
possible. In the mouse, surprisingly, inbreeding depression is not a serious issue as
long as the breeders stem from the same natural population of closely related indi-
viduals. This is probably explained by the fact that wild mice, trapped in the same
natural area, already have a relatively high percentage of consanguinity.
9.2.2 Inbred Mice are Genetically Stable in the Long Term
Around 230 different inbred strains were listed in the reference book by Michael
Festing (1979). In 1998, 426 strains, with a brief description for each of them,
were listed on the MGI website (http://www.informatics.jax.org/external/festing/
mouse/STRAINS.shtml), but it is more than likely that many of these strains have
been lost or terminated. However, among this impressive collection, two dozen
have become very popular.
The fact that all members of the same inbred strain are genetically identical
(isogenic) is certainly the major reason why they have become so prevalent in
biomedical research. Scientists working with the same inbred strain, but in differ-
ent laboratories or at different time periods, can perform experiments where the
variations in the experimental results, by definition, will not be the consequence
of possible differences in the genetic constitution of the animals. In Chap. 10,

devoted to the analysis of complex trait inheritance, we will provide examples
showing that this is indeed a huge advantage.
Being isogenic also provides the great advantage that one can define, in detail
and comprehensively, the phenotypic characteristics of each inbred strain by gath-
ering phenotypic data concerning this strain from several laboratories and storing
them in the same database. For example, The Jackson Laboratory has developed
a program to collect baseline phenotypic data on the most popular inbred strains
of mice through a coordinated international effort. Information collected in this
program (The Mouse Phenome Database) is freely available to the community
through the Internet (http://phenome.jax.org/) (Paigen and Eppig 2000). The estab-
lishment (and updating) of this database was made possible only because inbred
mice are isogenic and accordingly genetically stable in the long term (Table 9.1).
Taking a look at the descriptions and genetic profiles of inbred strains in this
database is always of great help when designing an experimental protocol. It may
also contribute to saving animal lives by avoiding the repetition of experiments
whose results are already known from experiments performed previously or else-
where (for example, the dosage of a particular metabolite or the evaluation of a
specific biological parameter).
Finally, being isogenic, mice of the same inbred strain are also histocompatible
(or syngeneic). This means that they permanently accept tissue transplantations
from any mouse of the same strain (and sex). Immunogeneticists have extensively
used this peculiarity, since it allows studying the fate of cells with an immuno-
logical function in different contexts (cellular cooperation). It has also been exten-
sively used (and still is) for the serial transplantation of malignant cells.
9.2.3 The Genetic Purity of Inbred Strains

Must be Regularly Monitored
Although considered relatively stable in the long term, the genetic profile of a
given inbred strain may change for two main reasons. The first results from acci-
dental contamination by another strain; the second results from the progressive
and insidious accumulation of novel mutations.
Genetic contamination resulting from the accidental mating of individuals of
one inbred strain with another strain is by far the most important cause of altera-
tion of the genetic profile. Such contaminations always result in a sudden and mas-
sive exchange of alleles and generally occur between strains that have the same
or similar coat color (i.e., albino (Tyrc/Tyrc), agouti (A/A), or non-agouti (a/a)).
These accidental crosses also occur between interstrain hybrid F1s and one of the
parental strains, and between inbred and outbred strains (albino in particular). As
a rule, accidental crosses result in an abrupt increase in breeding performances of
the colony; such a change must always be considered suspicious and suggestive of
a genetic contamination!
Table 9.1 This is part of a table listing in the Mouse Phenome Database

♀ Females Mean SD N ♂ Males Mean SD N
129S1/SvlmJ 10.4 +0.494 N = 8 129S1/SvlmJ 10.6 +0.481 N = 7
A/J 9.85 +0.495 N = 8 A/J 10.0 +0.489 N = 8
AKR/J 10.2 ±0.470 N = 5 AKR/J 9.36 +0.495 N = 8
ALR/LtJ 8.82 ±0.495 N = 8 ALR/LtJ 8.75 ±0.484 N = 7
ALS/LtJ 10.1 ±0.463 N = 8 ALS/LtJ 9.81 +0.449 N = 7
BALB/cByJ 10.3 +0.515 N = 8 BALB/cByJ 10.0 +0.489 N = 6
BPH/2 J 9.98 +0.471 N = 7 BPH/2 J 10.0 +0.445 N = 7
BPL/1 J 11.7 +0.439 N = 6 BPL/1 J 11.3 +0.438 N = 7
BPN/3 J 10.8 ±0.437 N = 8 BPN/3 J 10.5 +0.454 N = 7
BTBR/J 8.85 ±0.546 N = 8 BTBR/J 9.82 ±0.514 N = 6
BUB/BnJ 9.58 ±0.505 N = 7 BUB/BnJ 8.72 +0.480 N = 4
C3H/HeJ 9.25 +0.495 N = 7 C3H/HeJ 9.79 +0.470 N = 7
C57BL/6 J 10.2 +0.584 N = 114 C57BL/6 J 10.3 +0.548 N = 88
C57BR/cdJ 10.7 +0.439 N = 7 C57BR/cdJ 10.7 +0.443 N = 8
C57L/J 10.7 ±0.481 N = 4 C57L/J 10.8 ±0.533 N = 7
PWK/PhJ 10.3 ±0.482 N = 7 PWK/PhJ 10.3 ±0.493 N = 8
RBA/DnJ 10.7 +0.444 N = 7 RBA/DnJ 10.6 +0.449 N = 8
RBF/DnJ 9.78 ±0.541 N = 9 RBF/DnJ 9.84 ±0.525 N = 8
RF/J 11.2 +0.454 N = 7 RF/J 11.4 +0.447 N = 8
Rill S/J 10.0 ±0.438 N = 8 RIIIS/J 10.3 ±0.432 N = 7
SB/LeJ 8.63 ±0.433 N = 7 SB/LeJ 7.52 ±0.437 N = 8
SEA/GnJ 10.9 +0.438 N = 6 SEA/GnJ 11.1 +0.436 N = 7
SF/CamEiJ 9.06 +0.489 N = 6 SF/CamEiJ 9.14 ±0.463 N = 8
SJL/J 9.84 +0.549 N = 8 SJL/J 10.6 +0.481 N = 4
SKIVE/EiJ 10.1 ±0.441 N = 7 SKIVE/EiJ 10.2 ±0.449 N = 8
SM/J 9.65 +0.546 N = 8 SM/J 9.62 ±0.484 N = 7
SOD1/EU 10.8 +0.495 N = 8 SOD1/EiJ 11.7 +0.464 N = 7
SPRET/EiJ 10.8 ±0.443 N = 7 SPRET/EiJ 11.6 ±0.430 N = 4
The red blood cell counts (measured in n/μL ± 1 SD) were performed on male and female mice of
72 different inbred strains (only 28 strains are represented here). Arrows indicate the most extreme
values <http://phenome.jax.org/db/qp?rtn=views/measplot&brieflook=31802&projhint=CGDph
eno1> Many other phenotypic parameters are stored in this Mouse Phenome Database for the most
common inbred strains, allowing selection of the “best strain” when outlining an experimental pro-
ject. Consultation of this database avoids wasting animals by repeating measurements uselessly
Mouse strains A2G and C57BL6/Ks are two well-known examples of inbred
strains for which genetic contamination has been reported. A2G was consid-
ered to be a substrain of strain A until it was discovered that it probably origi-
nated from an “illegitimate” mating with an unknown partner. Mice of the A2G
strain exhibit natural resistance to myxovirus (influenza), a peculiarity uncommon
in most other laboratory strains, and it makes sense to believe that this character-
istic is a “memory” of the illegitimate mating that occurred when the strain was
developed. Strain C57BL/Ks (now C57BLKS) is another interesting case. The
strain derives from strain C57BL/6 but was contaminated with up to 25 % from
the DBA/2 genome, 4 % from C57BL/10 J, from a 129 source and possibly some
other undefined source. These untraced (and successive) contaminations were sus-
pected for two reasons: because C57BL/Ks mice have a haplotype at the H2 his-
tocompatibility complex, which is not the one normally found in C57BL/6 mice
(C57BLKS mice are H2d, like strain DBA/2, instead of H2b like strain C57BL/6);
and because congenic mice for the same obese (Lepob) mutation in these two back-
grounds (C57BL/6J and C57BLKS) exhibited a different phenotype (Coleman and
Hummel 1973). The suspicion of genetic contamination has now been molecularly
documented and even cleverly used in an attempt to unravel the genetic causes of
the background effect on Lepob phenotypic expression (Mao et al. 2006).
It is likely that many genetic contaminations have occurred in the past that have
been rapidly detected and eliminated, but it is feared that the enormous increase in
numbers of genetically engineered mouse (GEM) strains we are witnessing nowadays
will exacerbate the threat of genetic contamination due to overcrowding of the breed-
ing facilities. Commercial breeders are extremely sensitized to the risk linked with
genetic contamination and perform regular monitoring of their stocks and strains.
Most of them also have backups (archives) of their stocks cryopreserved in an embryo
bank, allowing the rapid development of a fresh strain when necessary. At present,
genetic monitoring of inbred strains is based on the use of molecular techniques at the
DNA level and provides quick and highly reliable answers (See Box 9.1).
Box 9.1: Genetic monitoring of inbred strains and their derivatives

A variety of techniques has been described in the past to assay the genetic
quality of inbred strains. All these techniques were based on the postulates
that each inbred strain, as previously mentioned, is expected a priori to be
homozygous at all loci. These techniques were designed following the pro-
gress of the genetic tools available for the species and consisted of analyzing
a few traits, controlled by a set of specific alleles, and defining a specific pat-
tern for each strain. Reciprocal skin grafting, for example, was extensively
used in the 1960s because histocompatibility is controlled by many genes
and requires complete genetic identity between the donor and the recipient.
Skin grafting was a relatively inexpensive procedure, but unfortunately it
was often influenced in both directions by environmental factors, yielding
false-positive and false-negative results.
Analysis of the electrical charge of enzymatic proteins (isozymes) by
electrophoresis in gels became popular in the mid-1970s because the tech-
nique was highly reliable and relatively easy to handle. However, this tech-
nique had the major drawback of being expensive to apply because each test
required the use of specific and costly reagents.
Currently, most of the genetic monitoring techniques applied to inbred
strains are based on DNA analysis and are extremely powerful. Most of
these tests are based on the analysis of strain-specific DNA sequences,

revealed by routine techniques such as polymerase chain reaction (PCR)
amplification of microsatellites (also known as simple sequence length poly-
morphism, SSLP), or SNP genotyping.
However, it must be kept in mind that the control of genetic purity should be
undertaken in a broader context, considering parameters of very different nature
(coat color, behavior, spontaneous diseases, breeding performance, etc.), and
not only by applying sophisticated molecular techniques. Among these param-
eters, a careful observation of individuals from the same inbred strain, even if it
may appear rather subjective, is always a very important source of information.
Genetic monitoring using microsatellite markers

Microsatellite markers are very popular because they are extremely easy to type
at a very low cost (Benavides 1999; Mashimo et al. 2006). The technique con-
sists of the amplification of short repeated sequences, in general dinucleotides
of the type (CA)n or (TA)n, with flanking primers using genomic DNA. There is
an enormous number of microsatellite loci in the mouse and rat genomes (prob-
ably around 105), and it is generally not a problem to find a set of such molecu-
lar markers whose amplification products define a strain-specific pattern. This
strain-specific pattern may be assayed on a sample of animals of the strain and
compared to a reference DNA that is archived in the laboratory. Routine analy-
sis of DNA samples with microsatellite markers will confirm isogenicity and,
provided the markers have been carefully selected, it could also guarantee that
the strain whose DNA is assayed indeed corresponds to its designation. One of
the advantages of microsatellites is the fact that these are multiallelic markers,
meaning that when tested in different inbred strains a marker will show several
alleles (band sizes) for some of these strains. The use of fluorescently labeled
primers for microsatellite loci combined with capillary electrophoresis repre-
sent a new, fast, and automated system for genetic monitoring (Mashimo et al.
2006). With this method, the resulting PCR products can be distinguished from
one another by both their size and the fluorescent dye associated with them.
The availability of different dyes allows the possibility of developing multiplex
PCR (i.e., the combination of primers for multiple loci in one reaction) and
pooling several PCR products into one capillary (Bryda and Riley 2008).
Genetic monitoring based on the use of single nucleotide polymorphisms

(SNPs)
Genotyping for SNPs is an alternative approach that is now very popular
for genetic quality control. SNP genotyping is inexpensive and can be per-
formed in most research institutions or ordered to external companies. SNPs
are the most common type of genetic variation observable at the DNA level
and are found in both coding and non-coding regions. When localized in
coding sequences, if the variant leads to an amino acid change or the crea-
tion of a stop codon, the SNP is said to be non-synonymous. On the other
hand, if the SNP does not change the protein sequence it is considered syn-
onymous. Almost all SNPs are bi-allelic, presenting one of only two pos-
sible nucleotides (e.g., homozygous G/G or T/T) or both (e.g., heterozygous
G/T) in an individual. Petkov and coworkers from The Jackson Laboratory
(Maine, USA) have described the allelic distribution of 235 SNPs in 48
mouse strains and selected a panel of 28 such SNPs, enough to character-
ize most of the almost 300 inbred, wild-derived, congenic, consomic, and
recombinant inbred strains maintained at The Jackson Laboratory (Petkov
et al. 2004a). This set of markers encompassing all mouse chromosomes
is an excellent tool for detecting genetic contaminations in mouse facilities
by way of automated PCR systems. The same laboratory developed a new
set of 1,638 informative SNPs selected from the publicly available data-
bases and tested 102 inbred strains using Amplifluor genotyping (Myakishev
et al. 2001). The selected SNPs are distributed approximately ~1.5 Mb
apart across the mouse genome and, on average, 37 % will be polymorphic
between any two inbred strains. Interestingly, these markers revealed sub-
tle differences between closely related inbred strains and substrains, a result
that was independently confirmed for the most popular C57BL/6 substrains:
C57BL/6J from The Jackson Laboratory and C57BL/6N from the National
Institutes of Health (Mekada et al. 2009; Zurita et al. 2011; Simon et al.
2013). SNP genotyping assays are currently based on allele-specific PCR
(including KASPar fluorescent technology) (Nijman et al. 2008), real-time
PCR (TaqMan®), direct sequencing, or DNA arrays (Moran et al. 2006). For
those interested in the allele distribution of SNPs in different inbred strains,
the Mouse Phenome Database presents the most comprehensive collection
of SNPs, with more than 8 million unique loci and numerous inbred strains
genotyped (see http://phenome.jax.org/db/q?rtn=docs/genonav).
Box 9.2: Preserving the genetic purity of inbred strains

As discussed above, efficient techniques exist to monitor the genetic quality
of inbred strains. However, once a strain is recognized as contaminated, the
situation is irreversible. All individuals of the strain must be discarded, and
another strain must be developed from a fresh set of breeders. The dramatic
consequences of a genetic contamination imply that measures should be taken
to prevent it. There are two efficient ways of preserving an inbred nucleus
from genetic contamination: embryo freezing and complete physical isola-
tion. Embryo freezing is, theoretically, the most efficient way of preservation
because, once frozen, genomes are insensitive to mutations (DNA does not
replicate), and of course contamination cannot occur. Sperm freezing may, in
some circumstances, be used as an alternative to embryo freezing, but is of no
use when a diploid genome must be preserved (Glenister and Thornton 2000).
Complete isolation of a breeding nucleus, for example, into a plastic isolator

is a very efficient way of preserving genetic integrity and it is also an elegant
way of preserving, at the same time, the health status of a rodent colony. Most
of the commercial breeders of laboratory mice have chosen this strategy, which
combines several advantages at a reasonable cost.
Box 9.3: Breeding protocols for the maintenance of an inbred strain

The system represented in a leads to the establishment of three (and not one)
independent inbred strains, which are progressively divergent from one another
due to genetic drift. In addition, if a strain stops breeding (a common situa-
tion in practice), it is then permanently lost. The system represented in b is
certainly the best one, since, at each generation, three new pairs are established
from generation N to breed mice of generation N+1. However, when the prog-
enies are very small in size (a situation that is also common), it is not always
possible to set the three new pairs of brothers and sisters. Finally, the system
represented in c is the one that is generally used in practice.
(a)
(b)
(c)
Mutations are another source of genotypic change and are important to consider
for two reasons: first, because their occurrence is completely beyond the control of
the colony manager; and second, because they are insidious and in general impos-
sible to detect by simple phenotypic observation or routine genetic monitoring. As
reported in Chap. 7, the spontaneous mutation rates are quite low. They have been
estimated to be in the range of 0.1 to 0.5 × 10−6 per locus per gamete for muta-
tions towards a dominant allele and in the range of 0.6 to 0.8 × 10−6 per locus
per gamete for mutations towards a recessive allele (Schlager and Dickie 1967).
However, while a proportion of these new mutant alleles are effectively eliminated
by inbreeding, another proportion may become progressively fixed in the homozy-
gous state, replacing the original allele; this is one aspect of what geneticists call
genetic drift. Genetic drift is a very slow and insidious process that is unavoidable.
It contributes inexorably to strain divergence (and to the generation of substrains)
when the same strain is propagated independently in different places.
Recently collected data concerning single nucleotide polymorphisms (SNPs) in
different C57BL/6 substrains kept independently for a few years at The Jackson
Laboratory indicated that the mutation rate for generating SNPs is very low (Wade
et al. 2002). In addition, and assuming that only one SNP out of seven is trans-
lated into a functional polymorphism (see Chap. 7), this would suggest that the
occurrence of new mutations is not a serious issue in the generation of sub-line
divergence. The problem, however, is that the consequences of a novel mutation
are not predictable. Mutations which are hidden in the genomes of substrains and
can affect the outcome of an experiment are sometimes referred to as “passen-
ger mutations” (Kenneth et al. 2012). There are many examples in the literature
where substrains, although stemming from the same original inbred strains, have
acquired new and unique phenotypic characteristics as a consequence of genetic
drift (Bulfield et al. 1984; Stevens et al. 2007; Mattapallil et al. 2012). Mice of the
C57BL/6J/OlaHsd substrain, for example, are homozygous for a deletion of the
Snca locus (encoding for α-synuclein) on chromosome 6 (Specht and Schoepfer
2001). This deletion has modest phenotypic effects but might interfere in an
unpredictable manner with other mutations if, for example, the C57BL/6J/OlaHsd
substrain is used as a background strain for the production of knockout. In addi-
tion, a few spontaneous mutations have been reported to segregate differentially
in the most popular substrains of C57BL/6 mice (C57BL/6J from The Jackson
Laboratory and C57BL/6 N from the National Institutes of Health (separated in
1951), including a retinal degeneration mutation (Crb1rd8) present in the N sub-
strain and a deletion in the Nnt gene present only in the J substrain. The most com-
prehensive comparative phenotypic and genomic analysis of these popular strains
has been recently published (Simon et al. 2013).
Similarly, if mice of substrain C3H/HeJ are experimentally infected with Gram-
negative bacteria they may react very differently from mice of substrain C3H/OuJ.
This is explained by the occurrence of a spontaneous mutation at the Tlr4 locus
(encoding for a Toll-like receptor) in the substrain C3H/HeJ, where all mice are
homozygous for the defective allele Tlr4Lps-d (Poltorak et al. 1998). A very similar
comment could be made for mice of the CBA/NJ substrain (CBA/CaHN-Btkxid/J)
which, unlike mice of all other CBA substrains, are homozygous for an X-linked
mutation (Btkxid) producing a syndrome of immunodeficiency homologous to
Bruton disease in humans (Berning et al. 1980).
What we have just said concerning the insidious and unavoidable occurrence
of new mutations in an inbred strain also explains and justifies the recommenda-
tion by the International Committee on Standardized Genetic Nomenclature for
Mice that inbreeding should never be relaxed. Inbreeding is inefficient in prevent-
ing mutations from occurring, but it contributes to the elimination of a substan-
tial proportion of the new mutant alleles, and accordingly helps to preserve the
genetic profile of a given strain in the long term. Similarly, the same international
committee on nomenclature has decided that two strains with the same origin but
separated in different colonies by 20 or more generations (for example, 12 in labo-
ratory A and 10 in laboratory B) should be considered as two different substrains
and designated appropriately (Davisson 1996; Wotjak 2003).
9.2.4 Most Inbred Strains are Derived from a Small Number

of Ancestors
In 1982, two independent observations were published (Ferris et al. 1982;

Yonekawa et al. 1982) reporting the remarkable structural homogeneity of mito-
chondrial DNA (mtDNA) among different strains of laboratory mice. This was
quite surprising because the mitochondrial genome, a 16-kb double-stranded cir-
cular DNA molecule, was known in most species (including humans) to have a
relatively high evolution rate. Considering that the mtDNA is inherited exclusively
from the female parent, the most likely explanation to account for these observa-
tions is that all classical inbred strains share a common maternal lineage, probably
inherited from a female albino mouse, bred as a pet, around 200 years ago. These
observations, essentially based on the analysis of mtDNA restriction fragment
length polymorphisms, have been recently confirmed and refined after sequencing
and it has been confirmed that the laboratory strains are all derived from female
progenitors of the Mus musculus domesticus subspecies, while Mus musculus mus-
culus does not appear to have made any contribution to the mtDNA (Goios et al.
2007).
A few years after the observations concerning the mtDNA were published,
similar experiments were made concerning the Y chromosome, a totally pater-
nal contribution. Using five probes identifying Y chromosome-specific restriction
fragments, six distinct Y chromosomes were identified among 39 standard inbred
strains of mice, indicating that a minimum of six male mice contributed to the
formation of the common inbred strains. Three Y chromosome types, distributed
among 31 strains, were found to be of Asian (M. m. musculus) origin, while the
remaining three Y chromosome types (8 strains) were of M. m. domesticus origin
(Bishop et al. 1985; Tucker et al. 1992). All these findings are completely consist-
ent with the historical records: they confirm that the laboratory inbred strains were
all derived from one or a few related females of the Mus musculus domesticus
subspecies while the inter-strain polymorphisms represent the contribution of six
males, all of them of the Mus musculus musculus subspecies.
9.2.5 Laboratory Inbred Strains have a Polyphyletic Origin
Inbred strains are often said to be artificial populations because their genetic consti-
tution (isogenicity and homozygosity) has no natural equivalent. In fact, they could
also be considered artificial populations because we now know, from historical
records confirmed by extensive molecular data collected at the DNA level (sequenc-
ing), that they do not stem from one and a single subspecies of the Mus genus but
from at least two: Mus musculus domesticus and Mus musculus musculus (Guénet
and Bonhomme 2003). The finding of this polyphyletic origin was no real surprise
if we recall the observations reported above concerning the origin of the mtDNA
molecule and of the Y chromosome. This was also suspected for quite a long time,
because it was the only way to explain that some electrophoretic variants of plas-
matic proteins (for example, the esterase-2 allele c (Es2c) or the phosphoglucomutase
1 allele b (Pgm1b)), which are frequently found in laboratory strains as well as in
mice of the M. m. musculus subspecies, are extremely rare in the genome of wild
mice of M. m. domesticus subspecies (Bonhomme 1986; Bonhomme et al. 1987).
The polyphyletic origin was confirmed and substantiated further after the com-
plete high-resolution sequencing of the genomes of a large panel of inbred strains
(Waterston et al. 2002; Wade et al. 2002; Yalcin et al. 2004; Frazer et al. 2007; Yang
et al. 2011). In short, one can say that the genomes of laboratory inbred strains are a
mosaic of chromosomal regions with distinct subspecific origins (Fig. 9.3).
On average, and according to the most recent estimates, the genetic contribu-
tions of the different Mus musculus subspecies is as follows: M. m. domesticus
68 %, M. m. musculus 6 %, M. m. castaneus 3 %, and M. m. molossinus 10 %. The
remaining 13 % of haplotypes are of unknown ancestral origin.
It is also important to note that the distribution of diversity is markedly non-ran-
dom among the chromosomes, with large regions of extremely low diversity and hot
spots of diversity (Frazer et al. 2007; Church et al. 2009; Yalcin et al. 2011). This
observation is particularly interesting because it results in an increase in genetic pol-
ymorphisms, making each inbred strain different from the other, and much more dif-
ferent from each other than we would have expected if mutations and genetic drift
were the only source of diversity. Studies on the genetic determinism of complex
traits benefit from this unique situation, as will be discussed in Chap. 10.
9.2.6 Inbred Strains Recently Derived from Wild Specimens
Over the last 20 years a variety of strains, derived from small nuclei of wild
specimens trapped in well-defined geographical regions and belonging to well-
characterized taxonomic groups, have been established in various laboratories.
(a)
a - Mus m. musculus b - Mus m. castaneus
a
a b
d - Mus m. domesticus c - Mus m. molossinus

d c
Most common laboratory strains
(b)
Fig. 9.3 Origin of classical inbred strains of the laboratory mouse. a Historical data, confirmed
by sequence data, indicate that modern laboratory inbred strains derive from a small number of
ancestors belonging to several different subspecies of the genus Mus. Today’s classical laboratory
inbred strains must be regarded as recombinant strains derived from four parental components (in
unequal percentages): M. m. domesticus, M. m. musculus, M. m. castaneus, and M. m. molossi-
nus. For this reason it would probably be more appropriate to designate them as Mus “labora-
torius”! This polyphyletic origin explains (partially) the interstrain polymorphism segregating
among the different laboratory strains. b The figure represents four mouse chromosomes in which
some segments derive from one of the four ancestor subspecies (based on Frazer et al. 2007)
A list of these strains was published in the book Genetic Variants and Strains of
the Laboratory Mouse (Bonhomme and Guénet 1996) and many of these strains
are described on the internet at http://jaxmice.jax.org/list/cat481389.html. Most of
these strains are now fully inbred with, in general, well over the required 20 gen-
erations of brother × sister matings.
Amongst all these inbred strains, special mention must be made of those derived
from the Mus spretus species (for example, SEG/Pas, SPRET/Ei, and STF/Pas)
because this species is one of those most distantly related to the laboratory strains
(from the evolutionary point of view) that can still produce fertile hybrids. The
production of these inter-specific hybrids results, in most instances, from natural
matings between laboratory strain females and Mus spretus males, although some
hybrids have also been produced with the opposite cross either by artificial insemi-
nation or by in vitro fertilization. F1 males with Mus spretus are sterile, as a conse-
quence of the Haldane rule, but F1 females are fertile and can be used to produce
backcross progeny by mating them to males of either the laboratory strain or of

the wild-derived strain. Segregating backcross progeny born to such F1 females,
because of the very large number of allelic differences involved, have been instru-
mental for generating high-density/high-resolution genetic maps (see Chap. 4).
With the increasing use of techniques based on PCR amplification for the
detection of genetic polymorphisms at the DNA level, the laboratory strains
derived from M. m. musculus, M. m. molossinus, and M. m. castaneus have also
been found to be of great value. Using a large set of oligonucleotides for the
amplification of microsatellites, it has been demonstrated that 70 % of these prim-
ers yield PCR products polymorphic in length between strain PWK (an inbred
strain derived from M. m. musculus) and the laboratory strain C57BL/6. By this
criterion, the PWK strain appeared almost as distantly related as Mus spretus to
the common laboratory strains (70 % vs. 74–84 %). With even more refined tech-
niques, such as those based on the analysis of discrete structural variation that are
capable of detecting a single base replacement (single strand conformation poly-
morphism or SSCP and denaturating gel gradient electrophoresis or DGGE), or
even regional sequencing, it is likely that virtually any DNA stretch from a non-
coding region that is more than 100 bp long would be found polymorphic between
any inbred strain of wild origin and a reference laboratory strain. These observa-
tions should encourage scientists involved in gene mapping and/or positional clon-
ing of genes to use the strains derived from M. m. musculus, M. m. molossinus or
M. m. castaneus more intensely than those derived from M. spretus. They are eas-
ier to breed and frequently have the considerable advantage of allowing the pro-
duction of F2 progeny where each individual results from two informative meioses
and not just one—as is the case for the offspring of backcrosses.
The development and use of inbred strains recently derived from wild progeni-
tors has been a great addition to the resources available to mouse geneticists and
may prove even more useful in the future for the analysis of quantitative (com-
plex) traits. However, one must be careful when using these new strains because
the abundance of polymorphisms (SNPs or indels) sometimes makes difficult the
establishment of causal correlation between the structural variations at the DNA
level and their consequences in terms of gene function and expression. Examples
of the advantages of using wild-derived inbred strains can be found in reviews on
the subject (Guénet and Bonhomme 2003; Dejager et al. 2009).
9.2.7 Phylogenic Relationships Between Inbred Strains
Based on genotyping data collected by using a set of informative SNP markers and
using an appropriate computer program for the optimal neighbor-joining method
under the principle of maximum parsimony, a diagram has been established by
researchers at The Jackson Laboratory (Petkov et al. 2004b), which represents the
phylogenic relationships of the most commonly used inbred strains of the labora-
tory mouse (Fig. 9.4).
Fig. 9.4 A mouse family tree. The 60 inbred strains represented in this figure have been gen-
otyped for a set of 1,465 informative SNP markers, evenly distributed over the whole genome
(spaced on average <1.5 Mb). Applying the neighbor-joining method to the data, the authors
constructed a family tree that could be organized into three groups: group 1, BALB/c, 129,
and DBA-related strains; group 2, Swiss mice and Asian strains; group 3, wild-derived strains.
The length and angle of the branches have been optimized for printing and do not reflect the
actual evolutionary distances between strains. This family tree is in good agreement with most
other existing genealogies (from Petkov et al. 2005). Using more markers for genotyping would
increase the resolution of the phylogenetic tree (see, for example, Petkov et al. 2004b)
This diagram is in good agreement with the historical data previously collected
(Beck et al. 2000) and can be used, for example, for the selection of closely or
distantly related strains. This information is of primary importance for the design
of an experimental protocol aiming to study the genetic determinism of inter-strain
phenotypic differences. Indeed, selecting more distantly related parental strains
when setting up a cross offers a greater chance of obtaining a higher resolution in
the genetic analysis (Frazer et al. 2007).
9.3 Interstrain F1 Hybrids
Resulting from the cross of two inbred strains, F1 hybrids are heterozygous at
all loci for which the parental strains have different alleles, but they are geneti-
cally uniform (isogenic) like their parents. Pairs of the same sex are equivalent to
monozygotic (identical) twins or to cloned mice. They are also histocompatible

and permanently accept tissue transplantations from either parental strain, from
their littermates, and from all their offspring; however, the parental strains will not
accept a graft from the F1 hybrids.
F1 mice also exhibit the legendary hybrid vigor (heterosis), the opposite of
inbreeding depression, making them the material of choice in many experimen-
tal protocols. This is common, for example, in the protocols aimed at the produc-
tion of genetically engineered animals, where F1 hybrids are often used because
of their high production of pre-implantation embryos that are highly resistant to
manipulation (e.g., DNA pronuclear microinjection or for the creation of robust
chimeras). However, a major drawback is that, when intercrossed, their progeny
(F2) are genetically heterogeneous, since the alleles at all polymorphic loci start
segregating due to recombination events in the F1 gametes during meiosis.
Interstrain hybrids can also be used to generate genetically heterogeneous
populations. This is the case when, for example, F1 hybrids between strain A and
strain B (abbreviated ABF1 or AXBF1) are crossed with F1 hybrids between strain
C and strain D (CDF1 or CXDF1) to generate a four-way heterogeneous stock.
In this case the basic ingredients of such a genetically heterogeneous stock (i.e.,
the original inbred strains A, B, C, and D) are perfectly identified, and similar,
although not identical stocks can be produced at will when necessary. As we will
explain later in this chapter, genetically heterogeneous stocks with an even more
complex structure (for example, eight-way crosses stemming from eight different
and unrelated inbred strains) have also been bred on a large scale for research in
quantitative genetics (Threadgill and Churchill 2012) (see Chap. 10).
9.4 Co-isogenic and Congenic Strains
9.4.1 Co-isogenic Strains
When a mutation occurs in the breeding nucleus of an inbred strain, and if we

assume that the new mutant allele has substituted the original one then the inbred
strain in question differs from the original strain at one and only one specific
locus. If the new mutation is viable and does not impair fertility, one can prop-
agate the new strain by mating brother × sister mutant mice or, preferably, by
mating, at each generation, a non-mutant mouse of the original inbred strain to
an animal of the new mutant strain. These two strains are said to be co-isogenic
strains or segregating inbred strains (Fig. 9.5).
Co-isogenic strains are extremely useful for gene annotation because they allow
a comparison of the phenotypes of two allelic forms of a particular gene under opti-
mal conditions (i.e., with no influence from the genetic background). A large number
of such strains are stored worldwide in the major genetic repositories. Some inbred
strains, like the famous C57BL/6, have several co-isogenic “companion” strains seg-
regating for a variety of allelic forms involved, for example, in the determinism of
9.4 Co-isogenic and Congenic Strains 339
Fig. 9.5 Co-isogenic strains.
The figure represents two
mice of the same highly
inbred strain DW/JPas. The
obese mouse is homozygous
for a short-sized duplication
of the gene encoding the
extracellular domain of the
leptin receptor (Lpr). The
(Lprdb-Pas) mutant allele is
inactive and the co-isogenic
mouse grows to be obese
coat color. Mice of the C57BL/6-Tyrc (albino) co-isogenic strain have become pop-
ular for the production of easily recognizable C57BL/6- +/+ ↔ C57BL/6-Tyrc/Tyrc
chimeric mice from C75BL/6 ES cells injected into albino C57BL/6-Tyrc/Tyrc blas-
tocysts (Schuster-Gossler et al. 2001).
Other strains, co-isogenic for mutations with detrimental effects on develop-
ment or metabolism are also very interesting models because they can help in the
analysis of pathophysiology, providing both the experimental animal and its con-
trol. Using such strains it is possible, for example, to attempt phenotypic rescues
by grafting normal cells into a co-isogenic partner as a preliminary study for the
design and development of possible therapies for human diseases. Co-isogenic
strains, when developed in parallel to the background strain, may accumulate other
genetic differences over time as a consequence of genetic drift. Thus, to mini-
mize the effects of this drift, they must be periodically backcrossed to the original
parental strain, or be cryopreserved.
Co-isogenic strains have two major drawbacks that are inherent in their origin
and seriously limit their use: (i) they appear mainly as a consequence of a rare
and fortuitous event (a mutation); and (ii), although they can appear in any inbred
strain, it is in general not the strain that we would have been primarily interested in.
For these two reasons, the use of co-isogenic strains is rather limited (see below).
9.4.2 Transgenic Strains are Equivalent but not Identical

to Co-isogenic Strains
Genetically engineered mice can also be considered co-isogenic strains when the
genetic modification is done in a way such that the targeted locus or transgene is the
only difference from the wild-type animals. In the case of classical transgenic mice
(additive or pronuclear transgenesis), this can be achieved by performing the pronu-
clear DNA (transgene) microinjection using embryos derived from an inbred strain
(e.g., FVB/N or C3H/He). Transgenic lines must be developed independently from

each founder animal (microinjected embryos) and are normally kept by backcross-
ing the transgenic carriers (hemizygous Tg/0) with wild-type animals from the back-
ground strain and by selecting, at each generation, the new carriers (typically by PCR
genotyping). One important difference between classical transgenic (by pronuclear
microinjection) and co-isogenic strains is that the structure of the transgenic insertion
can change with time: for example, in terms of copy number or DNA methylation. It
can also be lost, leaving behind a micro-rearrangement, in general a micro-deletion.
9.4.3 Congenic Strains
Congenic strains are an alternative to co-isogenic strains with the advantage that
any allele of the genome may be moved (geneticists would say “introgressed”) into
any inbred background. The disadvantage, as we will explain, is that the situation
is not as pure, from the genetic point of view, as it is in the case of co-isogenics.
Congenic strains are produced by crossing two strains: the first one carries the
allele or chromosome region of interest (i.e., spontaneous, induced or targeted
mutations, as well as transgenes), and is referred to as the donor strain; the sec-
ond strain is referred to as the recipient strain or background strain. The F1 off-
spring generated by crossing the above-mentioned two strains are backcrossed to
the background strain, and the offspring that carry the allele of interest (i.e., the
one originating from the donor strain) are crossed again to the background strain
and so on, typically for ten or more successive generations.
During this succession of backcrosses, the chromosomes of the background
strain progressively replace those of the donor strain, except for the one that carries
the allele of interest. For this particular chromosome, the segment containing the
selected or targeted allele is reduced in size only when a recombination event occurs
that replaces a piece of chromosome of the donor strain with the homologous seg-
ment of the background strain. Since the occurrence of this sort of event depends
upon the size of the segment, one then realizes that the chromosome carrying the
targeted allele is gradually “eroded” on both sides, generation after generation, but
in a nonlinear manner. The chromosomal segments flanking the selected locus have
a tendency to remain associated with this locus, and this is the major difference
between congenic and co-isogenic strains. In other words, while co-isogenic strains
differ from the background strain at a single locus, congenic strains differ by a short
chromosomal segment flanking the targeted locus, with the size of this segment
being progressively reduced during the successive backcross generations.2
Since, on average, at each generation, an equivalent proportion of the back-
ground strain replaces one half of the genome of the donor strain, the pro-
gression of genome substitution is given by the formula 1/2N, where N is the
number of backcross generations. This means that, theoretically, after 10 backcross
2 The reduction in size of the introgressed chromosomal segment is in steps instead of linear.
generations only 1/210 (<1/1,000) of the donor genome remains in the congenic
strain. It is clear that this assumption is, again, purely statistical and the actual per-
centage of donor genome is subject to variations at each generation. In addition,
and as we already pointed out, this estimation stands only for the chromosomes that
do not carry the allele of interest (the selected or targeted allele). In the latter case,
the reduction in size is a much slower process. According to Johnson (1981), if two
loci A and B are distant by c Morgans, the probability that no recombination occurs
between these two loci is e–c per generation and, therefore, e–nc after n genera-
tions. In the case of congenics, if A is the targeted locus and B a gene in the vicinity
(located, for example, 10 cM from A), the probability that the two loci remain in the
same parental configuration after 10 generations is ~0.37 (=37 %). If A and C are
5 cM apart, the probability increases to 60.6 %, and it increases to ~90 % for two
loci separated by 1 cM (0.01 Morgan). Stated differently, this means that there is
only a 10 % (=100–90) chance that the segment harboring the introgressed gene
will be smaller than 2cM (1 cM on each side) after a series of 10 backcrosses. This
is not negligible since, as we discussed in Chap. 5, 1 cM of the mouse genome may
contain up to 30–40 genes or even more, depending on the region (Fig. 9.6).
9.4.3.1 Marker-Assisted Congenics or Speed Congenics
The use of polymorphic and easy-to-score DNA markers has allowed a much more
rapid and rigorous process of congenic strain development: the the so-called marker-
assisted breeding (or backcrossing), also referred to as speed congenics methodology.
The principle that underlies the speed congenics process is based on the fact that one
can select the breeders, at each generation of backcrossing, based on the percentage
of donor genome they have, by using either microsatellites or SNPs to distinguish the
two parental strains. Obviously, the mouse with the lowest percentage of donor DNA
is the one to select as a breeder for setting up the next backcross. Doing this greatly
reduces the number of generations necessary to reach full congenicity (for example,
from N10 to N5), and the strain development time, approximately by half.
At this point, it is important to note that, although a large number of molecu-
lar markers are necessary to perform efficient and reliable genotyping during the
first backcross generation (in general 80–100 evenly distributed over the whole
genetic map, for the N2 generation), this number decreases rapidly because, once
a marker is typed “homozygous” for the allelic form of the background strain, it
is no longer necessary to genotype the offspring of the future generations for this
marker—it is permanently fixed (Markel et al. 1997; Wakeland et al. 1997). In
order to fix the background Y chromosome, it is recommended to mate the female
F1 hybrid to a male of the recipient strain early in the breeding scheme. Using
molecular markers helps in the selection of breeders with the smallest amount of
“flanking” or “hitchhiking” DNA, helping to alleviate the “flanking gene” concern
(Wolfer et al. 2002; Chen et al. 2004). This requires the breeding of a large num-
ber of offspring, but these mice can be genotyped at an early age and discarded if
considered unnecessary for future matings (Figs. 9.7 and 9.8).
F M
Background B Donor
strain strain
B F1 (50 % B)
B N2 (75 % B)
Selection of
animals carrying
the allele of
interest
B N3 (87.5 % B)
B N4 (93.8 B)
N…
Fig. 9.6 Congenic strains 1. This scheme represents the successive steps in the establishment
of a congenic strain. The initial step is a cross between two strains: a donor strain (black in the
example) carrying the gene of interest (e.g., the targeted locus that can be a transgene or another
allele) and a recipient or background strain (white in the example). At each generation, a breeder
carrying the gene of interest (*) is backcrossed to a partner of the recipient (or background, B)
strain. The degree of gray color indicates that, after each backcross generation, the offspring have
an increased amount of the background genome. When the targeted gene has no easily recogniz-
able phenotype, molecular genotyping is necessary. This genotyping is based on an easily detect-
able structural alteration (in most instances by PCR) within the locus in question. Closely linked
markers may also be used
Everything described so far about how to establish a speed congenic strain cor-
responds to a standard protocol that can be applied in virtually any laboratory. In this
strategy, the geneticist chooses the most “interesting” breeders for the intended pur-
pose and mates them with an inbred partner of the background strain, then nature
does the rest. In this context, the length of pregnancy and the time to reach sexual
maturity are the only limits in the progress towards full congenicity. However, one
can substantially accelerate the production of congenic strains by combining the
efforts of geneticists and those of embryologists. One can choose, for example,
3-week-old females as heterozygous (carriers) breeders, superovulate them, collect
their oocytes and perform in vitro fertilization with sperm from the background strain
(as discussed in Chap. 2). The fertilized eggs (zygotes) can then be implanted into
(a) Congenic Strain A-B
D7Mit86
D7Mit58
D7Mit25
D7Mi327
Targeted locus
(b)
F1
N2
N4
N6
Fig. 9.7 Congenic strains 2. a After each backcross generation, 50 % of the genomic DNA of
the donor strain (black chromosomes), on average, is replaced by the equivalent proportion of the
genomic DNA of the background strain (grey chromosomes). With an appropriate genotyping
assay, one can quantify the percentage of loci that are still heterozygous versus those that have
become homozygous in the offspring of the backcrossed progeny (i.e., the mice that exhibit the
lowest percentage of heterozygosity—boxed in the picture). Systematically selecting the breeders
for the next (N + 1) generation among those with the lowest possible number of heterozygous
loci is advantageous and speeds up the establishment of a congenic strain. The strategy can be
used with any species and any markers. This is often called marker-assisted selection (MAS). b
The chromosomal segments flanking the targeted allele are irrelevant and may generate difficulties
in the interpretation of some experimental results. Genotyping with molecular markers allows the
quality of the congenic strains to be increased by reducing the amount of irrelevant flanking DNA.
For this, it is sufficient to retain as breeders the rare offspring with a recombination event between
closely flanking markers and the targeted locus, as indicated in the figure. This selection can be
perfectly applied after the two first backcross generations. A congenic strain with flanking regions
of the “donor type” smaller than 1 cM is of top quality. The example shows microsatellite markers
(polymorphic between the parental strains) flanking the gene (locus) of interest on chromosome 7
N2 (15 %) N3 (3 %) N4 (<1 %)
60 60 60
Number of mice
50 50 50
40 40 40
30 30 30
20 20 20
10 10 10
0.3 0.5 0.7 0.05 0.15 0.25 0.01 0.03 0.05
Heterozygosity
Breeders to be intercrossed
Fig. 9.8 Speed congenics. Selecting the breeder with the lowest percentage of introgressed
(donor) DNA at each backcross generation requires the use of a great number of markers during
the first generations of the breeding program. However, it is important to note that once a marker
is typed “homozygous”, it is no longer necessary to type it in the forthcoming generations. The
bench work (genotyping) is then progressively reduced (from Wakeland et al. 1997)
pseudo-pregnant females and, when these females deliver their progeny, one can pro-
ceed with another round of selection with molecular markers. With an efficient proto-
col, the time to implement a new backcross generation can be reduced to 7–8 weeks,
and a new congenic strain can then be established in no more than 10 months (super-
speed congenics). In this regard, Japanese scientists have established a new record by
injecting round spermatid nuclei from immature males (only 17 days old) into mature
oocytes in vitro. With this technique called ROSI (for ROund Spermatid Injection),
they were able to develop a full-congenic strain (N3 mice genotyped with 86 DNA
markers) in only 106 days, a true high-speed congenic strategy (Ogonuki et al. 2009).
9.4.3.2 Congenic Strains and the Influence of Genetic Background
It is increasingly recognized that the genetic background (i.e., all genomic sequences
other than the gene of interest) can influence the phenotype of an animal affected by
a mutation. It has been shown that mutations (spontaneous and induced), transgenes,
and targeted alleles (knock-outs and knock-ins) that are “moved” (introgressed) into
a different background can exhibit a change in phenotype (Linder 2001; Doetschman
2009). This is mainly the result of the effect of several modifier genes. One of the
first cases involved the classical diabetes (Leprdb) mutation that presented transient
diabetes in a C57BL/6 background but overt diabetes in C57BLKS (Hummel et al.
1972). Other examples include background effects on survival rate in Egfr– (epider-
mal growth factor receptor) knockout mice (Threadgill et al. 1995) and effects on
tumor incidence and spectrum in Trp53 and Pten knockout mice (Kuperwasser et al.
2000; Freeman et al. 2006), to name only a few. In order to avoid confounding or
unreliable experimental results, particularly with the increasing number of mouse
strains, attention to genetic background is crucial (Banbury 1997; Linder 2001).
A Genetic Background Resource Manual by The Jackson Laboratory is freely
available at: https://secureweb.jax.org/jaxmice/literature/geneticBackground.html.
This 12-page booklet contains a series of examples where the genetic background
has been misleading and explains how to take this into account in experiments
involving mice. We strongly recommend it.
9.4.3.3 Congenic Strains and the Genetic Determinism

of Complex Traits
As we will discuss in Chap. 10, congenic strains have been extensively used since the
early days of mouse genetics and still are. They are particularly suited for the genetic
analysis of phenotypes that are controlled by several genes, and it is precisely by devel-
oping such strains that George D. Snell and his colleagues from The Jackson
Laboratory could elucidate the genetic determinism of histocompatibility (Snell 1948).3
As we already mentioned at the beginning of the present chapter, tissue trans-
plantations performed between mice belonging to unrelated populations—for
example, mice from two different inbred strains—are rejected. On the other
hand, the same transplantations performed between any two mice of the same
inbred strains (and the same sex) are permanently accepted. The problem is that,
in the case of tissue transplantations, the rejection, which is the observed phe-
notype, is controlled by several loci, each of them independently triggering the
same phenotype. To clarify the situation, Snell bred a series of strains with the
same C57BL/10 genetic background, but congenic for a single Mendelian unit
inducing tissue incompatibility. To simplify the analysis of the phenotype and to
save time, Snell injected tumor cells into mice segregating for the histocompati-
bility gene (all symbolized by H). At each generation, only the mice that survived
were “selected” and accordingly were “resistant” to the (tumoral) tissue trans-
plantation. He called these congenic mice congenic-resistant (CR) and developed
a very clever protocol to characterize each of these strains, thus avoiding duplica-
tions (CR strains congenic for the same H locus just by chance). By doing this,
Snell succeeded in making an inventory of many of the H loci segregating among
the laboratory strains. This strategy could be adapted with almost no change to
the genetic analysis of any trait that is under polygenic control; for example,
resistance to infectious diseases. When a congenic strain has been established,
there is still a lot of work to do to finally characterize the gene involved in the
3 G.D. Snell, J. Dausset and B. Benacerraf were awarded the Nobel Prize in 1980 “for their
discoveries concerning genetically determined structures on the cell surface that regulate immu-
nological reactions”.
Fig. 9.9 Reciprocal Reciprocal Congenic Strains A-B

congenics. Reciprocal
congenic strains allow for
comparisons to be made
with a high degree of
standardization because
epistatic interactions may be
controlled by this procedure.
If, for example, a given
allele in the background of
a congenic strain interferes
with the expression of the
introgressed gene, this might
Strain A Strain B
be detected when comparing
the two reciprocal congenic
strains or the F1 of these
congenic strains with the
parental inbred strain, as
indicated in the figure
Strain A-B Strain B-A
(BxA-B)F1 (AxB-A)F1
phenotype, given that the chromosomal fragment is sometimes quite large.

Nevertheless, congenic strains are certainly of great help in these investigations.
A few more comments are necessary to complete our description of the con-
genic strains. The first refers to the fact that a pair of congenic strains can be per-
fectly established even if the donor strain is not inbred. For example, if a mouse is
identified with an interesting characteristic segregating in a non-inbred population,
it is possible to derive one or more strains congenic for this trait, following the
same protocol described above.
Another interesting possibility is to develop reciprocal congenic strains by
introgressing a specific locus of strain A into the background strain B and, recipro-
cally, the homologous locus of strain B into the background strain A. At the end
of the experiment, one has a total of four strains: the two parental inbred strains A
and B on the one hand, and the reciprocal congenic strains AB and BA on the other.
One can then compare the F1 between strain A and the congenic strain BA with
the reciprocal F1 hybrid between strain B and the congenic strain AB. This type of
experiment, making use of F1, has the advantage of eliminating the side effects of
possible epistatic interactions with the genetic background and is likely to provide
more reliable answers (Fig. 9.9).
Finally, a comment is warranted on the use of congenic strains as tools for the
analysis of quantitative (or complex) traits. When we discussed the experiments
by Snell regarding the genetic analysis of histocompatibility, we mentioned that
the derivation of CR strains made possible the individual identification of several
H loci. Of course, this identification exclusively concerns the genes that are in
a different allelic form in the congenic partners; those that are non-polymorphic
remain undetected. This may appear to be a truism, but keeping in mind that the
classical inbred strains of laboratory mice were all derived from a small pool of
ancestral progenitors, it is clear that the experiments by Snell made possible the
discovery of only a small proportion of all the H genes of the mouse species.
Many other loci remained undetected, and it is likely that the derivation of new
CR strains from wild mouse specimens would certainly be very rewarding. This
comment applies, of course, to all situations where many genes (and many alleles)
are involved in the determinism of a complex or quantitative trait (See Chap. 10).
9.5 Consomic Strains
Consomic strains, also called chromosome substitution strains (CSS), are a variation
of the congenic strains concept in which the introgressed DNA is a complete chro-
mosome, rather than a piece of chromosome flanking a given gene (Nadeau et al.
2000). These strains have been very useful for the rapid mapping of phenotypic
traits to a specific chromosome. They are also useful for the detection of chromo-
somal regions (the so-called quantitative traits loci, QTLs) having an influence in
the determinism of a particular phenotype (for example, the resistance to or suscep-
tibility for carcinogenesis). This point will be explained in some detail in Chap. 10.
Only a few sets of consomic mouse strains are available, but it is likely that other
sets will be developed in the future to accompany the development of investigations
in multifactorial inheritance (Gregorova et al. 2008; Mattson et al. 2008) (Fig. 9.10).
Using a marker-assisted protocol, consomic strains are easy to produce. However,
one must keep in mind that tiny pieces of chromosomes of the donor strain might
escape the marker-assisted selection process if, by chance, they are not identi-
fied by a marker. In the same way, there is no guarantee that the telomeric region
of a given chromosome pair is transferred intact since there is, in most instances,
no distal marker to check this. Finally, and according to the available information,
attempts to develop a full set of inter-specific consomic mouse strains from distantly
related mouse species or subspecies (for example, Mus spretus as a donor strain and
Consomic Strain A-B
Strain A Strain B-Chr 4 A Strain B-Chr 3 A
Strain B-Chr 2 A Strain B-Chr 1 A Strain B
Fig. 9.10 Consomic strains. A consomic strain is an inbred strain in which one of the chromosome
pairs has been replaced by the homologous chromosome pair of another inbred strain after a series
of marker-assisted backcrosses. A complete panel of consomic strains consists of 21 strains, each
derived from the same donor and host strains but having each a different chromosome pair (Chr
1–19, X or Y) of the host strain replaced by its homolog from the donor. A reciprocal panel can be
produced by inverting the donor and host strains. One can never be sure that two strains are fully
consomic for the telomeric ends because telomeric markers are often missing
C57BL/6 as a background strain) have proved difficult or even unsuccessful, pre-

sumably because of deleterious epistatic interactions between genes (or alleles) that
have been separated by evolution for a long time (over 1.5 Myr). For example, some
hybrid sterility genes have been reported that will result in complete sterility of the F1
male offspring between Mus spretus (any strain) and laboratory strains (Forejt 1996).
9.6 Recombinant Inbred Strains and Recombinant

Congenic Strains
Recombinant inbred strains (RIS) are developed by crossing two parental inbred
strains to generate F1 hybrids and then intercrossing these F1s to generate F2s.
Finally, randomly chosen F2 animals are then brother × sister mated for 20 or
more generations to develop a group of related inbred strains (Bailey 1971). RIS
9.6 Recombinant Inbred Strains and Recombinant … 349
are grouped by sets (also referred to as panels): a collection of RIS derived from the
same parental strains. For example, the C57BL/6 × DBA/2 (BXD) is, at the moment,
the largest mouse RI panel with ~90 strains. These are true inbred strains, meaning
that they are homozygous at all loci but have the additional characteristic that each
RIS has a unique fixed combination of the parental alleles in a 50:50 proportion (on
average). For example, each strain of the set of 33 AXB-BXA strains, derived from
the initial cross of a C57BL/6 mouse with a A/J mouse, carries either the B6 allele or
the A allele at each locus of its genome; by typing all of these allelic forms, one can
establish a strain distribution pattern (SDP) for each of the strains, which lists the col-
lection of alleles inherited from either the parental strain A or the parental strain B6.
Of course, this SDP is fixed forever in each strain (not taking into account the rare
mutations that inevitably occur), and new data are constantly added to it, allowing
correlations to be made between genotypes and phenotypes simply by scanning, gen-
erally with the help of a simple computer program, the co-segregation of a new phe-
notype (or genotype) with the existing SDP. RIS have proved very helpful when used
for gene mapping, in particular for the rapid regional assignment of microsatellites
on a given chromosome, when these markers were cloned by the thousands for the
establishment of high-density genetic maps (see Chap. 4). They have also been used
for the mapping of chromosomal regions (QTLs) involved in the genetic determinism
of some behavioral characteristics (for example, taster/non-taster for a chemical com-
pound, alcohol intake, etc.) or of some immunological responses, and they will very
likely still be of great help in many other experiments where the phenotype is meas-
ured on a group of animals rather than on individuals (Zou et al. 2005) (Fig. 9.11).
Recombinant congenic strains (RCS) are similar to RIS in their genomic structure
except that the proportion of the parental alleles in a given strain is not 50:50 but
75:25 or 87.5:12.5, depending on the set (Demant and Hart 1986). This is achieved
by inbreeding mice of the first or second backcross generation to one of the parental
inbred strains (the background strain). As we will explain in Chap. 10, RCS are help-
ful for identifying genes associated with polygenic inheritance, especially when the
number of genes is high. RCS with a small percentage of introgressed genome in
a background strain have a greater power of resolution, and their use increases the
likelihood of zero or only one single locus governing the studied phenotype (QTL)
being isolated in a given RCS. For example, RCS have been very helpful for unrave-
ling the genetic determinism of colon cancer in the mouse (Demant 2003).
Interspecific recombinant congenic strains (IRCS) have also been developed
from the parental strains C57BL/6JPas and SEG/Pas (Mus spretus) (Burgio et al.
2007). This set of strains has proved particularly useful for the analysis of the
genetic determinism of some anatomical traits (Burgio et al. 2009). The differences
between congenic strains and recombinant congenic strains is that, in the case of
congenic strains, the introgresssed region(s) is unique, with the smallest possible
size, and chosen a priori by the investigator, while there is in general more than
one region in the case of RCS, with these regions being of variable size and not
selected by the investigator. This being taken into account, and provided the strain
combination is appropriate it is clear that it may sometimes be advantageous to
choose a specific RCS as a donor strain for the development of a congenic strain.
Recombinant Inbred Strain AXB
Strain A Strain AXB1 Strain AXB2 Strain AXB5
Strain AXB8 Strain AXB9 Strain AXB15 Strain B
Fig. 9.11 Recombinant inbred strains. This diagram represents a panel of six recombinant

inbred strains (RIS) flanked by the parental strains A and B. These strains derive from the same
initial cross but each have a unique combination of loci derived by recombination of the alleles
present in the original parental strains. Since RIS are inbred and each strain has a unique geno-
type, RIS have a number of advantages over F2 or backcross mouse populations as tools for map-
ping genes or quantitative trait loci (QTLs)
9.7 The Collaborative Cross
The panels of RIS described in the preceding section represent a first-rate resource
for the identification and analysis of the genetic determinants of complex traits.
Since all the mice within a given strain have the same genotype, phenotyping
can be carried out on groups of varying sizes, yielding a phenotype that can be
expressed in terms of percentage with a confidence interval that is only depend-
ent on the size of the sample. Using RIS allows assessing the genetic determinism
of susceptibility to certain drugs, to certain forms of cancers, and to experimen-
tal infections with pathogens. These types of experiments would be difficult, if
not impossible, to achieve by the mere genetic analysis of F2 or backcross pop-
ulations. Another advantage of the RIS is to reduce the cost of the experiments.
Indeed, given that most of the existing strains are already genotyped for a large
number of genetic markers it is in general easy to detect co-segregation of one or
a few specific marker(s) with the data collected from phenotyping. Unfortunately,
because they are all derived from a handful of classical inbred strains, the dif-
ferent panels of available RIS display a relatively low level of genetic diversity
9.7 The Collaborative Cross 351
when compared to the diversity found in the Mus genus as a whole and this often
appears as a limitation in the use of RIS.
Considering these advantages and drawbacks in the use of the RIS panels stim-
ulated discussions among a group of researchers interested in quantitative genetics
(The Complex Trait Consortium), and these discussions led to the idea to develop
a new resource, better adapted to the analysis of complex traits. Nowadays, this
resource is being actively developed and it is known as Collaborative Cross (Fig.
9.12 a, b). The Collaborative Cross (CC) is an extension of the recombinant inbred
strain concept with however a much higher power of resolution and a much higher
level of genetic diversity (Churchill et al. 2004; Chesler et al. 2008; Threadgill
et al. 2011). The Collaborative Cross is derived from a panel of eight carefully
selected founder inbred strains that consist of: (i) three classical, traditional
inbred strains (A/J, C57BL/6J, 129S1/SvImJ); (ii) two inbred strains affected by
a genetically complex pathology (diabetes/obesity) NOD/LtJ, NZO); and (iii)
three inbred strains derived from wild progenitors of the three main subspecies
of the Mus genus (CAST/Ei derived from Mus m. castaneus; PWK/PhJ derived
from Mus m. musculus and WSB/Ei derived from Mus m. domesticus). The eight
founder strains were first crossed pairwise to generate all [= (8 x 7) / 2 = 28]
Fig. 9.12 The Collaborative (a)

Cross. a This is a randomized A B C D E F G H
cross of eight unrelated G0 X X X X
mouse inbred strains selected
by the members of the
Complex Trait Consortium. G1 X X
The strains are first crossed

pair-wise to make all G2 X
((8 x 7) / 2 = 28) possible

G1. A set of possible four- G3 X
way crosses is performed,

keeping Y-chromosome G4 X
and mitochondrial balance.
Finally, all eight genomes are
brought together in G2:F1, X
and the offspring of this cross G>20

are inbred. The Collaborative
Cross is a community
resource that was initially (b)
designed for the purpose A B C D E F G H
of mapping complex traits.
b The initial plans were to
breed around 1,000 inbred
strains where all the alleles
of the initial inbred strains
would be associated with a
wide and unique variety of
combinations. The illustration
presents the 19 autosomal
chromosomes plus X and Y
of a hypothetical RIS from
the Collaborative Cross
possible G1 hybrids, then a balanced subset of non-overlapping 8-way progeny

was selected and brother x sister mated for several (i.e. ≥ 20) generations to pro-
duce a very large set of RI lines with variable proportions of the genomes of the
parental strains. These RI lines will allow for the detection of biologically rel-
evant correlations between thousands of traits with an unprecedented power of
resolution. Such a panel of CC lines is currently being developed in three labo-
ratories: the University of North Carolina in Chapel Hill, the Tel Aviv University,
and the University of Western Australia in Perth. For most of these CC lines the
progress of inbreeding is monitored by PCR genotyping of the SNPs segregat-
ing among the parental strains. Nowadays a few tens of lines, displaying > 90 %
homozygosity, have already been made available to the community in particu-
lar to some “Mouse Clinics”, for extensive phenotyping. For example, a com-
prehensive and comparative phenotypic analysis is under development at the
German Mouse Clinic (Helmholtz Zentrum München-in Neuherberg-http://www.
mouseclinic.de/), with more than 500 parameters being tested including param-
eters that characterize allergy, behavior, blood chemistry, bone and cartilage struc-
ture, energy metabolism, eye and vision, immunology, lung function, neurology,
nociception, and pathology. An example of the power of the CC lines for the
analysis of complex traits was reported in the case of experimental infections with
the Ebola virus (Rasmussen et al. 2014). Unexpectedly, while none of the classi-
cal laboratory strains display the whole range of symptoms commonly associated
with human Ebola haemorrhagic fever, after experimental infection, strains from
the Collaborative Cross exhibited a variety of phenotypes ranging from complete
resistance to a severe (lethal) haemorrhagic syndrome. These observations indicate
that the genetic background strongly determines susceptibility of mice to Ebola
hemorrhagic fever and this opens avenues for the development of a better animal
model for the study of human infection.
At its conception the project of the Complex Trait Consortium, was to generate
1,000 RI lines recombinant for variable genomic proportions of the eight parental
strains enabling detection of biologically relevant correlations between thousands
of measured traits with an unprecedented power of resolution (135,000 recombina-
tion events!). Practical concerns related to the facilities needed to host all these
strains, their distribution, their health status and the difficulties to raise funding to
preserve them have led researchers to consider some reduction in the generation of
new CC lines. At the end of 2014 around 100 strains were available for research
purpose. These CC strains are all, at least, 90 % homozygous and derive from at
least 6 and in many cases from all 8 founder strains. In parallel to the breeding of
these strains a genotyping project involving more than 77,000 maximally inform-
ative SNPs is under development. This project, known as the MegaMUGA (for
Mega Mouse Universal Genotyping Array), is to cover the genome of every one
of the CC lines with a very high density of SNP markers (average spacing of 33
kb). Even if it may take some time before the Collaborative Cross project is com-
pleted, it is now a reality with multiple applications. For more information con-
cerning Mega MUGA refer to: http://csbio.unc.edu/CCstatus/Media/MegaMUGA
Flyer.pdf.
9.8 Outbred and Random-Bred Stocks 353
9.8 Outbred and Random-Bred Stocks
Outbred and random-bred stocks are populations of laboratory animals that are radi-
cally different from those we considered above in the sense that they are genetically
heterogeneous, or heterogenic as we might say to keep the same sort of terminol-
ogy. According to the official definition, outbred mouse stocks are “closed popula-
tions (for at least four generations) of genetically variable animals that are bred to
maintain maximum heterozygosity”. Compared with inbred strains, F1 hybrids, or
congenic strains, the genetic constitution of a given animal, taken randomly from an
outbred stock, is not known a priori and must be defined when necessary.
Outbred mice represent the bulk of laboratory animals sold by commercial ven-
dors for the purpose of experimentation. These animals are usually bred according
to a system that minimizes (or, more exactly, reduces) inbreeding, and accord-
ingly contributes to the maintenance of a certain amount of heterozygosity in the
population (Hartl 2001). A classical breeding scheme for these populations would
consist, for example, of the mating in room C and D of n males originating from
room A with the equivalent number of females taken from room B, with n being as
great as possible. For the production of the next generation (G + 1), the breeding
scheme would be similar with n males from room C being mated with n females
of room D, and so on. Doing this, generation after generation, the polymorphic
alleles that were segregating in the population at generation G have the greatest
chance of still being represented at generation G + 1 in roughly the same propor-
tion. The greater the samples of breeders used for the production of G + 1, the
smaller the variations in frequency at each generation (Poiley 1960).
The degree of genetic heterogeneity in outbred colonies depends greatly on their
history. It can be very low, for example, as a consequence of genetic drift (or the bottle-
neck effect), when the pool of breeders has been accidentally or intentionally reduced
to a few individuals (this is common when a new breeding facility is created and a
small group of breeders is imported). In contrast, genetic heterogeneity can be much
higher when the stock has been recently outcrossed. Some commercial breeders prob-
ably monitor the polymorphisms segregating in their stocks with DNA markers, but
the methodology they use and the results they get are not always made public. Being
genetically heterogeneous, outbred and randombred stocks have a greater fertility
index than inbred strains and, accordingly, they are sold at a much lower price per unit.
Because outbred colonies are heterogeneous populations, like human populations,
they are often considered as being the most appropriate category of laboratory ani-
mals to use in toxicology, and pharmacology research. However, several geneticists
have disputed this point of view and it has even been considered that, in many stud-
ies, outbred mice were used inappropriately, wasting animals’ lives and resources on
suboptimal experiments (Chia et al. 2005; Festing 2010). In fact, any outbred stock
can be replaced by a “synthetic” population obtained by intercrossing classical inbred
strains. As we already said, crossing two inbred strains to produce an F1 progeny
and then crossing two independent F1 generates a four-way polymorphic population.
This population is heterogenic, in the sense that individuals are genetically different.
In addition, the population often carries a greater number of allelic forms, which is
generally considered an advantage compared to a classical outbred population.
Recently, however, researchers have considered that outbred stocks might be
useful to refine the identification of QTLs, because these heterogeneous stocks
accumulate in their genome many recombination breakpoints over time that split
their chromosomes into “fine-grained mosaics”, facilitating the high-resolution
mapping of complex traits (Mott et al. 2000; Flint et al. 2005; Yalcin et al. 2010).
Finally, random-bred stocks are of very limited interest to geneticists. These
stocks are bred with no specific rules, paying almost no attention to the genetic
diversity in the population. Since they are in general of relatively small size, they
drift rapidly towards a moderately inbred but still undefined population.
References
Aylor DL, Valdar W, Foulds-Mathes W, Buus RJ, Verdugo RA, Baric RS, Ferris MT, Frelinger
JA, Heise M, Frieman MB, Gralinski LE, Bell TA, Didion JD, Hua K, Nehrenberg DL,
Powell CL, Steigerwalt J, Xie Y, Kelada SN, Collins FS, Yang IV, Schwartz DA, Branstetter
LA, Chesler EJ, Miller DR, Spence J, Liu EY, McMillan L, Sarkar A, Wang J, Wang W,
Zhang Q, Broman KW, Korstanje R, Durrant C, Mott R, Iraqi FA, Pomp D, Threadgill
D, de Villena FP, Churchill GA (2011) Genetic analysis of complex traits in the emerging
Collaborative Cross. Genome Res 21:1213–1222
Bailey DW (1971) Recombinant-inbred strains. An aid to finding identity, linkage, and function
of histocompatibility and other genes. Transplantation 11:325–327
Banbury (1997) Mutant mice and neuroscience: recommendations concerning genetic back-
ground: banbury conference on genetic background in mice. Neuron 19:755–759
Beck JA, Lloyd S, Hafezparast M, Lennon-Pierce M, Eppig JT, Festing MF, Fisher EM (2000)
Genealogies of mouse inbred strains. Nat Genet 24:23–25
Benavides FJ (1999) Genetic contamination of an SJL/J mouse colony: rapid detection by PCR-
based microsatellite analysis. Contemp Top Lab Anim Sci 38:54–55
Berning AK, Eicher EM, Paul WE, Scher I (1980) Mapping of the X-linked immune deficiency
mutation (xid) of CBA/N mice. J Immunol 124:1875–1877
Bishop CE, Boursot P, Baron B, Bonhomme F, Hatat D (1985) Most classical Mus musculus domes-
ticus laboratory mouse strains carry a Mus musculus musculus Y chromosome. Nature 315:70–72
Bonhomme F (1986) Evolutionary relationships in the genus Mus. Curr Top Microbiol Immunol
127:19–34
Bonhomme F, Guénet JL (1996) The laboratory mouse and its wild relatives. In: Lyon M, Rastan
S, Brown DM (ed) Genetic variants and strains of the laboratory mouse. Oxford University
Press, New York. pp 1577–1596
Bonhomme F, Guénet JL, Dod B, Moriwaki K, Bulfield G (1987) The polyphyletic origin of
laboratory inbred mice and their rate of evolution. Biol J Linn Soc 30:51–58
Bryda EC, Riley LK (2008) Multiplex microsatellite marker panels for genetic monitoring of
common rat strains. J Am Assoc Lab Anim Sci 47:37–41
Bulfield G, Siller WG, Wight PA, Moore KJ (1984) X chromosome-linked muscular dystrophy
(mdx) in the mouse. Proc Natl Acad Sci USA 81:1189–1192
Burgio G, Baylac M, Heyer E, Montagutelli X (2009) Genetic analysis of skull shape variation
and morphological integration in the mouse using interspecific recombinant congenic strains
between C57BL/6 and mice of the mus spretus species. Evolution 63:2668–2686
References 355
Charlesworth D, Willis JH (2009) The genetics of inbreeding depression. Nat Rev Genet 10:783–796
Chen S, Kadomatsu K, Kondo M, Toyama Y, Toshimori K, Ueno S, Miyake Y, Muramatsu
T (2004) Effects of flanking genes on the phenotypes of mice deficient in basigin/CD147.
Biochem Biophys Res Commun 324:147–153
Chesler EJ, Miller DR, Branstetter LR, Galloway LD, Jackson BL, Philip VM, Voy BH, Culiat
CT, Threadgill DW, Williams RW, Churchill GA, Johnson DK, Manly KF (2008) The col-
laborative cross at Oak ridge national laboratory: developing a powerful resource for systems
genetics. Mamm Genome 19:382–389
Chia R, Achilli F, Festing MF, Fisher EMC (2005) The origins and uses of mouse outbred stocks.
Nat Genet 37:1181–1186
Church DM, Goodstadt L, Hillier LW, Zody MC, Goldstein S, She X, Bult CJ, Agarwala R,
Cherry JL, DiCuccio M, Hlavina W, Kapustin Y, Meric P, Maglott D, Birtle Z, Marques AC,
Graves T, Zhou S, Teague B, Potamousis K, Churas C, Place M, Herschleb J, Runnheim R,
Forrest D, Amos-Landgraf J, Schwartz DC, Cheng Z, Lindblad-Toh K, Eichler EE, Ponting
CP (2009) Lineage-specific biology revealed by a finished genome assembly of the mouse.
PLoS Biol 7:e1000112
Churchill GA, Airey DC, Allayee H, Angel JM, Attie AD, Beatty J, Beavis WD, Belknap JK,
Bennett B, Berrettini W, Bleich A, Bogue M, Broman KW, Buck KJ, Buckler E, Burmeister M,
Chesler EJ, Cheverud JM, Clapcote S, Cook MN, Cox RD, Crabbe JC, Crusio WE, Darvasi A,
Deschepper CF, Doerge RW, Farber CR, Forejt J, Gaile D, Garlow SJ, Geiger H, Gershenfeld
H, Gordon T, Gu J, Gu W, de Haan G, Hayes NL, Heller C, Himmelbauer H, Hitzemann
R, Hunter K, Hsu HC, Iraqi FA, Ivandic B, Jacob HJ, Jansen RC, Jepsen KJ, Johnson DK,
Johnson TE, Kempermann G, Kendziorski C, Kotb M, Kooy RF, Llamas B, Lammert F,
Lassalle JM, Lowenstein PR, Lu L, Lusis A, Manly KF, Marcucio R, Matthews D, Medrano JF,
Miller DR, Mittleman G, Mock BA, Mogil JS, Montagutelli X, Morahan G, Morris DG, Mott
R, Nadeau JH, Nagase H, Nowakowski RS, O’Hara BF, Osadchuk AV, Page GP, Paigen B,
Paigen K, Palmer AA, Pan HJ, Peltonen-Paloti L, Peirce J, Pomp D, Pravenec M, Prows DR, Qi
Z, Reeves RH, Roder J, Rosen GD, Schadt EE, Schalkwyk LC, Seltzer Z, Shimomura K, Shou
S, Sillanpaa MJ, Siracusa LD, Snoeck HW, Spearow JL, Svenson K, Tarantino LM, Threadgill
D, Toth LA, Valdar W, de Villena FP, Warden C, Whatley S, Williams RW, Wiltshire T, Yi N,
Zhang D, Zhang M, Zou F, The Complex Trait Consortium (2004) The collaborative cross: a
community resource for the genetic analysis of complex traits. Nat Genet 36:1133–1137
Coleman DL, Hummel KP (1973) The influence of genetic background on the expression of the
obese (Ob) gene in the mouse. Diabetologia 9:287–293
Davisson MT (1996) Rules for nomenclature of inbred strains. In: Lyon MF, Rastan S, Brown
SDM (eds) Genetic variants and strains of the laboratory mouse. Oxford University Press,
Oxford, pp 1532–1536
Dejager L, Libert C, Montagutelli X (2009) Thirty years of Mus spretus: a promising future.
Demant P (2003) Cancer susceptibility in the mouse: genetics, biology and implications for
human cancer. Nat Rev Genet 4:721–734
Demant P, Hart AA (1986) Recombinant congenic strains–a new tool for analyzing genetic traits
determined by more than one gene. Immunogenetics 24:416–422
Doetschman T (2009) Influence of genetic background on genetically engineered mouse pheno-
types. Methods Mol Biol 530:423–433
Ferris SD, Sage RD, Wilson AC (1982) Evidence from mtDNA sequences that common labora-
tory strains of inbred mice are descended from a single female. Nature 295:163–165
Festing MF (1979) Inbred strains in biomedical research. Macmillan, London
Festing MF (2010) Inbred strains should replace outbred stocks in toxicology, safety testing, and
drug development. Toxicol Pathol 38:681–690
Flint J, Valdar W, Shifman S, Mott R (2005) Strategies for mapping and cloning quantitative trait
genes in rodents. Nat Rev Genet 4:271–286
Forejt J (1996) Hybrid sterility in the mouse. Trends Genet 12:412–417
Frazer KA, Eskin E, Kang HM, Bogue MA, Hinds DA, Beilharz EJ, Gupta RV, Montgomery
J, Morenzoni MM, Nilsen GB, Pethiyagoda CL, Stuve LL, Johnson FM, Daly MJ, Wade
CM, Cox DR (2007) A sequence-based variation map of 8.27 million SNPs in inbred mouse
strains. Nature 448:1050–1053
Freeman D, Lesche R, Kertesz N, Wang S, Li G, Gao J, Groszer M, Martinez-Diaz H, Rozengurt
N, Thomas G, Liu X, Wu H (2006) Genetic background controls tumor development in
PTEN-deficient mice. Cancer Res 66:6492–6496
Glenister PH, Thornton CE (2000) Cryoconservation–archiving for the future. Mamm Genome
11:565–571
Goios A, Pereira L, Bogue M, Macaulay V, Amorim A (2007) mtDNA phylogeny and evolution
of laboratory mouse strains. Genome Res 17:293–298
Gregorova S, Divina P, Storchova R, Trachtulec Z, Fotopulosova V, Svenson KL, Donahue LR,
Paigen B, Forejt J (2008) Mouse consomic strains: exploiting genetic divergence between
Mus m. musculus and Mus m. domesticus subspecies. Genome Res 18:509–515
Grüneberg H (1952) The genetics of the mouse, 2nd edn. Martinus Nijhoff, The Hague
Guénet JL, Bonhomme F (2003) Wild mice: an ever-increasing contribution to a popular mam-
malian model. Trends Genet 19:24–31
Hartl DL (2001) Genetic management of outbred laboratory rodent populations. Charles River
Genetic Literature
Hummel KP, Coleman DL, Lane PW (1972) The influence of genetic background on expres-
sion of mutations at the diabetes locus in the mouse. I. C57BL-KsJ and C57BL-6J strains.
Biochem Genet 7:1–13
Johnson LL (1981) At how many histocompatibility loci do congenic mouse strains differ? J
Hered 72:27–31
Kenneth NS, Younger JM, Hughes ED, Marcotte D, Barker PA, Saunders TL, Duckett CS (2012)
An inactivating caspase 11 passenger mutation originating from the 129 murine strain in
mice targeted for c-IAP1. Biochem J 443:355–359
Kuperwasser C, Hurlbut GD, Kittrell FS, Dickinson ES, Laucirica R, Medina D, Naber SP, Jerry
DJ (2000) Development of spontaneous mammary tumors in BALB/c p53 heterozygous
mice. A model for Li-Fraumeni syndrome. Am J Pathol 157:2151–2159
Linder CC (2001) The influence of genetic background on spontaneous and genetically engi-
neered mouse models of complex diseases. Lab Anim (NY) 30:34–39
Mao HZ, Roussos ET, Peterfy M (2006) Genetic analysis of the diabetes-prone C57BLKS/J
mouse strain reveals genetic contribution from multiple strains. Biochim Biophys Acta
1762:440–446
Markel P, Shu P, Ebeling C, Carlson GA, Nagle DL, Smutko JS, Moore KJ (1997) Theoretical and
empirical issues for marker-assisted breeding of congenic mouse strains. Nat Genet 17:280–284
Mashimo T, Voigt B, Tsurumi T, Naoi K, Nakanishi S, Yamasaki K, Kuramoto T, Serikawa T
(2006) A set of highly informative rat simple sequence length polymorphism (SSLP) markers
and genetically defined rat strains. BMC Genet 7:19
Mattapallil MJ, Wawrousek EF, Chan CC, Zhao H, Roychoudhury J, Ferguson TA, Caspi RR
(2012) The Rd8 mutation of the Crb1 gene is present in vendor lines of C57BL/6 N mice and
embryonic stem cells, and confounds ocular induced mutant phenotypes. Invest Ophthalmol
Vis Sci 53:2921–2927
Mattson DL, Dwinell MR, Greene AS, Kwitek AE, Roman RJ, Jacob HJ, Cowley AW Jr (2008)
Chromosome substitution reveals the genetic basis of Dahl salt-sensitive hypertension and
renal disease. Am J Physiol Renal Physiol 295:837–842
Mekada K, Abe K, Murakami A, Nakamura S, Nakata H, Moriwaki K, Obata Y, Yoshiki A (2009)
Genetic differences among C57BL/6 substrains. Exp Anim 58:141–149
Moran N, Bassani DM, Desvergne JP, Keiper S, Lowden PA, Vyle JS, Tucker JH (2006)
Detection of a single DNA base-pair mismatch using an anthracene-tagged fluorescent probe.
Chem Commun 48:5003–5005
References 357
Moriwaki K, Shiroishi T, Yonekowa H (1994) Genetics in wild mice: its application to biomedi-
cal research. Japan Scientific Societies Press
Morse HC III (1978) Origins of inbred mice. Academic Press, New York
Mott R, Talbot CJ, Turri MG, Collins AC, Flint J (2000) A method for fine mapping quantitative
trait loci in outbred animal stocks. Proc Natl Acad Sci USA 97:12649–12654
Myakishev MV, Khripin Y, Hu S, Hamer DH (2001) High-throughput SNP genotyping by allele-
specific PCR with universal energy-transfer-labeled primers. Genome Res 11:163–169
Nadeau JH, Singer JB, Matin A, Lander ES (2000) Analysing complex genetic traits with chro-
mosome substitution strains. Nat Genet 24:221–225
Nijman IJ, Kuipers S, Verheul M, Guryev V, Cuppen E (2008) A genome-wide SNP panel for
mapping and association studies in the rat. BMC Genom 9:95
Ogonuki N, Inoue K, Hirose M, Miura I, Mochida K, Sato T, Mise N, Mekada K, Yoshiki A, Abe
K, Kurihara H, Wakana S, Ogura A (2009) A high-speed congenic strategy using first-wave
male germ cells. PLoS ONE 4:e4943
Paigen K, Eppig JT (2000) A mouse phenome project. Mamm Genome 11:715–717
Petkov PM, Cassell MA, Sargent EE, Donnelly CJ, Robinson P, Crew V, Asquith S, Haar RV,
Wiles MV (2004a) Development of a SNP genotyping panel for genetic monitoring of the
laboratory mouse. Genomics 83:902–911
KA, Robinson P, Scott VE, Wiles MV (2004b) An efficient SNP system for mouse genome
scanning and elucidating strain relationships. Genome Res 14:1806–1811
Petkov PM, Graber JH, Churchill GA, DiPetrillo K, King BL, Paigen K (2005) Evidence of a
large-scale functional organization of mammalian chromosomes. PLoS Genet 1(3):e33
Poiley SM (1960) A systematic method of breeder rotation for non-inbred laboratory animals
colonies. Proc Anim Care Panel 10:159
Poltorak A, He X, Smirnova I, Liu MY, Van Huffel C, Du X, Birdwell D, Alejos E, Silva M, Galanos
C, Freudenberg M, Ricciardi-Castagnoli P, Layton B, Beutler B (1998) Defective LPS signaling
in C3H/HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene. Science 282:2085–2088
Rader K (2004) Making mice: standardizing animals for American biomedical research, 1900–
1955. Princeton University Press, New Jersey
Rasmussen AL, Okumura A, Ferris MT, Green R, Feldmann F, Kelly SM, Scott DP, Safronetz
D, Haddock E, LaCasse R, Thomas MJ, Sova P, Carter VS, Weiss JM, Miller DR, Shaw GD,
Korth MJ, Heise MT, Baric RS, Manuel de Villena FP, Feldmann H, Katze MG (2014) Host
genetic diversity enables Ebola hemorrhagic fever pathogenesis and resistance. Science. pii:
1259595. [Epub ahead of print]
Schlager G, Dickie MM (1967) Spontaneous mutations and mutation rates in the house mouse.
Schuster-Gossler K, Lee AW, Lerner CP, Parker HJ, Dyer VW, Scott VE, Gossler A, Conover JC
(2001) Use of coisogenic host blastocysts for efficient establishment of germline chimeras
with C57BL/6 J ES cell lines. Biotechniques 31:1022–1026
Simon MM, Greenaway S, White JK, Fuchs H, Gailus-Durner V, Wells S, Sorg T, Wong K, Bedu
E, Cartwright EJ, Dacquin R, Djebali S, Estabel J, Graw J, Ingham NJ, Jackson IJ, Lengeling
A, Mandillo S, Marvel J, Meziane H, Preitner F, Puk O, Roux M, Adams DJ, Atkins S, Ayadi
A, Becker L, Blake A, Brooker D, Cater H, Champy MF, Combe R, Danecek P, di Fenza
A, Gates H, Gerdin AK, Golini E, Hancock JM, Hans W, Hölter SM, Hough T, Jurdic P,
Keane TM, Morgan H, Müller W, Neff F, Nicholson G, Pasche B, Roberson LA, Rozman
J, Sanderson M, Santos L, Selloum M, Shannon C, Southwell A, Tocchini-Valentini GP,
Vancollie VE, Westerberg H, Wurst W, Zi M, Yalcin B, Ramirez-Solis R, Steel KP, Mallon
AM, de Angelis MH, Herault Y, Brown SD (2013) A comparative phenotypic and genomic
analysis of C57BL/6 J and C57BL/6 N mouse strains. Genome Biol 14(7):R82
Snell GD (1948) Methods for the study of histocompatibility genes. J Genet 49:87–108
Specht CG, Schoepfer R (2001) Deletion of the alpha-synuclein locus in a subpopulation of
C57BL/6 J inbred mice. BMC Neurosci 2:11
Stevens JC, Banks GT, Festing MF, Fisher EM (2007) Quiet mutations in inbred strains of mice.
Trends Mol Med 13:512–519
Strong LC (1978) Inbred mice in science in origins of inbred mice. In: Morse III HC (ed)
Academic Press—Adapted for the Web by: mouse genome informatics. The Jackson
Laboratory, Bar Harbor, Maine USA
Threadgill DW, Churchill GA (2012) Ten years of the collaborative cross. Genetics 190:291–294
Threadgill DW, Miller DR, Churchill GA, de Villena FP (2011) The collaborative cross: a recom-
binant inbred mouse population for the systems genetic era. ILAR J 52:24–31
Threadgill DW, Dlugosz AA, Hansen LA, Tennenbaum T, Lichti U, Yee D, LaMantia C, Mourton
T, Herrup K, Harris RC et al (1995) Targeted disruption of mouse EGF receptor: effect of
genetic background on mutant phenotype. Science 269:230–234
Tucker PK, Phillips KS, Lundrigan B (1992) A mouse Y chromosome pseudogene is related to
human ubiquitin activating enzyme E1. Mamm Genome 3:28–35
Wade CM, Kulbokas EJ 3rd, Kirby AW, Zody MC, Mullikin JC, Lander ES, Lindblad-Toh K, Daly
MJ (2002) The mosaic structure of variation in the laboratory mouse genome. Nature 420:574–578
Wakeland E, Morel L, Achey K, Yui M, Longmate J (1997) Speed congenics: a classic technique
in the fast lane (relatively speaking). Immunol Today 18:472–477
Waterston RH, Lindblad-Toh K, Birney E, Rogers J, Abril JF, Agarwal P, Agarwala R, Ainscough
R, Alexandersson M, An P, Antonarakis SE, Attwood J, Baertsch R, Bailey J, Barlow K, Beck
S, Berry E, Birren B, Bloom T, Bork P, Botcherby M, Bray N, Brent MR, Brown DG, Brown
SD, Bult C, Burton J, Butler J, Campbell RD, Carninci P, Cawley S, Chiaromonte F, Chinwalla
AT, Church DM, Clamp M, Clee C, Collins FS, Cook LL, Copley RR, Coulson A, Couronne
O, Cuff J, Curwen V, Cutts T, Daly M, David R, Davies J, Delehaunty KD, Deri J, Dermitzakis
ET, Dewey C, Dickens NJ, Diekhans M, Dodge S, Dubchak I, Dunn DM, Eddy SR, Elnitski L,
Emes RD, Eswara P, Eyras E, Felsenfeld A, Fewell GA, Flicek P, Foley K, Frankel WN, Fulton
LA, Fulton RS, Furey TS, Gage D, Gibbs RA, Glusman G, Gnerre S, Goldman N, Goodstadt
L, Grafham D, Graves TA, Green ED, Gregory S, Guigó R, Guyer M, Hardison RC, Haussler
D, Hayashizaki Y, Hillier LW, Hinrichs A, Hlavina W, Holzer T, Hsu F, Hua A, Hubbard T,
Hunt A, Jackson I, Jaffe DB, Johnson LS, Jones M, Jones TA, Joy A, Kamal M, Karlsson EK,
Karolchik D, Kasprzyk A, Kawai J, Keibler E, Kells C, Kent WJ, Kirby A, Kolbe DL, Korf I,
Kucherlapati RS, Kulbokas EJ, Kulp D, Landers T, Leger JP, Leonard S, Letunic I, Levine R,
Li J, Li M, Lloyd C, Lucas S, Ma B, Maglott DR, Mardis ER, Matthews L, Mauceli E, Mayer
JH, McCarthy M, McCombie WR, McLaren S, McLay K, McPherson JD, Meldrim J, Meredith
B, Mesirov JP, Miller W, Miner TL, Mongin E, Montgomery KT, Morgan M, Mott R, Mullikin
JC, Muzny DM, Nash WE, Nelson JO, Nhan MN, Nicol R, Ning Z, Nusbaum C, O’Connor
MJ, Okazaki Y, Oliver K, Overton-Larty E, Pachter L, Parra G, Pepin KH, Peterson J, Pevzner
P, Plumb R, Pohl CS, Poliakov A, Ponce TC, Ponting CP, Potter S, Quail M, Reymond A, Roe
BA, Roskin KM, Rubin EM, Rust AG, Santos R, Sapojnikov V, Schultz B, Schultz J, Schwartz
MS, Schwartz S, Scott C, Seaman S, Searle S, Sharpe T, Sheridan A, Shownkeen R, Sims S,
Singer JB, Slater G, Smit A, Smith DR, Spencer B, Stabenau A, Stange-Thomann N, Sugnet C,
Suyama M, Tesler G, Thompson J, Torrents D, Trevaskis E, Tromp J, Ucla C, Ureta-Vidal A,
Vinson JP, Von Niederhausern AC, Wade CM, Wall M, Weber RJ, Weiss RB, Wendl MC, West
AP, Wetterstrand K, Wheeler R, Whelan S, Wierzbowski J, Willey D, Williams S, Wilson RK,
Winter E, Worley KC, Wyman D, Yang S, Yang SP, Zdobnov EM, Zody MC, Lander ES (2002)
Initial sequencing and comparative analysis of the mouse genome. Nature 420:520–562
Wolfer DP, Crusio WE, Lipp HP (2002) Knockout mice: simple solutions to the problems of
genetic background and flanking genes. Trends Neurosci 25:336–340
Wotjak CT (2003) C57BLack/BOX? The importance of exact mouse strain nomenclature. Trends
Genet 19:183–184
Yalcin B, Fullerton J, Miller S, Keays DA, Brady S, Bhomra A, Jefferson A, Volpi E, Copley RR,
Flint J, Mott R (2004) Unexpected complexity in the haplotypes of commonly used inbred
strains of laboratory mice. Proc Natl Acad Sci USA 101:9734–9739
References 359
Yalcin B, Nicod J, Bhomra A, Davidson S, Cleak J, Farinelli L, Osterås M, Whitley A, Yuan W,

Gan X, Goodson M, Klenerman P, Satpathy A, Mathis D, Benoist C, Adams DJ, Mott R,
Flint J (2010) Commercially available outbred mice for genome-wide association studies.
PLoS Genet 6(9):e1001085
Yalcin B, Wong K, Agam A, Goodson M, Keane TM, Gan X, Nellaker C, Goodstadt L, Nicod J,
Bhomra A, Hernandez-Pliego P, Whitley H, Cleak J, Dutton R, Janowitz D, Mott R, Adams
DJ, Flint J (2011) Sequence-based characterization of structural variation in the mouse
Yang H, Wang JR, Didion JP, Buus RJ, Bell TA, Welsh CE, Bonhomme F, Yu AH, Nachman
MW, Pialek J, Tucker P, Boursot P, McMillan L, Churchill GA, de Villena FP (2011)
Subspecific origin and haplotype diversity in the laboratory mouse. Nat Genet 43:648–655
Yonekawa H, Moriwaki K, Gotoh O, Miyashita N, Migita S, Bonhomme F, Hjorth JP, Petras ML,
Tagashira Y (1982) Origins of laboratory mice deduced from restriction patterns of mito-
chondrial DNA. Differentiation 22:222–226
Zou F, Gelfond JA, Airey DC, Lu L, Manly KF, Williams RW, Threadgill DW (2005)
Quantitative trait locus analysis using recombinant inbred intercrosses: theoretical and empir-
ical considerations. Genetics 170:1299–1311
Zurita E, Chagoyen M, Cantero M, Alonso R, Gonzalez-Neira A, Lopez-Jimenez A, Lopez-
Moreno JA, Landel CP, Benitez J, Pazos F, Montoliu L (2011) Genetic polymorphisms
among C57BL/6 mouse inbred strains. Transgenic Res 20:481–489
Chapter 10
Quantitative Traits and Quantitative
Genetics
10.1 Introduction
In contrast with qualitative or dichotomous traits, quantitative traits are measured

using quantitative or semi-quantitative variables and their inheritance is controlled
by multiple genes acting independently or in association. Quantitative traits are
also influenced to varying degrees by the environment and this explains why they
are often designated complex traits, with the adjective “complex” referring more
to the determinism of the phenotypic expression than to the trait itself.
Plasma cholesterol level, blood glucose level, daily water intake, body weight
when adult, susceptibility to certain forms of cancer or to a particular infectious
agent, etc. are all examples of quantitative traits because, while marked differences
often exist between mice of different inbred strains, denoting an obvious genetic
control for the traits in question, individual measures may also vary between ani-
mals of the same strain, although they are all genetically identical, indicating non-
genetic sources of variation.
Because of these differences from qualitative traits, it has been suggested in
the past that quantitative traits might obey non-Mendelian patterns of inheritance.
Nowadays, it is established that the inheritance of quantitative traits is based on
the same Mendelian principles as qualitative traits, but their inheritance cannot be
analyzed with the same methodologies.
Understanding the mechanisms of inheritance of quantitative traits and ulti-
mately identifying the genes that influence such traits is one of the major chal-
lenges geneticists must address nowadays because many human disorders with
high prevalence (obesity, hypertension, cancers, etc.) as well as susceptibility to
many common diseases in humans and animals are inherited as complex traits.
Similarly, the selection of the best breeders in domestic animals is essentially
based on a judicious exploitation of quantitative traits (Mackay 2009; Mackay
et al. 2009). The genetics of complex traits is the topic of the present chapter, to
which we will add a brief comment about the genes that have a modifying effect
on the phenotypic expression of Mendelian traits.
This part of mammalian genetics makes use of various models for the analysis
of the experimental data, often requiring a solid background in mathematics and

362 10 Quantitative Traits and Quantitative Genetics
statistics. In this chapter, we have deliberately chosen to minimize the mathemati-

cal developments. However, references to relevant textbooks or publications will
be provided for those readers willing to gain a wider competence in the matter.
10.2 Mean and Variance: Two Essential Parameters

for the Characterization of a Population
The main characteristic of quantitative traits is that, even when all known param-
eters influencing the trait and its measurement are perfectly controlled, trait values
are still subject to inevitable and incontrollable fluctuations. This applies to repeated
measurements of the same trait performed on the same individual, and to measure-
ments performed on a group of individuals that share the same genetic makeup and
environmental exposure from the moment of conception to the moment of analysis.
If one wishes to assess very robustly the phenotype of a particular individual, one
must repeat the same measurement several times under exactly the same experimen-
tal conditions and use the mean of all values as the best estimate. For example, blood
pressure can be measured on mice using an inflatable tail cuff but, although the reli-
ability of this technique has dramatically improved, it is recommended to perform
multiple measurements under the same conditions (same hour, same operator, same
apparatus, etc.). The mean of all collected values will be used as the most accurate
phenotypic assessment. The variance or its derivatives (standard deviation, SD;
standard error of the mean, SEM) will provide a useful estimate of the repeatability
of the measurement. These fluctuations generally remain in a quite narrow range and
represent the individual variability reflecting some transient, within-animal changes
not necessarily related to its metabolism.
The same issue arises at the population level. If one wishes to establish the
blood pressure of male mice of the C57BL/6J inbred strain at 12 weeks of age
under standard diet, it is absolutely necessary to make this measurement on a
group of mice that have been bred under exactly the same conditions and environ-
ment. Values measured on individual mice of the group will be slightly different
from one another and none of them can be taken as the “true estimate”: this is
referred to as inter-individual variations. The best estimate is again the mean of all
values. In this case, it is usually not necessary to repeat the measurement several
times on the same individuals, since the main source of variation will be between
individuals. Here again, the variance is an important parameter for evaluating the
fluctuations of the trait within a group of genetically and environmentally homoge-
neous individuals (Table 10.1 and Fig. 10.1).
These two types of variations are part of the concept called residual variance.
Residual variance represents the part of the total variance of the individual val-
ues of a parameter, measured in a group of individuals, which cannot be explained
either by differences in the genetic factors, by variations in the environmental fac-
tors, or by variations in the measurement methodology.
10.2 Mean and Variance: Two Essential Parameters for the Characterization 363
Table 10.1 The total plasma cholesterol levels of inbred mice
The data on this table represent the mean plasma cholesterol levels (in mg/dL ± SD) in female
mice of 12 commonly used inbred strains (aged 15–19 weeks). The inter-strain variations for
this parameter are relatively large while the intra-strain variations are (in most cases) relatively
limited. This indicates that this trait is clearly under genetic control. The data are from the
Mouse Phenome database http://phenome.jax.org/db/qp?rtn=views/measplot&brieflook=2920
&projhint=Paigen1
Frequency
inbred 2
inbred 1
outbred
stock
Phenotypic value for a trait
Fig. 10.1 Distribution of a quantitative trait in inbred and non-inbred populations. The pheno-

typic variance of a quantitative trait measured in a population of individuals is generally higher in
outbred populations than in inbred strains due to genetic heterogeneity. As a consequence, inbred
strains offer a greater power of resolution for detecting a difference in mean phenotypic value
induced, for example, by a drug, a mutation or a transgene. The values observed in two different
inbred strains sometimes partially overlap, illustrating the lack of absolute correlation between
genotype and phenotype when dealing with quantitative characters
As a consequence of inter-individual fluctuations, it is frequently observed

that two genetically different populations which show a different mean value for
a trait exhibit a partially overlapping distribution of the phenotypic values. This
illustrates that not only is it impossible to assign precisely a trait value to a gen-
otype, but it is also impossible to infer the genotype of an individual from its
phenotype.
10.3 Why Study the Genetics of Complex Traits

in Laboratory Mice?
Many of the quantitative traits studied in laboratory animals are directly related
to human physiology and pathology (for example, hypertension, diabetes, obesity,
etc.) and with the recent advances in human genetics and genomics, one could
consider that these traits would be better studied in human populations, where the
results can be directly exploited, rather than being explored in a model organism
whose biology may differ from human biology. However, this approach is chal-
lenged by the difficulties inherent in the identification of genes controlling quan-
titative traits, especially if we remember that each trait is under the control of
an unknown number of genetic factors whose effects are variable in intensity. In
most instances, none of these genetic factors is sufficient, in itself, to induce the
observed phenotype but each of them contributes, to some extent, to its expres-
sion. In addition, these multiple genetic factors are often involved in complex epi-
static interactions making it difficult to tease apart their individual effects. Finally,
and most importantly, environmental factors can modulate the biological effects of
genetic factors, making their analysis even more complex.
Laboratory rodents, namely mice and rats, offer very potent means of analyz-
ing the genetic control of complex traits in highly standardized and controlled
conditions. Crossing inbred strains with established phenotypic differences offers
the possibility of investigating gene–phenotype associations. Moreover, the exist-
ence of populations that are highly standardized from the genetic point of view,
such as recombinant inbred strains, recombinant congenic strains, congenic strains
or strains from the Collaborative Cross (described in Chap. 9), provide exceptional
tools for gene detection and the evaluation of allelic effects. In addition, the genomic
sequence is available for several strains of the mouse species, and a wide range of
strategies is available to induce genetic alterations and study their effect, as described
in the previous chapters. For all these reasons, the mouse offers unparalleled oppor-
tunities for exploring the genetics of quantitative traits of biomedical interest.
10.4 The Genetic Determinism of Quantitative Traits
Based on their determinism one can consider that quantitative traits are of two cat-
egories. The first category, the simplest, is when individual alleles that participate
in the definition of a phenotype act independently by merely adding up the pri-
mary effect of each of them with no other form of interaction. This situation, which
is rather rare, corresponds to what geneticists call the additive model. Figure 10.2
10.4 The Genetic Determinism of Quantitative Traits 365
Individual effect of genotypes Mean phenotypic values for some genotypes

Genotype Value
A/A A/B B/B 1 1A/A 2A/B 3A/A -6
locus 1 -5 0 +5 2 1B/B 2A/B 3A/B +6
locus 2 -8 +3 +7 3 1A/A 2B/B 3A/A -2
locus 3 -4 -2 +6 4 1A/B 2A/A 3B/B -2
Fig. 10.2 Relationship between genotype and phenotype. The figure represents a case of mul-
tigenic inheritance where the loci have only additive effects. In this example, a quantitative trait
is controlled by three independent polymorphic loci (1, 2, and 3). The left panel indicates the
effects of the genotypes at the three loci (A/A, A/B or B/B) on the average value in the population.
The right panel shows the average phenotypic values associated with some (actually four) geno-
typic combinations, calculated as the (algebraic) sum of the effects of genotypes at each locus.
One can see that genotypes 3 (A/A; B/B; A/A) and 4 (A/B; A/A; B/B), although different, are asso-
ciated with the same average phenotypic value (-2). This example illustrates that one cannot infer
the genotype of an individual from its phenotype as a quantitative trait
provides an example of such an additive effect, where a phenotype result from all
possible combinations (9) of two alleles (A or B) at three loci (1, 2 or 3). This
example also illustrates the important notion that individuals with a different genetic
make-up may nonetheless exhibit the same phenotype (ex: 3 and 4 in Fig. 10.2-right
box). In this simple situation, the identification of genetic factors depends on the
strength of gene effects and on the size of the population analyzed. In most cases
of quantitative inheritance, the additive model does not explain all experimental
observations and one must then make the assumption that epistatic interactions oper-
ate among the different genes with the effect of the different alleles at a given locus
depending on the genotype at one or several other loci: a complex situation indeed!
10.5 The Concept of Quantitative Trait Locus (QTL)
The genetic determinants that are responsible for quantitative traits are in general
numerous and, for this reason, they have been designated polygenes in the past.
Nowadays, they are known as quantitative trait loci (QTLs). A QTL is defined
as a locus or haplotype whose different alleles are associated with different aver-
age phenotypic values. For example, if individuals homozygous for the a allele at
locus X (Xa/Xa) are on average significantly heavier than those which are homozy-
gous for the b allele (Xb/Xb) (in the absence of any other difference between the
two groups, such as sex, age, food, genetic background, etc.), we can conclude
that there is, at locus X or in its vicinity, a gene that controls body weight. Locus X
is called a QTL, a locus controlling a quantitative trait.
Note that this effect can be assessed only on groups of individuals since, once
again, no conclusion can be drawn from single individuals. The difference between the
body weight of the animals differing in their genotype at locus X must be statistically
significant, which may require using large groups of animals if the effect of the QTL
(the body weight difference) is small. Using fewer animals may not reveal this differ-
ence and the QTL may be missed. Therefore, the capacity to detect QTLs is directly
related to the experimental design, in particular to the number of animals analyzed, as
well as to the strength of the QTLs segregating in the population.
Most quantitative traits are determined by several QTLs with a wide range of
effect size, and these QTLs together control part of the phenotypic variation in a
population. As previously mentioned, environmental parameters also contribute to
this variation, as well as other sources, namely the interactions between the geno-
type of an individual and its environment (often designated G×E, reflecting the
fact that genes’ effects can vary in different environments). Finally, uncontrolled
errors in measuring the phenotype of interest can also occur.
When deciphering the genetic control of a trait, it is important to quantify the
contribution of genetic factors to the phenotypic variations. Heritability measures
this contribution and is defined as the ratio of genotypic variance to phenotypic
variance in the population that was analyzed. It has been refined into two more
precise estimates. Broad-sense heritability takes into account the variance due to
all types of genetic effects: additive effects, dominance effects, and epistatic inter-
actions. Narrow-sense heritability considers only additive effects. Heritability (in
both senses) is therefore a variable between 0 and 1. Higher values correspond to
traits that are under stronger dependence of genetic factors. For example, the herit-
ability of a fully penetrant Mendelian mutation is 1. Quantitative traits of medical
significance have very variable heritability, and can be as low as 0.2. It is impor-
tant to note that the heritability value is not an intrinsic characteristic of a trait, but
depends on the population from which it was estimated, since it is conditioned by
the number and nature of genetic variants segregating in this population.
10.6 Positioning QTLs on the Genetic Map
The genetic mapping of genes controlling a quantitative trait is based on the iden-
tification of differences in the average phenotype between groups of individuals
depending on their genotype at a particular genomic location. Although this may
resemble the procedure used for the mapping of qualitative traits, there are how-
ever important differences that result from the poor genotype–phenotype correla-
tion at the individual level.
A first major difference is that, since the genetic alteration causing a Mendelian
or qualitative trait generally involves only one locus, the genotyping of progeny
can be interrupted when significant evidence of linkage has been detected between
the locus of the mutant allele and one or a few flanking markers. For the localiza-
tion of quantitative traits the situation is radically different because, in general, one
does not know the number of QTLs involved in the determinism of the phenotype
and for this reason the genotyping of a progeny must be carried out until the entire
genome is covered with many evenly spaced markers.
10.6 Positioning QTLs on the Genetic Map 367
Second, while strategies have been developed to identify the chromosome

carrying a mutation using less than a few dozen animals, QTL mapping always
requires the analysis of a large cohort, of at least one or two hundred animals, usu-
ally resulting in relatively imprecise location with large confidence intervals.
While the fine mapping of a Mendelian mutation can be achieved through the identi-
fication of individuals whose chromosomes have recombined between a genetic marker
and the mutant locus, refining the location of a QTL requires the analysis of a group of
animals carrying the same haplotype in order to estimate their average phenotype.
Finally, although data analysis is straightforward for qualitative traits segregat-
ing in crosses between two inbred strains, QTL detection requires sophisticated
statistical models to account for gene interactions, especially when they are com-
plex. For these reasons, QTL mapping, and even more so QTL characterization, is
a much more difficult and lengthy endeavor than finding the gene responsible for a
Mendelian trait.
10.6.1 Using Two-Generation Crosses for the Detection

and Positioning of QTLs
The crosses that are most frequently used for the genetic localization of mouse
QTLs are backcrosses (BC) or F2 intercrosses (F2) bred from parental inbred
strains. Most of the time, these crosses involve strains where large phenotypic dif-
ferences exist for the trait being measured. This would be the case, for example, of a
cross between a hypertensive and a normotensive strain or between any two strains
with marked differences in daily food intake. This situation applies to any phenotype
that can be measured in individual animals using a quantitative variable. In some
cases, several traits are measured on each animal, as a way of better describing the
status of the individual for a given condition. Each trait can then be submitted to
genetic analysis independently. Alternatively, several traits can be combined into a
composite variable derived from mathematical combination of the original measure-
ments to challenge the hypothesis that genetic (epistatic) interactions possibly occur.
A special situation applies to the genetic predisposition to develop a certain dis-
ease. Recording only the death or the absence of the disease in every animal is a
binary trait that is poorly informative. A better quantitative measurement would
be the age of the onset of the disease and a much more refined evaluation of the
susceptibility would be to measure phenotypes at the cell or the organ levels that
reflect pathophysiological processes characteristic of the disease.
In the mouse, unfortunately, not many strains spontaneously develop a disease
faithfully modeling a homologous human condition. When they exist, these strains
have been crossed with a wide variety of normal (resistant or healthy) strains for the
purpose of QTL mapping. An example is the NOD (non-obese diabetic) strain of mice,
which spontaneously develops type I diabetes mellitus. This strain has been frequently
used to study the genetic determinism and pathology of diabetes in crosses with a vari-
ety of diabetes-resistant strains (e.g., the non-obese normal inbred strain or NON).
The choice of the parental strains for making a particular cross has a major
influence on the number and position of the QTLs that will be identified. Whatever
the situation, the greater the phenotypic differences, the greater the chance of
detecting QTLs involved in the determinism of the trait being studied.
Choosing the most appropriate type of cross (BC or F2) is another important deci-
sion that is often guided by the phenotype of the F1s. One has also to take into account
the interactions (dominant, recessive, additive or epistatic) of the different alleles. In
some circumstances it may be wiser to analyze an F2 rather than a backcross because,
in this case, all sorts of genotypes (a/a, a/b, b/b) appear in the progeny. In a backcross
progeny, on the other hand, only two classes of genotypes occur, a/a and a/b, and this
may hamper the detection of a QTL in which the b allele would be recessive.
Before making the cross, it is also very important to establish the mean value and
the variance of the phenotype for the two parental strains and their F1. Since all the
mice within each parental strain or within their F1s are genetically homogeneous,
the observed variances should be of the same order of magnitude in the three groups
since phenotypic variations originate from non-genetic factors. In all cases, the
knowledge of the average values of the different inbred strains and their F1 progeny
is important for deciding the best cross to make. When the average value of the F1 is
close to the average value of one of the parental strain, it is recommended to make a
backcross by crossing the F1 with the other parental strain. When the average value
of the F1 is intermediate, deciding on the F2 is a sound choice (Fig. 10.3).
Fig. 10.3 Distribution of a (a)
quantitative trait in a cross A F1 B
between two inbred strains.
The phenotypic variance
in the F1 is of the same
frequency
order of magnitude as that F1xA

observed in the parental lines,
due to the lack of genetic
variability in these groups.
a If the average phenotypic
value of the F1 is close to
one of the two parental lines
(line B in this case) then it parameter
would be more informative
to breed a backcross progeny (b)
A F1 B
by crossing with strain A
(F1 × A). b If the average
phenotypic value of the F1 is
intermediate between those
frequency
F2
of the two parental strains,
choosing to breed an F2 is a
better option
parameter
10.6.2 Point-by-Point Analysis of the Progeny
The F2 or BC progeny bred for the genetic localization of QTLs must be carefully
phenotyped using the same protocol as for the parental strains. Any increase in the
phenotypic variance in the BC or F2 population—compared with that of the paren-
tal strains and F1—results by definition from an increase in the genetic variability
of the population in question and reflects the action of the genetic factors segregat-
ing in the cross on the phenotypic variance.
Phenotyping must also be performed in a very standardized manner because,
if genotyping errors can be detected when analyzing data and easily corrected by
retyping, phenotypic values can in general be assessed only once on every animal,
and no longer after its death, should it occur. Accurate phenotyping counts at least
as much as accurate genotyping in the mapping of QTLs.
The genotypes of the animals are established by typing genetic markers evenly
distributed over the genetic map. Nowadays, these markers are microsatellites or
SNPs selected in order to achieve an average spacing of 10–15 cM. Before any
QTL mapping analysis is performed, it is highly recommended to check that the
observed genotypes are consistent with the known position of the markers on the
chromosome map. Some computer programs offer features for detecting genotyp-
ing errors.
The first level of analysis consists of seeking an association between each geno-
typed marker and the phenotype. The aim is to identify markers for which individu-
als carrying different genotypes (a/a or a/b in a backcross; a/a, a/b or b/b in an F2)
show different average phenotypes. For each marker, the offspring are sorted accord-
ing to their genotype and the mean phenotypic values of the different genotypic
classes are compared using Student’s t-test or analysis of variance (ANOVA) (if the
phenotypic values follow a normal distribution, either as raw values or after appro-
priate transformation) or a non-parametric test. A significant difference suggests the
existence of a QTL in the vicinity of the marker (Fig.10.4). By repeating this analysis
for all markers genotyped, one can identify all chromosomal regions that are playing
a role in the genetic control of the trait. In a given chromosomal region, the QTL is
most likely located close to the marker with the strongest association (based on the
p-value). To refine the likely position of the QTL, additional markers can be geno-
typed in the region of interest, but it is not helpful to perform mapping with a high
density of markers in a backcross or an F2 population (5 cM spacing is sufficient).
10.6.3 The Concept of LOD Score
To define the statistical significance of a QTL, geneticists use the LOD score (log-
arithm of the odds). The LOD score is a statistic that compares the likelihoods
of two alternative hypotheses referring to the phenotypic difference observed
between two classes of genotypes at a particular marker. The first hypothesis is
that the observed difference is indeed due to the presence of a QTL in the vicinity
Strain A Strain B
F1 Strain A
(AxB)F1xA
backcross progeny
Genotype Student's
Markers p
a/a a/b t test
127 ± 41 132 ± 38
Locus X 0.63 > 0.5
(N = 45) (N = 53)
122 ± 29 140 ± 27
Locus Y 3.17 0.002
(N = 51) (N = 47)
Fig. 10.4 Pointwise statistical analysis in the case of a backcross between two strains. The
backross progeny was produced by crossing F1 with strain A. The phenotypic values in the back-
cross population were distributed according to a Gaussian (normal) distribution. Genotyping was
then achieved by typing the backcross individuals for marker loci whose position is known and
evenly distributed over all chromosomes. To test the effect of the genotype at a given locus, the
average phenotypic values of homozygous (a/a) and heterozygous (a/b) offspring at this locus
were compared using Student’s t test. Here, we compare the average phenotypic values of indi-
viduals homozygous (a/a) and heterozygous (a/b) for two loci X and Y. In the case of locus X,
there is no significant difference between the two groups. In contrast, mice homozygous a/a at
the Y locus exhibit an average phenotypic value significantly lower (122 ± 29) than that of het-
erozygous individuals (a/b) (140 ± 27). We conclude that there is a QTL controlling the trait
studied in the proximity of the Y locus
of the marker while the second hypothesis considers that the difference results
only from random fluctuations. The LOD score computes the ratio between the
likelihoods of these two hypotheses and expresses this ratio as a base-10 loga-
rithm. The higher the LOD score, the more likely the presence of a QTL in the
region in question. A LOD score value of 3 calculated for a marker indicates that
the association between the phenotype and the genotype at this marker is 103
(1,000) times more likely to be due to the existence of a QTL close to this marker
than to random fluctuations.
10.6.4 Threshold of Significance
Determining the actual level of statistical significance when performing QTL map-
ping is an issue. If 200 genetic markers have been genotyped, 200 statistical tests
will be performed and there is a risk that some of them will lead to p-values below
the standard 0.05 threshold just by chance. In fact, this level of significance means
that the difference observed could happen in 1 out of 20 tests by chance, i.e. in the
absence of any effect of the marker on the phenotype. With 200 markers tested,
one would expect to get 10 markers associated by chance with the phenotype with
a p-value of 0.05, in the absence of any true QTL. Therefore, a more stringent
threshold must be adopted to avoid these false positives.
An abundant literature has addressed this issue. Appropriate significance lev-
els depend on the type of cross, phenotype distribution, and marker density.
Nowadays, it is generally accepted that the optimal strategy for estimating sig-
nificance thresholds is phenotypic data permutation. Animal genotypes remain
unchanged but phenotypic data are reshuffled between animals, to break all true
causative genotype–phenotype associations. When permuted data are submitted to
QTL analysis, all detected associations are false-positive. For each permutation,
the highest LOD score observed is considered. By performing hundreds or thou-
sands of such permutations, one can calculate the frequency at which LOD scores
of 3, 4, 5, etc. were observed, all of which are false positives. One can also deter-
mine the LOD score value that has been observed in exactly 5 % of the permuta-
tions. This LOD score value is taken as the true 0.05 threshold. All recent QTL
mapping programs incorporate data permutation.
10.7 Assessing the Strength of a QTL on the Trait Studied
Once a QTL has been identified close to a genetic marker, one can evaluate the
strength of its effect on the trait by calculating the proportion of the phenotypic
variance controlled by the QTL in the population studied. We will consider the
case where the effects of the genotypes are not influenced by environmental fac-
tors (no gene × environment interactions).
The total phenotypic variance (VT) in a F2 or backcross (BC) population is the
sum of the phenotypic variance of genetic origin (VG) and of the phenotypic vari-
ance due to individual and environmental factors (VE). VE can be estimated by the
phenotypic variance measured in the parental lines or in the F1. VG is therefore the
difference between the phenotypic variance of the F2 or backcross population and
the phenotypic variance of the F1 population.
When considering a particular marker, VG can be decomposed into two frac-
tions: VQ, the genetic variance explained by the genotype at the marker, and VNQ,
the genetic variance explained by other genetic factors (Table 10.2). These two
fractions are calculated from an ANOVA with the genotype at the marker as the
main factor. The higher the VQ/VG ratio, the stronger the effect of the QTL.
Table 10.2 Origin of the phenotypic variance in different populations (the case of a backcross)
Animals are classified into two groups: Xa/Xa and Xa/Xb, according to their genotype at the
marker locus X near which a QTL has been detected. VE is the variance due to individual and
environmental factors; VT is the total variance in the backcross population; VG is the variance
of genetic origin in the backcross population; VQ is the part of the genetic variance explained by
the QTL; VNQ is the fraction of the genetic variance due to other genetic factors (other QTLs).
VG = VQ + VNQ
10.8 Interval Mapping
The locus-wise analysis estimates the LOD score at each genotyped marker, i.e.
the likelihood of the existence of a QTL at this position. These markers are usu-
ally separated by 10–15 cM and it is often useful to estimate the LOD score at
intermediate positions. A simple approach would be to genotype additional mark-
ers to increase marker density. In fact, this is useful for the regions found to be
associated with the phenotype, for refining the most likely position of a QTL, and
this is the most accurate method. However, it is possible to interpolate genotypes
between genotyped markers and compute LOD scores at intermediate position
without genotyping additional markers. This method is called interval mapping.
Interval mapping consists of guessing the genotype of each animal at positions
between two flanking markers, from the genotype of the animal at these markers
and the recombination fractions between the position being assessed and the two
flanking markers. Inferring the genotype at an intermediate position is straightfor-
ward when the two flanking markers are close and the animal has the same geno-
type at both markers. In this case, it is more than likely that the animal also carries
the same genotype at all intermediate positions. In other cases, the algorithm con-
siders all possible options, with their probability, to compute the LOD score. This
results in maximizing the likelihood of existence of a QTL at this position. By
performing this analysis at all positions between genotyped markers, one obtains
a continuous LOD score curve for each chromosome. This curve is anchored at
genotyped markers that provide reliable genotypes. LOD scores are less reliable at
intermediate positions, and one should be very cautious if the flanking markers are
separated by more than 20 cM. If a QTL is suspected in such a region, additional
markers should be genotyped at intermediate positions.
10.8 Interval Mapping 373
LOD score
18 cM
Position on chromosome
10 cM
Fig. 10.5 Determination of the confidence interval of a QTL using the LOD score curve estab-
lished by interval mapping. The X-axis represents the chromosome with the position of each ana-
lyzed marker (microsatellites or SNPs in general). The curve indicates the LOD score associated
with the presence of a QTL at each position along the chromosome. The peak of the curve deter-
mines the position that represents the maximum likelihood for the presence of a QTL. The line
corresponding to one log10 unit under the maximum LOD score (the second upper horizontal
dotted line) is then drawn and the points of intersection of this line with the LOD score curve
gives the confidence limits of the interval (18 cM in the case illustrated). In this case, it is recom-
mended to genotype more markers in the QTL region to refine the LOD score curve and better
define the confidence interval
The most likely position for a QTL is the one corresponding to the highest
value for the LOD score (often called the peak of the curve) with a certain confi-
dence level.
It is important to keep in mind that the existence of a QTL at a given position
of the genetic map is associated with a certain probability of being right. However,
in no way it is possible to conclude that a QTL exists with absolute certainty. As
the LOD score falls below the significance threshold, the chances increase that the
association is due to sampling fluctuations and not to the effect of a specific gene.
In the same way, the position of a QTL is not accurate. The precision of the posi-
tioning of a putative QTL along a chromosome is expressed as an interval that con-
tains the QTL with a certain level of statistical confidence (for example 95 or 90 %
confidence interval). Several methods exist to calculate the confidence interval (C.I.)
associated to a QTL location. The simplest is based on the likelihood ratio test (Lander
and Botstein 1989) and consists of moving sideward (left and right) of the estimated
position to the locations corresponding to a decrease in the LOD score of either one or
two units. The total width corresponding to one or two LOD drop-off can then be con-
sidered as the 96.8 or 99.8 % confidence interval, respectively. Another method uses
Bayesian statistics and provides more relevant estimates (Fig. 10.5).
10.9 Searching for Multiple QTLs Simultaneously
The strategy outlined so far works under the assumption that each QTL is detect-
able independently from the others. However, there are frequent situations where
the phenotypic effect of a QTL depends on the genotype of the animal at other
genomic locations. In this case, scanning the genome one locus at a time misses
these associations. Various statistical frameworks have been proposed to tackle
this problem and fall into multiple QTL mapping approaches. One can use the
genotype at a marker as a covariate in locus-wise analysis. For example, consider-
ing the genotype at a first QTL might help identifying others whose effect depends
on the first QTL.
It is also possible to scan the genome for all pairs of genomic locations and test
whether pairs of QTLs, acting either additively or in epistasis, can be detected.
A number of models have been proposed and implemented in various statistical
packages. This method can be time-consuming since many combinations of posi-
tions must be considered (especially when combined with interval mapping and
data permutations). Moreover, because a huge number of tests are performed, one
must use very stringent significance thresholds, which often precludes finding sig-
nificant locus pairs. However, mapping QTLs controlling a trait cannot ignore the
possibility of epistatic interactions.
10.10 Using Recombinant Inbred and Recombinant

Congenic Strains
10.10.1 Recombinant Inbred Strains
One of the limitations of two-generation crosses (F2 or backcross) is that each

progeny is unique, which does not allow replicating the assessment of the phe-
notype associated with a given genotype. Moreover, the density of recombination
breakpoints among the experimental population, which ultimately determines the
resolution of QTL mapping, is quite low. This has prompted many investigators to
study the inheritance of complex traits by using recombinant inbred strains (RIS),
which were described in Chap. 9. As discussed, each of these strains is inbred and
accordingly all animals of a given strain are genetically alike. Given these con-
ditions, one can establish the phenotype of each strain using an optimal number
of individuals of each sex and thus obtain a mean phenotypic value and variance
associated with a particular genetic make-up.
The strategy used for detecting QTLs when using RIS consists of the compari-
son of the average phenotypic value corresponding to one or the other parental
allele at each locus whose genotype is established, then searching for possible
genotype–phenotype associations. This comparison makes use of the ANOVA
(analysis of variance) and compares the intra-strain variance (which is equivalent
10.10 Using Recombinant Inbred and Recombinant Congenic Strains 375
Phenotypic value
A B 1 2 3 4 5 6 7 8 9 10 11 12 13
Locus 1 a b b a b b a b b a a b a a b
Locus 2 a b b b a a b b b a b a b a a
Fig. 10.6 Using recombinant inbred strains for QTL location. Parental strains A and B differ
by a specific phenotypic trait. For this particular phenotypic trait a group of animals from each
recombinant inbred strain has been phenotyped. The graph shows, for each strain, the mean value
and standard deviation observed. For each marker analyzed, one looks (in general with the help
of computer software) for a possible association between one parental allele and a high or low
value for the phenotypic trait being studied. For locus 1, there is no obvious association between
genotype and phenotype. However, for locus 2, strains that have a high average phenotypic value
for the trait in question are all homozygous for the a allele, while those with low value are all
homozygous for the b allele. Provided that the statistical test is significant, one can deduce the
presence of a QTL for the trait studied in the vicinity of locus 2
to the residual variance since it concerns individuals which are all genetically
identical), the inter-strain variance for those strains that have the same genotype at
the marker locus (which is related to the genetic heterogeneity between the differ-
ent lines), and the variance between the two groups of strains (which is controlled
by the specific effect of the genotype at the marker locus) (Fig. 10.6).
10.10.2 Advantages and Disadvantages of RIS
The use of RIS offers certain advantages over two-generation crosses:

• Mice of RIS allow the very precise assessment of the phenotype associated with
each genotype, since this assessment can be performed on a group of geneti-
cally identical animals. Replication improves the power of QTL detection and
the precision of their positioning on the genetic map by reducing the phenotypic
variance.
• Most RIS have already been genotyped for a great number of markers distrib-
uted throughout the whole genome; it is then only necessary to perform careful
phenotyping for the trait under investigation.
• The genome sequences of most of the parental strains of existing RIS sets are
now available. When these complete sequences are combined with haplotype
reconstructions it is possible to inspect the full genome sequence for each RIS.
This is a potentially very rich source of information.
• The genome of a RIS contains on average four times more recombination

breakpoints than that of a backcross offspring, which provides finer resolution
(reduced confidence interval) for QTL mapping (see Chap. 4).
• It is possible to compare results obtained in different laboratories on the same
set of RIS, either in order to correlate two independent phenotypes or simply
to assess the reproducibility of a study. Nothing like this can be done using the
offspring of two-generation crosses that are all different.
However, RIS have some weaknesses:
• Sets of RIS exist only for a few pairs of parental inbred strains and the opti-
mal set for the character under investigation may not be readily available. Note,
however, that it is not necessary for the identification of QTLs that the two
parental strains of an RIS set diverge for the phenotype of interest. In fact, the
two parental strains may exhibit the same average phenotype though carrying
different alleles at QTLs, while the strains of the recombinant inbred set may
have inherited different associations of these alleles and then display a wide
range of phenotypes, allowing for QTL identification and mapping. An example
of this situation was published a few years ago (Grisel et al. 1997).
• When using RIS, the accurate location of a QTL depends on the number of strains
analyzed, exactly as the accuracy of the location of a QTL depends on the number
of offspring analyzed when using a two-generation cross. However, unlike in two-
generation crosses, in which one can decide a priori the number of animals which
will be analyzed and increase this number if desired, each set of RIS has a fixed
number of strains that cannot be increased unless one decides to re-develop new
RIS from the same parental strains and to genotype them, which is a considerable
amount of work requiring additional crosses and technical investment. However,
one can study several sets of RIS for the same phenotype and compare the QTLs
identified. When common QTLs are found in two sets, data from the two sets can
be combined and the mapping resolution is consequently increased.
• The analysis of the allelic interactions at a given locus (i.e. assessing the effects
of recessivity/dominance between alleles) requires making additional crosses,
whereas this is not the case in an F2 which features animals from all three
classes of genotypes (a/a, a/b, and b/b) at each marker locus.
10.10.3 Recombinant Congenic Strains
Understanding the genetic control of complex traits require to isolate and analyze
the individual effects of every QTL identified. When a character is controlled by
several QTLs, RIS have a limitation that is a direct consequence of the equal contri-
bution of both parental genomes in the genome of each RIS. Indeed, each RIS differs
from its two inbred parental strains, on average, by half of the QTLs that are segre-
gating. If we assume that a given trait is controlled by six independent QTLs, a given
RIS will differ on average from each of its parental strain by three QTLs. In these
circumstances it is difficult to study the individual effect of one particular QTL.
10.10 Using Recombinant Inbred and Recombinant Congenic Strains 377
26 RIS 26 RCS
9 14
8 12
7
Number of strains
Number of strains
10
6
5 8
4 6
3
4
2
1 2
0 0
0 1 2 3 4 5 0 1 2 3 4 5
Number of QTL differing between a Number of QTL differing between a
particular RIS and parental strain A or B particular RCS and parental strain B
Fig. 10.7 Advantage of recombinant congenic strains (RCS) for QTL analysis. Comparison

between a set of 26 RIS and a set of 26 RCS derived from the same parental strains A and B.
If we consider that five QTLs control a characteristic making strain A different from strain B,
around 16 (8 + 8) of the 26 RIS differ from the parental line A (or B) by two or three of these
five QTLs and only four differ by a single one. On the other hand, 10 out of the 26 RCS differ
from the background strain B by a single QTL. It is therefore easier to analyze the individual
effects of each of the five QTLs by using RCS
It is mainly for this reason that other genetic populations, called recombinant
congenic strains (RCS), were developed (Groot et al. 1992). These strains are also
inbred, just like RIS, and they are derived from the offspring of a cross in which
a donor inbred strain is previously backcrossed two or three times with another
inbred line considered as recipient (see Chap. 9).
Each RCS differs from the background strain for one eighth (12.5 %) of the
genome of the donor line. Using such RCS, it is then easier to find at least one
strain that differs from the recipient strain by only one QTL, which allows assess-
ment of its individual effect on the phenotype. Several sets of RCS have been
developed for the analysis of complex traits and have been genotyped with hun-
dreds of genetic markers. However, they have never reached the same popularity
as RIS (Fig. 10.7).
10.11 Using Congenic Strains
When the phenotypic analysis of the progeny of a cross suggests the presence of a
QTL in a particular chromosomal region, other experiments are required. First, it is
necessary to confirm its existence since the presence of the QTL is not certain but
only associated with a certain probability (assessed by the LOD score). Its position
and the boundaries of the candidate interval must also be refined because the meth-
ods used for QTL detection result in ill-defined edges. Finally, and most importantly,
the QTL should be isolated in a specific strain to assess its individual effects with no
interference from the other QTLs possibly segregating in the same cross.
2 2 8 8 14 14 2 2 8 8 14 14
2 2 8 8 14 14
Strain A Strain B
F1
Several backcrosses x strain B
(with marker assisted selection for the presumptive QTL)
N2 N4 N5 N10 N10F2
2 2 8 8 14 14 2 2 8 8 14 14 2 2 8 8 14 14 2 2 8 8 14 14 2 2 8 8 14 14
Fig. 10.8 Congenic strains. A congenic strain (described in Chap. 9) is different from the back-
ground strain by only a short chromosomal segment containing the presumptive QTL. This is
achieved by selectively introgressing the chromosomal segment expected to contain the QTL
from the parental strain A into parental strain B. At every generation, progeny are genotyped
for a few markers flanking the confidence interval of the QTL, and animals heterozygous at all
markers are kept for further breeding. Note that, if these two markers are distant by more than
15–20 cM, it is recommended to genotype additional markers within the interval to make sure
that the animals selected for further breeding have retained the entire interval. The speed congen-
ics strategy (see Chap. 9) can be used to accelerate the process of congenic strain development
To reach these three objectives, geneticists breed a strain congenic for this
QTL. A congenic strain (as described in Chap. 9) is different from the background
strain by only a chromosomal segment containing the presumptive QTL. This is
achieved by introgressing selectively the chromosomal segment expected to con-
tain the QTL from the parental strain B into parental strain A. At every genera-
tion, progeny are genotyped for a few markers flanking the confidence interval of
the QTL, and animals heterozygous at all markers are kept for further breeding
(Fig. 10.8). Note that, if these two markers are distant by more than 15–20 cM, it
is recommended to genotype additional markers within the interval to make sure
that the animals selected for further breeding have retained the entire interval. The
speed congenics or high-speed congenics strategies can be used to accelerate the
process of congenic strain development (see Chaps. 2 and 9).
To confirm a QTL, one can either develop a strain congenic of parental strain A
for the B allele at the QTL (denoted A.B-QTL1) or the opposite (B.A-QTL1). In
some cases, it may be useful to produce both.
10.11 Using Congenic Strains 379
If the parental strain A and its congenic partner A.B-QTL1 show a distinct
p henotype, it is possible to conclude that the genomic region transferred in the
congenic strain harbors one or more genetic factors controlling the trait. However,
it should be noted that the absence of difference does not rule out the existence of
a QTL in the region. Once isolated in the A background, the effect of the B allele
may be too weak to change the phenotype of strain A. This is observed for exam-
ple in the case where a trait is controlled by several QTLs with moderate indi-
vidual effects. The analysis of congenic strains may fail to confirm these QTLs. In
this case, it is recommended to intercross the congenic strains to produce strains
carrying B alleles at two QTLs simultaneously.
One can then refine the location of the QTL by again crossing the A-QTL1−B
strain with strain A to break the original interval into a collection of smaller, par-
tially overlapping, sub-intervals (Fig. 10.9). Comparing the phenotype of each of
these sub-congenic strains with their genetic structure provides strong evidence to
narrow down the location of the QTL. This process can be repeated to reduce as
much as possible the size of the physical interval harboring the QTL.
Congenic strains are invaluable biological tools for investigating the nature,
structure, and function of QTLs because they allow one to manipulate a single unit
at a time. Among their many advantages, one is exceptional: working with con-
genic strains allows the study of the individual components of any trait. For exam-
ple, diabetes or hypertension, which are two intensively studied complex traits
(a) Confidence interval for the QTL

Effect (b) Confidence interval for the QTL
Effect
+ +
CG CG
Congenic strains Congenic strains
_
+
_
+
_
+
_
+
Interval harboring a single QTL Interval harboring two QTLs
CG CG
Fig. 10.9 Using a series of overlapping sub-congenic strains to confirm and refine the location
of a QTL. a The picture represents a set of sub-congenic strains homozygous for a chromosome
fragment (grey rectangles) of different size. By matching the chromosomal regions encompassed
by the congenic segment with the phenotype of the different sub-congenic strains, with respect
to the phenotype studied, it is possible to reduce the interval harboring a QTL. b The offspring
of crosses set up between different sub-congenic strains (or RIS) are also useful for shortening
the interval harboring a QTL. Sometimes this sort of experiment discloses the existence of two
closely linked QTLs instead of only one. (CG = candidate gene)
in human, mice, and rats, result from the additive and interactive actions of an as
yet undefined, although probably large, number of QTLs, having each a moder-
ate effect (a sub-phenotype so to say). Isolating each of these QTLs in a congenic
strain allows the study of their function and importance in the expression of the
sub-phenotype even if the latter is modest.
Finally, congenic strains can be very useful for revealing the effect of weak
QTLs otherwise masked by the strong effect of a major QTL. For example, sus-
ceptibility to Theiler’s virus is strongly influenced by the major histocompatibility
complex (H2 locus). In a first cross between strains C57BL/10 and SJL/J, Brahic
and Bureau (1998) identified this H2 haplotype as a major factor. They made a
second cross between SJL/J and the B10.S strain, which is congenic of C57BL/10
for the H2 haplotype of the SJL strain. In this second cross, all progeny carry the
same H2 haplotype and the effect of other QTLs can be revealed more efficiently.
10.12 Using Other Strains or Stocks for the Mapping

of QTLs
10.12.1 Consomic Strains (CS)
Consomic strains or chromosome substitution strains (CSS) are similar to con-

genic strains with the exception that the introgressed genomic segment is a com-
plete chromosome (Singer et al. 2004). They are useful tools for the identification
and analysis of QTLs but, unfortunately, only a few sets are available. When it
is possible to use such a set, it allows rapid association of a phenotype with a
particular chromosome for which the donor and background strains differ. If,
for example, one observes that strain C57BL/6J differs from its consomic part-
ner strain C57BL/6J-Chr 1A/J (carrying chromosome 1 from A/J on an otherwise
C57BL/6J background) for a specific trait, this means that chromosome 1 contains
at least one QTL controlling the trait.
Further characterization of a QTL identified on a chromosome requires the pro-
duction of a series of sub-consomic strains (analogous to sub-congenic strains) to
refine the location of the QTL(s) and evaluate its (their) individual effect(s). One
major limitation of consomic strains is that they miss QTLs located on different
chromosomes that operate in epistatic interaction.
Specific chromosome substitution strains (PSCSS) are a variant of consomic
strains where one specific chromosome of the recipient mouse strain C57BL/6J
has been substituted by the homologous counterparts from several different inbred
strains and not just one as in the case of consomics. These PSCSS form a spe-
cial population that has an identical genetic background to the recipient strain
C57BL/6 and differs only in the donor chromosomes (Xiao et al. 2010).
10.12 Using Other Strains or Stocks for the Mapping of QTLs 381
10.12.2 The Collaborative Cross (CC): A Novel, Powerful

Tool for Studying the Genetics of Complex Traits
The Collaborative Cross (commonly abbreviated CC) has been described in

Chap. 9, along with the other kinds of genetically standardized strains (Threadgill
et al. 2002; Churchill et al. 2004; Threadgill and Churchill 2012), and is already
considered a next-generation tool for the analysis of complex traits.
The CC consists of a large panel of mouse RIS descending from the ran-
dom and unique association of an equal proportion of the genome of eight unre-
lated founder strains, comprising five classical inbred strains (A/J, C57BL/6J,
129S1/SvImJ, NOD/LtJ, and NZO/H1LtJ) and three wild-derived strains rep-
resenting the three major Mus musculus subspecies (CAST/EiJ, PWK/PhJ, and
WSB/EiJ). To date, over 400 strains have been developed and a few dozen are con-
sidered complete (homozygosity > 90 %).
The CC panel shares many advantages with the RIS and has additional
advantages related to the greater number of genetic polymorphisms segregating
among the different parental strains, to the much greater number of independ-
ent strains that will be available, and to the higher number of recombination
breakpoints per strains. Computer simulations have indicated for example that
500 RIS of the Collaborative Cross would be adequate to map a single additive
locus that accounts for only 5 % of the phenotypic variation to within 0.96 cM.
Even if this mapping resolution will not be sufficient to identify single genes
unambiguously, it represents nevertheless a considerable leap forward in the
power of resolution.
10.12.3 Interspecific Recombinant Congenic Strains (IRCS)
IRCS are a variety of the RCS (mentioned above) with parental strains belonging to
two different mouse species. They were developed from the parental strains C57BL/6
(the background strain) and an inbred strain derived from the Mus spretus species
(SEG/Pas) as the donor strain (Burgio et al. 2007, 2012). These strains are equiva-
lent to RCS discussed above with, however, some important differences. First, the
introgressed component is of very remote origin and accordingly contributes to
an important amount of polymorphism. Second, the genomic contribution of each
parental strain is very unequal since each IRCS strain carries up to eight SEG/Pas
chromosomal segments with an average size of 11.7 Mb, totalizing 1.37 % of the
genome. Finally, when adding up the individual contributions of all 55 strains the
SEG/Pas genome covers 39.7 % of the total genome. IRCSs are useful to unravel
QTL with small effects and gene interactions.
10.12.4 Diversity Outbred (DO) Stock
The major limitation in QTL mapping is the resolution, and resolution itself
depends on the density of recombination breakpoints in the individuals (or
strains) used for genetic analysis. When the density is low, this results in wide
QTL peaks with large confidence intervals. Much better resolution, down to the
gene level, can be reached by analyzing populations (like the human popula-
tion for example) that have accumulated huge densities of recombination break-
points over many generations of random crosses. This observation led a group
of geneticists of The Jackson Laboratory to develop the Diversity Outbred
(DO) stock, by continued random mating of 144 partially inbred lines of the
Collaborative Cross. Each mouse of this stock is genetically unique, and once
genotyped by using high-density genotyping arrays (Li et al. 2005; Churchill
et al. 2012), it allows unparalleled resolution for QTL mapping. Groups of DO
mice approximate the genetic diversity and level of heterozygosity found in
human populations (i. e. an average of 390 recombination events per genome
at G10) and can be used to validate previously identified QTLs. Groups of DO
mice approximate the genetic diversity and level of heterozygosity found in
human populations.
10.13 Cloning QTLs
Once a QTL has been identified, confirmed and assigned to a small chromosomal
region, identifying the quantitative trait gene (QTG) responsible for the effect
observed is the ultimate goal. Even though substantial progress in the knowledge
of the mouse genomic sequence has been made in recent years (see Chap. 5), this
last step remains a difficult enterprise. There is no unique strategy to go from a
genomic region to the gene and it is generally a combination of approaches that
will provide clues which, confronted and interconnected, will point at candidate
genes eventually submitted to functional analysis.
When the QTL location has been narrowed to a region of a few Mb using con-
genic and sub-congenic strains, which may require the production and phenotyp-
ing of large numbers of animals, the strategies for identifying the causative gene
resemble those used for Mendelian traits. They include, in particular, the com-
parison of whole-genome or whole-exome sequences, the production and in sil-
ico analysis of gene expression data, and thorough literature review and database
searching to collect detailed information on gene function. One should also care-
fully look for data coming from other animal models or human conditions. These
investigations should lead to a limited number of candidate genes that must be
submitted to functional evaluation. The most appropriate testing depends on the
nature of the trait and the phenotype that best characterizes the QTL effect.
10.13 Cloning QTLs 383
10.13.1 Analyzing the DNA Polymorphisms in the QTL

Region
When the interval is significantly reduced (i.e. less than 1 or 2 Mb), which may
require breeding, phenotyping, and genotyping thousands of mice, it is then pos-
sible to look at the genome structure focusing on genetic issues. The first thing to
establish, when possible, is an exhaustive list of the genes (10–30 on average) which
map within the interval, with the likely function of each of them, when this is known
from genome annotation. Nowadays, this step of QTL analysis is made somewhat
easier if we remember that the genome of several inbred strains has been completely
sequenced, making the alignments between the parental strains easier and faster.
While comparing these alignments, it is important to check for the possible existence
of indels and more generally the integrity of the different genes in the two parental
strains. Small-sized deletions and insertions are common findings in the mamma-
lian genome and even though many of them exhibit no clear effect on the phenotype
when homozygous they may nevertheless entail slight phenotypic variations.
Gene copy number variations (CNVs) are also important structural differences,
which may account for quantitative phenotypic differences (see Chap. 5 for com-
ments; Cutler and Kassner 2008). Finally, SNPs are very important structural vari-
ations to look at for two main reasons. (i) First, because among the most recently
published results reporting the successful positional cloning of a QTL (whatever the
species) a majority indicate that SNP differences have been the starting point, with one
of the non-synonymous SNPs being associated with a conformational change often
leading to a difference in activity of the encoded protein. (ii) SNPs are also important
polymorphisms to look at because they can help in the determination of the ancestral
origin of the haplotype containing the QTL under investigation and accordingly can
suggest comparisons to be made with strains unrelated to the parental strains but seg-
regating for the same QTL. Any SNP that might be causative of a missense mutation
or a splicing defect would require special attention. Nonsense mutations, generating
null or hypomorphic allele, are candidates for qualitative mutations but have not been
often recognized as being responsible for quantitative phenotypic differences.
Based on the information collected in several species (including plants), the
genetic alterations that are the best candidates to account for phenotypic differences
in quantitative trait inheritance are those that result in proteins slightly modified in
their structure, expression level or stability in time but not in loss or gain of function.
SNP analysis is a logical and straightforward approach but it can sometimes be
extremely difficult when, for example, the QTL encompasses a SNP-rich region.
In this case thousands of SNPs must be analyzed, with many of them being irrel-
evant or outside the coding regions.
In conclusion, in many instances the structural variations that can be observed
at the sequence level are insufficient to provide an answer in terms of gene identi-
fication. Other investigations are necessary to unravel the biology of the candidate
genes: for example where, when, and at what level they are expressed.
10.13.2 Quantitative Complementation
Validation of a candidate gene is generally achieved by performing trans-comple-

mentation with another allelic form of the candidate gene, which can be either a
mutant allele that already exists somewhere in a genetic repository or can be engi-
neered especially for this purpose.
A strategy that seems to be efficient, when applicable, is known as QTL-knockout
interaction (Darvasi 1998). The experimental protocol requires no less than four
strains and four crosses. Strain A, encoding for the “high” allelic form at the QTL
(Qh) in question is mated with a strain carrying a null (knockout) allele of the gene
of interest (m–) and with the co-isogenic strain carrying the wild-type allele of the
same gene (m+). Similarly, two other crosses are made with the strain B encoding
the “low” allelic form (Ql) and the same two strains as above, i.e. (m–) and (m+). A
greater phenotypic difference between the (Qh)/(m–) and (Ql)/(m–) genotypes than
between the (Qh)/(m+) and (Ql)/(m+) genotypes provide some evidence of quantita-
tive failure of the mutation to complement the QTL alleles and validate the candidate
gene. Theoretically, the QTL/knockout interaction test requires the use of strains co-
isogenic for the knockout (m–) and wild type (m+) alleles to avoid potentially con-
fusing interactions of other genes in the background.
In summary, although only a few dozen rodent QTLs have been cloned to date,
it appears from studies conducted in other species that QTLs are generally made
out of non-null allelic variants. These variants are generally characterized by dif-
ferences in gene expression level or by subtle structural variations that translate
into differences in terms of activity for the encoded proteins. This is a major dif-
ference from qualitative or Mendelian traits, where null mutations are quite
common.
Thousands of QTLs have been mapped onto the mouse genome during the last
two decades. However, only a few genes underlying complex traits have been suc-
cessfully identified, and fine mapping of QTL genes still remains a challenge for
mouse geneticists.
10.14 The Analysis of Expression QTLs (eQTLs)
The analysis of gene expression, for example by using expression arrays or RNA
sequencing, allows the discovery of quantitative differences, sometimes important,
between strains or individuals. Gene expression level is a quantitative phenotype,
controlled by expression QTLs (eQTLs), amenable to QTL mapping using the
methodologies described for other phenotypes.
A number of published studies have shown that most eQTLs are located in cis,
i.e. in the vicinity of the expressed gene. They most likely correspond to classical
regulatory elements such as promoters, enhancer, 3′UTRs, etc (see Chap. 5). In
this case, one eQTL influences the level of expression of a single gene. However,
a small fraction of eQTLs appear to control the expression of multiple genes
10.14 The Analysis of Expression QTLs (eQTLs) 385
located on different chromosomes. Although the resolution of eQTL mapping can-

not rule out the possibility that these genes are controlled by several, independent
but tightly linked, regulatory factors, it is hypothesized that they correspond to key
elements of regulatory networks.
Notably, RIS have proved very valuable in mouse studies, in particular the
BXD set established between C57BL/6J and DBA/2J strains, which currently
comprises around a hundred strains. Gene expression levels can be measured and
compared in every inbred strain comprising a given set of multiple individuals,
in males and females, in multiple tissues and organs, at different ages and under
multiple conditions. Given that these strains have been extensively genotyped, this
experimental design allows quick identification of eQTLs, either sex-, age- or tis-
sue-specific. It also provides optimal power to identify correlations between the
expression levels of different genes or between organs across a series of geneti-
cally different inbred strains. Most interestingly, the BXD set has been heavily
investigated for a wide variety of phenotypes including metabolic, behavioral,
immunological, and many other traits. It is now possible to search in silico for
associations between phenotypic data and gene expression levels, and identify, at
a genome-wide scale, genes whose expression correlates, positively or negatively,
with a trait of interest. This approach, which will certainly reveal unsuspected
relationships, has been made possible and publicly available by the develop-
ment of centralized databases and analytical tools, in particular by GeneNetwork
(http://www.genenetwork.org).
10.15 The Case of Modifier Genes
When studying the phenotype of mouse mutations, variations in phenotypic

expression are common observations. Whereas, for example, mice homozygous
for the Tyrc allele are always completely albino, mice homozygous for the same
Leprdb-Pas mutation (leptin receptor non-functional allele) have a phenotype that
is very different when the mutant allele is on the PWK/Pas background or the
DW/J background. In the former case, the mice are moderately fat but mostly dia-
betic and secrete enormous amounts of urine (sometimes up to 50 ml per day!).
On the DW/J background, on the contrary, the mice carrying the same allele grow
to become very fat (up to 80 g) but urinate almost a normal quantity. This obser-
vation of great variations in the phenotypic expression is more the rule than the
exception in mammals and for some mutations the situation is sometimes extreme.
Many dominant bone mutations have probably been lost because they could not be
transmitted from one generation to the next for a lack of expressivity while others
were so severely expressed that they appeared to be incompatible with life.
This aspect of background-dependent phenotypic variation also exists in human
populations where it has a great importance making a specific syndrome more or
less compatible with every-day life. The same situation is also common with cer-
tain forms of cancer with a clear genetic determinism and great variations in terms
of severity depending on the affected person. Colon cancers triggered by mutant

alleles at the APC gene are a good example, as are the deficiencies in the enzyme
ferrochelatase (FECH) producing a syndrome with extreme variations depending
on the patient.
All the variations reported above are the consequence of multiple genetic inter-
actions between the mutant gene and a number of unknown genes modulating the
expressivity of the phenotype. The effect of these modifier genes can be measured
by quantitative variables and the genes can be identified using the general strate-
gies described above. This research is extremely difficult in humans but the mouse
can provide relevant models. The genes identified in mice may provide at least
an indication of metabolic pathways and physiological processes that have a sig-
nificant probability of being conserved between the two species. Interesting and
promising results have been obtained recently.
10.16 Conclusions
The susceptibility of humans and domestic animals to certain forms of cancer or

to most infectious diseases, just like the variations in expression of some meta-
bolic diseases, is often influenced by genetic factors. To date, the nature of these
factors, although obvious, is unknown and this is unfortunate because it might be
used as a way of influencing the prognosis of many diseases. For this reason, one
can predict that studies on the genetic determinism of complex traits will expand
in the years to come. In such studies, the mouse will undoubtedly play a pivotal
role because in this species, more than in any other, powerful tools and strategies
are available. This will certainly help researchers to know better the composition
of the QTLs at the molecular level.
References
Brahic M, Bureau JF (1998) Genetics of susceptibility to Theiler’s virus infection. BioEssays

20:627–633
Burgio G, Baylac M, Heyer E, Montagutelli X (2012) Exploration of the genetic organization
of morphological modularity on the mouse mandible using a set of interspecific recombi-
nant congenic strains between C57BL/6 and mice of the Mus spretus species. G3 (Bethesda)
2:1257–1268. doi: 10.1534/g3.112.003285
Churchill GA, Airey DC, Allayee H, Angel JM, Attie AD, Beatty J, Beavis WD, Belknap JK,
Bennett B, Berrettini W, Bleich A, Bogue M, Broman KW, Buck KJ, Buckler E, Burmeister
M, Chesler EJ, Cheverud JM, Clapcote S, Cook MN, Cox RD, Crabbe JC, Crusio WE,
Darvasi A, Deschepper CF, Doerge RW, Farber CR, Forejt J, Gaile D, Garlow SJ, Geiger
H, Gershenfeld H, Gordon T, Gu J, Gu W, de Haan G, Hayes NL, Heller C, Himmelbauer
H, Hitzemann R, Hunter K, Hsu HC, Iraqi FA, Ivandic B, Jacob HJ, Jansen RC, Jepsen KJ,
References 387
Johnson DK, Johnson TE, Kempermann G, Kendziorski C, Kotb M, Kooy RF, Llamas B,
Lammert F, Lassalle JM, Lowenstein PR, Lu L, Lusis A, Manly KF, Marcucio R, Matthews
D, Medrano JF, Miller DR, Mittleman G, Mock BA, Mogil JS, Montagutelli X, Morahan
G, Morris DG, Mott R, Nadeau JH, Nagase H, Nowakowski RS, O’Hara BF, Osadchuk
AV, Page GP, Paigen B, Paigen K, Palmer AA, Pan HJ, Peltonen-Palotie L, Peirce J, Pomp
D, Pravenec M, Prows DR, Qi Z, Reeves RH, Roder J, Rosen GD, Schadt EE, Schalkwyk
LC, Seltzer Z, Shimomura K, Shou S, Sillanpää MJ, Siracusa LD, Snoeck HW, Spearow
JL, Svenson K, Tarantino LM, Threadgill D, Toth LA, Valdar W, de Villena FP, Warden
C, Whatley S, Williams RW, Wiltshire T, Yi N, Zhang D, Zhang M, Zou F; Complex Trait
Consortium (2004) The collaborative cross, a community resource for the genetic analysis of
complex traits. Nat Genet 36:1133–1137
Churchill GA, Gatti DM, Munger SC, Svenson KL (2012) The diversity outbred mouse popula-
tion. Mamm Genome 23:713–718
Darvasi A (1998) Experimental strategies for the genetic dissection of complex traits in animal
models. Nat Genet 18:19–24
Grisel JE, Belknap JK, O’Toole LA, Helms ML, Wenger CD, Crabbe JC (1997) Quantitative
trait loci affecting methamphetamine responses in BXD recombinant inbred mouse strains. J
Neurosci 17:745–754
Groot PC, Moen CJ, Dietrich W, Stoye JP, Lander ES, Demant P (1992) The recombinant con-
genic strains for analysis of multigenic traits: genetic composition. FASEB J 10:2826–2835
Lander ES, Botstein D (1989) Mapping mendelian factors underlying quantitative traits using
RFLP linkage maps. Genetics 121:185–199
Li R, Lyons MA, Wittenburg H, Paigen B, Churchill GA (2005) Combining data from multi-
ple inbred line crosses improves the power and resolution of quantitative trait loci mapping.
Genetics 169:1699–1709
Mackay TFC (2009) Q&A: genetic analysis of quantitative traits. J Biol 8:23
Mackay TFC, Stone EA, Ayroles JF (2009) The genetics of quantitative traits: challenges and
prospects. Nat Rev Genet 10:565–577
Singer JB, Hill AE, Burrage LC, Olszens KR, Song J, Justice M, O’Brien WE, Conti DV, Witte
JS, Lander ES, Nadeau JH (2004) Genetic dissection of complex traits with chromosome
substitution strains of mice. Science 304:445–448
Threadgill DW, Hunter KW, Williams RW (2002) Genetic dissection of complex and quantitative
traits: from fantasy to reality via a community effort. Mamm Genome 13:175–178
Threadgill DW, Churchill GA (2012) Ten years of the collaborative cross. G3 (Bethesda)
2:153–156
Xiao J, Liang Y, Li K, Zhou Y, Cai W, Zhou Y, Zhao Y, Xing Z, Chen G, Jin L (2010) A novel
strategy for genetic dissection of complex traits: the population of specific chromosome sub-
stitution strains from laboratory and wild mice. Mamm Genome 7–8:370–376
Publications of General Interest
Falconer DS, Mackay TFC (1996) Introduction to quantitative genetics, 4th edn. Longman, Harlow
Flint J, Valdar W, Shifman S, Mott R (2005) Strategies for mapping and cloning quantitative
traits in rodents. Nat Rev Genet 6:271–286
Lynch M, Walsh B (1998) Genetics and analysis of quantitative traits. Sunderland, MA Sinauer
Mackay TF (2001) The genetic architecture of quantitative traits. Ann Rev Genet 35:303–339
Xu S, Atchley WR (1996) Mapping quantitative trait loci for complex binary diseases using line
crosses. Genetics 143:1417–1424
Zeng ZB (1994) Precision mapping of quantitative trait loci. Genetics 136:1457–1468
Computer softwares
Broman KW, Wu H, Sen S, Churchill GA (2003) R/qtl: QTL mapping in experimental crosses.
Bioinformatics 19:889–890
Conover WJ, Iman RL (1981) Rank transformations as a bridge between parametric and non-
parametric statistics. Am Stat 35:124–129
GeneNetwork: http://www.genenetwork.org
Kruglyak L, Lander ES (1995) A nonparametric approach for mapping quantitative trait loci.
Genetics 139:1421–1428
Lander ES, Green P, Abrahamson J, Barlow A, Daly MJ, Lincoln SE, Newberg LA, Newburg
L (1987) MAPMAKER: an interactive computer package for constructing primary genetic
linkage maps of experimental and natural populations. Genomics 1:174–181
Sen S, Churchill GA (2001) A statistical framework for quantitative trait mapping. Genetics
159:371–387
Glossary
5'-untranslated region (5'-UTR) or leader sequence A sequence measuring

100 to several thousands bp, between the transcription initiation site and the initial
AUG. This sequence usually contains a ribosome binding site (RBS), known as the
Kozak sequence (gcc)gcc(A/G)ccAUGG, which includes the AUG initiation codon.
3'-untranslated region (3'-UTR) or trailer sequence A sequence, downstream
of the stop codon, which is required for the processing of mRNAs. The polyad-
enylation signal, composed of sequences like AAUAAA, is an important compo-
nent of the 3'-UTR.
Acrocentric A chromosome in which the centromere is located near one end. The
40 chromosomes (19 pairs of autosomes, X and Y) of the normal laboratory mouse
are all acrocentric (see Chap. 3).
Additive inheritance A situation where the observed phenotype results from the
cumulative, individual effects of the alleles carried at a group of loci, with no epi-
static interactions and no allele-dependent effect of the environment.
Allele (allelomorph) One of the alternative forms of a gene, which may differ
from the others by a variety of polymorphisms ranging from a single nucleotide
(an SNP) to a sequence of several nucleotides.
Alternative splicing This expression means that not all exons in a gene are
assembled to form the maturated mRNA, on the contrary, some are deleted (or
spliced out). It is estimated that ~95% of multiexonic genes of the mammalian
genomes are alternatively spliced. However, not all genes have introns.
Anchor locus A locus whose location in the genome is very precisely known,
used as a landmark in the construction of genetic maps. A gene that is, at the same
time, characterized phenotypically and at the DNA level (sequence) and that has
orthologs in several other species is an ideal anchor locus.

J.L. Guénet et al., Genetics of the Mouse, DOI 10.1007/978-3-662-44287-6
390 Glossary
Aneuploidy Any variations from the normal 2n (euploid) chromosome number.

Triploidy (3n), haploidy (n), trisomy (2n+1), monosomy (2n−1) are classical
aneuploidies. Aneuploidies sometimes involve only a portion of chromosome.
Annotation Gene annotation consists of establishing the structure and the likely
function(s) of a given gene or DNA sequence.
Anonymous locus A sequence of DNA with no known function but with at least
two allelic forms that can be followed generation after generation through some
form of DNA analysis.
Analysis of variance (ANOVA) ANOVA is a collection of statistical methods,
developed by R.A. Fisher, and used to analyze the robustness of observed differ-
ences between groups of individuals. Using ANOVA, a statistically significant
result would be when a probability (p-value) is less than a threshold (significance
level), and justifies the rejection of the null hypothesis (the observed difference
between two samples may result from chance only) (see chi-square (χ2) test).
Anti-sense mRNA An RNA molecule or oligonucleotide that is complementary
to an mRNA molecule and can form a duplex with it. Antisense mRNAs interfere
with mRNAs translation (in mammals) and provoke their destruction by ribonucle-
ases specific for double-stranded molecules.
Autosome Any chromosome other than the X or Y chromosome. The normal
mouse karyotype consists of 19 pairs of autosomes.
B1/B2 repeats Also designated Short Interspersed Nuclear Elements or SINE

(see Chap. 5). B1/B2 are the two most prominent classes of repetitive elements in
mammalian genomes, with a size of 500/600bp. The mouse B2 repeats are equiva-
lent to the human Alu sequences.
BAC (Bacterial artificial chromosome) A cloning vector that uses the origin of
replication of the functional fertility plasmid (F plasmid) of E. coli. BACs can take
inserts with a size up to 150–350 kb and can be used for the production of trans-
genic mice. BACs are more reliable vectors than YACs.
Backcross Literally, a backcross is the cross of a hybrid with one of its parents or
an individual genetically similar to it. In practice, it is a cross between an animal
heterozygous for two different alleles and an animal homozygous for one of the
two alleles in the heterozygous parent (for example A/a × a/a or A/a × A/A). In
the particular case where the cross is between a heterozygous mouse and a mouse
homozygous for the recessive allele (A/a × a/a), the backcross is called a testcross
(see Chaps. 2 and 4).
BLAST (basic local alignment of sequences tool) A popular computer pro-
gram that searches for similarity of a selected sequence to sequences stored in a
database.
Blastocoel(e) or blastocele The fluid-filled cavity of a blastocyst.
Glossary 391
Bin A bin is a group of syntenic genetic markers that have not been separated
(ordered) by meiotic recombination in a given cross (see syntenic).
CAGE (Cap-analysis of gene expression) A technique based on the preparation

and sequencing of concatamers of DNA tags deriving from the initial 20 nucleo-
tides from the 5′ end of mRNAs allowing high-throughout analysis of the capped
transcripts population in a biological sample. CAGE detects the transcriptional
activity of each promoter.
Candidate gene A gene thought to be likely responsible for an observed pheno-
type because of its genetic localization and/or likely function. Candidate genes
are, in most instances, suggested by positional cloning or association studies.
CAT (or CCAAT, or CAAT)-box A consensus sequence GGCCAATCT, inserted
75−80 bp upstream from the transcription start site. Some genes with relatively
ubiquitous expression do not have this GGCCAATCT sequence.
Centimorgan (cM) The unit used to describe genetic distances. A centimorgan
is the distance between two genes that will recombine with a frequency of exactly
1% (see Chap. 4). Genetic distances are additive, while recombination fractions
are not. Genetic distances are computed from recombination fraction using a map-
ping function.
Centromere The centromere is the part of a chromosome that links sister chro-
matids and attaches to the spindle fibers (through the kinetochore) (see Chap. 3).
Chi-square A statistical test developed by Karl Pearson and commonly used by
geneticists to compare observed data from a given experiment (for example, the
different proportions of phenotypes in a progeny) with the theoretical data that
would be obtained according to a specific hypothesis. The chi-square (χ2) test
appreciates whether the differences between observed and expected proportions
could result from chance (the null hypothesis) or if they are more likely to be due
to an explanatory factor. (see Chap. 4).
Chiasma (plural: chiasmata) The point where two homologous non-sister chro-
matids exchange genetic material during meiosis.
Chimera (or Chimaera) An organism that is composed of two (or more) popula-
tions of genetically distinct cells originating from two or more embryos. Chimeras
generally result from human intervention (see Chap. 2—see also Mosaics).
Chromatin The complex of DNA, RNA and protein that makes up the core of
chromosomes. There are two sorts of chromatin: euchromatin, whose structure is
open, allowing the transcription of the DNA component, and the heterochromatin,
which is made of repetitive sequences that are not transcribed.
Chromosome banding A staining process that produces a discrete, reproducible
pattern of light- and dark bands that can be used to identify individual chromo-
somes and chromosomal regions. See G-bands and Q-bands (see Chap. 3).
392 Glossary
Clone-by-clone A strategy used for sequencing the human genome. The genome
in question is cloned into BACs, the clones are ordered and shotgun-sequenced
(see shotgun sequencing). Finally, the sequence is assembled by ordering head-to-
tail the sequences from adjacent BAC clones.
Coding sequence A stretch of DNA or RNA whose sequence ultimately determines
the sequence of a protein (see Chap. 5). The coding sequence excludes introns.
Codominance A kind of allelic interaction in which an animal heterozygous for
two alleles (A1 and A2) at the A locus, expresses at the same time, the phenotypes
that would be observed in the two corresponding homozygotes (A1/A1 and A2/A2).
Codominance is more the rule than the exception in mammals.
Coisogenic A strain of mice that differs from an established inbred strain by a
single point mutation at a given locus (see Chap. 9).
Collaborative cross (CC) A panel of recombinant-inbred strains, generated by
randomizing the genetic diversity of existing inbred mouse resources from the
three major Mus musculus subspecies (M. m. musculus, M. m. domesticus, and M.
m. castaneus). A useful tool for mapping multigenic traits (see Chaps. 9 and 10).
Comparative genomic hybridization (CGH) CGH is a molecular method for
assessing possible copy number variations (CNVs), through independent labeling
of a reference sample and a test sample of denaturated DNA with fluorophores of
different colors (usually red and green) (see Chap. 5).
Complementary uniparental disomies/nullosomies A normal, euploid (2n)
embryo resulting from the fusion of two aneuploidy gametes. When one parent
contributes two chromosomes of the same pair and the other none, this results in
an embryo with complementary uniparental disomy/nullosomy (abbreviation UpD
or UPD). Some of these embryos are viable, others are not due to the differential
imprinting of the chromosomes in the gametes (see Chap. 6).
Complex diseases Diseases whose etiology consists of a mixture of environmen-
tal and genetic factors. In many instances the genetic factors are numerous and/or
of various “strength”.
Compound heterozygote An individual heterozygous for two mutated alleles at a
given gene (for example Am1/Am2).
Congenic A strain of mice that is formed by introgressing (i.e. backcrossing
repeatedly) a chromosomal segment carrying a locus of interest into an inbred
parental strain for ten or more generations (see Chap. 9). For example, B6.C-Tyrc
is a congenic strain with C57BL/6 (B6) background carrying a segment of chro-
mosome 7 from BALB/c (C) origin that harbors the albino mutation (Tyrc), result-
ing in albino B6 mice.
Conplastic Conplastic strains have the same nuclear genome but different mito-
chondrial genome. A conplastic strain is developed by transferring the nuclear
genome from one inbred strain into the cytoplasm of another (the donor parent
Glossary 393
is always the female parent during the backcrossing program). A minimum of 10

backcross generations is required for a strain to be 100% conplastic. Conplastic
strains are useful for studying the role of mtDNA.
Consensus sequence A theoretical sequence that represents the nucleotides most
often found at each position when a number of different sequences are compared.
Consomic strains Consomic strains are a particular case of congenic strains, in
which the transferred segment consists of an entire chromosome. Also called chro-
mosome substitution strains (CSS) (see Chap. 9).
Contig A DNA region represented by a set of (head-to-tail) overlapping genomic
clones (BACs, P1 or YACs) that together span a region of the genome larger than
that covered by any one clone (see Chaps. 5 and 9).
Copy number variants (or variation) (CNVs) Segments of DNA that are
present in an individual with a copy number that is different from the reference
genome (C57BL/6).
Cosmid A hybrid plasmid containing the cos site from the lambda phage, used
as a cloning vector combining the properties of a plasmid and a phage. Cosmids
allow recombinant DNA molecules of 30 to 50 kb to be packaged into phage par-
ticles in vitro. Upon infection into a host cell the recombinant DNA molecule rep-
licates as a plasmid.
Coverage The average number of times a same nucleotide is sequenced in
genome shotgun sequencing. Higher coverage (×8 - ×10) provides higher reliabil-
ity of sequence data.
CpG island Short DNA sequences (a few kb long) that are GC-rich, CpG-rich,
and predominantly non-methylated. CpG islands are associated with the 5'-end of
genes and most, perhaps all, of these sequences are sites of transcription initiation
(see Chap. 5).
Crossing-over A crossing-over is the reciprocal exchange of genetic material
between homologous chromosomes during meiosis (see Chap. 4).
Cytogenetics The part of Genetics that deals with chromosome structure, chro-
mosome behavior during meiosis and pathology resulting from chromosomal
breakage or imbalance.
Deleterious allele An allele with a more or less severe effect on the phenotype. A
missense allele can be a deleterious allele if the substituted amino acid impairs the
function of the protein. A null-allele is often deleterious.
Deletion mutations The loss of one or more nucleotides or, sometimes, a frag-
ment of chromosome.
Deme A breeding unit in natural populations of mice. A deme usually consists of
one dominant male with up to six-eight females (see Chap. 1).
394 Glossary
DGGE (denaturating gel gradient electrophoresis) A sensitive gel-based tech-

nique for detecting single nucleotide changes within orthologous or allelic PCR
products that have been denaturated and gel-fractionated as single strands.
Dinucleotide Two successive nucleotides on the same DNA strand, written with
the 5' nucleotide first.
Disjunction The normal process by which the two homologs of each chromo-
some in a meiotic cell separate and move to different gametes as a single unit (see
Chap. 2).
Distal A relative term meaning closer to the telomere; the opposite of proximal.
DNA marker A short sequence of DNA with allelic variation that can be fol-
lowed directly by a DNA-based assay such as hybridization techniques, PCR or
sequencing (see Chap. 2).
DNA methylation The addition of a methyl group to cytosine to form 5-methyl-
cytosine in CpG dinucleotides. Between 8 and 10% of the cytosine nucleotides in
the mouse genome are methylated. Methylation alters gene expression and plays a
major role in parental genomic imprinting.
Domain Part of a protein where the polypeptide chain folds to form a discrete
globular structure that has a defined function. Also apply to the sequence of DNA
encoding the protein domain in question.
Dominant allele Dominance describes the relationship of one allele to a second
at the same locus when an animal heterozygous for these alleles expresses the
same phenotype as an animal homozygous for the first allele. The second allele
of the pair is considered recessive. Mice homozygous for the agouti allele (A/A)
as well as mice heterozygous for the agouti and nonagouti alleles (A/a) carry hairs
with yellow pigment (phaeomelanin) and black pigments (eumelanin) and can-
not be distinguished, while nonagouti (a/a) mice have black hairs (eumelanin).
Hence A is dominant, whereas a is recessive. True dominant alleles (such as
Caracul—Krt71Ca or Rex—Krt25Re) are uncommon in the mouse. In many cases
mice homozygous for the dominant allele are recessive lethal (Ay/+ mice are yel-
low while Ay/Ay genotypes die in utero).
Dominant negative mutation A mutation that prevents a wild-type allele in
the same cell from functioning. Dominant negative mutation commonly acts by
producing an altered polypeptide that prevents the assembly of a multimeric pro-
tein. A number of mutations at the Kit locus, that encodes the KIT tyrosine kinase
receptor act in a dominant-negative manner. Mutations in the gene Col26a1—
encoding collagen–are dominant negative.
Double heterozygotes Animals heterozygous at two loci on the same chromo-
some. Depending on the alleles assortment, double heterozygotes can be in cou-
pling (AB/ab) or in repulsion (Ab/aB).
Downstream A region of the DNA molecule that lies 3' to the point of reference.
Glossary 395
Draft sequence Preliminary form of a sequence with gaps and errors.
Electrophoretic variant Allelic variant of an enzyme (allozyme) with altered

electrophoretic mobility due to a charge-changing amino-acid substitution.
Embryonic stem (ES) cells Pluripotent cells taken from the inner cell mass of
blastocyst stage embryos, which are cultivated in vitro as a permanent cell line.
These cells can be genetically modified in a number of ways and, once reinserted
into a developing embryo (blastocyst), they are capable of participating in the for-
mation of the germline (see Chap. 8).
Embryonal carcinoma EK cells Pluripotential cells derived from spontane-
ous or experimental teratocarcinomas (EC) or from embryos (EK cells) that were
used for studying some aspects of early mouse embryogenesis. These cells were
“precursors” of ES cells. They were named after Evans and Kaufman, who made
extensive use of them.
ENCODE The ENCODE project (ENCyclopedia Of DNA Elements) has set as
its major aim to establish a catalog of all the structural and functional elements of
the genome.
Endogenous retroviruses (ERVs) A DNA sequence resulting from the integra-
tion of more or less complete retroviral copies into the mouse genome. ERVs are
flanked by two long terminal repeats (LTR).
Endonuclease An enzyme that, unlike exonucleases, cleaves RNA or DNA mol-
ecules at an internal position rather than progressively from either the 5' or 3' end.
Enhancer A (50–1500 bp) DNA sequence that increases the expression of a gene
when bound to a transcription factor (activator). Enhancers are generally cis-acting
and in most cases they exert their influence irrespective of their location or orienta-
tion. Enhancers are numerous in mammalian genomes.
Ensembl Genome browser and database of sequenced genomes jointly main-
tained by the European Bioinformatics Institute and the Wellcome Trust Sanger
Institute. Accessible on the web by the URL www.ensembl.org.
Epistasis Epistasis characterizes the interaction between non-allelic genes in which
one gene suppresses or enhances the expression of another. When homozygous the
albino allele (Tyrc) at the Tyr locus is epistatic over all other genes with an action on
coat color. Epistasis more generally describes cases where the phenotype controlled
by two loci cannot be predicted by considering the two loci independently. Epistasis
is a very common situation among genes that control complex traits.
Estrous cycle The estrous cycle consists of cyclic physiological changes induced
by reproductive hormones in mammalian females. In the mouse, the estrous cycle
lasts 4 to 6 days and is arbitrarily divided into four stages: proestrus, estrus, metes-
trus, and diestrus.
Ethyl methane sulfonate (EMS) A potent alkylating agent that is active on post-
meiotic germ cells and ES cells.
396 Glossary
Ethyl nitrosourea (ENU) A highly potent alkylating agent used to introduce ran-
dom mutations (mostly base pair changes) in the mouse DNA. ENU is active on
pre- and post-meiotic germ cells.
Euchromatin The main fraction of chromosomal DNA that is uncoiled dur-
ing interphase, and contains transcriptionally active regions. The other fraction is
heterochromatin.
Exon trapping Special technique used in the past to search for coding sequences
(exons).
Exon The part of a gene sequence that remains present within the messenger
RNA (mRNA) after introns have been removed by RNA splicing. The word was
coined from “expressed region”.
Exonuclease An enzyme that, unlike endonucleases, degrades progressively RNA
or DNA molecules from either the 5' or 3' end rather than at an internaI position.
Expressed sequence tag (EST) ESTs are short sub-sequences (~350 to 500 bp)
of a cDNA sequence, starting in general from the 3' end, sometimes from the 5'
end. ESTs can be used as molecular probes to retrieve the complete transcript of a
gene.
Expressivity A genotype exhibits variable expressivity when individuals with that
genotype differ in the extent to which they express the phenotype normally associ-
ated with that genotype. Mice heterozygous for the brachyury mutation (T/+) are
usually characterized by short tails, but the length of their tail is highly variable
from one mouse to the other. The T mutation exhibits variable expressivity. Such
variations can be caused by environmental factors, by modifier genes or by chance
(developmental noise).
F1 The offspring of a cross between two different inbred strains (see also hybrid
F1).
FANTOM research project Functional Annotation of the Mouse Genome
(FANTOM) is an international research consortium founded in 2000 by Dr.
Hayashizaki and his colleagues at RIKEN in Tokyo, Japan with the aim to func-
tionally annotate the mouse DNA sequence. FANTOM has since developed and
expanded over time to encompass the regulation of genes, networks of genes and
their impact in disease.
Fingerprinting Any method that identifies unique features of a clone that can
be used to determine overlaps between this clone and other clones in a library.
Restriction sites are useful tools for DNA fingerprinting.
Finished sequence The final form of a sequence from the Mouse Genome con-
taining less than 1 error in 10,000 bp.
FISH Fluorescent in situ hybridization (see Chap. 4).
Glossary 397
Fluorophore A chemical substance that can re-emit light upon stimulation by

light of a particular wavelength (excitation light).
Forward genetics (positional cloning) A strategy whose aim is to character-
ize the structural alteration(s) at the genome level that is (are) associated with (or
responsible for) a specific phenotype. The strategy proceeds from phenotype to
genotype and, for this reason, it is often referred to as forward genetics. It is the
opposite of reverse genetics.
Frameshift mutation Deletion or insertion of a few base pairs (not a multiple of
3!) that alters the reading frame downstream of it.
Functional genomics The study of the function of genes in a genome.
Genetic drift The unavoidable evolution of the genetic structure of a population

over generations. In fully inbred strains, genetic drift results from spontaneous
neutral mutations that disappear or become fixed throughout generations. Residual
heterozygosity in partially inbred strains is often responsible for genetic drift.
Genetic marker Any gene or short DNA sequence whose chromosomal location
is precisely known and whose structure exhibits variations among the individu-
als of the same strain or species. A genetic marker can be a phenotypic marker or
molecular marker (see Chap. 4).
Genome The total genetic information present within a single cell nucleus of an
animal. The haploid genome of the mouse is 3 × 109 bp and encodes 41,968 genes
(genes with nucleotide sequence data—as of November 2014) (see Chap. 5).
Genomic Library A collection of DNA clones large enough to guarantee that
any sequence of interest in the genome is likely to be present in at least one clone.
Genotype The set of alleles present at one or more loci of a given individual.
Giemsa A stain used to accentuate visually the difference between bands and
interbands on metaphase chromosomes (see Chap. 3).
Golden path The set of minimally overlapping cloned DNA that covers a large
segment (and ideally all) of the genomic DNA of a given chromosome.
Haldane’s rule In interspecific hybrids, the heterogametic sex is more severely
affected by traits that concern viability or sterility. For example, hybrids between
Mus spretus and Mus musculus domesticus are male sterile but female fertile.
Haplotype Pertaining to a particular set of alleles that are found together at linked
loci. In linkage studies, haplotypes provide very reliable data for determining the
order of loci (see Chaps. 4 and 9).
HAVANA (for Human and Vertebrate Analysis and Annotation of the
genome): A program undertaken by a team at the Sanger Institute for the system-
atic and careful annotation of the human, mouse and zebrafish genomes.
398 Glossary
Hemizygous Describes an individual who has only one member of a chromo-

some pair or chromosome segment rather than the usual two. X-linked genes in
males are hemizygous. Chromosomal deletions and transgenic insertions are often
hemizygous.
Heterozygote An animal with two different alleles at a particular locus. In this
case, the locus is considered heterozygous.
Hierarchical shotgun sequencing (HSS) A sequencing strategy that makes use
of cloned DNA with large inserts previously assembled into a series of overlap-
ping contigs.
High-resolution/high density map A genetic map with a great number of genetic
(mostly molecular/DNA) markers accurately mapped. High resolution and high
density maps are constructed based on large sized progenies of crosses between
strains with a high level of genetic polymorphism. Such maps have been instru-
mental for the complete sequencing of the mouse genome.
Histocompatible Pertaining to a genetic state in which cells from two animals
can be cross-transplanted without triggering rejection. Histocompatibility is con-
trolled by many genes. H2, on chromosome 17, is the major histocompatibility
complex or MHC.
Histones A class of alkaline proteins whose function is to package (protect ? iso-
late ?) the nuclear DNA. Histones are the major components of chromatin.
Homolog A gene that shares with another gene a common ancestry. The term
homolog may apply to the relationship between genes separated by the event of
speciation (see ortholog) or genetic duplication (see paralog).
Homozygote An animal with two identical alleles at a particular locus.
Hotspot, recombinational A region of chromosome, usually less than a few kilo-
bases in length, that participates in crossover events at a very high rate relative to
neighboring “cold” regions of chromosome (see Chap. 4).
Hybrid F1 The offspring of two homozygous individuals (e.g., inbred strains).
For example (C57BL/6 x C3H)F1 mice come from a cross between a female
C57BL/6 and a male C3H.
IBD/S Identical by Descent/State two genes or DNA segments with identical
nucleotide sequences in two or more individuals are said to be identical by state
(IBS). If they are identical because they were inherited from a common ancestor
then they are said to be identical by descent (IBD).
Ideogram A schematic representation of chromosomes indicating their relative
size, the position of the centromere and their banding patterns.
Imprinting A genetic mark that alters the expression of a gene. Imprinting var-
ies depending on the parent a given gene is inherited from. Only a small subset of
Glossary 399
genes in the mammalian genome is imprinted. The biological meaning of imprint-

ing is not yet known but abnormal imprinting can result in pathology.
Inbred Strain A strain that is essentially homozygous at all loci, typically pro-
duced by brother-sister matings for at least 20 successive generations. BALB/c
and C57BL/6J are popular mouse inbred strains.
Intercross A cross between two identical hybrid individuals (A/a x A/a).
Interference When a crossover occurs in a region, this affects the likelihood that
another crossover event occurs in the adjacent region. This interaction is called
interference. Interference varies in intensity among species.
Intra Cytoplasmic Sperm Injection (ICSI) (or micro-insemination) A proce-
dure that consists in the injection of a sperm head into the cytoplasm of the oocyte
in general by using a piezo-driven micromanipulator (see Chap. 2).
Introgression The introduction of one or several alleles of foreign origin into a
different gene pool. The word introgression is sometimes used when a segment
of chromosome containing a gene of interest is selectively transferred by sexual
reproduction, from its original background into another strain. Artificial introgres-
sion of a given gene, from one strain into another, results in a strain congenic for
the gene in question. Introgression of a complete chromosome from a donor strain
into a recipient strain results in a consomic strain.
Introns Introns correspond to the nucleotide sequences within a gene that are
removed by RNA splicing while the final mature RNA product is being gener-
ated. The word was coined from “intragenic region”. Some mammalian genes,
such as those encoding histones or tRNAs, have no introns. Some introns contain
sequences that are transcribed in non-protein coding RNAs.
Isogenic Individuals sharing identical genes (alleles). Identical twins, clones,
and individuals from an inbred strain are isogenic. Inbred strains are isogenic and
homozygous at all of their loci.
Junk DNA Coined by the geneticist Susumo Ohno, this expression referred to the
non protein-coding fraction of genomic DNA. Nowadays, geneticists consider that
the proportion of “junk DNA” in a mammalian genome is limited to only a few
percent and while most of the genomic DNA is transcribed (see Chap. 5).
Karyotype The number and appearance of the chromosomes in a eukaryotic cell.

Kilobase A stretch of 1,000 base pairs of DNA (abbreviation: kb).
Knock-in The targeted insertion of a (cloned) exogenous gene into the mouse genome
with the aim to disrupt an endogenous gene while expressing the transgenic one.
Knockout (KO) An animal with one of its gene inactivated by genetic engineer-
ing. A knockout gene can also result from a knock-in (see Chap. 8).
400 Glossary
Kozak sequence (gcc)gcc(A/G)ccAUGG (see Chap. 5).

Linkage group A set of loci in which all members are linked either directly or
indirectly to all other members of the set. A linkage group is equivalent to the
genetic information associated to a single chromosome.
Linkage map Genetic or meiotic map. A linkage map is based on linkage data.
Linkage Pertaining to the situation where two loci are close enough to each other
on the same chromosome that recombination frequency between them is reduced
to a level significantly less than 50%.
Locus Any genomic site, whether functional or not, that can be mapped through
formal genetic analysis.
LOD score The logarithm (base 10) of odds. The lod score is a statistical test
developed by Newton E. Morton that is used in linkage analysis. It compares the
likelihood of obtaining the test data if the two loci were indeed linked, to the like-
lihood of observing the same data in the absence of linkage. A LOD score of 3 or
more is traditionally considered significant to confirm linkage between two loci.
Long Conserved Non-Coding Sequences (see Ultraconserved Elements - UCE)
Long Interspersed Nuclear Elements (LINE) An important category of trans-
posable elements in mammals. Among these LINEs, the L1 family is the most fre-
quent (17–20% of mouse genomic DNA).
Mapping function A mathematical function that converts non additive recom-

bination fractions (because of multiple crossing-overs) into additive genetic dis-
tances. Several mapping functions have been proposed (Haldane, Kosambi, etc…)
to account for various levels of interference.
Megabase A stretch of 1,000,000 base pairs of DNA (abbreviation: Mb).
MegaMUGA (Mega Mouse Universal Genotyping Array) A genotyping tool

involving more than 77,000 informative SNP markers and covering the whole
mouse genome with an average spacing of 33 kb between markers. For informa-
tion concerning Mega MUGA refer to: http://csbio.unc.edu/CCstatus/Media/Mega
MUGAFlyer.pdf
Meiosis The process by which diploid germ cell precursors segregate their chro-
mosomes into the haploid nuclei of the gametes.
Meiotic product A single haploid genome within an egg or sperm cell.
Mendelian inheritance/proportions A pattern of segregation for a given pheno-
type that is (statistically) in agreement with Mendel’s laws of inheritance.
Metacentric A chromosome in which the centromere is in the middle of the
structure and the two arms of roughly the same size. When the centromere is
Glossary 401
shifted towards one end, the word sub-metacentric is used. Sub-metacentric chro-
mosomes have a long arm (symbol q) and a short arm (symbol p).
MicroRNA or miRNA A short sized (21–25 nt long) single stranded, non-coding
RNA molecule which functions in RNA silencing and post-transcriptional regula-
tion of gene expression (see Chap. 5).
Microsatellites A very short unit sequence of DNA (2–6 bp) that is repeated
multiple times in tandem. Microsatellites (also called simple sequence repeats or
SSRs) are highly polymorphic and have been very useful in linkage analysis (see
Chaps. 4 and 5). A polymorphism at a microsatellite locus is also referred to as a
simple sequence length polymorphism (SSLP) or Short Tandem Repeat (STR).
Minisatellites A highly polymorphic type of locus containing tandemly repeated
sequences having a unit length of 10–40 bp. Minisatellite polymorphisms can be
assessed by restriction fragment length polymorphism (RFLP) analysis or by poly-
merase chain reaction (PCR). Also referred to as variable number of tandem repeat
(VNTR) loci (see Chap. 5). These sequences are the base of the original “DNA
Fingerprinting” used in forensics.
Missense mutation A non-synonymous substitution in a codon that results in
the substitution of an amino acid for another (see Chap. 7). The Eiche’s dominant
spotting mutation at the Kit locus (KitW-ei) results from the replacement of the Gly
amino acid at position 597 by an Ala residue in the KIT receptor kinase receptor.
Model organism Any organism with a phenotype reminiscent of, or similar to
a human phenotype. Some mutant genotypes of the mouse are faithful (homolo-
gous) models of human diseases, others are much less faithful (analogous). Both
models are useful.
Monobrachial homology A mouse heterozygous for two Robertsonian translo-
cations of different origins with one arm in common. For example Rb(16.17) and
Rb(5.17), are said to be heterozygous with monobrachial homology for chromo-
some 17.
Monosomic A karyotype with 2n-1 chromosomes. Monosomy can be primary,
when one complete chromosome is missing or tertiary if only a fragment of chro-
mosome is missing.
Mosaics Mosaics are organisms composed of cells with a different genetic consti-
tution, although deriving from one and a single conceptus (see Chap. 2). Because
one of their two X-chromosomes is randomly inactivated, mammalian females
heterozygous for different X-linked alleles, are mosaics.
Mouse Clinic Large-scale phenotyping platforms where mouse mutants or strains
are thoroughly analyzed for the greatest possible number of parameters using a
panel of highly standardized protocols.
mtDNA Mitochondrial DNA (see Chap. 5).
402 Glossary
Multifactorial A trait controlled by at least two factors, which may be genetic or

environmental (see Chap. 10). Behavioral differences between inbred strains, such
as anxiety, are multifactorial traits.
Mutant allele A mutant allele at a locus is associated with a phenotype distinct
from that observed in individuals carrying the most common, so-called wild-type,
allele. Non-mutant alleles are often designated wild-type allele i.e. the most com-
mon form present at a given locus.
Mutation A new allele that arose abruptly and is present in the genome of an ani-
mal but not in the genome of either of its parent(see Chap. 7).
N2, N3, N4 etc. Symbols used to describe the generation of backcrossing and the
offspring that derive from it. The N2 generation describes offspring from the initial
cross between an interstrain FI hybrid and one of the parental inbred strains. Each
following backcross generation is numbered in sequence (see Chap. 9).
Neutral allele An allele with no noticeable effect on the phenotype. A missense
allele can be neutral if the change in nucleotide sequence does not affect the amino
acid sequence, or if the amino acid substitution has no effect on the protein func-
tion or stability.
Non-coding RNAs RNA molecules that are transcribed from the genome and do
not encode protein sequences. The Encyclopedia of DNA Elements (ENCODE)
project suggested that over 80% of the DNA in the mammalian genome is tran-
scribed and have an important biological function even if the function in question
is not yet elucidated.
Non-disjunction An accident occurring during the meiotic process leading to an
abnormal distribution of the chromosomes in the daughter cells (see Chap. 3).
Non-sense mutation The mutation of any codon towards a stop codon. Such a
mutation can truncate the protein.
Oligo-nucleotide A chain of nucleotides (nt) usually 10 to 500 nt long.
Oligonucleotides are often used as primers for polymerase chain reaction (PCR)
amplification.
ORF—Open reading frame The part of a (protein coding) DNA sequence that
contains no stop codons.
Orthologs Orthologs are genes in different species that evolved from a common
ancestral gene by speciation. Orthologous genes in general retain the same func-
tion in the course of evolution. Identification of orthologs is instrumental for reli-
able prediction of gene function in newly sequenced genomes.
Outcross A cross between genetically unrelated animals.
Overdominance A rrare condition in which the heterozygotes (M/m) have a phe-
notype that is more pronounced than that of either homozygotes (M/M and m/m)
Glossary 403
(see Chap. 6). Mice homozygous for the Mplhlb219 mutation in the thrombopoietin
(TPO) receptor MPL (Cys → Arg) have a 80% decrease in the number of platelets
in comparison to the wild-type mice. However, mice heterozygous for the same
Mplhlb219 allele show an overdominance effect with a significant increase in platelet
number.
p-arm The short arm of a sub-metacentric chromosome (“p” stands for petit—
small in French).
Paralog Paralogs are genes related by duplication within a genome. While ortho-
logs retain the same function in the course of evolution, paralogs evolve new
functions, even if these are related to the original one. The Keratin (Krt) and
Homeobox (Hox) genes have many paralogs in the mouse.
Pedigree A schematic representation of the filiation relationship in a family.
When the family is small the term micro-pedigree is often used.
Penetrance The fraction of individuals of a given genotype that effectively
exhibit the expected phenotype. Penetrance is usually expressed as a percentage.
Where less than 100% of genotypically mutant animals are phenotypically mutant,
the phenotype is said to be incompletely penetrant. The determinism of penetrance
is not known. In most cases it results from chance (developmental noise) but can
also be influenced by modifier genes.
Pericentric In the vicinity of the centromere or involving the centromere – exam-
ple: a pericentric inversion (see Chap. 3).
PFGE or Pulsed-field gel electrophoresis A technique for separating large DNA
molecules from each other (see Chap. 5).
Phenotype The physical manifestation of a genotype within an animal. A mutant
phenotype is caused by a mutant genotype and is manifested as an alteration
within an animal that distinguishes it from the wild-type. Phenotypes range from
severe malformations leading to death or debility to extremely subtle changes in
the physical properties of a biological molecule (for example its electrophoretic
charge).
Phenotypic marker Phenotypes for which the variation observed in a population
is entirely explained by a single “mendelian” factor.
Phylogenetic tree A diagram showing the postulated evolutionary relationships
that exist among related species in terms of their divergence from a series of com-
mon ancestors at different points in time (see Chap. 1).
Physical map A map based on a great number of minimally overlapping cloned
DNAs.
Pleiotropy Pleiotropy describes a situation where a mutant allele has an effect
on different (apparently unrelated) phenotypic traits. Mice homozygous for the
404 Glossary
piebald allele (Ednrbs) have defects in pigmentation, are deaf and often die from
megacolon. Piebald has pleiotropic effects.
Poly-A tail A stretch of poly(A) added at the 3' end of mRNAs during transcript
maturation and before splicing. The poly-A string ensures the stability of the
transcript.
Polygenic A phenotype resulting from the interactions of two or more genes with
alternative alleles (see Chap. 2).
Polymorphic A term formulated by population geneticists to describe loci at
which there are two or more alleles that are each present at a frequency of at least
1 % in a population of animals. Then, a polymorphism is a genotypic variation
within a population.
Polytypic species A species where several subspecies or geographical/morpho-
logical races are recognized. Mus m. domesticus is typically a polytypic species.
Position effect Corresponds to the variations in the expression of a gene when its
molecular environment is changed either after translocation or through transgen-
esis (see Chap. 8).
Positional cloning See Forward genetics.
Primary RNA The RNA molecule before splicing.
Primers Short oligonucleotides, which anneal to template DNA to prime PCR.
Promoter See TATA box; CAT-box and 5'UTR.
Proximal A relative term meaning closer to the centromere; the opposite of distal.
Pseudogene A DNA sequence that closely resembles a functional gene but is not
expressed. Processed pseudogenes do not have introns or promoters. They are cop-
ied from mRNA and incorporated into the genome. Unprocessed pseudogenes, orig-
inate from the retrotranscription of messenger RNAs back into the genomic DNA in
more or less random locations (see Chap. 5). Pseudogenes are sometimes extremely
difficult to differentiate from real genes and some of them even have a function.
q-arm The long arm of a sub-metacentric chromosome (q stands for queue)

QTL—quantitative trait locus (plural: Quantitative trait loci—QTLs) are
sequences of DNA containing or linked to the genes that determine a quantitative
trait.
Quantitative trait A phenotype that can vary in a quantitative manner when mea-
sured among different individuals (see Chap. 10). The variation in expression can
be due to combinations of genetic and environmental factors, as well as chance.
Radiation hybrids Somatic cell hybrids with a full set of hamster chromosome
and fragments of mouse chromosomes, generated by X or γ-irradiation, randomly
Glossary 405
inserted into the hamster chromosomes. These interspecific cell hybrids have been
very helpful for the (non-meiotic) chromosomal assignment of cloned genes in the
mouse.
Recessive allele A recessive allele expresses its characteristic phenotype only
when homozygous.
Reciprocal translocations Reciprocal (or balanced) translocations are rearrange-
ments resulting from a reciprocal exchange between the telomeric ends of two
non-homologous chromosomes with no change in the total genomic information
content. Reciprocal translocations are the most common form of structural rear-
rangements of the mouse karyotype.
Recombinant congenic strain (RCS) A variation on recombinant inbred strains
in which the initial outcross is followed by several generations of backcrossing
prior to inbreeding (see Chap. 9).
Recombinant inbred (RI) strain A special type of inbred strain formed from an
initial outcross between two inbred strains followed by at least 20 generations of
inbreeding (see Chap. 9).
Recombinant The result of a crossing-over in a doubly heterozygous parent such
that alleles at two loci flanking the crossing-over that were present on opposite
homologs are put together on the same homolog.
Restriction fragment length polymorphism (RFLP) A DNA variation that
affects the distance between contiguous restriction sites (most often a nucleotide
change that creates or suppresses a site) within or flanking a DNA fragment that
hybridizes to a cloned probe (see Chap. 4). RFLPs are detected upon Southern
blot hybridization. This polymorphism has been extensively exploited as a genetic
polymorphism.
Retrotransposon or retroposon An inserted genomic element that originated
from the reverse transcribed mRNA produced from another region of the genome
(see Chap. 5).
Reverse genetics A strategy whose aim is to characterize the function of a gene
by analyzing the consequences, at the phenotypic level, of alterations occurring
spontaneously or engineered at the DNA level (the opposite of forward genetics).
Robertsonian translocation A fusion between the centromeres of two acro-
centric chromosomes producing a single metacentric element (see Chap. 3).
Robertsonian translocations reduce the number of centromeres but do not alter the
number of chromosome arms.
Round spermatid injection (ROSI) The fertilization of super-ovulated oocytes
with the nucleus of round spermatids. When the round spermatid is removed from
young males, this technique dramatically reduces the time required for the devel-
opment of fully congenic mouse strains (see Chap. 2).
406 Glossary
Satellite DNA A discrete fraction of DNA visible in a cesium chloride density-

gradient as a “satellite” to the main DNA band. The term refers to all simple
sequence DNA having a centromeric location (see Chap. 5).
Semidominant An allele is said to be semidominant when the phenotype of the
heterozygotes is intermediate between the phenotype for the dominant allele and
the recessive allele. The KitW-f mutant allele is typically semidominant: KitW-f/+
mice have a fuzzy coat, while KitW-f/W-f and Kit+/+ mice are white and fully pig-
mented respectively.
Shotgun sequencing A method of sequencing DNA that does not require the
physical mapping of large sized cloned fragments. For shotgun sequencing, the
DNA if fragmented mechanically (i.e. randomly) into segments with a size rang-
ing from 100 to 1,000 bp. The fragments are then sequenced using the chain ter-
mination method. Finally the individual sequences are ordered in silico into a
continuous sequence based on the sequence overlapping (see Chap. 5).
Silencer A DNA sequence capable of binding regulatory sequences.
Silent (or synonymous) substitution A mere SNP with no effect on the protein
sequence. For example, the codons CCT and CCC both code for the proline amino
acid.
Simple sequence length polymorphism (SSLP) The polymorphism at a micro-
satellite locus. Also called SSR for “simple sequence repeats”.
SINE Short interspersed element. Families of selfish DNA elements that are a few
hundred base pairs in size and dispersed throughout the genome (see Chap. 5).
Single nucleotide polymorphism (SNP) A one base-pair difference between two
DNA sequences that is either natural or induced. SNPs can be used as molecular
markers.
Spliceosome A set of highly specific molecules (at least five small nuclear RNAs
and around 150 proteins) that is essential for RNA splicing (see Chap. 5).
Splicing A mechanism leading to the excision of the introns, i.e. the regions
flanked by a splicing donor site and a splicing acceptor site (see Chap. 5).
spretus Abbreviated form of Mus spretus, a mouse species commonly used in
interspecific matings for the generation of high-density/high-resolution linkage
maps (see Chaps. 2 and 9). This species is common in the western Mediterranean
border.
SSCP (single strand conformation polymorphism) A sensitive gel-based tech-
nique (different from denaturing gradient gel electrophoresis, DGGE) for detect-
ing single nucleotide changes within allelic polymerase chain reaction (PCR)
products that have been denatured and gel-fractionated as single strands.
SSLP Simple Sequence Length Polymorphism; SSR: Simple Sequence Repeat;
and STR: Short Tandem Repeat.; see microsatellite.
Glossary 407
Strain distribution pattern (SDP) The distribution of the segregating alleles at

a single locus across a group of animal used for analysis in a linkage study (see
Chap. 9). Used in the context of backcross data and data obtained from recombi-
nant inbred (RI) strains.
Strain Refers to a population of mice with known lineage that are bred within a
closed colony in order to maintain certain defining characteristics. Inbred strains
are produced by brother-sister matings. Random-bred colonies are called outbred
stocks (see Chap. 3).
Superovulation Superovulation of female mice allows harvesting large numbers
of fertilized eggs or unfertilized oocytes for the purpose of experimental manipula-
tion. It requires the injection of females aged 3 to 5 weeks with gonadotropin hor-
mones that artificially induces ovulation.
Sympatric Refers to related species that have overlapping ranges in nature but do
not interbreed. In different parts of its range, Mus musculus is sympatric with Mus
macedonicus, Mus spicilegus, and Mus spretus (see Chap. 2).
Syngenic Literally “of the same genotype.” Used most frequently by immunolo-
gists to describe interactions between cells from the same inbred strain.
Syntenic Describes the physical co-localization of genetic loci on the same chro-
mosome within an individual or species. Conserved synteny refers to the situation
where two linked loci in one species have homologs that are also linked in another
species.
Targeted Mutation A type of mutation in which a DNA construct assembled in

vitro substitutes a gene in the genome. The constructs designed to eliminate gene
function (loss of function) are often referred to as knockouts (KO), while con-
structs designed to introduce a mutation in the gene sequence are often referred to
as knock-ins (KI).
TATA box The first core promoter sequence identified in eukaryotic protein-cod-
ing genes. The TATA box is the binding site of transcription factors or histones.
Only 25% of mammalian genes contain a TATA box.
Taxon (plural taxa) Any recognized level in the systematical nomenclature (e.g.
species, subspecies…).
Telocentric A chromosome in which the centromere is at one end. Many cytoge-
neticists consider that telomeric chromosomes, in fact, are acrocentric with a very
short arm. True telocentric chromosomes are probably instable structures.
Telomere The distal end of a chromosome. Telomeres consist of repetitive
sequences and are considered as insulators whose role is to prevent the chromo-
some ends of being damaged. Telomeres may play a fundamental role in the con-
trol of senescence.
408 Glossary
Transgene A fragment of foreign DNA that has been incorporated (randomly)

into the genome through the manipulation of pre-implantation embryos. The indi-
viduals carrying a transgene are called transgenic (Tg) (see Chap. 6).
Translocation Pertaining to a novel chromosome formed by breakage and
reunion of DNA molecules into a new configuration (see Chap. 5).
Transposons Short DNA sequences that have the capacity to change their posi-
tion within the genome. Some transposons have been engineered to serve as tool
for the generation of tagged mutations (see Chap. 7).
Triploïd A conceptus with 3n chromosomes instead of 2n.
Trisomic A conceptus with 2n + 1 chromosomes instead of 2n − symbol Ts. All
trisomic mice (except those for chromosome 19) are inviable. Models for human
chromosome 21 trisomy have been developed in the mouse (see Chap. 3).
Twins Monozygotic twins are extremely rare in mice, if even they exist.
Ultraconserved Elements (UCE) UCEs are highly conserved DNA sequences

shared among evolutionary distant taxa. In most cases, the function(s) of these
UCEs is unknown but might not be essential.
Unequal crossover A crossover event that occurs between non-allelic sites.
Unequal crossover can lead to the duplication of sequences on one homolog and
the deletion of sequences on the other (see Chap. 5). The deleted haplotype is, in
most cases, eliminated while the duplicated haplotype generate a CNV.
Variant Literally, an alternative form. Used in conjunction with locus, phenotype,

or mouse strain. A ‘DNA variant’ is equivalent to an alternative DNA allele. A
variant mouse usually refers to an animal that carries a mutant allele or expresses a
mutant phenotype.
VNTR “Variable number of tandem repeats” locus; see Minisatellite.
Whole-genome shotgun sequencing (WGS) A sequencing strategy that consists

of the mechanical fragmentation (e.g., by sonication) of the mammalian DNA into
short segments, which are sequenced from both ends using the chain termination
method. WGS is fast and less expensive than hierarchical sequencing.
Wild type An allele that functions normally and is commonly found in wild
populations.
YAC Yeast artificial chromosome. A vector for cloning genomic inserts from
300 kb to 1 Mb in length. YACs are relatively unreliable vectors, some being
deleted and others chimerical. They must then be analyzed with great care.
Zygote The fertilized egg containing pronuclei from both the mother and the
father.

Genetics of The Mouse

Uploaded by

Copyright:

Available Formats

Genetics of The Mouse

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Genetics of The Mouse

Uploaded by

Copyright:

Available Formats

Jean Louis

Genetics of the Mouse

ISBN 978-3-662-44286-9 ISBN 978-3-662-44287-6 (eBook)

Library of Congress Control Number: 2014945779

Springer Heidelberg New York Dordrecht London

© Springer-Verlag Berlin Heidelberg 2015

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

1 Origins of the Laboratory Mouse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2 Basic Concepts of Reproductive Biology and Genetics. . . . . . . . . . . . 19

2.3.3 Epistasis and Pleiotropy. . . . . . . . . . . . . . . . . . . . . . . . . 41

4.4 High-Resolution, High-Density Genetic Maps . . . . . . . . . . . . . . . 110

5 The Mouse Genome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

6 Epigenetic Control of Genome Expression. . . . . . . . . . . . . . . . . . . . . . 187

6.3 Parental Imprinting of Autosomal Genes. . . . . . . . . . . . . . . . . . . . 196

7 Mutations and Experimental Mutagenesis. . . . . . . . . . . . . . . . . . . . . . 221

8 Transgenesis and Genome Manipulations . . . . . . . . . . . . . . . . . . . . . . 267

9 The Different Categories of Genetically Standardized

9.2.3 The Genetic Purity of Inbred Strains Must be

10 Quantitative Traits and Quantitative Genetics. . . . . . . . . . . . . . . . . . . 361

10.11 Using Congenic Strains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377

1.1.1 Phylogenetic Relationships of Laboratory Mice

Figure 1.2 represents a phylogenetic tree of a sample of 28 vertebrate species,

© Springer-Verlag Berlin Heidelberg 2015 1

hence the importance of the binominal nomenclature.

Fig. 1.2 Phylogenetic tree representing the relationships between 28 vertebrate species. Branch

species. Figure 1.4 summarizes the phylogenetic relationships within the genus

1.1.2 How the House Mouse Became a Domestic Species

The beginning of human/mouse commensalism is very ancient and probably dates

biological consequences of an increase in air pressure. Much later, Joseph Priestley

1.1.3 How the House Mouse Became a Model for Geneticists

1.1.3.1 From Louis-Théodore Coladon to Gregor Mendel and Lucien

Once rediscovered by H. de Vries, C. E. Correns, and E. von Tschermark-

1.1.3.2 The Origins of Laboratory Mouse Strains

a pathologist working at the University of Pennsylvania), she made interesting

1.1.3.3 Mice Have Been Instrumental in Research in Biology

1.1.4 The Community of Mouse Geneticists

1.1.5 The Main Institutions Involved in Mouse Genetics

The Jackson Laboratory (internationally known as the JAX-Lab), which was

1.1.6 Books and Other Sources of Information Concerning

1.1.7 The Future of Mouse Genetics

The state of mouse genetics, as expected, changed dramatically at the turn of

Artzt K (2012) Mammalian developmental genetics in the twentieth century. Genetics

2.2 Reproduction in the Laboratory Mouse

2.2.1 The Estrous Cycle and Pregnancy

1 The website http://informatics.jax.org/ is a fundamental database resource for the laboratory

© Springer-Verlag Berlin Heidelberg 2015 19

2.2.2 Inducing Ovulation in the Mouse (Superovulation)

The information provided in the preceding section concerning mouse reproduction

(prepubertal females) are treated by injection of gonadotropin hormones and

2.2.4 In Vitro Fertilization in the Mouse

When a mutant or transgenic female is potentially fertile (i.e., when it produces

2.2.6 Intra Cytoplasmic Sperm Injection

Intra cytoplasmic sperm injection (ICSI, also known as micro-insemination)

2.2.7 Cryopreservation of Mouse Embryos and Spermatozoa

2.2.8 Twinning in the Mouse

occasionally observed in utero at very low frequency, between embryonic days 8

2.2.9 Cloning Laboratory Mice

Cloning is an asexual method of reproduction that is commonly used in plants

1.1.1 Phylogenetic Relationships of Laboratory Mice

1.1.2 How the House Mouse Became a Domestic Species

1.1.3 How the House Mouse Became a Model for Geneticists

1.1.3.1 From Louis-Théodore Coladon to Gregor Mendel and Lucien

1.1.3.2 The Origins of Laboratory Mouse Strains

1.1.3.3 Mice Have Been Instrumental in Research in Biology

1.1.4 The Community of Mouse Geneticists

1.1.5 The Main Institutions Involved in Mouse Genetics

1.1.6 Books and Other Sources of Information Concerning

1.1.7 The Future of Mouse Genetics

2.2 Reproduction in the Laboratory Mouse

2.2.1 The Estrous Cycle and Pregnancy

2.2.2 Inducing Ovulation in the Mouse (Superovulation)

2.2.4 In Vitro Fertilization in the Mouse

2.2.6 Intra Cytoplasmic Sperm Injection

2.2.7 Cryopreservation of Mouse Embryos and Spermatozoa

2.2.8 Twinning in the Mouse

2.2.9 Cloning Laboratory Mice

2.2.10 Mosaics and Chimeras

2.3 Basic Notions of Genetics

2.3.1 Genes and Alleles

2.3.3 Epistasis and Pleiotropy

2.3.4 Penetrance and Expressivity

2.4 Phenotyping Laboratory Mice: The Mouse Clinics

3.2 The Chromosomes of the Mouse

3.3 Identifying the Chromosome Pairs: The Normal

3.4 Meiosis and Gametogenesis

3.5 Variations in Chromosome Number

3.5.1 The Euploid Heteroploidies

3.5.2 The Aneuploid Heteroploidies

3.5.2.1 Nullisomies and Monosomies

3.5.2.2 Trisomies, Tetrasomies, Double Trisomies etc

3.5.2.3 Aneuploidies of the Sex Chromosomes

3.6 Variations in Chromosome Structure

3.6.1 The Structural Rearrangements Resulting from a

3.6.2 The Structural Rearrangements Resulting from Two

3.6.2.1 The Structural Rearrangements Resulting from Two Breaks in

Reciprocal or balanced translocations, as the name indicates, result from a r eciprocal