Series Editor. Editorial Board. Founding Editors. Paul M. Wassarman. Olivier Pourquié. Blanche Capel. B. Denis Duboule. Anne Ephrussi.

Transcription

1

2 Series Editor Paul M. Wassarman Department of Developmental and Regenerative Biology Mount Sinai School of Medicine New York, NY USA Olivier Pourquié Howard Hughes Medical Institute Stowers Institute for Medical Research Kansas City, MO USA Editorial Board Blanche Capel Duke University Medical Center Durham, NC, USA B. Denis Duboule Department of Zoology and Animal Biology NCCR Frontiers in Genetics Geneva, Switzerland Anne Ephrussi European Molecular Biology Laboratory Heidelberg, Germany Janet Heasman Cincinnati Children s Hospital Medical Center Department of Pediatrics Cincinnati, OH, USA Julian Lewis Vertebrate Development Laboratory Cancer Research UK London Research Institute London WC2A 3PX, UK Yoshiki Sasai Director of the Neurogenesis and Organogenesis Group RIKEN Center for Developmental Biology Chuo, Japan Cliff Tabin Harvard Medical School Department of Genetics Boston, MA, USA Founding Editors A. A. Moscona Alberto Monroy

3 Academic Press is an imprint of Elsevier 525 B Street, Suite 1900, San Diego, CA , USA 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA 32, Jamestown Road, London NW1 7BY, UK Linacre House, Jordan Hill, Oxford OX2 8DP, UK First edition 2009 Copyright # 2009 Elsevier Inc. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) ; fax (+44) (0) ; permissions@elsevier.com. Alternatively you can submit your request online by visiting the Elsevier web site at http: //elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made ISBN: ISSN: For information on all Academic Press publications visit our website at elsevierdirect.com Printed and bound in USA

4 CONTRIBUTORS Jeremy S. Dasen Smilow Neuroscience Program, Department of Physiology and Neuroscience, New York University School of Medicine, New York, NY, USA Nicolas Denans Stowers Institute for Medical Research, Kansas City, Missouri, USA Jacqueline Deschamps Hubrecht Institute, Developmental Biology and Stem Cell Research, Uppsalalaan, Utrecht, The Netherlands Walter J. Gehring Department of Cell Biology, Biozentrum, University of Basel, Klingelbergstrasse Basel, Switzerland Tadahiro Iimura Tokyo Medical and Dental University, Global Center of Excellence (GCOE) Program, International Research Center for Molecular Science in Tooth and Bone Diseases, Tokyo, Japan Thomas M. Jessell Departments of Neuroscience, and Biochemistry and Molecular Biophysics, Howard Hughes Medical Institute, Kavli Institute for Brain Science, Columbia University, New York, NY, USA Rohit Joshi Department of Biochemistry and Molecular Biophysics, Columbia University, New York, NY, USA François Karch Department of Zoology and Animal Biology and NCCR Frontiers in Genetics, University of Geneva, Geneva, Switzerland Urs Kloter Department of Cell Biology, Biozentrum, University of Basel, Klingelbergstrasse Basel, Switzerland Robb Krumlauf Stowers Institute for Medical Research, Kansas City, Missouri, USA, and Department of Anatomy and Cell Biology, Kansas University Medical School, Kansas City, Kansas, USA ix

5 x Contributors Katherine M. Lelli Department of Genetics and Development, Columbia University, New York, NY, USA Robert K. Maeda Department of Zoology and Animal Biology and NCCR Frontiers in Genetics, University of Geneva, Geneva, Switzerland Richard S. Mann Department of Biochemistry and Molecular Biophysics, Columbia University, New York, NY, USA Yuichi Narita Friedrich Miescher Institute for Biomedical Research, Basel, Switzerland, and Institut de Génétique et de Biologie Moléculaire et Cellulaire, Illkirch Cedex, C.U. de Strasbourg, France Olivier Pourquié Stowers Institute for Medical Research, Howard Hughes Medical Institute, Kansas City, Missouri, USA, and Department of Anatomy and Cell Biology, The University of Kansas School of Medicine, Kansas City, Kansas, USA Filippo M. Rijli Friedrich Miescher Institute for Biomedical Research, Basel, Switzerland, and Institut de Génétique et de Biologie Moléculaire et Cellulaire, Illkirch Cedex, C.U. de Strasbourg, France Hiroshi Suga Department of Cell Biology, Biozentrum, University of Basel, Klingelbergstrasse Basel, Switzerland, and Present address: Barcelona Science Park, Universitat de Barcelona, Barcelona, Spain Stefan Tümpel Stowers Institute for Medical Research, Kansas City, Missouri, USA, and Present address: Institut für Molekulare Medizin und Max-Planck-Forschungsgruppe Stammzellalterung, Universität Ulm, Ulm, Germany Deneen M. Wellik Department of Internal Medicine, Division of Molecular Medicine & Genetics, and Department of Cell and Developmental Biology, University of Michigan Medical Center, Ann Arbor, Michigan, USA Leanne M. Wiedemann Stowers Institute for Medical Research, Kansas City, Missouri, USA, and Department of Pathology and Laboratory Medicine, Kansas University Medical School, Kansas City, Kansas, USA Teddy Young Hubrecht Institute, Developmental Biology and Stem Cell Research, Uppsalalaan, Utrecht, The Netherlands

6 PREFACE Hox genes were discovered almost 30 years ago as one of the very first unifying principles in development of Bilaterian species. These genes code for a family of conserved transcription factors which, in most species, are organized in clusters along chromosomal territories. Their action and distribution along the anteroposterior axis of the embryo, as well as their deployment in time, exhibit a striking order which reflects their linear organization on the chromatin. This peculiar arrangement, termed colinearity, was first recognized by Ed Lewis in the fly embryo. Despite very active research during the past three decades, Hox gene regulation and function remain extremely mysterious. Particularly, the molecular mechanism underlying the fundamental property of colinearity found in organisms ranging from flies to humans remains unknown. Hox mutations exhibit spectacular homeotic properties, whereby the identity of a body segment can be changed in that of a different segment. Whereas these effects are now well characterized in vertebrates and invertebrates at the phenotypic level, the molecular details of the targets and functions of Hox proteins underlying these identity changes are poorly understood. Thus, despite the wealth of research focused on Hox genes, major questions are still to be answered. A large body of literature on Hox genes has been published since the seminal paper from Ed Lewis in 1978, but strikingly, very few monographs have been devoted to this fascinating topic. Thus, the goal of this book is to provide a comprehensive and up-to-date summary of recent developments in the field of Hox biology. This is a large field and due to space limitations, some areas might be covered more extensively than others. In this book, we cover some history of the characterization of the Hox complexes in the fly, as well as discussions of the organization, regulation, and function of Hox genes in patterning the body axis in invertebrates (essentially Drosophila) and in vertebrates. The book begins with a chapter by Robert K. Maeda and François Karch who recapitulates the history of the discovery of the BX-C complex in the fly and describes the striking colinear organization of the cis-regulatory elements controlling expression of the Ubx, AbdA, and AbdB genes initially recognized by Ed Lewis. This organization is strikingly different from that of vertebrate Hox clusters where no such colinear distribution of the cis-regulatory sequences is observed. Genes involved in the early patterning of the embryo, such as the gap and pair rule genes, control the initiation of Hox xi

7 xii Preface gene expression, whereas their maintenance requires the genes of the Polycomb and Trithorax complexes that act on chromatin to respectively maintain the repressed or activated configurations of Hox genes. The chapter also describes our current understanding of the role of specific chromatin domains and regulators in the colinear regulation of the BX-D cis-regulatory elements. The next chapter by Walter Gehring, Urs Kloter, and Hiroshi Suga presents genetic and phylogenetic arguments, supporting the notion that the second thoracic segment in the fly (T2), which is specified by Antenapedia, corresponds to a developmental and evolutionary ground state in Bilaterians. In vertebrates, this ground state would correspond to the thoracic level patterned by the Hox6 group. Based on these arguments, the authors further argue about the origin of the Hox clusters by duplication and unequal crossing-over, leading to the progressive addition of genes in between the two extremities of the cluster. In various circumstances in flies and vertebrates, the posterior Hox genes have been shown to be functionally dominant over the anterior ones, a property called posterior prevalence or phenotypic suppression. The arguments developed and the model proposed by Gehring and colleagues in this chapter clearly challenge this notion for the genes expressed anterior to T2 in flies. Strikingly, very little is known about the mode of action and the targets regulated by Hox proteins. A longstanding paradox in the field is the relative lack of specificity of the Hox binding sequences compared to the exquisite developmental functions assumed by individual Hox proteins. In the third chapter, Richard Mann, Katherine Lelli, and Rohit Joshi discuss this question and argue in favor of different modes of Hox regulation. They survey the different kinds of Hox targets characterized, and distinguish very specific targets recognized by a single paralog, from targets showing less specificity and, hence, recognized by several Hox factors. Hox proteins can act with cofactors, such as the TALE proteins, that help cooperative binding to DNA. This cooperative binding induces conformational changes revealing novel, specific binding properties of the complexes to their DNA targets. Finally, they discuss how Hox proteins also interact with collaborators, forming what they call a Hoxasome, which influences the transcriptional outcome of the Hox-based regulation. In the fly and vertebrates, Hox genes are not expressed in the anteriormost part of the brain (telencephalon, diencephalon, and mesencephalon), which is patterned by other Homeobox-containing genes, such as Otx in vertebrates or orthodenticle in the fly. In the vertebrate central nervous system, Hox genes are involved in the patterning of the hindbrain and the spinal cord. The hindbrain or rhombencephalon corresponds to the posterior part of the brain, which is transiently segmented into seven rhombomeres. These segments define compartments that acquire distinct functional identities during development, and which control a variety of physiological functions such as respiration. In the hindbrain, Hox genes are expressed

8 Preface xiii segmentally, with their expression boundaries respecting the rhombomeric frontiers. Mutations in the mouse demonstrated that Hox genes play a key role in the control of the identity of the rhombomeres. The regulation and role of Hox genes in patterning the vertebrate nervous system has been extensively studied in the hindbrain and more recently in motoneurons. The fourth chapter by Stefan Tumpel, Leanne Wiedemann, and Robb Krumlauf summarizes our current understanding of the role of anterior Hox genes in early patterning of the hindbrain in vertebrates. It describes the cis-regulatory codes and regulatory networks established during hindbrain differentiation by the Hox1 to Hox4 paralog genes which are involved in initiating and maintaining Hox gene expression at the appropriate segmental level. Chapter 5, by Yuichi Narita and Filippo Rijli, focuses on the later functions of Hox genes in hindbrain development and, more specifically, on their role in the establishment of the complex neuronal connectivity that underlies various important physiological functions and behaviors. In chapter 6, Jeremy Dasen and Tom Jessell discuss our understanding of the role of Hox transcription factors in the patterning of the spinal cord motoneurons. Recent elucidation of the role of these factors in the establishment of the various levels of motoneuron organization such as columns and pools is detailed, as well as the recognition of key cofactors such as FoxP1 in this process. In vertebrates, aside from the nervous system, the role of Hox genes in axial patterning has been examined in great detail at the level of the vertebral column. The spine is progressively formed in a head-to-tail direction during embryogenesis by the rhythmic addition of vertebral precursors, termed somites. The somitic columns formed during embryogenesis become subsequently patterned into different anatomical regions when somitic derivatives differentiate to form the vertebrae and associated muscles. Mouse knock-out experiments have shown that somite regional identity is largely controlled by Hox genes. In chapter 7, Tadahiro Iimura, Nicolas Denans, and Olivier Pourquié describe the early colinear activation of Hox gene expression in the precursors of the vertebrate spine (the paraxial mesoderm), which roughly positions the Hox expression domains at the appropriate axial level. They discuss how this temporal expression sequence is translated into the characteristic colinear expression domains along the forming axial skeleton. A second repositioning phase involving the segmentation machinery is also required for the definitive positioning and subsequent maintenance of the Hox expression domains in the somites. Interestingly, such a two-step regulation of Hox expression in the developing embryo is also described for fly embryos in chapter 1 and for the hindbrain in chapter 4. Strikingly, in these different systems, it largely relies on different mechanisms. Related issues are also discussed in chapter 8 by Teddy Young and Jacqueline Deschamps, which is furthermore concerned with the role of

9 xiv Preface the Cdx genes in the regulation of Hox genes in the embryo. Cdx genes belong to the ParaHox cluster, which was proposed to share a common evolutionary origin with the Hox cluster. In chapter 9, Deneen Wellik details the later function of Hox genes in patterning the vertebrate axis once the complex nested expression patterns are established in the embryo. The regional patterning of vertebrae was originally proposed to be dependent on the combinatorial action of all Hox proteins in the vertebral precursors the Hox code. However, this idea was subsequently challenged by the concept of posterior prevalence, which assumed that only the posteriormost genes expressed in a given segment are involved in patterning this segment. The knockout of entire paralog groups in the mouse somehow reconciles these two ideas, demonstrating that whereas posterior genes are clearly dominant over anterior ones, some level of combinatorial functions of adjacent paralog groups are required for the appropriate patterning of vertebrae. While this book is expected to meet the expectations of Hox aficionados, it is also intended to provide a survey of the field to newcomers. We hope that this book will take its place as a useful tool for those working in the ever growing field of Hox biology. I am indebted to all the authors for their excellent contributions. I also thank Tara Hoey at Elsevier for her continuous help and support. I am also grateful to Joanne Chatfield for her most valuable editorial assistance and to Silvia Esteban for the cover illustration.

10 CHAPTER ONE The Bithorax Complex of Drosophila: An Exceptional Hox Cluster Robert K. Maeda and François Karch Contents 1. Pseudoallelism and the History of the BX-C 2 2. The Ed Lewis Model 3 3. Molecular Genetics of the BX-C 6 4. Initiation and Maintenance Phases in BX-C Regulation 9 5. Initiators, Maintenance Elements, and Segment-Specific Enhancers Organization of the Cis-Regulatory Regions into Chromosomal Domains Chromatin Boundaries Flank the Parasegment-Specific Domains Boundaries Versus Insulators and Long-Distance Interactions Mixing the Old and the New Colinearity in the BX-C 24 References 27 Abstract In his 1978 seminal paper, Ed Lewis described a series of mutations that affect the segmental identities of the segments forming the posterior two-thirds of the Drosophila body plan. In each class of mutations, particular segments developed like copies of a more-anterior segment. Genetic mapping of the different classes of mutations led to the discovery that their arrangement along the chromosome paralleled the body segments they affect along the anteroposterior axis of the fly. As all these mutations mapped to the same cytological location, he named this chromosomal locus after its founding mutation. Thus the first homeotic gene (Hox) cluster became known as the bithorax complex (BX-C). Even before the sequencing of the BX-C, the fact that these similar mutations grouped together in a cluster, lead Ed Lewis to propose that the homeotic genes arose through a gene duplication mechanism and that these clusters would be Department of Zoology and Animal Biology and NCCR Frontiers in Genetics, University of Geneva, Geneva, Switzerland Current Topics in Developmental Biology, Volume 88 ISSN , DOI: /S (09) # 2009 Elsevier Inc. All rights reserved. 1

11 2 Robert K. Maeda and François Karch conserved through evolution. With the identification of the homeobox in the early 1980s, Lewis first prediction was confirmed. The two cloned Drosophila homeotic genes, Antennapedia and Ultrabithorax, were indeed related genes. Using the homeobox as an entry point, homologous genes have since been cloned in many other species. Today, Hox clusters have been discovered in almost all metazoan phyla, confirming Lewis second prediction. Remarkably, these homologous Hox genes are also arranged in clusters with their order within each cluster reflecting the anterior boundary of their domain of expression along the anterior-posterior axis of the animal. This correlation between the genomic organization and the activity along the anteroposterior body axis is known as the principle of colinearity. The description of the BX-C inspired decades of developmental and evolutionary biology. And although this first Hox cluster led to the identification of many important features common to all Hox gene clusters, it now turns out that the fly Hox clusters are rather exceptional when compared with the Hox clusters of other animals. In this chapter, we will review the history and salient features of bithorax molecular genetics, in part, emphasizing its unique features relative to the other Hox clusters. 1. Pseudoallelism and the History of the BX-C The term homeotic was first introduced by William Bateson more than a century ago (1894) to describe phenotypic variations in which something is changed into the likeness of something else (Bateson, 1894). The first isolated homeotic mutation was described in 1915 by Calvin Bridges (in Bridges and Morgan, 1923). Like all insects, Drosophilae have three thoracic segments (T1, T2, and T3). The landmarks of these thoracic segments are pairs of legs emanating from each of thoracic segments, a pair of wings that develop from the dorsal part of T2, and a pair of flight organs, called halteres, that develop from T3. In Bridges mutant, the anterior part of T3 develops like a copy of the anterior part of T2. This is visible on the fly as a transformation of the anterior haltere to a structure resembling the anterior part of the wing. As T2 is the most prominent thoracic segment, Bridges named his mutant bithorax (bx). In 1919, Bridges isolated a second homeotic mutation showing a somewhat similar homeotic transformation of posterior haltere toward posterior wing. This mutation, which he named bithoraxoid (bxd), maps to approximately the same region of the Drosophila third chromosome as bx. Because of the similarity in phenotypes and map location of bx and bxd, Bridges and Morgan (1923) were surprised to observe that the two mutations complemented. In 1934, a third mutation affecting the identity of T3 was discovered by Hollander (1937). In this case, the effect of the mutation is dominant, with heterozygous flies harboring swollen halteres, a sign of a

12 Bithorax Complex of Drosophila 3 weak transformation toward wings. Although this mutation has been given many names, it acquired its definitive name, Ultrabithorax (Ubx), in Unlike the bx or bxd mutations, Ubx homozygotes die as first instar larvae. Interestingly, although bx and bxd mutations complement each other, Ubx mutations fail to complement both the bx and the bxd mutations (bx/ubx animals look similar to bx homozygous flies and bxd/ubx animals are similar to bxd homozygous flies). These complex genetic interactions, where two or more genes appear to occupy the same locus under certain conditions, and different loci under other conditions is called pseudoallelism. Lewis began his undergraduate training in 1937, with the conviction that understanding pseudoallelism was crucial in defining the nature of genes. For more than 30 years, he devoted his research to understanding pseudoallelism using the BX-C as model system (for details, see two excellent perspectives written by Duncan and Montgomery, 2002a,b in Genetics). During this time, he identified hundreds of mutations in the BX-C, among which was a deletion of the entire region around bx. This deletion displayed an astonishing phenotype that changed the way people thought about the BX-C. Larvae homozygous for this deficiency die at the first instar stage with T3 and all eight abdominal segments (A1-A8) developing like a copy of T2. This phenotype indicated that the bithorax locus contained not only genes specifying T3, but also other genes responsible for the identities of all the abdominal segments. In his 1978 paper, Lewis describes the series of mutations that affect each of these segments. The actual names of these mutations are abx/bx, bxd/pbx, and iab-2 through iab-8 (Lewis, 1978). Phenotypic analysis indicated that each class of mutation defined an element that was required for the identity of a single segment. Remarkably enough, these elements mapped to the chromosomes in an order that corresponded to the body segment in which they acted. This correspondence between body axis and genomic organization is referred to as colinearity (see Figs 1.1 and 1.2). Because these different mutations formed a series of pseudoalleles, it was not entirely clear if they defined individual genes. Thus the term segment-specific function was commonly used to refer to the elements of this allelic series. 2. The Ed Lewis Model Because embryos deficient for the whole BX-C have all their segments posterior to T2 developing as copies of T2, Lewis proposed that T2 represents the ground state of development (i.e., the default state) and that each class of mutation represents a segment-specific function that allows a more-posterior segment to differentiate away from the ground state. Furthermore, the fact that mutations in individual segment-specific functions

13 4 Robert K. Maeda and François Karch abx/bx bxd/pbx iab-2 iab-3 iab-4 iab-5 iab-6 iab-7 iab A BC g Ubx abd-a Abd-B Figure 1.1 Organization of the BX-C. The 300-kb-long genomic DNA of the BX-C is displayed as the multicolored horizontal line. Map coordinate corresponds to the numbering of the original Drosophila genome project sequence of Martin et al. (1995). The structures of the three transcription units Ubx, abd-a, and Abd-B are depicted below the genomic bar with the arrows indicating the polarity of transcription. The extents of each of the nine segment-specific cis-regulatory domains are indicated by the different colors of the genomic DNA. The orange and red regions (abx/bx and bxd/pbx) regulate the expression of Ubx. The regions shaded in blue regulate and-a and correspond to iab-2, iab-3, and iab-4. Finally, the regions in green (iab-5 through iab-8) regulate Abd-B. The corresponding adult segments affected by the mutations in each cis-regulatory domain are indicated in the same color on the drawing of the fly. Reproduced with permission of the Company of Biologists. always caused homeotic transformations toward the last unaffected moreanterior segment (and not always to T2), meant that everything required for more-anterior segment development had to be present in more-posterior segments. For example, iab-3 homozygous flies have their A3 segment developing like a copy of A2. Thus, the role of iab-3 þ function must be to assign segmental identity to A3. However, because A3 is transformed into a copy of A2 instead of T2 (as in the BX-C deficiency), the abx/bx þ, bxd/pbx þ, and the iab-2 þ segment-specific functions required for A2 specification must normally be present in the developing A3 segment. Lewis summarized these findings into two rules:...a [segment-specific function]

14 Bithorax Complex of Drosophila 5 T1 T2 T3 A1 A2 A3 A4 A5 A6 A7 A8 300abx/bx bxd/pbx iab-2 iab-3 iab-4 iab-5 iab-6 iab-7 iab A BC g Ubx abd-a Abd-B Figure 1.2 The model of Ed Lewis revisited. The diagram of a Drosophila larva is depicted on the left (the y-axis). The three thoracic segments (T1-T3) and eight abdominal segments (A1-A8) are indicated (as well as the correspondence with the parasegments). The genomic organization of the BX-C is represented in the X-axis. In his original version, the model of Ed Lewis was represented as a matrix with more and more dots in the posterior segments, symbolizing the on or off status of the segment-specific functions. In this most updated version, activation along the chromosome of the segment-specific functions (abx/bx through iab-8) is envisioned as a sequential opening of chromosomal domains (also referred as to the open for business model ; Peifer et al., 1987). In PS2/T1 and PS3/T2, none of the BX-C function is active, as represented by the black ovals, which symbolize inactive chromosomal domains. The junctions between the adjacent ovals represent the chromosomal boundaries. The domains are maintained inactive by the products of the Pc-G genes. In PS5/pT2aT3, the abx/bx domain opens, allowing the enhancers residing in the domain to regulate Ubx expression in a pattern specific for that parasegment. In parasegment 6/pT3aA1, the adjacent bxd/pbx domain opens up to regulate Ubx in a pattern specific for that parasegment. Like in its original version, this model envisages that the more posterior a segment is along the anteroposterior axis, the more segment-specific functions are active in it. The Abd-B transcription unit positioned 5 0 from the iab-5 through iab-7 regulatory domains poses a problem to the chromosomal domain hypothesis. Indeed, it is unclear how iab-5, iab-6, oriab-7 can regulate Abd-B in their respective segments, while their Abd-B target promoter still resides in a closed domain. While there is no answer to this apparent discrepancy, recent evidences regarding Pc-G regulation suggest that, contrary to the classical picture of their role, Pc-G complexes do not set a repressed chromatin state that is maintained throughout development but have a much more dynamic role. Pc-G target genes can become repressed or be reactivated or exist in intermediate states (see for instance Schwartz and Pirrotta, 2008).

15 6 Robert K. Maeda and François Karch derepressed in one segment is derepressed in all segments posterior thereto..., and...the more posterior a segment...the greater the number of BX-C [segment-specific functions] that are in a derepressed state (Lewis, 1978). Lewis believed that the segment-specific functions acted in an additive fashion to progressively differentiate segments away from T2, an idea supported by the fact that some mutations in anterior segment-specific functions also caused slight changes in more-posterior segments (Fig. 1.2; see below). He synthesized all these findings, along with the peculiar colinearity of the BX-C segment-specific functions into model of where genes opened along the chromosome in a segmentally regulated fashion from anterior to posterior (see Fig. 1.2 for a modern visualization of the Lewis model). It is important to note that all the mutations affecting individual segment-specific functions are viable as homozygotes. Lewis also reported the existence of lethal mutations within the BX-C. The Ubx class mentioned above that failed to complement the bx, bxd, and pbx mutations being one of them. In 1985, the groups of Gines Morata in Madrid and Robert Whittle in Sussex independently published two papers describing a different route to isolate mutations in the BX-C (Sanchez-Herrero et al., 1985; Tiong et al., 1985). Using the whole BX-C deficiency chromosome mentioned above, they performed a screen aimed at identifying mutations that fail to complement the lethality of the BX-C deficiency. Three independent complementation groups (each giving rise to homozygous lethality) were identified. One of these corresponded to the Ubx mutation. However, the two other complementation groups were new. The first, abdominal-a (abd-a), affected segments A2 A4, while the second, Abdominal-B (Abd-B), affected abdominal segments A5 A Molecular Genetics of the BX-C During the course of these genetic screens, the cloning of the whole BX-C was reported in two successive papers (Bender et al., 1983; Karch et al., 1985). The cloning provided the molecular basis to explain much of the genetic data gathered by over the preceding decades. Overall, the BX-C was found to cover 300 kb of DNA. All the mutations affecting the segment-specific functions were found to be associated with rearrangement breakpoints (such as translocations, inversions, deficiencies, or insertions of transposable elements). The lesions associated with a given class of mutations always clustered in a relatively small part of the BX-C, and different 1 It should be noted that abd-a mutations are truly recessive lethal, while Abd-B heterozygous flies are sterile (this dominant sterility explains why Abd-B is written with a capital A, unlike abd-a alleles).

16 Bithorax Complex of Drosophila 7 classes of mutations never overlapped. The collinear arrangement of the segment-specific functions along the chromosome and the body segments they specify along the AP axis was also confirmed (Fig. 1.1). The observation that all the mutations in each class are associated with rearrangement breakpoints not only helped to localize them on the DNA map (more than hundred mutations have been localized) but also suggested that the segment-specific functions were probably not simple proteincoding regions (otherwise point mutations would have been recovered during the numerous screens performed). Further evidence to support this hypothesis came from the Hogness laboratory. With the help of Arthur Kornberg s laboratory, the Hogness lab used overlapping probes to scan developmental Northern blots to identify transcripts. From this analysis, they determined that only about 12 kb of the 300 kb of DNA from the BX-C are present as mature poly(a) þ transcripts (Hogness et al., 1985). By mid-1983, a cdna spread across a 70 kb span of the DNA was isolated. This 70 kb span of DNA corresponded to the genomic region associated with Ubx mutations, and thus, the cdna was soon identified as the Ubx gene product. At the same time, cloning of the Antennapedia complex (Antp-C ) by Rick Garber and Matthew Scott in the Gehring and Kaufmann laboratories, respectively, led to the identification of the Antp transcription unit (covering 100 kb of DNA; Garber et al., 1983; Scott et al., 1983). It was not long before sequence comparisons between the two genes revealed a region of DNA similar in both genes. This sequence became known as the homeobox (McGinnis et al., 1984; Scott and Weiner, 1984). The homeobox accelerated the identification of the remaining Drosophila Hox genes. Very quickly, two other homeobox genes were identified within the BX-C in the regions where the abd-a and Abd-B mutations had been mapped (Regulski et al., 1985). These molecular studies suggested that the whole BX-C encodes only three homeotic genes: Ubx, abd-a, and Abd-B. The first genetic confirmation of this was published in 1987 by Casanova et al. (1987), who showed that a Ubx;abd-A;Abd-B triple mutant embryo harbored the same phenotype as an embryo carrying a complete deletion of the entire BX-C. This hypothesis was later confirmed when the whole region was sequenced (Martin et al., 1995). But what are the nine segment-specific functions identified by Ed Lewis if genetic and molecular analysis indicates that the BX-C only encodes three homeotic proteins? The description of the expression patterns of Ubx, abd-a, and Abd-B brought an answer to this apparent paradox (Beachy et al., 1985; Celniker et al., 1990; Karch et al., 1990; Macias et al., 1990; Sanchez-Herrero, 1991; White and Wicox, 1985). Figure 1.3 shows the central nerve cord of wild-type and various mutant embryos (see below) stained with an antibody directed against Abd-B. Like Ubx and abd-a, though in a more-posterior area, Abd-B is expressed in an intricate pattern

17 8 Robert K. Maeda and François Karch Fab-7 WT iab-7 Sz PS10 PS12-like PS12 PS13 PS14 PS10 PS11 PS12 PS13 PS14 PS10 PS11 PS11-like PS13 PS14 Figure 1.3 Abd-B expression in the embryonic central nervous system of WT and mutant embryos. After staining, the central nervous systems were dissected out from 12-h-old embryos. In wild type, the typical Abd-B expression pattern is characterized by an anterior-to-posterior gradient from PS10 to 14 in the number of expressing nuclei per parasegment, as well as by the intensity in each nucleus. Note that the Abd-B protein expressed in PS14 is an isoform derived from alternatively spliced transcripts initiating from the B, C, and g promoters (see Fig. 1.1). In Fab-7 mutant embryos, the PS11-specific expression pattern is replaced by the pattern expressed in PS12, resulting into the homeotic transformation of PS11/A6 into PS12/A7. In iab-7 Sz, the whole iab-7 regulatory domain is deleted. As a consequence, the PS12-specific expression pattern is replaced by the pattern specific for PS11. that is finely tuned from one parasegment (PS) to the next. 2 By staining various mutant embryos, it was finally understood that the segment-specific functions corresponded to cis-regulatory regions that regulate the expression of Ubx, abd-a, or Abd-B in a segment-specific fashion. Mutations in any of the segment-specific regulatory regions alter the expression of its relevant target gene. For example, flies homozygous for the iab-7 Sz mutation have their seventh abdominal segment transformed into a copy of the sixth. Consistent with this, the embryonic Abd-B expression pattern characteristic for PS12/A7 is replaced by the pattern normally present in PS11/A6 (Fig. 1.3; Galloni et al., 1993). 2 Homeotic gene expression in Drosophila does not exactly respect segmental borders. They are shifted, being composed of the posterior part of one segment and the anterior part of the next segment. This unit of expression is called the parasegment (PS) and explains why the adult phenotypes observed in homeotic mutations often affect the posterior part of one segment and the anterior part of the next. For example, the bxd mutation mentioned above actually transforms the posterior part of T3 and the anterior part of A1 into copies of the posterior part of T2 and the anterior part of T3, respectively. This is less visible in the abdominal segments, where the anterior portion of each segment is hidden underneath the posterior part of the preceding segment.

18 Bithorax Complex of Drosophila 9 The finding that the segment-specific functions correspond to cisregulatory domains helped to explain the phenomenon of pseudoallelism in the BX-C. In Fig. 1.1 the cis-regulatory region of the BX-C is schematically detailed. The regulatory regions interacting with the Ubx gene are shown in red and orange. They include the abx/bx and bxd/pbx regions that regulate Ubx expression in PS5 and PS6, respectively (Beachy et al., 1985; Little et al., 1990; White and Wicox, 1985). 3 As explained above, bx and bxd mutations fully complement, but mutations in Ubx fail to complement both the bx and bxd mutations. This can now be explained by the fact that these segment-specific functions require Ubx function for their activity. For example, if we look at the contribution of each chromosome to Ubx expression independently, a chromosome carrying a bx mutation fails to produce Ubx protein in PS5 (where the bx cis-regulatory element is normally active), but produces the normal amount of Ubx protein in PS6 (where the bxd/pbx cis-regulatory element is active). The Ubx mutant chromosome in trans, however, does not produce a functional Ubx product in PS5 or PS6. The resulting trans-heterozygote is therefore Ubx / in PS5 but Ubx þ/ in PS6. Because segment-specific functions behave as recessive mutations, bx/ubx mutants resemble bx mutant flies. 4. Initiation and Maintenance Phases in BX-C Regulation How the cis-regulatory elements control BX-C gene expression has been the focus of much research for the past 20 years. Through this work, it now seems clear that the regulation of homeotic gene expression can be divided into two phases: initiation and maintenance. The initial determination of the AP axis during Drosophila embryogenesis is under the control of three classes of transcription factors that are deployed in a cascade and lead to the subdivision of the embryo into 14 parasegments (the maternal, gap, and pair-rule genes; for reviews see, e.g., DiNardo et al., 1994; Hoch and Jackle, 1993; Ingham, 1988; Kornberg and Tabata, 1993). It is now known that these proteins interact with elements in each of the cisregulatory regions of the BX-C, to determine the ultimate homeotic gene pattern (Casares and Sanchez Herrero, 1995; Irish et al., 1989). For example, 3 abx allele stands for anterobithorax. Ed Lewis distinguished these alleles from bx because they primarily affect the dorsal part of anterior T3. bx mutations, on the other hand, affect anterior part of T3 without affecting the dorsal region. However, both types of enhancers are part of the same regulatory region that is active in PS5 (mostly, anterior T3). A similar distinction can be made for the bxd and pbx elements that are both active in PS6 but in different regions. pbx is mostly active in the anterior part of PS6 (mostly corresponding to posterior T3) while bxd is mostly active in the posterior part of PS6 (corresponding to anterior A1 in the adult; see Fig. 1.5).

19 10 Robert K. Maeda and François Karch the combination of the gap and pair-rule gene products present in PS12 allows iab-7, but not iab-8, to control Abd-B expression in PS12/A7. However, because the gap and pair-rule genes are only transiently expressed in the early embryo, and the activity states of the segment-specific cisregulatory regions is fixed for the life of the fly, a system to maintain homeotic gene expression is required in each cis-regulatory domain (Struhl and Akam, 1985). This maintenance system has been shown to require the products of the Polycomb group (Pc-G) and trithorax group (trx-g) genes. While the Pc-G products function as negative regulators, maintaining the inactive state of the cis-regulatory regions not in use, the trx-g products function as positive regulators, maintaining the active state of the active regulatory regions (Kennison, 1993; Paro, 1990; Pirrotta, 1997; Simon, 1995). Both the Pc-G and trx-g products are thought to maintain the active or inactive state of each parasegment-specific cis-regulatory region by modifying the chromatin structure of each region. Indeed, both Pc-G and trx-g proteins contain members that bind, modify or move histones. Thus, the current model suggests that Pc-G proteins compact chromatin to prevent activators from binding to the cis-regulatory regions, while trx-g proteins open the chromatin to keep the cis-regulatory domains accessible to activators. It should be noticed that recent evidences regarding Pc-G regulation suggest that, contrary to the classical picture of their role, Pc-G complexes do not set a repressed chromatin state that is maintained throughout development but have a much more dynamic role. Pc-G target genes can become repressed or be reactivated or exist in intermediate states (see for instance Schwartz and Pirrotta, 2007, 2008). The distinction between the initiation of expression and the maintenance of expression has led to the identification of DNA elements that mediate these distinct phases, cleverly called initiators and maintenance elements (also known as Pc-G or trx-g response elements PREs/TREs: see next). 5. Initiators, Maintenance Elements, and Segment-Specific Enhancers Confirmation of the segment-specific and biphasic nature of BX-C gene regulation has been done using reporter gene constructs. In these experiments, DNA fragments from the various regulatory regions were cloned upstream of a lacz gene reporter. By making transgenic flies carrying these reporter constructs and studying their resulting patterns of expression, scientists have been able to identify specific DNA fragments from the BX-C that are required for initiating segment-specific expression, maintaining a restricted pattern of expression, and allowing segment-independent,

20 Bithorax Complex of Drosophila 11 attb PRE Ubx-lacZ PS6 initiator Initiation A Head Tail C Maintenance B PS6 D PS6 With PRE@58A Without PRE@58A Figure 1.4 Identification of initiation and maintenance elements with lacz reporter constructs. The structure of the lacz reporter constructs used in these studies is shown on top of the figure. The plasmids have been transformed into flies using the new FC31 recombinase system, and are located in the 58A platform (Bischof et al., 2007). All embryos contain the plasmid with a PS6 initiator element that activates the lacz reporter gene in PS6 and more-posterior parasegments, following a pair-rule pattern (Qian et al., 1991). Early embryos at initiation phase are shown in panels (A) and (C). At this particular stage called germband extension, parasegments 8-14 curve around toward the dorsal side (as indicated by the curved arrow). At later stages of development (panels B and D), the germband has retracted such as the posterior parasegments are at the posterior pole of the embryo. In panel (B), the presence of the maintenance element (PRE) maintains the initial anterior lacz expression pattern. In absence of the PRE (panel D) the initial expression pattern degenerates and b-galactosidase is detected anteriorly to the initial PS6 anterior border. The use of the FC31 recombinase system allows refuting any positional effect as both constructs are inserted within the same platform. The PRE used in these studies is derived from the bxd/pbx regulatory domain (Horard et al., 2000; Sipos et al., 2007). cell-type-specific expression. Figure 1.4 provides examples to highlight the differences between each type of element. BX-C initiator elements can be defined as specific types of enhancers that confer a parasegmentally restricted pattern of expression to a reporter gene during early embryogenesis (Barges et al., 2000; Busturia and Bienz, 1993; Mihaly et al., 2006; Muller and Bienz, 1992; Qian et al., 1991; Shimell et al., 2000; Simon et al., 1990; Zhou et al., 1999). For example, Fig. 1.4A and C shows the lacz expression pattern of an early embryo where lacz expression is driven by a DNA element derived from the bxd/pbx region is normally responsible for Abd-B expression in PS6. Likewise, this element