UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Functional dissection of the Drosophila melanogaster bithoraxoid Polycomb response element Argiropoulos, Bob 2002

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2002-74986X.pdf [ 9.04MB ]
Metadata
JSON: 831-1.0099654.json
JSON-LD: 831-1.0099654-ld.json
RDF/XML (Pretty): 831-1.0099654-rdf.xml
RDF/JSON: 831-1.0099654-rdf.json
Turtle: 831-1.0099654-turtle.txt
N-Triples: 831-1.0099654-rdf-ntriples.txt
Original Record: 831-1.0099654-source.json
Full Text
831-1.0099654-fulltext.txt
Citation
831-1.0099654.ris

Full Text

FUNCTIONAL DISSECTION OF THE DROSOPHILA M E L A N O G A S T E R BITHORAXOID P O L Y C O M B RESPONSE E L E M E N T by BOB ARGIROPOULOS B.Sc , University of Western Ontario, 1995 M.Sc., University of Western Ontario, 1997 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY In THE F A C U L T Y OF G R A D U A T E STUDIES (Department of Zoology) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A September 2002 © Bob Argiropoulos, 2002 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract During the development of Drosophila melanogaster, the Polycomb Group (PcG) proteins act through complex, modular elements termed Polycomb response elements (PREs) to maintain the silent state of the homeotic and other loci. The bithoraxoid (bxd) 5.1 PRE of the homeotic gene Ultrabithorax (Ubx) is capable of maintaining the correct Ubx expression pattern throughout embryogenesis in a PcG-dependent manner. The bxd5.l PRE is also capable of conferring pairing-sensitive repression to the mini-white gene located in the same transposon. In order to understand how PcG proteins are recruited to PREs, a gel mobility shift assay was used to identify four fragments within the bxd5.l PRE that bind protein complexes from nuclear extracts that contain the PcG protein Polyhomeotic (PH). Chapter 2 of this thesis examines the in vivo contribution of these four PH binding sites in embryonic silencing and pairing-sensitive repression. I show, using a germline transformation assay, that deletion of each PH binding site, in the context of bxd5.\, disrupts embryonic PRE activity but not pairing-sensitive repression. Double mutant analysis of sites with related binding activities indicate that sites MHS-70 and MPA-168 constitute one functional unit of PRE activity, which is disabled by either mutation. By contrast, sites MHN-90 and S1HB-90 act synergistically to promote PRE activity. Furthermore, mutation of two d(GA)3 repeat elements within MHS-70 destabilizes PH complex formation in vitro and partially abrogates PRE activity in vivo indicating that these repeat elements are essential for PRE-mediated silencing. Chapter 3 of this thesis explores the modular structure of the core maintenance element within bxd5.l, the bxd\.5 PRE. The results indicate that the bxdl.5 PRE is a complex element built up of at least three modules, UPS, PSR and DPS, that make distinct contributions to silencing by the bxd PRE. The UPS and DPS modules directly repress the Ubx promoter in a parasegment-specific and developmental stage-specific manner. The PSR and DPS modules are capable of pairing-sensitive repression. Genetic analyses reveal that each module depends on the function of a subset of PcG and trxG genes that are required specifically for embryonic or pairing-sensitive repression or for both processes. The results clearly demonstrate that embryonic and pairing-sensitive repression are separable functions of the bxdl .5 PRE. Taken together, these studies provide insight into how the bxd PRE is built and into the nature of the functional components that read and interpret the information encoded by this complex cis-regulatory element. ii Table of Contents Abstract ii Table of Contents i i i List of Tables vi i i List of Figures ix Acknowledgements xi Chapter 1: General Introduction 1 I. Development 1 II. Epigenetic Regulation 5 III. The Polycomb Group (PcG) Genes 5 IV. Polycomb Group Complexes 6 V. Polycomb Group Homologues 8 VI. The polyhomeotic (ph) Locus 11 VII. Initiation of Silencing 12 VIII. Mechanisms of Polycomb Group Action 13 A. Heterochromatin Model 13 B. The Histone Deacetylation Model 17 C. The D N A Looping Model 17 D. Interference With Basal Transcription Machinery 18 E. Nuclear Compartmentalization Model 18 IX. The Trithorax Group (trxG) Genes 19 X . Polycomb Group Response Elements (PREs) 20 XI. Conserved PRE Sequence Motifs 21 XII. PRE Functional Assays 23 iii XIII. Regulation of Ultrabithorax (Ubx) XIV. bithoraxoid Polycomb Response Element 24 30 Chapter 2: Functional role of the Polyhomeotic-binding sites within the bithoraxoidSA Polycomb Response element I. Introduction 34 A. Site-Specific Interactions of PH with Sub-Fragments o f W 5 . 1 34 II. Results 44 A. What is the Expression Pattern of the bxd5A Control? 44 B. What is the Effect of bxd5.1 Deletions? 49 i . tad5.1AMHS-70andfoo/5.1AMPA-168 49 i i . W5.1AMHN-90 52 i i i . &of5.1ASlHB-90 52 C. Do the MHS-70/MPA-168 and MHN-90/S1HB-90 Fragments Function Independently? 56 i . W5.1AMHS-70+AMPA-168 56 i i . W5.1AS1HB-90+AMHN-90 57 D. What is the Role of d(GA)3 Sequences in the MHS-70 Fragment? 57 E. Are G A F and PSQ Needed for PcG-mediate Silencing? 59 III. Discussion 60 A. Role of PH Binding Sites in bxd5.1 PRE Activity 60 B. Initiation of Silencing 64 C. bxdSA PRE Design 67 D. Embryonic Silencing can be Uncoupled from Pairing-Sensitive Repression 68 iv E. Role of G A F and Psq in 6xJ-Mediated Silencing 69 F. Role of POU homeodomain proteins in facd-mediated silencing 71 Chapter 3: Modular structure of the bithoraxoid 1.5 Polycomb Response element I. Introduction 72 A. Properties of bxd\.5 PRE 72 B. Structural Organization of the bxdl.5 PRE 73 i . PREs and TREs within the PSR fragment ofbxdl .5 consist of closely situated but separable sequences 76 i i . bxdl.5 PRE-mediated silencing in imaginal discs requires the PSR subfragment 77 i i i . The bxdl.5 PRE is a compound element composed of sequences with different PRE-like properties that interact in vitro with different PcG complexes 78 C. Chromatin Boundary Elements 79 D. Overall Rationale 81 II. Results 82 A . What is the Basal Ubx Expression Pattern? 82 B. What is the Overall Function of the bxdl.5 PRE? 87 C. Do the bxdl.5?RE Modules Retain PRE Activity? 88 i . UPS+PSR and PSR+DPS 88 i i . UPS 91 i i i . PSR 91 iv. DPS 92 D. Are Embryonic Silencing and Pairing-Sensitive Repression Separable Functions of the bxdl .5 PRE? 95 v E. What are the Functional Components that Contribute to the Activity of the Different bxdl .5 PRE Modules? 98 III. Discussion 106 A. PRE-Mediated Gene Silencing and Basal Transcription 110 B. Embryonic function of the bxdl .5 PRE modules 112 C. Mechanisms of bxdl.5 PRE-Mediated Silencing 114 D. Is pairing-sensitive repression essential? 118 Chapter 4: General Discussion 121 I. Main conclusions 121 II. Modularity in the bxdl.5 PRE 121 III. PRE-Promoter Interactions 123 IV. Final Thoughts 125 Chapter 5: Materials and Methods 126 A. Plasmids and P-element Transformation Vectors 126 B. Polymerase Chain Reaction (PCR) 128 C. D N A Manipulations 128 i . Generation of bxd5. IAS 1-HB90 130 i i . Generation of bxd5.lAMHS-70 130 i i i . Generation of bxd5.l AMP A-168 133 iv. Generation of W5.1AMHN-90 133 v. Generation of bxd5.1 AS 1-HB90+AMHN-90 133 vi. Generation of W5.1AMHS-70+AMPA-168 136 vii . Generation of W 5 . 1 - L S 1/9 136 vii i . Generation of U T D 136 E. Drosophila Embryo Transformation 136 vi F. In Situ Localization of Transposons on Polytene Chromosomes 138 G. P-element Mobilization 139 H. Immunohistochemical Staining 140 I. Scoring Misexpression of Ubx/lacZ Reporter in Embryos 141 J. Scoring Pairing-Sensitive Repression of the mini-white gene 142 K. Microscopy and Photography 143 L. Drosophila Culture 143 References 145 Abbreviations 165 vn List of Tables Table 1 -1 PcG homologues and their prominent protein domains 10 Table 2-1 Trl and psq enhance the homeotic phenotypes of ph and Pc 62 Table 2-2 Summary of bxd5.\ PRE deletion analysis 63 Table 3-1 Effects of homozygous PcG and trxG mutations on the embryonic PRE activity of the bxd\.5 PRE and its modules 103 Table 3-2 Effects of heterozygous PcG and trxG mutations on the pairing-sensitive repression activity of the bxd\.5 PRE and the PSR and DPS modules 107 Table 3-3 Summary of the PRE activities associated with the bxdl .5 PRE and its modules 111 Table 5-1 List and description of Drosophila strains used in this thesis 144 Vll l List of Figures Figure 1-1 Relationship Between Embryonic Parasegments and Adult Segments 3 Figure 1-2 Drosophila PcG Complexes 9 Figure 1-3 Mechanisms of PcG Action 14 Figure 1-4 The Bithorax Complex , 26 Figure 1-5 U B X expression pattern inwildtype and PcG mutant embryos 28 Figure 1-6 The bithoraxoid 5.1kb Polycomb response element of the Ubx gene 32 Figure 2-1 PH Binding Sites within the bxd5.\ PRE 37 Figure 2-2 Shared sequence elements between the MHS-70 and MPA-168 sites are the d(GA) 3 repeats 40 Figure 2-3 Nucleotide Sequence of sites S1HB-90 and MHN-90 42 Figure 2-4 Summary of the gel mobility shift analysis of MHS-70 linker-scanning mutations 43 Figure 2-5 Diagrammatic representation of the Ubx/lacZ reporter construct and the wildtype and mutant bxd5.1 fragments tested 45 Figure 2-6 Silencing mediated by the bxd5.l PRE 48 Figure 2-7 Deletion of sites MHS-70 and MPA-168 disrupt bxd5.\ PRE-mediated silencing in embryos. 51 Figure 2-8 Metameric misexpression pattern of fc«/5.1AMHN-90 53 Figure 2-9 Deletion of sites MHN-90 and S1HB-90 disrupt bxd5.\ PRE-mediated silencing in embryos 55 Figure 2-10 Mutations of d(GA) 3 repeat elements within MHS-70 reduce bxd5.l PRE activity 58 Figure 2-11 Trl and psq mutations enhance the extra sex combs phenotype of ph and Pc mutations. 61 Figure 2-12 Model of bxd5.\ PRE-mediated silencing 66 ix Figure 3-1 Summary of the biochemical and genetic analyses of the bxd\.5 PRE 75 Figure 3-2 Model of the regulation of Ubx expression by the bxdl .5 PRE 80 Figure 3-3 Diagrammatic representation of the p{y+, Ubx/lacZ, w+} transformation vector and the constructs tested 84 Figure 3-4 The basal Ubx/lacZ construct does not have PRE activity 86 Figure 3-5 Embryonic silencing activities of the bxdl.5 PRE modules at germ band extension 90 Figure 3-6 The DPS module silences the Ubx promoter in a metameric pattern in germ band retracted embryos 94 Figure 3-7 Embryonic silencing and pairing-sensitive repression are separable functions of the bxdl .5 PRE 97 Figure 3-8 The bxdl .5 PRE responds differently to ph mutations than all other mutations tested 101 Figure 3-9 Effects of homozygous PcG mutations on the embryonic PRE activity of the bxdl .5 PRE modules 105 Figure 3-10 The ability of the bxdl .5 PRE and the PSR module to exhibit pairing-sensitive repression is abolished in pho and Pc mutant backgrounds 109 Figure 3-11 Model of bxdl.5 PRE-mediated silencing 117 Figure 5-1 Diagram of the transformation vector pCaSpeR4-t/foc/7acZ 111 Figure 5-2 Diagram of the transformation vector P{Su(Hw) yellow* Su(Hw) Ubx/lacZ white" Su(Hw)} 129 Figure 5-3 Flowchart for the generation of bxd5.IAS 1-HB90 131 Figure 5-4 Flowchart for the generation of bxd5.1AMHS-70 132 Figure 5-5 Flowchart for the generation of bxd5.l AMP A-168 134 Figure 5-6 Flowchart for the generation of bxd5.1AMHN-90 135 x Acknowledgements First and foremost I would graciously like to thank my supervisor Dr. Hugh Brock for providing me with the opportunity to work in a great laboratory. I sincerely appreciate the effort he has put into forwarding my career and for this I am forever indebted to him. His comments and criticisms have elevated the quality of my research and of this thesis many fold. He has been patient, encouraging and perpetually enthusiastic. Thank you Hugh. I would like to thank the members of my advisory committee, Drs. Vanessa Auld, Carolyn Brown, Don Moerman and Ivan Sadowski for their guidance and assistance throughout my PhD training. I extend my appreciation to members of the Brock lab, past and present: Jacob Hodgson, Ester O'Dor, Sebastien Bloyer, Jack Chevalier, Cynthia Fisher, Michael Kyba, Bryan Andrews, Yong-Jun Wang, Tom Milne, B i l l Waples, Afua Osei and Joyce Tan. Their constructive criticisms, stimulating discussions and helpful ideas were refreshing and always appreciated. The work presented in this thesis was, in large part, a collaboration with Dr. Jacob Hodgson and for this I owe him an immense debt of gratitude. I would like to thank Dr. Grigliatti and members of his laboratory for their hospitality and encouragement throughout my years at UBC. I wish to express my sincerest gratitude to the Gall family for their love and encouragement. I also thank the Guimonds who were hugely supportive and Joost Schulte and Katharine Sepp for introducing me to mountaineering and telemark skiing. I would like to express my deepest gratitude to my parents, Panos and Helen, and to my siblings, along with their respective spouses, for their constant support, understanding and love. Thank you for believing in me. Finally, I am extremely indebted to my wife Kimberly. Kim, you are my guiding light, my ethical compass, and my sunshine. This thesis and degree are dedicated to you and to the baby we lost. Thank you for brightening my days and for offering support and encouragement at times when my worries were the least of your worries. I could not have made it this far without you. I look forward to spending an eternity with you. May the road rise up to meet you May the wind be ever at your back May the sun shine warm upon your face And rains fall soft upon your fields. And until we meet again, May God hold you in the hollow of His hand. Irish blessing All men dream: but not equally. Those who dream by night in the dusty recesses of their minds wake in the day to find that it was vanity: but dreamers of the day are dangerous men, for they may act their dreams with open eyes, to make it possible. T.E. Lawrence, The Seven Pillars of Wisdom XI Chapter 1 General Introduction I. Development The molecular basis of determination of cell fate is of great importance in understanding the development of higher organisms. Determination is controlled by a complex network of transcription factors whose expression patterns are tightly regulated. To understand determination it is essential to understand the molecular mechanisms of this transcriptional network. Much of the information that determines the spatial and temporal expression pattern of transcription factors is encoded in an organism's D N A via sophisticated cis-regulatory sequences. Deciphering how these cis-regulatory sequences control gene expression is a key to understanding development. Drosophila melanogaster provides a model system for studying the control of cell fate determination during embryonic development. Many of the genes that regulate Drosophila development have been identified and characterized (Akam, 1987). Indeed, many of the genes involved in embryonic development of Drosophila are known to control development in other animals (Carroll et al., 2001). Therefore, knowledge obtained about how these genes are controlled in Drosophila will most likely be applicable to other higher eukaryotes. A common characteristic of embryos of many higher organisms is the formation of repeated morphological units called segments or metameres. Drosophila is segmented in two distinct registers: the embryonic parasegmental register and the larval or adult segmental register. Parasegments are embryonic metameric units that include the primordial cells of the posterior part of one post-embryonic segment and the anterior part of the next more posterior post-embryonic segment (Martinez-Arias and Lawrence, 1985). The relationship between the embryonic parasegments and the larval or adult segments is illustrated in Figure 1-1 A . Systematic genetic screens have identified many genes that encode components required for proper segmentation (Nusslein-Volhard and Wieschaus, 1980; Wieschaus et al, 1984). Study of the interactions between these genes has shown that Drosophila development is controlled by hierarchical interactions between sets of genes and by cross-regulatory interactions between members within these sets. These hierarchical interactions are best exemplified in the process of anterior-posterior body axis formation that is controlled by three general classes of genes. 1 Figure 1-1. A . A diagram illustrating the relationship between the embryonic parasegments and the larval/adult segments. The parasegments are out of phase with the segments by the width of a single compartment. Each adult segment is divided into an anterior compartment, labelled A , and a posterior compartment, labelled P. The adult thoracic segments are labelled T l to T3 and the adult abdominal segments are labelled A l to A8. B. The homeotic genes of the bithorax complex, Ultrabithorax (Ubx), abdominal-A (abd-A) and Abdominal-B (Abd-B) are expressed in unique domains along the body axis of the embryo. The Ubx gene specifies the identity of parasegments (PS) 5 and 6 that corresponds to the shaded region in the adult and extends from the posterior compartment of the second thoracic segment (T2) to the anterior compartment of the first abdominal segment (Al ) . Figure adapted from Hogness etal, 1985. 2 0 > The maternal coordinate genes are expressed during oogenesis and are responsible for determining the axes of the egg (Nusslein-Volhard et al, 1987). The segmentation genes are expressed from the zygotic genome, respond to the positional information provided by the maternal coordinate genes and are responsible for dividing the embryo into segments (Nusslein-Volhard and Weischaus, 1980; Wieschaus et al, 1984). Finally, the homeotic genes are responsible for determining the unique identity of each segment (Lewis, 1978). Most of the genes belonging to these three classes have been extensively described at the genetic, structural and evolutionary levels. Although we have much knowledge about how homeotic gene expression patterns are initiated, very little is known about how they are maintained throughout development. This thesis addresses how transcription of a particular homeotic gene, Ultrabithorax (Ubx) is regulated throughout embryonic development. Determination of parasegmental identity in Drosophila is mediated by the activities of the homeotic genes of the Antennapedia and Bithorax complexes (Lewis, 1978; Duncan, 1987; Kaufman et al, 1990). Expression of the homeotic genes is confined to unique domains along the body axis whose anterior and posterior boundaries are maintained throughout development. The restricted expression domains of the Bithorax complex genes are shown in Figure 1-1B. These unique domains of homeotic gene expression are actively maintained by repressive elements. Homeotic genes contain enhancer elements that lack positional information and have the potential to be active in all parasegments. Misexpression of homeotic genes outside these boundaries results in posterior homeotic transformations in all segments of the body axis. Initial activation of the homeotic genes occurs in response to the developmental information provided by the maternal coordinate and segmentation genes mentioned above. Because the products of the segmentation genes are only transiently available, the maintenance of homeotic gene expression throughout development depends on other factors. These other factors include the gene products from the Polycomb (PcG) and trithorax group (trx-G) genes. In general, members of the Polycomb group are required to maintain the transcriptionally silent state in those cells in which the homeotic genes were initially repressed and the genes of the trithorax group are required for the continued expression in those cells in which the homeotic genes were originally active (reviewed in Francis and Kingston, 2001; Poux et al, 2002). The PcG proteins and the cis-regulatory elements they interact with are the focus of this thesis, and will be described in detail in subsequent sections. The hallmark feature of gene regulation by PcG and trxG proteins is that it can lead to a mitotically stable pattern of gene expression, often referred to as epigenetic regulation. 4 II. Epigenetic Regulation Epigenetic regulation involves the maintenance of a particular state of gene expression, most commonly repression, throughout repeated mitosis, and frequently meiosis. Changes in such heritable expression states occur without an alteration of the primary D N A sequence but do involve the assembly of specialized structures of chromatin. It is becoming increasingly evident that epigenetic regulation of gene transcription plays a critical role in the regulation of gene expression in many biological processes. Examples of epigenetic silencing include: position-effect variegation (PEV) in Drosophila (reviewed in Grewal and Elgin, 2002); cosuppression in plants, fungi and Drosophila (Jorgenson, 1995; Pal-Bhadra et al, 2002); gametic imprinting in mammals (reviewed in Brannan and Bartolomei, 1999); dosage compensation by X chromosome inactivation in mammals (reviewed in Boumil and Lee, 2001); telomeric position-effect (TPE) and mating type silencing in yeast (reviewed in Gasser and Cockell, 2001). Repression of target loci by the PcG in Drosophila is another example of epigenetic silencing. A common theme that emerges amongst these diverse processes is the complex integration of epigenetic regulatory pathways with alterations in chromatin structure over target D N A loci. III. The Polycomb Group (PcG) Mutations in the PcG genes of Drosophila cause posterior homeotic transformations in all segments of the body axis (Lewis, 1978). The PcG is named for the presence of ectopic sex combs on the second and third legs of mutant male adult flies. The phenotypes of PcG mutations are similar to phenotypes of gain-of-function mutations of the homeotic genes. This similarity results from ectopic expression of the homeotic genes beyond their proper spatial boundaries in PcG mutants (Struhl and Akam, 1985; Jones and Gelbart, 1990; Soto et al, 1995). In embryos mutant for genes in the PcG, homeotic gene expression is correctly established, but it subsequently spreads beyond the normal boundaries resulting in posterior transformations, showing that PcG genes are not required for initial repression but maintain the silent state of the homeotic loci. The homeotic genes are not the sole targets of the PcG. Mutations in PcG genes affect the expression pattern of the PcG member polyhomeotic (Fauvarque et al, 1995) and of the segmentation genes engrailed (Moazed and O'Farrell, 1992), even-skipped (McKeon et al, 1994), knirps, giant (Pelegri and Lehmann, 1994), hedgehog, patched, cubitus interruptus and decapentaplegic (Randsholt et al, 2000). There are 16 genetically characterized members of the PcG: Polycomb (Pc) (Lewis, 1978), extra sex combs (esc) (Struhl, 1981; Frei et al, 1985), Polycomblike (Pel) (Duncan, 1982), Enhancer of Polycomb (E(Pc)) (Sato et al, 1983), super sex combs (sxc) (Ingham, 1984), polyhomeotic (ph) (Dura et al., 1985), Sex combs on midleg (Scm) (Jurgens, 1985), Sex combs extra (See) (Breen and Duncan, 1986), Suppressor two of zeste (Su(z)2) (Adler et al, 1989), Enhancer of zeste {E(z)} (Jones and Gelbart, 1990), Posterior sex combs (Psc) (Jurgens, 1985; Adler etal, 1991), Additional sex combs (Asx) (Jurgens, 1985; Sinclair et al, 1992), pleiohomeotic (pho) (Girton and Jeon, 1994) formerly known as polycombeotic (pco) (Phillips and Shearn, 1990), multi sex combs (mxc) (Santamaria and Randsholt, 1995), cramped (crm) (Yamamoto et al, 1997) and Suppressor of zeste 12 (Su(z)12) (Birve et al, 2001). Deletion analysis of the genome has estimated that there may be up to 30 or 40 members in total (Jurgens, 1985; Landecker et al, 1994). A l l the PcG members identified thus far are ubiquitously expressed and are involved in properly maintaining the silent state of homeotic and non-homeotic genes. Another common characteristic of the PcG is that mutation of one PcG gene usually enhances the phenotype of other PcG mutants in double mutants (Jurgens, 1985; Adler et al, 1989; Cheng et al, 1994; Campbell et al, 1995). Such genetic interactions are not observed for every mutant combination and, in many cases, are allele-specific (Campbell et al, 1995). In addition, mutation of one PcG gene can, in some cases, be compensated by extra gene copies of another member (Cheng et al, 1994; Campbell et al, 1995). Taken together, these observations suggest that the function of the PcG is sensitive to the dosages of its members. The mass action model (Locke et al, 1988) has been proposed to explain this characteristic of PcG function. According to this model, each PcG gene encodes a different subunit of a multimeric complex. Reducing the amount of a particular subunit would result in a drastic effect on the stability and, ultimately, in the activity of the multimeric complex. IV. PcG complexes In addition to the genetic data, there is biochemical data supporting the notion that PcG proteins form multimeric protein complexes. First, immunohistochemical staining of third instar larval salivary gland polytene chromosomes with antibodies raised to PcG proteins showed that many PcG proteins are found bound to specific chromosomal sites. PH, PCL and PH bind at -100 sites in a completely overlapping pattern (Zink and Paro 1989; DeCamillis et al, 1992; Franke et al, 1992; Lonie et al, 1994); PSC, E(Z), A S X , SU(Z)2 and E(PC) share most of the PH-PC-PCL sites but bind to unique sites of their own (Martin and Adler, 1993; Rastelli et al, 1993; Carrington and Jones, 1996; Sinclair et al, 1998a; 1998b). The fact that there is only partial overlap between PcG binding sites on polytene chromosomes, as well as the fact that 6 different PcG mutants display different levels of homeotic and non-homeotic gene derepression (McKeon and Brock, 1991; Simon et al, 1992; Pelegri and Lehmann, 1994; Soto etal, 1995), suggests the existence of separate multimeric PcG complexes that may have different affinities for target genes. The second line of evidence for PcG multimeric complexes comes from in vitro protein binding assays that showed PSC interacts with PC and PH (Strutt and Paro, 1997; Kyba and Brock, 1998b), PH interacts with S C M (Peterson et al, 1997), and E(Z) interacts with ESC (Jones et al, 1998). Indeed, several protein domains required for protein-protein interactions characterize the PcG. The recurring domains such as the chromodomain of PC, the zinc finger and S A M domains in P H and S C M , the RING finger in PSC and SU(Z)2, the SET domain in E(Z), the WD40 repeats in ESC and PHD fingers in PCL, have all been implicated in mediating protein-protein interactions (reviewed in Orlando and Paro, 1995; Kyba and Brock, 1998a; 1998b; Satijn and Otte, 1999). It is probable that protein-protein interactions of this sort are responsible for the assembly of PcG complexes through the progressive and cooperative recruitment of PcG proteins as proposed for the formation of the SIR silencing complex at the silent mating type loci and telomeres in yeast (Moretti and Shore, 2001). More conclusive and direct evidence for the formation of PcG complexes comes from the biochemical purification of distinct PcG-containing complexes from Drosophila embryos (reviewed in Simon and Tamkun, 2002). The first example, termed PRC1 (Polycomb repressive complex 1), is >l-2MDa in size and contains PC, PSC, PH, S C M and dRINGl in association with 20 other proteins including components of GTFs (general transcription factors), TBP and the Zeste DNA-binding protein (Shao et al, 1999; Saurin et al, 2001). PRC1 inhibits remodelling of nucleosome arrays by SWI/SNF, a chromatin-remodelling complex that contains several trxG proteins. It has been proposed that the association of PRC 1 with TAFs (TBP-associated factors) represents a novel TAF-containing complex that has direct access to promoters and inhibits transcription initiation (Simon and Tamkun, 2002). A second complex, biochemically separable from PRC1, is approximately 600kD and contains ESC and E(Z) in association with the histone binding protein p55 and the histone deacetylase RPD3 (Ng et al, 2000; Tie et al, 2001). Identification of this complex implicates histone modification, in this case histone deacetylase (HDAC) activity, with PcG repression. Neither of these complexes have sequence-specific D N A binding activity and conspicuously absent from either of these complexes is the product of the pleiohomeotic gene, PHO, the only known D N A binding PcG member (Brown et al, 1998). Hodgson et al. (2001) have purified multiple Polyhomeotic (PH)-7 containing complexes with D N A binding activity that lack PC, showing that PH and PC are not obligate binding partners as previously thought. The evidence for multiple and distinct PcG complexes is compelling yet it remains unclear how the PcG silence their target loci. A diagrammatic representation of the known Drosophila PcG complexes is presented in Figure 1-2. V. PcG Homologues Since their initial discovery in Drosophila, PcG homologues have been found in many other organisms including human, mouse, chicken, Xenopus, Caenorhabditis elegans and plants (Table 1-1). The vertebrate homeotic genes (Hox genes) are functional homologues of the Drosophila homeotic genes (reviewed in Satijn and Otte, 1999). Like their Drosophila counterparts, the mouse PcG homologues are responsible for restricting Hox gene expression in spatially restricted domains along the anterior-posterior body axis. In mice, targeted gene replacements of the Psc homologues Bmi-1 (van der Lugt et al, 1996) and Mel-18 (Akasaka et al, 1996), of the Pc homologue M33 (Core et al, 1997), of the ph homologue rae28 (Takihara et al, 1997), and of the esc homologue eed (Schumacher et al, 1996) caused posterior transformations of the axial skeleton due to the ectopic expression of several Hox gene expression boundaries. When heterozygous PcG mice were crossed to obtain double mutant PcG mice the Hox gene expression profiles were more severely affected and the homeotic transformations became stronger (Bel et al, 1998). A l l PcG gene-deficient mice have provided evidence that these genes also play a crucial role in hematopoiesis (reviewed in Takihara and Hara, 2000). Like Drosophila, it has been shown that mammalian PcG proteins are members of large multimeric complexes and interact via conserved domains (Alkema et al, 1997a; 1997b; Hashimoto et al, 1998; van Lohuizen et al, 1998; Tie et al, 2001). The homology is not limited to common structural motifs as the mechanism of PcG action appears to be conserved in evolution. For example, a M33 transgene transformed into Drosophila partially rescues a Pc mutation (Miiller et al, 1995). Furthermore, experiments targeting the Drosophila PC, SU(Z)2 and PSC proteins to a reporter gene in mammalian cells, via fusion to the lexA D N A binding domain, resulted in silencing of the reporter (Bunker and Kingston, 1994). These results reveal that there has been remarkable conservation of the mechanism of PcG function between flies and mammals. 8 Figure 1-2. Drosophila PcG complexes. Local deacetylation together with PcG proteins i proposed to induce a change in chromatin conformation, allowing PcG complexes to bind and stably maintain repression. HB, G A F , Zeste, PHO and GTFs have all been implicated in recruiting PcG-containing complexes (denoted by arrows). See text for details. 9 Table 1-1. PcG homologues and prominent protein domains. PcG Gene Organism Protein Domains References Polycomb Pc XPc CHCB3 M33 MPc2 Cbx/HPCl HPC2 Posterior Sex Combs Psc XBmil Bmil Mel-18 BMU Polyhomeotic Ph Rae-28/MPhl MPh2 HPH1/HPH2 Sex Combs on Midleg Scm SCML1 Enhancer of Zeste E(Z) mes-2 Enxl/Enx2 EZH1/EZH2 Curly Leaf medea Extra Sex Combs Esc mes-6 eed EED fie Pleiohomeotic Pho YY1 Enhancer of Polycomb E(Pc) Epcl, Epc2 EPC1, EPC2 Additional Sex Combs Asx AsxLl, AsxL2 Suppressor of Zeste 12 Su(z)12 Emfl, Em/2 Chromodomain Drosophila Xenopus Chicken Mouse Mouse Human Human Drosophila Xenopus Mouse Mouse Human Drosophila Mouse Mouse Human Drosophila Human Drosophila C. elegans Mouse Human A. thaliana A. thaliana Drosophila C. elegans Mouse Human A. thaliana Drosophila Human Drosophila Mouse Human Drosophila Mouse Drosophila A. thaliana RING finger SAM (SPM) & C, zinc finger SAM (SPM) Zing finger SET WD-40 repeats Zinc finger DNA binding domain Paro etal., 1991 Brunk etal., 1991 Yamaguchi etal., 1998 Pearce etal., 1992 Alkema etal., 1997b Gecz etal., 1995 Satijnefa/., 1997 Martin etal., 1993 Brunk et al., 1991 HaupteJa/., 1991 Tagawa et al., 1990 Alkema etal., 1993 DeCamillis etal., 1992 Nomura et al., 1994 Hemenway et al., 1998 Gunster et al., 1997 Bornemann et al., 1996 Van de Vosse etal., 1998 Jones etal., 1990 Holdemann etal., 1998 Hobertef a/., 1996 Chen etal., 1996 Goodrich etal., 1997 Grossniklaus etal, 1998 Frei etal., 1985 Holdemanefa/., 1998 Schumacher et al., 1996 Schumacher et al., 1998 Ohadefa/., 1999 Brown etal., 1998 Shi etal., 1991 Stankunas e< a/., 1998 Stankunas etal, 1998 Stankunas etal., 1998 Sinclair et al., 1992 Unpublished Birve et al., 2001 Birve et al., 2001 V I . The polyhomeotic (ph) Locus Polyhomeotic (ph) has one of the strongest and most varied phenotypes of all PcG mutations. Embryos homozygous for null ph alleles are developmentally arrested midway through embryogenesis, approximately 12 hrs after egg laying (AEL). These embryos exhibit massive cell death in the ventral epidermis (Dura et al, 1987) indicating that ph is required for cell viability in the ventral epidermis. Embryos homozygous for hypomorphic ph alleles develop to the end of embryogenesis with most abdominal segments transformed towards the identity of A8 and the three thoracic segments transformed towards a more posterior fate (Dura et al, 1987). These homeotic transformations of the ventral cuticle are similar to those caused by other PcG mutations. In addition, some ph mutant alleles result in extensive central nervous system (CNS) defects, presumably as a result of misregulation of homeotic and segmentation genes in the CNS (Smouse et al, 1988; Dura and Ingham, 1988; Smouse and Perrimon, 1990). Further affirmation of its classification into the PcG came from genetic studies which showed that ph alleles enhance the mutant phenotype of other PcG mutations including Pc, esc, Psc, Pel, Asx, E(z), and Scm (Dura et al., 1985; Cheng et al., 1994). The polyhomeotic locus is complex. Genetic and molecular analyses show that the ph locus consists of two transcriptional units termed ph proximal (php) and ph distal (phD) (Dura et al, 1985; Dura et al, 1987; Deatrick et al, 1991). The php andphD units appear to be functionally redundant, because two mutational events (one in each transcriptional unit) are necessary to obtain the null lethal phenotype (Dura et al, 1987; Boivin et al, 1999); mutations in either transcription unit alone exhibit weak viable phenotypes (Dura et al, 1987). Hodgson et al. (1997), however, argue against functional redundancy as they show thatphp andphD mutations have differing effects on the regulation of a reporter under the control of a large 14kb Ubx regulatory region named the bithoraxiod fragment, bxdXA. Because PH localizes to ectopic bxdXA insertion sites in transgenic flies (DeCamillis et al, 1992), different expression patterns of the transgene probably reflect direct differences in PH binding. The php unit encodes two embryonic mRNAs of 6.4 and 6.1kb which encode proteins of 170 and 140 KDa (PHP-170 and PHP-140) respectively (Hodgson et al, 1997). The phD unit transcribes a 6.4kb embryonic mRNA which encodes a 135 KDa protein (PHD-135) (Hodgson et al, 1997). PHP-170 and PHD-135 share 92% identity over a large region. PHP-170 contains a unique 192 amino acid domain at its amino terminus and a smaller region near the carboxyl terminus that shares 42% identity with PHD-135) (Hodgson et al, 1997). The characterized protein domains of PH include a zinc-finger and the S A M / S P M domain, both protein-protein interaction domains. The 11 SAM/SPM domain mediates homotypic self-association of PH and heterotypic interactions with other SAM/SPM-containing proteins (Kyba and Brock, 1998a; Kim et al, 2002). As mentioned earlier, antibody staining of polytene chromosomes shows 100% colocalization of PH-PC-PCL at approximately 100 sites (Franke et al, 1992). PSC, E(Z), A S X , SU(Z)2 and E(PC) share most of the PH-PC-PCL sites but also bind to unique sites (Martin and Adler, 1993; Rastelli et al, 1993; Carrington and Jones, 1996; Sinclair et al, 1998). These data indicate that PH may be a constituent of many heterogeneous multimeric complexes. One PH-containing complex, PRC-1, is capable of inhibiting chromatin remodelling by SWI/SNF complexes (Shao et al, 1999; Saurin et al, 2001). The role of PH in PRC-1 remains unknown. It has recently been shown that a specific PH isoform (PHP-140) can be immunoprecipitated from nuclear extracts with Barren and D N A Topoisomerase II, proteins essential for chromosome condensation and segregation during mitosis (Lupo et al, 2001). ph null embryos show defects in chromosome segregation, the same phenotype observed for barren mutant embryos, confirming the role of PH in mitotic chromosome segregation (Lupo et al, 2001). It has been suggested that one possible role of PH in maintenance of gene expression may be to regulate the D N A topology of chromosomal domains. A primary objective of this thesis is to determine how the PcG, with special emphasis on PH, interact with their target loci and silence gene expression. VII . Initiation of Silencing The initiation of homeotic gene silencing is conferred by the collective action of the segmentation gene products. Since the products of these genes are only transiently available, how is early transient repression of homeotic genes converted into permanent silencing? The first problem is how do the PcG proteins, which are ubiquitously expressed, recognize specific regulatory targets only in domains where their repressive activity is required? Two possibilities exist. The most obvious hypothesis is that the early repressors directly recruit the PcG to their target loci via protein-protein interactions promoting the generation of a repressive multimeric complex at these sites (Zhang and Bienz, 1992; Bienz, 1992). Indeed, sequence examination of the cis-regulatory elements of some PcG target loci revealed that they contain consensus repressor protein binding sites (Simon et al, 1990; Muller and Bienz, 1991; Busturia and Bienz, 1993). It has been proposed that the product of the segmentation gene hunchback (HB), directly or indirectly recruits PcG proteins to target loci to establish PcG silencing of Ubx (Qian et al, 1993). Kehle et al (1998) have identified a protein, dMi-2, that directly interacts with HB and is specifically required for Ubx silencing in Drosophila. Interestingly, Xenopus and human Mi-2 12 were purified as subunits of a complex possessing histone deacetylase and ATP-dependent nucleosome remodeling activities (Wade et al, 1998; Zhang et al, 1998). These findings have led to the speculation that the HB-dMi-2 complex might recruit an HDAC-containing PcG complex. However, Ubx regulatory elements lacking known HB binding sites are able to establish silencing (Poux et al, 1996). Alternatively, the chromatin structure of the repressed transcriptional state of the homeotic genes may be recognized by a PcG member or a PcG-containing complex and subsequently recruited to these loci. Consistent with this hypothesis, Polycomb (PC) interacts with nucleosomal core particles in vitro and the main nucleosomal binding domain coincides with a region in the C-terminal part of PC previously identified as the repressive domain (Breiling et al, 1999). A common theme for both of these hypotheses is the recruitment of the PcG to their target loci. Is recruitment of the PcG sufficient to establish the transition from initiation to maintenance of silencing? Support for the recruitment model comes from studies which have shown that certain PcG proteins act as a transcriptional silencers in Drosophila embryos when tethered to reporter genes by a variety of D N A binding domains (Muller, 1995; Poux et al, 2001a; Roseman et al, 2001). The silencing effects observed in these studies were dependent on other PcG proteins, indicating that the targeted PcG protein recruits other members. Recruitment of PcG complexes to their target loci is definitely an important facet of PcG biology but there are other equally important aspects. V I I I . Mechanisms of P c G silencing Once recruited to their targets, what are the molecular mechanisms of PcG-mediated transcriptional silencing? How are the silencing complexes inherited through time and cell divisions? Many models have been proposed to explain PcG function and the evidence for any model is not concrete. These models are schematically represented in Figure 1-3 and will be discussed in turn. A. Heterochromatin Model Based on sequence similarity between PC and the heterochromatin protein HP1, encoded by the Su(var)205 locus, a modifier of position-effect variegation (PEV), it has been suggested that PcG proteins induce the formation of stable heterochromatin-like structures at their target loci (Paro and Hogness, 1991) (Figure 1-3A). Cytological studies have defined two types of a n d 15 Heterochromatin and Histone Deacetylation Models DNA Looping Model Figure 1-3. Mechanisms of PcG Action. See text for details. Interference with Basal Transcription Machinery Figure 1-3. Cont inued chromatin: euchromatin, which appears as an extended structure and is transcriptionally active, and heterochromatin, a late-replicating, highly-compacted, transcriptionally silent structure (Grunstein et al., 1995; Henikoff, 2000). The phenomenon of P E V is the mosaic expression that occurs when a chromosomal rearrangement juxtaposes a euchromatic gene near heterochromatin (reviewed in Wakimoto, 1998). PcG-induced higher order heterochromatin-like structures are presumed to spread along a region of the chromosome blocking the access of R N A polymerase II and transcriptional activators to their targets. In addition, these higher order chromatin structures are stably inherited, through D N A replication, through reassembly of their protein complexes (Paro, 1993). This model is based on the heterochromatin-mediated transcriptional silencing observed at the yeast silent mating-type loci (HMR and HML) and at yeast telomeres (reviewed in Gasser and Cockell, 2001). Transcriptional silencing at these loci depends on the spreading of a complex containing the silent information regulators SIR3 and 4 and histones H3 and H4 (Hecht et al, 1995). Strong support for the spreading model comes from chromatin immunoprecipitation (ChIP) experiments that showed PC and other PcG proteins are structural components of large, silent chromosomal domains spanning several kilobases (Strutt et ah, 1997). Indeed, like SIR3 and SIR4, PC is able to bind to nucleosomes in vitro (Breiling et al, 1999). However, a number of observations appear to argue against this model. Sequence similarities, such as the chromodomain, do not necessarily indicate functional relationships. High resolution ChIP experiments have shown that PC does not spread along the regulatory regions of the silenced homeotic genes Ubx and Abd-B, but is highly enriched at discrete sequence elements (Orlando et al, 1998). Examination of chromatin accessibility on PcG repressed loci yielded conflicting results. Schlopherr et al. (1994) showed that the accessibility of restriction enzymes or exogenous polymerases to loci repressed by the PcG was unaltered. Similarly, McCall and Bender (1996) found that a reporter gene containing a GAL4 upstream activating sequence (UAS) inserted within the Ubx transcription unit was efficiently silenced in parallel with Ubx itself while the T7 R N A polymerase could still transcribe from its target promoter in all segments of the embryo. These results suggest that PcG-mediated repression does not involve major changes in the structure of the chromatin fibre. Conversely, Boivin and Dura (1998) have shown that a PcG silenced gene was less accessible to Escherichia coli D N A methyltransferase. Similarly, in a more elegant study, the chromatin structure of the bithorax complex was probed with three separate assays for D N A accessibility: (i) activation of RNA polymerase II (Pol II) transcription by GAL4, (ii) transcription by the bacteriophage T7 RNA 16 polymerase (T7RNAP), and (iii) FLP-mediated site-specific recombination (Fitzgerald and Bender, 2001). A l l three processes were restricted or blocked in Polycomb-repressed segments. Taken together, these data suggest that, to a certain extent, silenced chromatin is structurally different than active chromatin, and does not simply exclude all proteins. B. The Histone Deacetylation Model A related hypothesis, but not necessarily mutually exclusive, posits that the PcG form complexes on their target loci and modify higher-order chromatin structure through deacetylation of acetyl-lysine moieties in the N-terminal tails of histones, thereby influencing gene expression (Figure 1-3 A). A n ever-increasing number of studies are reporting a mechanistic link between the PcG and the state of histone acetylation at target loci. There is a strong correlation between deacetylation of histone H4 and gene silencing at the Fab-7 regulatory element, a known PcG target (Cavalli and Paro, 1999). Accordingly, hyperacetylation of H4 was found to be associated with the switching of the Fab-7 element from an inactive to an active state during embryogenesis. Biochemical analyses have shown that histone deacetylases (HDACs) copurify with PcG complexes in Drosophila in vivo (Tie et al., 2001; Chang et al, 2001). In mammals, H D A C activity co-immunoprecipitates with the PcG protein EED and H D A C inhibitors relieve EED-mediated gene silencing (van der Vlag and Otte, 1999). In light of the increasing awareness of the interaction between chromatin and transcription, the involvement of deacetylation of histones in PcG silencing is bound to receive attention. C. The D N A Looping Model Another model of PcG function proposes that multiple PcG binding sites are strategically placed along the D N A such that cooperative binding between the PcG complexes formed at these sites results in looping of the D N A (Pirrotta, 1995) (Figure 1-3B). This model is also referred to the hop-and-skip model (Pirrotta, 1998). This looping would prevent enhancers from interacting with their respective promoters. Indeed, a number of weak PcG binding sites have been identified in the Ubx gene that when combined confer stronger maintenance of a reporter in vivo (Pirrotta and Rastelli, 1994). A precedent for repression based on D N A looping is found in the arabinose operon in E. coli (Lobell and Schleif, 1990). The repression of the araBAD promoter is mediated by D N A looping between AraC protein bound at two sites near the promoter separated by 210bp, aral and ara02. The addition of arabinose, which induces the operon, breaks the loop, and promotes transcription of the operon. This is an attractive model 17 for PcG function as it can explain the fact that the D N A remains essentially accessible to certain trans-acting factors and the dose sensitivity of the PcG since correct assembly of the looping structures depends on multiple cooperative interactions between the different members. Although multiple PcG binding sites have been detected throughout the fly genome, direct evidence of D N A looping does not exist. D. Interference of basal transcription machinery Another model proposed for PcG function is that PcG proteins interfere with the assembly or activity of the basal transcription machinery (Figure 1-3C). This model was initially proposed by Bienz (1992) and is presently receiving a great deal of attention owing to several key findings. ChIP experiments showed that PC binds to promoter fragments of silenced homeotic genes (Orlando et al., 1998). More recent experiments showed that binding of PcG proteins to repressed promoters does not exclude general transcription factors (GTFs), PcG proteins interact in vitro with GTFs (Breiling et al., 2001), and PRC1, a PcG-containing complex fractionated from Drosophila embryos, contains GTFs and antagonizes chromatin remodelling by trxG-related SWI/SNF complexes (Saurin et al., 2001). Taken together, these results strongly suggest that PcG complexes maintain silencing by operating at promoters to inhibit GTF-mediated activation of transcription. This model is strongly reinforced by the finding in yeast that showed cohabitation of SIR proteins with a transcriptional activator and components of the transcription preinitiation complex (PIC) at the promoters of two SIR-induced silenced loci (Sekinger and Gross, 2001). E. Nuclear compartmentalization model In the nuclear compartmentalization model, the chromatin fibre is left unaffected and inactivation by the PcG is achieved by sequestering the target loci to specific compartments within the nucleus which may be inaccessible to some transcription factors (Figure 1-3D). This model is based on the findings that yeast SIR3 and SIR4 proteins, which are required for the heritable silencing of genes within the silent mating type loci and telomeres, are required for the perinuclear localization of yeast telomeres (Palladino et al, 1993). In addition, the Ikaros family of repressors initiate silencing of target genes in mature B and T lymphocytes through a direct effect on the promoter with localization to pericentromeric heterochromatin rather than causing heterochromatinization of their target genes (Sabbattini et al., 2001). Support for the nuclear compartmentalization model comes from a confocal microscopy study in human cell lines which 18 showed that a subset of PcG 'bodies' are not randomly dispersed but appear clustered into defined areas within the nucleus (Saurin et al, 1998). However, more compelling evidence against this model is the observation in vivo that multiple PcG complexes appear randomly distributed in complex patterns of 100 or more loci throughout most of the interphase nuclei of whole developing Drosophila embryos (Buchenau et al, 1998). IX. The Trithorax Group (trxG) Whereas members of the PcG are required for the propagation of the silent state, members of the trxG are required to stably maintain the active state of homeotic gene expression patterns established by the segmentation genes during embryogenesis. Homozygous trxG mutations cause transformations of the first and third thoracic segments to the second and anterior transformations of the abdominal segments (Capdevila and Garcia-Bellido, 1981). The trxG genes exhibit antagonistic genetic interactions with PcG genes suggesting that trxG proteins function to counteract or modulate PcG-mediated silencing. Indeed, most trxG genes were identified in a genetic screen for suppressors of the dominant homeotic phenotypes of Pc (Kennison and Tamkun, 1988; Tamkun et al, 1992). Like the PcG, mutations in trxG genes show strong genetic interactions among themselves (Shearn, 1989) and trxG proteins form multimeric protein complexes (Dingwall et al, 1995; Papoulas et al, 1998; Kal et al, 2000; Petruk et al, 2001). Although the action of the trxG is varied and complex, a unifying theme for the role of several trxG proteins lies in their chromatin remodelling activities that enhance the accessibility of transcriptional activators to DNA. In yeast, Drosophila and mammals, several types of ATP-dependent nucleosome remodelling complexes have been identified (SWI/SNF, NURF and CHRAC) all of which contain homologues of the trxG protein Brahma or related proteins (Cote et al, 1994; Tsukiyama and Wu, 1995; Varga-Weisz et al, 1997). In addition to chromatin remodelling, trxG function is associated with histone modification. TAC1, a l M D a Trithorax (TRX) containing complex, contains a member of the CBP/p300 family of histone acetyltransferases (HATs) (Petruk et al, 2001). TAC1 action at trithorax-response elements (TREs) may increase local acetylation of histones, thereby increasing accessibility of transcriptional activators to homeotic genes. Accumulating evidence suggests that an intricate relationship between the PcG and trxG proteins determines the higher-order chromatin structures responsible for long-term maintenance of stable transcriptional states of their target loci. 19 X. Polycomb Response Elements (PREs) Using reporter gene constructs, several cw-regulatory elements necessary for maintaining the silent state of homeotic genes were identified (Simon et al, 1990; Muller and Bienz, 1991; Busturia and Bienz, 1993; Simon et al, 1993; Chan et al, 1994; Chiang et al, 1995). The ability of these elements to maintain the silent state of downstream genes depends on the establishment of a repressive complex involving PcG proteins and their targets, the PcG response elements (PREs). PREs are "cellular memory modules" that lock in transcription states (active or silent) determined early in embryogenesis and maintain them throughout development (Simon, 1995; Cavalli and Paro, 1999). In fact, removal of a PRE from a silenced gene has been shown to result in a loss of silencing, even if the PRE is removed late in development (Busturia etal, 1997). PREs have been identified in the regulatory sequences of the homeotic genes proboscipedia (pb) (Kapoun and Kaufman, 1995), Sex combs reduced (Scr) and Antennapedia (Antp) (Gindhart and Kaufman, 1995), Ubx (Chan et al, 1994; Chiang et al, 1995), abd-A (Chiang et al, 1995), abd-B (Hagstrom et al, 1997) and from the non-homeotic genes engrailed (en) (Kassis, 1994), polyhomeotic (ph) (Fauvarque and Dura, 1993) and hedgehog (Rob Saint, personal communication). PRE function is very diverse as different PREs respond to different PcG genes. For example, the en PRE responds to different alleles of Psc and Pel but does not respond to Pc mutations (Kassis, 1994), the ph PRE responds to ph and Pc mutations (Fauvarque and Dura, 1993) and the pb PREs do not interact genetically with any PcG genes (Kapoun and Kaufman, 1995); one Scr PRE is sensitive to esc but not E(z) mutations while the other Scr PRE is sensitive to E(z) but not esc mutations (Gindhart and Kaufman, 1995). These results strongly suggest that different PcG complexes function at different PREs and provide further support for the hypothesis of multiple PcG complexes in the nucleus. A l l the PREs identified thus far share three characteristics. First, when PREs are combined with parasegment-specific initiation elements, they are able to maintain segmentally restricted patterns of expression conferred on a reporter by the initiation elements (Simon et al, 1990; Muller and Bienz, 1991; Busturia and Bienz, 1993; Chan et al, 1994; Chiang et al, 1995). In these experiments, PRE-mediated silencing of the reporter gene is strongly influenced by the genomic insertion site as a given PRE construct is strongly silenced at some insertion sites but not at others. This position-effect implies that the silencing ability of a PRE depends on the sum of the interactions with other enhancer or silencer elements located near the chromosomal site of insertion. Second, using the larval salivary gland polytene chromosome assay, transposons 20 containing a PRE create a new binding site for PcG proteins at its insertion site on polytene chromosomes (Zink et al, 1991; DeCamillis et al, 1992; Chan et al, 1994), which suggests that PREs are physical targets of the PcG proteins. Third, PREs cause PcG-dependent pairing-sensitive repression of the mini-white gene (Kassis, 1994; Chan et al, 1994; Gindhart and Kaufman, 1995; Pirrotta, 1999). The mini-white gene is required for the pigmentation of the eye and is included in transposon constructs as a transformation marker. This pairing-sensitive silencing is an example of a phenomenon in Drosophila known as transvection, in which regulatory interactions occur in trans between elements on different chromosomes (reviewed in Kennison and Southworth, 2002; Kassis, 2002). The silencing activity of a PRE, determined by the degree of variegation of the mini-white gene contained in the transposon, is frequently enhanced if the sequences flanking the insertion site can pair with the homologous sequences on the other copy of the chromosome. Silencing is dramatically enhanced if two copies of the PRE-containing transposon are homologously paired. This silencing activity was first reported for a 2.4kb fragment of D N A from the Drosophila engrailed gene (Kassis, 1991). In some cases, trans-interactions occur between PREs inserted at different sites and even on different chromosomes (Sigrist and Pirrotta, 1997). These results suggest that the PcG complexes on the PRE are stabilized by the physical proximity of other complexes as well as by their flanking sequences. Pairing-sensitive repression has only been observed in transgenic analyses but is not limited to mini-white expression in the adult eye. To date, pairing-sensitive repression of the yellow and mini-white promoters have been documented in the adult epidermis and testis (Hagstrom et al, 1997; Mallin et al, 1998; Muller et al, 1999). Another characteristic of PREs is that transposons carrying a PRE frequently insert at chromosomal regions that already contain PREs (Fauvarque and Dura, 1993; Kassis, 1994). This is known as the homing phenomenon and is not unique to PRE-containing transposons (Taillebourg and Dura, 1999). Pairing-sensitive repression and the homing phenomenon suggest that PcG complexes formed on PREs tend to interact with other PRE complexes resulting in more stable PcG complexes and more efficient silencing. XI. Conserved P R E Sequence Motifs Sequence comparisons have revealed several conserved motifs among the characterized PREs (Mihaly et al, 1998; Dellino et al, 2002). Interestingly, the recently cloned PcG gene pleiohomeotic (pho) encodes the only sequence-specific D N A binding protein of the PcG identified thus far that is able to directly bind to a conserved PRE sequence motif (Brown et al, 21 1998). The pho gene is the Drosophila homologue of mammalian Ying-Yang 1 (YY1); Y Y 1 is a ubiquitously expressed zinc finger-containing D N A binding protein that is able to act as either a transcriptional repressor or activator (Austen et al, 1997). In fact, the conserved sequence motif of the PRE includes the PHO core consensus and extends beyond it. PHO binding sites have been shown to have an essential role in the silencing activity of at least five different PREs. These include the PREs of the engrailed locus (Brown et al., 1998), the bxd PRE of Ubx (Fritsch et al, 1999), the iab-2 PRE of the abd-A gene (Shimmel et al, 2000), the PRE within the Mcp element from the iab-5 regulatory region of the Abd-B gene (Busturia et al, 2001) and the iab-7 PRE also from the Abd-B gene (Mishra et al, 2001). These studies revealed that PHO binding sites by themselves are not sufficient to serve as PREs, suggesting that PHO may be the DNA-interacting component of an exclusive subset of PcG complexes. However, specific interactions between PHO and other PcG proteins have not been demonstrated. Furthermore, tethering PHO to a reporter gene via the LexA D N A binding domain is neither able to silence the reporter nor to interact with PC-containing complexes (Poux et al, 2001a). These studies suggest that additional sequence elements are required to recruit PcG complexes. Another possible PcG complex recruiter might be G A G A factor (GAF) whose binding sequence is a common sequence motif found in many, but not all, PREs. G A G A factor, product of the trxG gene trithorax-like (trl), was first identified as a transcription factor that bound a repetitive G A element upstream of the engrailed (Soeller et al, 1993) and Ubx promoters (Biggin and Tjian, 1988). Consistent with its classification as a trxG member, G A F aids in the local remodelling of chromatin templates in vitro by relieving its repressive effects, allowing for transcription factor access to the promoter and subsequently transcription initiation (Tsukiyama et al, 1994; Okada and Hirose, 1998). Curiously, in the case of the iab-7 and the Mcp PREs of the abd-B gene, both of which contain consensus G A G A binding sites, a trl mutation caused suppression of PcG-mediated silencing, an effect that is expected from a PcG mutation (Hagstrom et al, 1997; Busturia et al, 2001). However, in the case of the bxd PRE of Ubx, which also contain consensus G A G A binding sites, trl mutations had no effect on PcG mediated silencing (Hodgson et al, 2001). This is a difficult result to explain as ChIP experiments confirm that GAF and PC binding sites coexist in vivo at these PREs (Strutt et al, 1997). Furthermore, biochemical assays revealed that PcG binding to some bxd PRE fragments is dependent on consensus sequences for G A F and that GAF is a component of at least some types of PcG complexes (Horard et al, 2000; Hodgson et al, 2001). However, other bxd PRE fragments lack G A F binding sites and still bind PcG complexes in vitro. Although GAF is 22 intimately associated with some PREs and despite its general function in transcriptional activation, it is required for PcG function, probably by configuring the chromatin to favour a transcriptionally inactive state (Horard et al, 2000, Hodgson et al, 2001). Although consensus sequences have yet to be identified, the latest type of sequence-specific activity found at PREs is Trithorax (TRX) binding. Trithorax is the archetypal member of the trxG and is a positive regulator of homeotic gene expression in Drosophila (Breen and Harte, 1991). ChIP experiments were used to map T R X binding sites within the B X - C of embryonic chromatin at high resolution (Orlando et al, 1998). T R X showed strong association with all the B X - C PREs and simultaneously bound with PcG proteins to these PREs in both the repressed and the active state (Orlando et al, 1998). Tillib et al. (1999) have since corroborated the ChIP data by mapping the identified T R X binding sites within the B X - C PREs to D N A fragments of between 200 and 2000 bp. Their biochemical and genetic data show that the identified T R X binding sites are either within or very close to minimal PREs or PC binding sites identified previously and that they are required for PRE function. Indeed, trx mutations reduce expression of transgenes driven by the bxd PRE sequences (Petruk et al, 2001). More recently, Poux et al. (2002) have been able to immunoprecipitate subfragments of the bxd PRE using anti-TRX antibodies. The close proximity of PREs and Trithorax response elements (TREs) indicates that a tight relationship exists between the PcG and the trxG and that their functions are somehow coordinated. In fact, six genes originally classified as PcG genes have now been reclassified as Enhancers of trxG and PcG (ETP) genes because they are required for both silencing and activation (Gildea et al, 2000). This reclassification of members of the trxG or the PcG to the ETP group was challenged by the findings of Bajusz et al (2001) who concluded that reassignment depended on the test system used and that in some cases the unexpected phenotypes resulting from specific mutant combinations may be due to the tissue- and allele-specific alterations of a global balance between activators and repressors of homeotic genes. Nevertheless, an intimate relationship exists between the PcG and trxG and this relationship is best expressed by Brock and van Lohuizen (2001) who have proposed renaming PREs to 'maintenance elements' in order to account for the dual function of these regulatory elements. XII. P R E Functional Assays Currently three assays exist in Drosophila that permit testing for PRE function in vivo; an embryonic assay, a larval assay and an adult assay. First, transposons carrying PREs placed adjacent to parasegmental enhancers and upstream of a homeotic gene promoter regulate the 23 expression of a 6-galactosidase (lacZ) reporter gene in spatially restricted domains in Drosophila embryos. Depending upon the enhancers and promoter contained within the transposon, a PRE can restrict expression in different anterior-posterior domains. The ability of a PRE to maintain the silent state of a reporter gene is dependent on the function of the PcG. This characteristic provides a genetic assay to determine i f a given fragment is a PRE and to determine the dependence of PRE function on particular PcG members. For the embryonic assay, developmentally staged embryos are collected, formaldehyde fixed, stained for LacZ expression and mounted. PRE activity is measured by the degree of misexpression relative to that of a positive control. For example, if a bxd PRE-containing transgene expresses LacZ beyond the boundary of PS6 (anteriorly) this constitutes misexpression as the bxd regulatory region of Ubx confers a spatially restricted expression pattern limited to PS6-13 (Simon et al, 1990). The ability to look for failure of maintenance of silencing at different times during embryogenesis provides information about the silencing capabilities of the element being tested. In addition, tissue-specific misexpression can be observed with this assay as all three germ layers and structures within them can be discerned and examined in post-gastrulation embryos. Secondly, because transposons containing PREs create new binding sites for PcG proteins at their insertion sites, polytene chromosomes of third instar larval salivary glands can be stained with an antibody raised against a PcG protein. This assay localizes the position of newly integrated PREs and provides an efficient means to determine which PcG proteins recognize the PRE within the transgene. Thirdly, PREs cause variegated expression and confer pairing-sensitive repression on the mini-white reporter gene. Both of these properties are sensitive to mutations in PcG genes. If pairing-sensitive repression occurs, then the extent of repression will be greater when the transgene is homozygous than when it is heterozygous. The pairing-sensitive effects of a PRE can be assayed by visual examination of the adult eyes. XIII. Regulation of Ultrabithorax The Ultrabithorax (Ubx) gene of the Bithorax complex (BX-C) of Drosophila is the best-characterized homeotic gene both genetically and molecularly (reviewed in Duncan, 1987). Ubx specifies the identities of PS5 and 6 which correspond to the region extending from the posterior compartment of the second thoracic segment (T2) to the anterior compartment of the first abdominal segment (Al ) , the haltere and third legs respectively (Figure 1-4) (Lewis, 1978; Morata and Kerridge, 1981). The Drosophila haltere is a T3 structure and is a much reduced and specialized hind wing, which functions as a balance organ. Ubx mutations are homozygous 24 Figure 1-4. Regulatory elements of the bithorax complex. The three transcription units that encode the three homeotic proteins of the bithorax complex are shown in relation to their respective regulatory elements. Each of these genes is transcribed right to left. The exons are shown as dark boxes, the introns as dashed lines. The regulatory regions of the Ultrabithorax gene are expanded and shown below the fly. The anterobithorax (abx) and bithorax (bx) regulatory domains program the spatial distribution of Ubx gene expression in parasegment 5 while the bithoraxoid (bxd) and postbithorax (pbx) regulatory domains restrict Ubx expression to parasegment 6 (after Peifer and Bender, 1986; Beachy, 1990; Casares and Sanchez-Herrero, 1995). 25 lethal; the embryos die at the first larval instar stage with the most obvious phenotype being the homeotic transformation of parasegments 5 and 6 (PS5 and 6) into PS4 (Lewis, 1978). Ubx also plays a minor role in specifying the identities of the abdominal parasegments, PS7 through 13 where it acts in conjunction with the other homeotic genes of the Bithorax complex, abdominal-A (abd-A) and Abdominal-B (abd-B) (Lewis, 1978; Sanchez-Herrero et al, 1985; Struhl and White, 1985). The activity of U B X is required throughout most of development for the specification of embryonic, larval and adult structures (Lewis, 1978). While Ubx specifies the identities of epidermal and neural structures in thoracic and abdominal parasegments (Lewis, 1978), it acts in the somatic mesoderm only in the abdominal parasegments and in the visceral mesoderm in PS7 (Bienz and Tremml, 1988). The spatial and temporal expression pattern of Ubx during development has been analyzed in great detail. Ubx mRNA transcripts are first detected at the syncytial blastoderm stage in a broad domain that includes the primordial of the posterior thorax and abdomen (Akam, 1983). By the cellular blastoderm stage, Ubx transcripts accumulate preferentially in PS6. During germ band extension, Ubx transcripts accumulate in the anterior compartments of the ectoderm at high levels in PS6 and decrease in intensity posteriorly from PS7 to 12 and weakest in PS5 and 13 (Figure 1-5A) (Akam et al, 1985). This expression pattern persists until germ band retraction and then drops. Following germ band retraction, Ubx transcripts are detected at high levels in the CNS in a pattern like that observed in the ectoderm. In addition, during dorsal closure, Ubx transcripts are detected in the visceral mesoderm around the forming midgut that corresponds to the anterior compartment of PS7. U B X protein first becomes detectable at early germ band extension and from this stage on it is distributed essentially as described for the Ubx transcripts (White and Wilcox, 1984, 1985; Beachy et al, 1985; O'Connor et al, 1988). The distribution of U B X in larval imaginal discs is identical to the distribution of Ubx transcripts, which are detected throughout the T3 haltere and leg discs. Imaginal discs are epithelial infoldings in the larvae of holometabolous insects (eg. Lepidoptera, Diptera) that rapidly develop into adult appendages (legs, antennae, wings etc.) during metamorphosis from larval to adult form. U B X is also weakly detected in the posterior portions of the T2 wing and leg discs and, as expected, absent in the T l leg, eye-antenna and genital discs. A major characteristic of the Drosophila homeotic genes is their large size. The Ubx gene spans approximately 125kb, of which 75kb corresponds to the Ubx transcription unit (Bender et al, 1983; Beachy et al, 1985; Hogness et al, 1985). By comparison, a whole mouse homeotic complex occupies approximately 120 kb, just about the size of the Ubx gene 27 A PcG mutant Figure 1-5. U B X expression pattern in wildtype and PcG mutant embryos. Germ band extended embryos are oriented anterior to the left and dorsal side up. In these embryos U B X was detected immunohistochemically using a polyclonal anti-UBX antibody and a secondary antibody conjugated to horse radish peroxidase. The anterior boundary of parasegment 6 (PS6) is indicated with an arrowhead. (A) Lateral view of a germ band extended embryo. U B X accumulates in the anterior compartments of the ectoderm at high levels in PS6 and decrease in intensity posteriorly from PS7 to 12 and weakest in PS5 and 13. (B) In PcG mutants, exemplified here by the polyhomeotic mutation ph2, U B X is ectopically expressed in the anterior parasegments indicating that the PcG are repressors of Ubx expression. 28 (Krumlauf, 1994). The regulatory regions of Ubx are scattered over a region of lOOkb and include both upstream and downstream sequences. This feature of Ubx led Lewis (1978) to overestimate the number of genes in the B X - C ; he proposed nine genes as the B X - C specifies nine distinct segment types (see Figure 1-4). In fact, the segment-specific elements he defined genetically turned out to be distinct regulatory elements, not structural genes (Simon et al, 1990; Muller and Bienz, 1991; Simon et al, 1993). Lewis's principal Ubx regulatory mutations have been mapped to four regions, each with a characteristic pattern of defects in specific parasegments. The anterobithorax (abx) and bithorax (bx) mutations, located about 35 kb downstream of the Ubx promoter (Peifer and Bender, 1986), show loss of the PS5 expression pattern in the embryonic central nervous system (White and Wilcox, 1984) and result in PS5 homeotic transformations in the adult (anterior haltere to wing) (Lewis, 1982; Casanova et al., 1985). The bithoraxoid (bxd) or postbithorax (pbx) mutations (Bender et ah, 1983), which are located about 25kb upstream of the Ubx promoter, show a reduction in Ubx expression in PS6 resulting in the transformation of PS6 into PS5 (posterior haltere into the posterior wing, respectively) (Lewis, 1978; Bender et al., 1985). The regulatory regions of Ubx contain cell-specific enhancers and silencers that drive Ubx expression in parasegment-specific domains. Indeed, germ line transformation assays, using a lacZ reporter gene, have confirmed the presence and complexity of these regulatory regions and have shown that many distinct elements contribute to the overall pattern of Ubx expression (Simon et al., 1990; Muller and Bienz, 1991; Busturia and Bienz, 1993; Simon et al, 1993; Chan et al, 1994; Chiang et al, 1995). The Ubx expression domain is established by several trans-regulatory genes and is achieved in two mechanistically distinct steps; the initiation and maintenance phases. In the initiation phase Ubx is activated in response to the segmentation gene products. The products of these genes are transcriptional regulators such as Fushi tarazu (FTZ) that activate Ubx in PS5 to 13 by binding to upstream and downstream enhancer elements. Ubx expression is restricted in the anterior PS (1-4) by the Hunchback (HB) segmentation protein whose expression pattern is opposite to Ubx; HB is expressed in the anterior PS (1-4) of the developing embryo. HB represses Ubx expression by competing with the activators for binding to overlapping binding sites within the parasegmental enhancers (Qian et al, 1991, 1993; Zhang et al, 1991). H B protein is available only for the first four hours of development however the Ubx gene must be restricted to its initial domain of expression for the remainder of development. To maintain this unique expression pattern during development, mechanisms of cellular memory have evolved that preserve transcription patterns in an epigenetic manner. The proteins of the PcG and the 29 trxG are part of such a mechanism and their involvement in Ubx regulation constitutes the maintenance phase. This second phase of Ubx expression requires the PcG that are necessary to silence Ubx expression in those cells where Ubx was initially OFF (Figure 1-5B) and the trxG that are necessary to maintain Ubx expression in those cells where Ubx was initially ON (Capdevila and Garcia-Bellido, 1981; Shearn et al, 1987). Determining how the PcG mediates the silencing of Ubx through the PRE is the major goal of this thesis. XIV. bithoraxoid Polycomb Response Element As mentioned earlier, the bithoraxoid regulatory region, located 25kb upstream of the Ubx promoter, is responsible for programming the spatial distribution of U B X protein in parasegment 6. Germline transformation assays have revealed that the minimal bxd fragment that is capable of mimicking the endogenous bxd expression pattern is a 14kb fragment, named bxdlA (Simon et al, 1993). Two subsequent studies have located two largely overlapping fragments within bxdlA, bxd6.5 (Chan et al, 1994) and bxd5.6 (Chiang et al, 1995) that are capable of silencing the Ubx/lacZ transgene in PS 1-5 for up to 15 hours of embryogenesis. The bxd6.5 and bxd5.6 PREs contain both embryonic and imaginal regulatory sequences. A search for the embryonic regulatory sequences within these two fragments revealed a smaller 5.1kb fragment, named bxd5.\ (Chan et al, 1994; Hodgson et al, 2001). A finer dissection of the bxdS.l PRE revealed three regulatory elements, the SI (1.8kb) and S2 (1.8kb) parasegment-specific enhancers and the M-element (1.5kb), a silencer maintenance element that contains PcG dependent silencers and is referred to as the bxdl.5 PRE (Figure 1-6). The separate SI and S2 enhancer elements express initially in PS6, 8, 10 and 12 during germ band extension (Chan et al, 1994; Horard et al, 2001). Expression appears in all parasegments at the end of germ band extension, indicating that repression cannot be maintained by these constructs. The bxdl.5 PRE gives no expression pattern by itself but confers long-term and long-range repression of the SI and S2 enhancers in PS 1-5 in a PcG-dependent manner and is able to silence the mini-white gene present in the same transposon (Chan et al, 1994). The bxdl .5 PRE can be subdivided into three regions: a central 661 bp region known as the pairing-sensitive region (PSR) (Sigrist and Pirrotta, 1997), flanked upstream by a 354bp fragment termed UPS (Upstream of pairing-sensitive region) and downstream by a 485bp fragment termed DPS (Hodgson et al, 2001). Genetic experiments have identified fragments within the bxdl.5 PRE that differentially respond to PcG mutations (Tillib et al, 1999; Horard et al, 2001). Furthermore, in vitro immunoprecipitation experiments have shown that sub fragments of the bxdl.5 PRE are able to interact with different PcG complexes 30 Figure 1-6. The bithoraxoid 5.1kb Polycomb response element of the Ubx gene. The coordinates of Bender et al. (1983) are shown at the top. The bxd5.\ PRE is located approximately 25kb upstream of the Ubx transcription start site. It is composed of three regulatory elements, the SI and S2 parasegment-specific enhancers, and the M-element or the bxdl.5 PRE, a silencer maintenance element that contains PcG-dependent silencers and is necessary for pairing-sensitive repression. The bxdl.5 PRE can be subdivided into three regions: a central 661bp region known as the pairing-sensitive region (PSR) flanked upstream by a 354bp fragment termed UPS (Upstream of pairing-sensitive region) and downstream by a 485bp fragment termed DPS (Hodgson et al., 2001). 31 32 present in nuclear extracts (Horard et ah, 2001). Taken together, these studies indicate that the bxdl.5 PRE is built up of multiple interaction sites with differing functions, which recruit different DNA-binding activities that contribute to the overall function of the PRE. In order to gain insight on the mechanisms of PcG-mediated silencing our laboratory has focused on studying how the product of the polyhomeotic (ph) gene, interacts with the bithoraxoid5.1 PRE. Using gel mobility shift analysis, four D N A binding activities were identified that contain PH and interact site-specifically with restriction fragments spanning bxd5.\, S1HB-90, MHS-70, MPA-168 and MHN-90. In Chapter 2 of this thesis I show, using a germline transformation assay, that deletion of any of these four PH-interaction sites, in the context of the bxd5.l PRE, disrupts PRE activity demonstrating that these sequences are necessary for PRE function in vivo. Furthermore, combinatorial deletion analysis of sites with related binding activities indicate that MHS-70 and MPA-168 constitute one functional unit of PRE activity, which is disabled by either mutation. By contrast, sites S1HB-90 and MHN-90 interact synergistically to promote PRE function. Mutational analysis of MHS-70 and MHN-90 show that in vitro binding depends on d(GA)3 repeats and OCT elements respectively. I show that mutation of the d(GA)3 repeat elements in MHS-70 in the context of bxdS.l PRE destabilizes embryonic maintenance of silencing demonstrating that d(GA)3 elements are required for PRE function in vivo. Mutation of the OCT elements within MHN-90 does not disrupt bxdS.l PRE activity indicating that other sequences within MHN-90 are required for silencing. Biochemical fractionation of the MHS-70 D N A binding activity revealed the presence of PH in conjunction with G A F and Pipsqueak (Psq), two proteins known to bind to d(GA) 3 repeat elements. Genetic analysis shows that mutations in trl (encoding GAF) and psq enhance the homeotic phenotypes ofph implicating these two proteins in PcG-mediated silencing. In chapter 3 of this thesis I provide evidence for the modular structure of the bxdl.5 PRE and determine the contributions of the individual modules to bxd PRE function. I show that the bxdl.5 PRE modules are able to directly silence (UPS+PSR, PSR+DPS, UPS and DPS) or repress (PSR) the basal Ubx promoter as opposed to silencing adjacent enhancers. Silencing conferred by the UPS and DPS modules is parasegment-selective and stage-specific. I demonstrate that embryonic silencing and pairing-sensitive repression are conferred by different bxdl .5 modules and are thus separable PRE functions. Genetic analyses reveal the role of different PcG members at individual modules in the PRE and show that some PcG members are required for embryonic silencing or pairing-sensitive repression or affect both processes. 33 Chapter 2 Functional role of the Polyhomeotic-binding sites within the bithoraxoid5.\ Polycomb Response element I. Introduction A. Site-specific interactions of PH with subfragments of bxdSA PcG proteins form complexes in vivo with PREs. To understand how these complexes form, it is essential to identify the sequences required for PRE function and to subsequently characterize the protein complexes that bind these sequences. The binding specificity of PcG proteins with PREs has been demonstrated in vivo either by the localization of PcG proteins to PREs in the larval salivary gland polytene assay (Pirrotta, 1997) or by immunoprecipitation of cross-linked chromatin with anti-PcG antibodies (Strutt et al., 1997; Strutt and Paro, 1997). These studies have defined the broad specificity of PcG binding to PREs at specific developmental stages, but have not addressed the sequence-specificity or the mode of interaction of PcG proteins with PREs. Nor have these studies documented the developmental changes in the composition and assembly of PcG complexes on PREs. To determine the sequence-specificity and mode of interaction of a PcG protein with a PRE throughout development, the interaction of PH with bxd5.\ was assessed by gel mobility shift assays (Hodgson et al, 2001; Hodgson and Brock, unpublished). Restriction fragments spanning bxd5.\ were analyzed in a gel mobility shift assay using crude nuclear extract from the Kcl67 tissue culture cell line and from developmentally staged embryonic nuclear extracts. The Kcl67 cell line is derived from embryos, lacks detectable expression of endogenous Ubx, and expresses all known PcG proteins (Cherbas et al, 199'4; Hodgson and Brock, unpublished), and thus is suitable to examine PcG-mediated repression of Ubx. Embryonic nuclear extracts were prepared from embryos staged 0-3 hrs, 3-6 hrs, 6-18 hrs A E L , which roughly correspond to the pre-silencing, initiation and maintenance phases of Ubx silencing respectively (Zhang and Bienz, 1992). The bandshift assay identified D N A binding activities that recognize three distinct fragments, MHS-70, MPA-168 and MHN-90 within the M-element, or bxdl.5, and a fourth 34 fragment from the SI enhancer, S1HB-90; the letter M indicates that the fragment is from the M -element, S1 indicates that the fragment is from the S1 enhancer, the letters following represent the restriction fragment and the numbers denote nucleotide length in base pairs (Figure 2-1). These four DNA-binding activities supershift in the presence of antibodies to PH, indicating that PH is present in these activities. One goal of this chapter is to determine i f these fragments are required for PcG-mediated silencing in vivo. I determined the effect of deletion of each of these four fragments on bxd5.l activity in embryonic silencing, and on pairing-sensitive repression (discussed in the general introduction). The binding activities recognizing these four fragments are developmentally regulated (Hodgson and Brock, unpublished). PH-containing binding activities recognizing MHS-70 and MPA-168 were not detected in 0-3 hr A E L embryo extracts. In contrast, binding activities recognizing S1HB-90 and MHN-90 were detected in the 0-3 hr embryo extracts. Considering that expression of Ubx is activated at late cellular blastoderm stage (2.5-3 hrs A E L ) and derepression in ph mutants is seen as early as 3.5 hrs A E L , the presence of PH complexes on sites S1HB-90 and MHN-90 at 0-3 hrs suggests an involvement in the initiation of silencing. Binding to MHS-70 and MPA-168 was first detected in embryos 3-6 hrs A E L , and changes in mobility of these activities were detected in the extracts from 6-18 hr embryos. Similar changes in binding activities were detected with the MHN-90 and S1HB-90 fragments. The developmental changes observed in the binding activities may track the transition from initiation to maintenance of silencing discussed in the general introduction. Therefore, I monitored i f changes in silencing of wildtype and mutant bxd5.l PRE transgenes that occurred in vivo reflect the in vitro biochemical characterization. Inspection of the sequences of each of these four fragments demonstrates sequence conservation between MHS-70 and MPA-168, and between MHN-90 and S1HB-90. Competition studies show that the binding activities recognizing MHS-70 and MPA-168 are related (Hodgson et al, 2001), as are the binding activities recognizing MHN-90 and S1HB-90 (Hodgson and Brock, unpublished). The MHS-70 sequence showed a complex arrangement of three distinct repeat elements. These include two terminal direct repeats of d(GA)3 elements flanking a central inverted d(GA)3 element, which are binding motifs for G A G A Factor (GAF) (Biggin and Tjian, 1988) and Pipsqueak (Psq) (Lehman et al, 1998). Interspersed between these repeats are two d(A) n tracts that are binding motifs for Hunchback (HB) (Stanojevic et al, 1989), and three direct repeats of a d(AGAGC) element with unknown function. Like MHS-70, the MPA-168 fragment has four non-overlapping d(GA)3 sequences and one d(A) n tract, but the 35 Figure 2-1. PH binding sites within the bxd5.1 PRE. The coordinates of Bender et al. (1983) are shown at the top. Gel mobility shift in combination with antibody supershift analyses revealed the presence of four restriction fragments within the bxd5.\ PRE that were bound by PH-containing complexes (Hodgson et al, 2001). These fragments (black boxes) are named by the first letters of the restriction enzymes defining them, followed by their length in base pairs. Small triangles mark the location of PHO binding sites and gray diamonds mark the location of d(GA) 3 repeat elements. Figure is drawn to scale. 36 arrangement of these sequence differs from that MHS-70 (Figure 2-2). However, unlike MHS-70, MPA-168 contains three PHO consensus binding sites located at its distal end. Sites S1HB-90 and MHN-90 share two common elements: (1) consensus POU homeodomain binding sites consisting of the -ATAT- , - A A A T or - A T T A - octamer sequences (Llyod and Sakonju, 1991; Phillips and Luisi, 2000) and (2) a variant of the HB consensus site. Although these sites are similar, they are not identical (Figure 2-3). Site MHN-90 contains an NTF-1 consensus site (NTF-1 is the product of the grainyhead locus) (Dynlacht et al, 1989) and more consensus POU homeodomain binding sites than S1HB-90. Together, the competition analyses and the sequence conservation suggest that similar activities bind each pair of fragments. An additional goal of this chapter is to determine if these pairs of fragments are required independently for PRE function. To this end, I compared the expression patterns of transgenes mutant for individual sites, and for the appropriate pairs of sites. The specific sites within MHS-70 required for the formation of PH-containing complexes were determined by competition bandshift assays using linker substitution analysis (Hodgson et al, 2001). Linker scanning mutations of any single d(GA) 3 element reduced but did not eliminate binding. Mutations in the central A/T rich sequences (LS3/6-7) partially abolished binding whereas mutations in both terminal d(GA) 3 elements (LSI/9) completely abolished binding, showing that these sites are essential for the formation of PH-containing complexes (Figure 2-4). Previous analyses have shown that d(GA) 3 sequences are recognized by the G A G A factor (GAF) (Biggin and Tjian, 1988) and Pipsqueak (Psq) (Lehman et al, 1998). Supershift and western analyses showed that specific isoforms of PH, G A F and Psq were present in a partially purified MHS-70 binding activity. Furthermore, a synthetic double stranded oligomer containing multiple copies of the d(GA) 3 repeat element was recognized by a D N A binding activity that contained PH, G A F and Psq. In this chapter I determined the in vivo contribution of the terminal d(GA) 3 elements (LSI/9) of MHS-70 to PRE function. I also determined if mutations in trl and psq alter PcG-mediated silencing. Linker substitution analysis of MHN-90 revealed that mutations in any one POU octamer recognition sequence did not disrupt binding in vitro (Hodgson et al, in preparation). Double mutations in any combination reduced but did not completely abolish binding. Mutation of all four OCT sequences completely abolished binding. I determined the in vivo contribution of these four OCT sequences in the context of the bxd5.\ PRE. The results show that each of the four PH binding sites identified in vitro are required for PRE function. The MHS-70 and MPA-168 fragments constitute one functional PRE unit 38 Figure 2-2. Shared sequence elements between sites MHS-70 and MPA-168 are the d(GA)3 repeats. (A) Nucleotide sequence of MHS-70 highlighting the important elements. The d(GA) 3 repeat elements, the Hunchback consensus binding sites {the d(A)„ tracts} and the d(GCTCT) element of unknown function are indicated. The orientation of the consensus binding sites are shown by arrows. (B) Like MHS-70, site MPA-168 has four non-overlapping d(GA) 3 sequences and one d(A) n tract, but the arrangement of these sequence differs from that MHS-70. Site MPA-168 also contains three PHO consensus binding sites at its distal end. 39 I i "CD < a CD u EH :;i 3 00 o a. IP ilS. c? EH EH rt! EH CD CD L L I <CD CD U EH CD .1 1 o w PH 3 pa Figure 2-3. A a n d B. Nucleotide sequence of sites MHN-90 and S1HB-90. Sites S1HB-90 and MHN-90 share two common elements: (1) consensus POU homeodomain binding sites consisting of the -ATAT-, - A A A T or - A T T A - octamer sequences (labelled OCT-1 through -4) and (2) a variant of the HB consensus site. Site MHN-90 also contains an NTF-1 consensus binding site. 41 42 E o U CU E s-o I o r-'CM a o • C/3 O O S l-H 00 vo ri O N I w\ C/3 « M II 3 •S -+1 *S 1 43 essential for embryonic silencing, whereas MHN-90 and S1HB-90 contribute synergistically to PRE function. I show that d(GA) 3 sequences are required for PRE function, and that mutations in genes encoding G A F and Psq interfere with PcG-mediated silencing. I also show that POU octamer sequences are dispensable for PRE function. The implications of these results are discussed. II. Results The biochemical data suggest that four fragments within bxdS.X recruit PH-containing binding activities in vitro. Therefore, the primary objective is to determine if these fragments are required in vivo for the wildtype function of the bxd5.\ PRE. The wildtype fragment and deletion derivatives thereof were subcloned upstream of a Ubx promoter//acZ fusion gene in the pCaSper4-L/i>x//acZ vector and subsequently transformed into Drosophila embryos (Figure 2-5; see Materials & Methods). Wildtype foxd-containing transgenes express LacZ in a spatially restricted expression pattern (expressed in PS6 to PS 13). For these experiments, developmentally staged embryos were stained for LacZ to determine the effects of deletions of these four fragments on embryonic silencing. Expression of LacZ beyond the boundary of PS6 (anteriorly) provides evidence for derepression. The ability to look for failure of repression at different times during embryogenesis may provide information about the order in which the binding sites operate and to what regions of the embryo they confer repression. I will also examine if the deletions affect pairing-sensitive repression in adult eyes. A. What is the expression pattern of the bxdSA control? In order to determine the in vivo contribution of each PH-binding site identified in vitro the effects of deleting each PH-binding site were assayed in the context of bxd5.\. This 5.1kb element contains the appropriate parasegmental enhancer elements that initiate and maintain the correct ON state of the Ubx expression pattern and the appropriate parasegmental repressor/silencer elements that initiate and maintain the OFF state. The bxd5.\ expression pattern is stable throughout embryogenesis and is clearly identifiable. Thus, any spatial and temporal differences of LacZ expression between the intact element and the deletion derivatives thereof are easily identified. Expression from the bxd5.\ control transgene can be detected at late cellular blastoderm (2.5 hours of development) as a broad domain in the posterior region of the embryo. The 44 Ubx promoter ^ac^ mini-white Bamm EcoRI S2 M Styl SI Hindlll BamHl EcoRI BamUl EcoRI Styl Styl AMHS-70 0 Hindlll AS1HB-90 f Hindlll BamHl EcoRI Styl AMPA-168 t Hindlll BamHl EcoRI Styl AMHN-90 •I Hindlll Figure 2-5. Diagrammatic representation of the Ubx/lacZ reporter construct and the wildtype and mutant bxd5.\ fragments tested. The reporter construct consists of the Ubx promoter fused in frame with the lacZ gene and the mini-white transformation marker, flanked by P-element ends. The wildtype bxd5.l fragment and deletion derivatives thereof were cloned upstream of the Ubx/lacZ reporter gene. The bxd5.l fragments are drawn to scale. endogenous Ubx expression pattern is first detected one hour later at 3.5 hours of development (White and Lehmann, 1986). The discrepancy between the expression of the bxd5.1 transgene and the endogenous gene is presumably due to the length of the Ubx gene (75 kb). The endogenous gene requires one hour for complete transcription (Kornfeld et al., 1989), compared to the lacZ gene which is considerably smaller (3 kb), and is therefore more rapidly transcribed. At germ band extension (4.5 hrs), bxd5.\ directs the expression of LacZ strongly in even-numbered parasegments from PS6-12, and weakly in odd-numbered PS7-13 (Figure 2-6B). The anterior boundary of LacZ expression is at PS6, and expression is not detected in PS 1-5. This expression pattern was limited to the ectoderm and has been described for larger bxd fragments, bxdlA and bxd6.5 (Chan et al., 1994; Chiang et al., 1995). The parasegmental pattern observed in the bxd5.1 transgenic lines is consistent with earlier findings which show that the segmentation genes play a major role in the activation of the Ubx gene (Ingham and Martinez-Artias, 1986; Ingham, 1988). Complete maintenance of the PS6 anterior boundary is observed in 2 of 10 transgenic lines until late embryogenesis. Nearly complete maintenance is exhibited in 8 of 10 lines tested. These latter lines exhibit spotty ectopic expression in the head and thoracic regions during germ band retraction, but with these exceptions, the correct anterior boundary is substantially maintained (Figure 2-6C). Misexpression of LacZ in these latter lines is confined to the ectoderm. Mesodermal misexpression of LacZ was detected in the later stages of development of some wildtype and mutant bxd5A transgenic lines. A simple explanation for this is chromosomal position-effects; the transgene may have landed near mesodermal enhancer elements. To confirm the dependence of bxd5.\ regulation on the PcG, transgenic bxd5.l adults were crossed to ph2,phm, Pc4, Asxu, Scmm and Pscx heterozygotes (see Table 5-1 on page 144 for a description of the mutant alleles used). Embryos transheterozygous for the bxd5A transgene and a PcG mutant allele exhibited complete misexpression of LacZ as early as full germ band extension (4.5 hours A E L ) in the ectoderm (Figure 2-6D) and, at later stages (9 hours AEL) , in the ectoderm and some mesodermal tissues. The severity of the misexpression phenotypes of bxd5.\ was similar in all genetic backgrounds tested showing that these PcG genes have an equivalent role in bxdSA -mediated maintenance of silencing. Pairing-sensitive repression was difficult to discern in some bxd5.\ transgenic lines because the levels of red pigment in the adult eyes were almost undetectable in the heterozygotes. However, of those transgenic lines with visible pigment (6/10), all exhibited very 46 Figure 2-6. Silencing mediated by the bxd5.l PRE. (A) Schematic representation of the pCaSper4-Ubx/lacZ transformation vector containing the wildtype bxd5.\ PRE. The PH binding sites are shown below the vector and are indicated as black boxes. (B) Transgenic germ band extended embryo showing the bxd5.l PRE expression pattern. The bxdS.X PRE construct directs the expression of LacZ strongly in the even-numbered parasegments from PS6-12, and weakly in odd-numbered PS7-13. Expression in the anterior parasegments, PS 1-5, is silenced by the bxd5.\ PRE. An arrowhead points to the anterior boundary of PS6. (C) At late germ band retraction, the bxd5.\ PRE continues to maintain the correct anterior boundary set at PS6. (D) The ability of the bxd5.\ PRE to maintain the silent state of the Ubx/lacZ reporter is dependent on polyhomeotic function. Loss ofph function results in loss of bxd5.\ PRE-mediated silencing leading to severe misexpression in the anterior parasegments. The misexpression phenotype observed in this embryo is representative of all the PcG mutants tested. (E) The tW5.1 PRE confers pairing-sensitive repression on the mini-white gene. Heterozygous flies (+/-) have more white* pigment in their eyes than their homozygous siblings (+/+). 47 48 strong pairing-sensitive repression of the mini-white gene (Figure 2-6E) because homozygous adult eyes were almost completely white. The results obtained with the intact bxd5A transgenic lines indicate that the correct bxd regulatory elements required for Ubx initiation and maintenance phases during embryogenesis are contained within this element. Thus, the bxd5A PRE provides an ideal regulatory element to assay the in vivo contribution of the PH binding sites, which are presumable repressive elements, located within this element. B. What is the effect of bxd5A deletions? i . faJ5.1 AMHS-70 and foo/5.1AMPA-168 I examined the effects of deletion of MHS-70 in the context of the bxd5A PRE. Initial expression of LacZ in the transgenic lines containing bxd5A AMHS-70 is identical to that of bxd5A. Expression is first detected at late cellular blastoderm stage (2.5 to 3 hours AEL) as a broad domain in the posterior region of the embryo. As development proceeds and as the bxd5A expression pattern emerges, one major difference between the bxd5A control lines and the bxd5A AMHS-70 lines is observed. At germ band extension the anterior boundary of expression set at PS6 is never maintained in the majority of the mutant lines obtained (5 of 6) (Figure 2-7B). At this and later stages of development, misexpression of LacZ is detected equally in all anterior PS (1-5) and is limited to the ectoderm. The overall expression pattern within PS6-13 is unaffected by this deletion. The expression pattern in the remaining line (1 of 6) is identical to the wildtype bxdSA lines; it is able to maintain the silent state in PS1-5 for up to 12 hours of embryogenesis. These results show that MHS-70 is essential for embryonic maintenance of silencing in the context of bxd5A. Nevertheless, the level of LacZ detected in each anterior PS is considerably reduced compared to the level of LacZ per posterior PS. Therefore, the MHS-70 deletion does not completely abolish silencing. This result implies that other fragments contribute to silencing in the bxd5A PRE. Deletion of MHS-70, in the context of bxd5A, did not affect pairing-sensitive repression of the mini-white gene. Homozygous transgenic adults consistently displayed a lighter eye colour than their heterozygous siblings, similar to the bxdSA control lines (data not shown). This result was expected as the MHS-70 deletion is downstream of the characterized pairing-sensitive region of bxd5A. The misexpression phenotype caused by the deletion of MPA-168 was indistinguishable from the misexpression phenotype caused by the deletion of MHS-70 in embryos (4 of 6 lines 49 Figure 2-7. Deletion of sites MHS-70 and MPA-168 disrupt bxd5.l PRE-mediated silencing in embryos. A l l embryos are at the germ band extension stage. An arrowhead points to the anterior boundary of PS6. (A) Wildtype bxd5A PRE expression pattern. (B) Deletion of site MHS-70, in the context of bxd5A, abrogates PRE activity. Expression of LacZ is detected in PS 1-5. (C) Deletion of MPA-168 results in a misexpression pattern similar to that seem in 6x^5.1 AMHS-70 embryos. (D) Deletion of both MHS-70 and MPA-168 results in a similar misexpression pattern as either mutant indicating that the region encompassing these two sites constitutes one functional unit of PRE activity. 50 „ S2 M SI 3'P R Ubx Promoter LacZ mini-white 5'P st _ l _ M H N - 9 0 M P A - 1 6 8 M H S - 7 0 mum K • S 1 H B - 9 0 bx45.1 R St M H N - 9 0 M P A - 1 6 8 M H S - 7 0 S 1 H B - 9 0 B C AMPA-168 ^ K M H N - 9 0 M P A - 1 6 8 AMHS-70 S 1 H B - 9 0 M H N - 9 0 AMPA-,68 M H S - 7 0 S 1 H B " 9 0 M H N - 9 0 AM PA-168 AMHS-70 S 1 H B - 9 0 51 tested) (Figure 2-1C). The remaining two lines were able to maintain the silent state until 12 hours of embryogenesis. Similarly, deletion of MPA-168 did not affect pairing-sensitive repression (data not shown). These results show that MPA-168 is essential for embryonic maintenance of silencing in the context of bxd5.\, but is not required for pairing-sensitive repression. i i . fccc/5.1AMHN-90: Initial expression of LacZ in the transgenic lines containing 6x<s?5.1AMHN-90 is identical to that of bxd5.1. As germ band extension (4.5 hours AEL) proceeds, the majority of the te/5.1AMHN-90 transgenic lines obtained exhibit low levels of misexpression in PS2 followed by misexpression in PS4 (Figure 2-8). Later, misexpression of LacZ is detected simultaneously in PSI, 3 and 5 (5 of 6 lines tested). The remaining line showed no misexpression at any time during embryogenesis. By germ band retraction (9 hours A E L ) expression in the 5 of 6 lines was detected equally in all anterior parasegments of the embryo, but the expression pattern within PS6-13 was unaffected. Misexpression from this transgene is only detectable in the ectoderm. The level of LacZ detected in the anterior parasegments is considerably less than the expression of LacZ in the posterior parasegments, showing that this deletion does not completely abrogate PRE activity. The metameric misexpression pattern seen in foc</5.1AMHN-90 is unique and more complex than any other single site deletion mutant, indicating that the repressive complexes formed at this site are functionally different from the complexes formed at the other binding sites. Deletion of MITN-90 in the context of bxd5.l, had no effect on pairing-sensitive repression of the mini-white gene (data not shown). i i i . W5.1AS1HB-90: S1HB-90 is the only identified PH binding site within bxd5.l not located within the M -element. Initial expression of LacZ in the transgenic lines containing forri5.1ASlHB-90 is identical to that of bxd5.l. In 5 of 7 lines obtained, weak misexpression of LacZ beyond PS6 is detected in germ band extended embryos in PS 1-5 (4 hours AEL) (Figure 2-9). The intensity of misexpression increases as germ band extension proceeds and persists throughout embryogenesis. The remaining two lines maintained the silent state until 12 hours of embryogenesis. 52 Figure 2-8. Metameric misexpression pattern of bxd5A AMFfN-90. (A) Schematic representation of the fad5.1 AMFfN-90 Ubx/lacZ reporter construct. (B) The anterior boundary of parasegment 6 (PS6) is indicated with an arrowhead. At early germ band extension misexpression beyond PS6 is first detected in the even-numbered PS2 and 4 (indicated by dots). (C) As germ band extension proceeds, misexpression is detected in all anterior parasegments. Figure 2-9. Deletion of sites MHN-90 and S1HB-90 disrupt bxd5.1 PRE-mediated silencing in embryos. A l l embryos are at the germ band extension stage. An arrowhead points to the anterior boundary of PS6. (A) Wildtype bxd5.\ PRE expression pattern. (B) Deletion of site MHN-90, in the context of bxd5.\, abrogates PRE activity. Expression of LacZ is detected in PS 1-5. (C) Deletion of S1HB-90 results in a misexpression pattern similar to that seem in Z>xJ5.1AMHN-90 embryos. (D) Deletion of both MHN-90 and S1HB-90 results in a more severe misexpression phenotype than either mutant indicating that these two sites act synergistically to promote PRE activity. 54 AMHN-90+ AS1HB-90 R st i — ( ) O AMHN-90 MPA-168MHS-70 AS1HB-90 As expected, deletion of S1HB-90, in the context of bxd5.1, did not affect pairing-sensitive repression of the mini-white gene (data not shown). C. Do the MHS-707MPA-168 and MHN-90/S1HB-90 fragments function independently? Competition analyses suggest that sites MHS-70 and MPA-168 bound related complexes, as did sites MHN-90 and S1HB-90. These data raise the possibility that interaction between these pairs of sites is crucial for PRE activity. If interactions between the competing sites exist, the interactions could be either synergistic or additive. Analyzing and comparing the misexpression phenotypes of the double deletion mutants to the single deletion mutants tested these hypotheses. Interactions were classified as synergistic when the total phenotypic effect of the double mutant was greater or different than the sum of the single mutant phenotype. Interactions were classified as additive when the total phenotypic effect of the double mutant was the sum or superimposition of the single deletion phenotypes. i . bxd5.1AMHS-70+AMPA-168 Both MHS-70 and MPA-168 sites were deleted in the context of bxd5.l and the mutant phenotypes of the double mutant were compared to the single mutant phenotypes. Expression from this double mutant construct is first detected at late cellular blastoderm stage (2.5 to 3 hours AEL) as a broad domain in the posterior region of the embryo, similar to the single deletion mutants. To ensure unbiased scoring, embryos containing single and double deletion mutations were scored on the same day, using coded slides, so that the identity of the transgene was unknown. In six of six lines tested, the misexpression phenotype of 6x<i5.1AMHS-70+AMPA-168 is indistinguishable from the misexpression phenotypes of either single deletion (compare Figure 2-7A and B to C). Therefore, the bxd5A PRE activity conferred by sites MHS-70 and MPA-168 is neither synergistic nor additive. Single deletion of either MHS-70 or MPA-168 partially abrogates PRE activity, and deletion of both sites has no further effect. Therefore, these results show that the region encompassing MHS-70 and MPA-168 constitutes one functional unit of PRE activity. As expected, transgenic flies carrying this construct continue to exhibit pairing-sensitive repression of the mini-white gene (data not shown). 56 i i . foo/5.1 AS 1HB-90+AMHN-90 MHN-90 and S1HB-90 fragments were deleted in the context of bxd5. \ and the mutant phenotypes were compared to the single mutant phenotypes as described above. Initial expression from this construct is the same pattern as in the bxd5.l control and as the single deletion mutants. By germ band extension (3.5 hours A E L ) expression of LacZ is detected in all the anterior PS, very strongly in PS2-5 and weakly in PS1 (19 of 21 lines tested) (Figure 2-9D). At this and later stages of development, misexpression of LacZ is detected in all anterior PS (1-5) and is limited to the ectoderm. The metameric expression pattern of foa/5.1AMHN-90 is not observed in this double mutant. The overall expression pattern within PS6-13 is unaffected by this deletion. Blind studies consistently showed that the misexpression phenotype of fotci5.1ASlHB-90+AMHN-90 is dramatically more severe than that of either single deletion (compare Figures 2-9B and C to D). Therefore, sites S1HB-90 and MHN-90 act synergistically to promote Ubx silencing. Transgenic flies carrying this construct continue to exhibit pairing-sensitive repression of the mini-white gene (data not shown). D. What is the role of unique sequences in the MHS-70 and MHN-90 fragments? Because deletion of MHS-70 and MHN-90 result in a detectable loss of PRE activity, it should be possible to directly investigate the role of the unique sequences within MHS-70 and MHN-90 by linker substitution analysis. A gel mobility shift assay using substitution and deletion mutants of MHS-70 was used to determine the MHS-70 sequence elements required for PH complex formation (Figure 2-4) (Hodgson et al., 2001). The LSI/9 mutation, which affects the proximal and distal d(GA)3 repeats of MHS-70, completely abolished PH complex formation on this site in vitro. The in vivo contribution of these proximal and distal d(GA)3 repeats within MHS-70 was tested in the context of focci5.1. Embryos containing MHS-70-LS1/9 in the context of bxd5.\ were generated. Initiation of LacZ expression in these transgenic flies was similar to that observed in wildtype bxd5.\ embryos. However, by early germ band extension (4 to 4.5 hours A E L ) this mutant transgene did not faithfully silence LacZ expression in the anterior PS. In four of four lines tested, spotty misexpression was observed in PS 1 to 5 in a small number of cells in the ventral ectoderm in germ band extension (Figure 2-10). Unlike the single site deletion mutant phenotypes, the misexpression phenotype of bxd5.\ MHS-70-LS1/9 was more evident in lateral than medial sections. This misexpression phenotype is also less severe and distinct from the fo«/5.1AMHS-57 a ( O A ) 3 d ( O A > 3 Figure 2-10. Mutations of d(GA) 3 repeat elements within MHS-70 reduce bxd5.\ PPvE activity. (A) Schematic representation of the bxd5 AUbx/lacZ reporter construct showing the sequence elements within MHS-70. (B) A wildtype bxd5.1 Ubx/lacZ germ band extended embryo exhibiting spotty misexpression in PSI to 5 and in head region. The anterior boundary of PS6 (indicated with an arrowhead) is maintained. (C) An embryo mutant for the two terminal d(GA) 3 repeat elements within MHS-70 (LSI/9) in the context of bxd5A. Misexpression is detected in a distinct subset of cells in all anterior parasegments (indicated by dots). 70 misexpression phenotype and was not observed in any bxd5.\ control lines, including those which show spotty misexpression in the anterior parasegments. These results indicate a requirement for the d(GA)3 repeats in PRE function in vivo and imply that other sequences within MHS-70 contribute to silencing. As expected, mutation of the proximal and distal d(GA)3 repeats of MHS-70 in the context of bxd5.\ did not affect pairing-sensitive repression of the mim-white gene (data not shown). The OCT-T linker scanning mutant, mutation of all four OCT sequences (see Figure 2-3), completely abolished the formation of PH complexes on MHN-90 (Hodgson et al., in preparation). The in vivo contribution of these sequences within MHN-90 was tested in the context of bxdS. 1. Three lines were generated and tested. Misexpression beyond PS6 was not observed in any of these lines. In fact, no differences were observed between these three transgenic lines and the bxd5.\ control lines (data not shown). These results indicate that the OCT-T sequences within MHN-90 are dispensable for bxd5.\ PRE function and that other sequences are required for silencing. E . Are G A F and PSQ needed for PcG-mediate silencing? The biochemical and transgenic data show that the proximal and distal d(GA) 3 repeats located within MHS-70 are required for the in vivo function of the bxd5.\ PRE. The question that arises is: are the d(GA) 3 repeat binding proteins, G A F and Psq, required for the wildtype function of the bxd5.\ PRE? Trl and psq mutants were obtained and genetic tests were performed to determine whether the products encoded by these genes function in bxd5.\ PRE-mediated silencing. Two types of genetic tests were performed to confirm the involvement of GAF and Psq in bxd5.\ PRE-mediated silencing. The first test was identical to the assay that ascertained the dependence of the bxd5.\ PRE on the PcG. Embryos transheterozygous for the bxd5.l transgene and the mutant allele being tested were stained for LacZ expression. Misexpression of LacZ in PS 1-5 confirms that the wildtype product encoded by the gene being tested is required to maintain the silent state of the Ubx/lacZ transgene in the anterior PS. I examined the effect of trl and psq null or hypomorphic alleles on the maintenance of silencing by the bxd5.\ PRE (see Table 5-1 for description). Embryos transheterozygous for the trl alleles, trlnc, trf2 or trf5, and the bxd5.l transgene did not result in misexpression of the bxd5.l Ubx/lacZ transgene (data not shown). The LacZ expression pattern in these embryos was indistinguishable from the bxd5.\ 59 control embryos at all developmental stages. Similarly, embryos transheterozygous for the psq null alleles, psq™13 andpsq lola, and the bxd5.l transgene did not exhibit misexpression. These embryos were also indistinguishable from the bxd5.l control embryos at all developmental stages (data not shown). The second genetic test used to confirm the participation of G A F and Psq in PcG-mediated silencing takes advantage of a fundamental feature of PcG biology. Genetic interactions between PcG genes are monitored by the enhancement of PcG mutations. By extension, monitoring genetic interactions between PcG genes and putative genes involved in silencing provides a sensitive genetic assay to identify genes required or involved in PcG-mediated silencing. The ability of trl and psq mutations to enhance the extra sex combs phenotype of ph was monitored. Wildtype Drosophila males possess a comb-like structure, called a sex comb, made up of 10-12 heavy, curved, black bristles arranged in a row at the distal end of the first tarsomere of each prothoracic leg (2 sex combs per male). PcG mutants often transform second and third legs into the first leg, thus increasing the number of sex combs per male fly (Figure 2-11). Trf1 dramatically enhanced the extra sex combs phenotype of ph1 and phm alleles (Table 2-1). Trl mutations also enhance the extra sex combs phenotype of the Pc mutations (Strutt and Paro, 1997). Similarly, the psq alleles psq2A03', psq™11', psq0U5 and psqlola strongly enhanced the extra sex combs phenotype of ph2 and ph409 (Table 2-1). These results show that the d(GA) n repeat binding proteins, GAF and Psq, play a role in PcG-mediated silencing and suggest that these proteins may recruit PcG proteins to the bxd PRE. III. Discussion: A. Role of P H binding sites in bxd5.l P R E activity The gel mobility shift assay, used in conjunction with the antibody supershift and competition assays, has identified four PH binding sites within the bxd5.l PRE. The results obtained from the germline transformation assay, summarized in Table 2-2, show that all four PH binding sites identified in vitro are required for the in vivo function of the bxd5.l PRE. Deletion of S1HB-90, MHS-70, MPA-168 or MHN-90, in the context of bxd5.l, results in the partial loss of maintenance of silencing but does not affect pairing-sensitive repression of the mini-white gene. Because the single deletion mutants only partially reduce repression by the PRE, this suggests that multiple PH binding sites contribute to PRE function. Competition 60 Figure 2-11. Trl and psq mutations enhance the extra sex combs phenotype of ph and Pc mutations. Wildtype Drosophila males have two sex combs (marked by arrowheads): one on each first leg (prothoracic leg) (A) and absent on the second (mesothoracic) (B) and third (metathoracic) (C) legs. Males hemizygous for ph mutations exhibit posterior homeotic transformations where the second and third legs are frequently transformed into the first, thus increasing the total number of sex combs per fly. Male flies transheterozygous for trl or psq and ph mutations exhibit an enhancement of the extra sex combs phenotype. In these flies, sex combs are always seen in the first legs (D), frequently on the second legs (E) and sometimes on the third legs (F). Table 2-1. Trl and Psq enhance the homeotic phenotypes of polyhomeotic and Polycomb mutations. Cross Genotype Number of flies scored * Average number of legs with sex combs" +/Y; psqMa/CyO X ph409/FM7C; +/+ ph409/Y; CyO/+ 75 3.86 ph409/Y;psqMa/+ 55 5.72 +/Y; psqR™/CyO X phmIFM7C\ +/+ ph409/Y; CyO/+ 64 4.20 ph409/Y;psqRFU/+ 31 5.89 +/Y; psq0U5/CyO X ph409/FM7C; +/+ ph409/Y; CyO/+ 36 3.60 phm/Y;psq0U5/+ 40 5.70 +/F; psqiola/CyO X ph2/ph2; +/+ ph2/Y; CyO/+ 53 2.46 ph2/Y; psqlola/+ 44 4.02 +/T; psJ*l3/CyO X ph2lph2; +/+ ph2IY; CyO/+ 74 2.55 ph2/Y; psq™ul+ 70 5.25 +/F; psq0U5/CyO X ph2/ph2; +/+ ph2IY; CyO/+ 51 2.36 ph2/Y;psq0U5/+ 50 5.14 +/F; psq2m/CyO X ph2/ph2; +/+ ph2/Y; CyO/+ 84 2.48 ph2/Y;psq2403/+ 68 5.03 psqMa/CyO X Pc4/TM6B +/Y; CyO/+; Pc4/+ 48 2.23 +/Y; psqMa/+; Pc4l+ 50 4.04 psq2m/CyO X Pc4/TM6B +/Y; CyOI+\ Pcl+ 58 3.77 +/Y;psq2403/+;Pc4/+ 58 5.22 +IY;trf2ITM3 X ph409/FM7C;+/+ ph409/Y; TM3/+ 46 4.51 ph409/Y; trf2l+ 55 5.43 +/Y; trf2ITM3 X ph2lph2; +/+ ph2/Y; TM3/+ 89 2.15 ph2/Y; trf2l+ 76 4.33 a A chi-squared test confirmed that the differences observed between the control PcG mutants and the transheterozygotes are statistically significant at P<0.05. 62 Table 2-2. Summary of bxd5.l PRE deletion analysis. Construct Total number of independent transformed lines Parasegment 6 restriction at germ band extension Misexpression in anterior parasegments at germ band extension Capable of pairing-sensitive repression bxd5.1 10 10 0 Yes 6x</5.1AMHS-70 6 1 5 Yes fctrf5.1AMPA-168 6 2 4 Yes fco/5.1AMHN-90 6 1 5 a Yes Z>x</5.1ASlHB-90 7 2 5 Yes 6JC</5.1AMHS-70+ AMPA-168 6 0 6 Yes W5.1AMHN-90+ 21 2 19 Yes AS1HB-90 aThe metameric misexpression pattern of these lines was unique. analyses suggest that sites MHS-70 and MPA-168 bound related complexes, as did sites M H N -90 and S1HB-90. Tests of functional interaction using double mutant constructs revealed that the region encompassing sites MHS-70 and MPA-168 constitutes one functional unit of PRE activity and that sites S1HB-90 and MHN-90 interact synergistically to promote PRE function. The bxdSA PRE is a subfragment of a much larger regulatory region whose general purpose is to express Ubx in the ectoderm and to maintain an anterior boundary of expression at PS6. Because bxd5A is an ectodermal-specific element it is not surprising that misexpression beyond PS6 caused by a deletion or a mutation within this element is confined to the ectoderm. Functional repressive elements within bxd5A silence the ectodermal enhancers in PS 1-5 and non-silenced ectodermal enhancers direct ectodermal misexpression of LacZ in PS 1-5. In ph mutants, the onset of derepression of Ubx expression occurs in early germ band extension (approximately 3.5 hour AEL) . Similarly, misexpression of bxd5A in ph mutants is also detected at this time. Interestingly, misexpression of LacZ in each of the single deletion mutants coincides with misexpression of bxd5A in ph mutants. These findings suggest that an intact element and a wildtype dose of PH are required for PRE activity and that loss of either leads to identical consequences, misexpression in anterior parasegments. Alternatively, the misexpression phenotype detected in ph mutants and the misexpression phenotype caused by the PH binding site deletions are the same because binding of PH to the PRE is the root cause of the phenotype. One of the drawbacks of the deletion analysis I performed on the bxd5A PRE is that the cause of the misexpression phenotype could be due to one of two things. First, deletion of a particular PH binding site removes a bone fide PRE sequence that is absolutely required for PRE function. Second, deletion of a particular PH binding site may lead to misexpression because the deletion alters the requisite spacing between essential binding sites within this complex. It is reassuring that the fce<i5.1MHS70-LSl/9 transgene exhibits misexpression as this construct preserves the normal spacing. Thus, for MHS-70, the gel-shift assay has identified sequences necessary for PcG complex formation. B. Initiation of Silencing The double deletion of S1HB-90 and MHN-90 resulted in misexpression at an earlier stage than that observed in any other transgene tested (soon after gastrulation, 3 hours AEL) indicating that these sites are required to initiate the silent state of the Ubx/lacZ reporter. There is incontrovertible evidence that the mechanisms involved in the initiation or establishment of 64 homeotic gene silencing are different from those involved in maintenance of silencing. Segmentation gene products are required for the initiation of silencing whereas the PcG is solely required for maintenance of the silent state (Struhl and Akam, 1985; White and Wilcox, 1985; McKeon and Brock, 1991; Simon et al., 1992; Soto et al., 1995). The exception to this is ESC, which is unique amongst the PcG as it is required only in the first few hours of development implicating a role for its involvement in initiating the silent state of the homeotic genes (Struhl, 1991; Simon et al., 1995). Unlike ESC, an early requirement for PH in the initiation of silencing has yet to be demonstrated. However, it has been recently shown that in early embryo extracts (0-3 hours), PH transiently associates with ESC in a complex with other PcG proteins (Poux et al., 2001b). This result is consistent with the finding that two PH-containing complexes were detected in the 0-3 hour embryo extracts and only on sites S1HB-90 and MHN-90 (Hodgson and Brock, unpublished). Furthermore, the metameric misexpression pattern seen in bxd5.\AMHN-90 suggests an intimate relationship between the early segmentation genes and PH in silencing. Taken together, these experiments provide strong evidence for a role of PH in the initiation of silencing and implicate sites S1HB-90 and MHN-90 as the targets for this activity. The presence of ESC and segmentation gene products in the S1HB-90 and MHN-90 binding activities would strengthen this hypothesis. The role of sites S1HB-90 and MHN-90 in the initiation phase of silencing is challenged by the observation that neither single site deletion exhibits defects in the initiation of silencing. This is a surprising result as site S1HB-90 is located within the SI enhancer, which is one of many enhancer elements that determine the initial expression pattern of Ubx in the embryo. One possible model to explain this is that the complexes formed on sites S1HB-90 and MHN-90 functionally interact (discussed below) and this interaction is necessary for the initial stages of silencing. Deletion of either site, in the context of bxdS.l, does not perturb the initiation of silencing as a portion of the 'initiation' complexes are retained on the other site present. However, deletion of either site does abrogate the element's ability to maintain the silent state. These results imply that sites S1HB-90 and MHN-90 are recognized by at least two different sets of PH-containing complexes: those involved in the initiation of silencing and those involved in maintenance. This is consistent with the biochemical data that shows dynamic developmental changes in the composition of the PH-containing complexes formed at these sites (Hodgson and Brock, unpublished). I propose that the PH-containing complexes formed at these sites from 0-3 hours are 'initiation' complexes, and are required to promote the formation or recruit the 'maintenance' complexes that are present from 3-18 hours (Figure 2-12). The biochemical 65 A. Initiation of Silencing (0-3 hours) MHN-90 MPA-168 MHS-70 S1HB-90 I Transition to Maintenance MHN-90 MPA-168 MHS-70 S1HB-90 B. Maintenance of Silencing (3-18 hours) trxG Figure 2-12. Model of bxd5.l PRE-mediated silencing. (A) Early in embryogenesis, sites MHN-90 and S1HB-90 are bound by PH-containing complexes in the anterior portion of the embryo. These complexes may recruit the PcG maintenance complexes that form on sites MHS-70 and MPA-168. (B) Interactions between the different PH-containing complexes promotes D N A looping and occludes the transcriptional activators. purification of these D N A binding activities and the identification of the constituents of these complexes should provide insights into the mechanisms involved in the initiation and maintenance phases of silencing and of the transition between them. Although the presence of HB binding sites within MHS-70 indicates that this sequence may also be involved in the initiation phase of Ubx silencing, the biochemical evidence argues against this, as PH-containing complexes formed at this site are first detected in 3-6 hour embryo extracts. This coincides with the maintenance period and precedes the detection of a HB-containing activity (Hodgson and Brock, unpublished). The data obtained from the deletion analysis also argue against a role of MHS-70 in the initiation of silencing. The bxd5.\ AMHS-70 construct is able to initiate but is unable to maintain silencing indicating that MHS-70 is recognized by maintenance complexes. C. bxdS.l PRE Design Why did the loss of sites MHS-70 and MPA-168 show the same extent of LacZ misexpression as that of either single deletion? The most obvious explanation for this observation is that the region encompassing MHS-70 and MPA-168 is required for the assembly of a distinct protein assembly whose structure or stability is altered by any change to this functional element. Furthermore, i f this functional element interacts with other parts of the PRE, then these interactions would not be further compromised by the double deletion as compared to either single deletion. Alternatively, deletion of both or either of these sites results in changes in D N A topology that suppress the silencing activity of the PRE activity. How can the synergistic interaction observed between sites S1HB-90 and MHN-90 be explained? One possible model is that the assembly of the initiation complexes on these sites alters the element's chromatin configuration such that it promotes the formation or recruits maintenance complexes. Thus, loss of these sites results in the element adopting a chromatin configuration reminiscent of a TRE, the transcriptionally active state that is incompatible with the assembly of repressive complexes that would otherwise form on these sites. The presence of either S1HB-90 or MHN-90 in the transgene is sufficient to convert the element into a PRE albeit weaker than the intact element, hence the misexpression seen during germ band extension. An alternative model is that the initiation complexes formed on sites S1HB-90 and MHN-90 physically interact, via D N A looping, and this interaction is required for the recruitment of maintenance complexes on these sites and for the subsequent recruitment or formation of maintenance complexes on sites MHS-70 and MPA-168 (Figure 2-12). The structural 67 hypothesis as sites MHS-70 and MPA-168, located 80bp apart, are approximately equidistant from sites S1HB-90 and MHN-90. The ability of PH to self-associate through its S A M domain also strengthens this hypothesis (Kyba and Brock, 1998a; Kim et al, 2002). According to this model, loss of both S1HB-90 and MHN-90 results in the complete loss of PH-containing complexes formed on the bxd5A PRE and causes a very severe misexpression phenotype. In contrast, loss of either S1HB-90 or MHN-90 results in a less severe phenotype as a fraction of the initiation complexes are formed and are sufficient to recruit a subset of maintenance complexes but not all. I favour the latter model as it incorporates the biochemical competition analysis, the transgenic analysis and the structural organization of the 6xd5.1PRE. D. Embryonic silencing can be uncoupled from pairing-sensitive repression PHO consensus binding sites are found in many, but not all, PREs and it has been proposed that PHO is a major recruiter for PcG complexes (Brown et al, 1998). Furthermore, PHO is essential for PRE-dependent pairing-sensitive repression of the mini-white reporter (Tillib et al, 1999; Shimmel et al, 2000). Site MPA-168 is the sole PH binding site within bxdSA that contains PHO consensus binding sites; three PHO binding sites are located at its distal end. Deletion of MPA-168, hence deletion of the three PHO binding sites, did not affect pairing-sensitive repression of mini-white. This is not a surprising result as three PHO binding sites are located within the PSR fragment of bxd5A (immediately upstream of MPA-168) and are thus sufficient for pairing-sensitive repression of mini-white. However, deletion of MPA-168 did partially abrogate bxd5.\ PRE activity in embryos, suggesting that these two PRE activities may be separable. It is unlikely that the loss of PRE activity in the fac<i5.1AMPA-168 construct is due to the loss of the PHO binding sites. A l l of the identified PH binding sites were required for embryonic silencing yet only one of four sites contain PHO binding sites, suggesting that PHO is not the sole recruiter of PcG complexes. This is corroborated by the finding that when targeted to a reporter gene in embryos, PHO is neither able to silence the reporter nor interact with PcG-containing complexes (Poux et al, 2001a). Additional proteins may recognize the MPA-168 fragment outside of the PHO binding sites, perhaps the same activity recognizing MHS-70. The presence of PHO binding sites as a recurring sequence motif in PREs is consistent with PHO action through these sites, but evidence that PHO contributes to the embryonic silencing function of PREs remains elusive. PHO binding sites within an element from the engrailed locus (Kassis, 1994) and from the iab-2 regulatory region of the abd-A gene (Shimmel et al, 2000) are required for pairing-sensitive repression of the mini-white gene but are not required for embryonic silencing of a reporter gene. More recently, it has been shown that the PHO binding sites within the Mcp PRE of the abd-B gene are necessary to maintain the silent state of a Ubx/lacZ reporter gene during imaginal discs development (Busturia et al, 2001). Since PRE-mediated pairing-sensitive repression of mini-white is a larval phenomenon, presumably occurring in the larval eye imaginal discs, it is likely that the mechanisms and PRE sequences involved in PcG silencing in embryos differ from those involved in silencing in larval imaginal discs. Furthermore, each of these mechanisms may have a differential requirement for PHO. It would be interesting to see the effects of mutating the three PHO binding sites within MPA-168 on bxd5.\ PRE activity. E. Role of GAF and Psq in ZucJ-mediated silencing The in vitro experiments show that the d(GA) 3 repeats located within MHS-70 are required for PH complex assembly and that the d(GA)3 repeat binding proteins, GAF and Psq, coelute with this PH complex. The in vivo experiments demonstrate a direct role for the d(GA) 3 repeats in maintenance of silencing by the bxdS.X PRE. The genetic crosses demonstrate a role for GAF and Psq in PcG-mediated silencing. These results are confounded by the fact that the wildtype bxd5.\ transgenes were unresponsiveness to heterozygous trl and psq mutations though this could be explained by the inherent limitations of this type of genetic assay. First, because the flies are heterozygous for the mutant allele they may be producing sufficient wildtype product for normal bxd5.\ regulation. Second, maternally deposited product may compensate for the loss of zygotic product. The biochemical and genetic data presented here corroborate those of Horard et al. (2000) who show that the ability of PcG complexes in nuclear extracts to bind to select bxd PRE fragments is mediated by G A F and is abolished by mutations in the G A G A consensus sequences. The involvement of G A F in PRE-mediated silencing is not intrinsic to all PREs. The majority of GAF binding sites on polytene chromosomes do not overlap with PcG binding, arguing strongly that only a small fraction of GAF protein is involved in PRE-mediated silencing (Poux et al, 2001). Furthermore, trl mutations abrogate the silencing of some PREs and have no effect on others (Hagstrom et al., 1997; Hodgson et al, 2001). A plausible model for the role of GAF in fctt/-mediated silencing is that this protein binds the d(GA) n repeats and recruits PcG proteins to the MHS-70 and MPA-168 fragments facilitating the assembly of a functional repressive complex. Alternatively, G A F could first recruit SIN3 histone deacetylation 69 complexes through its interaction with SAP 18 (Espinas et al, 2000), which then might generate a chromatin state favorable for PcG complex binding. These models, however, do not account for the fact that G A F is a known activator both in vitro and in vivo. In addition, targeting GAF to a reporter gene via a fusion with the LexA D N A binding domain does not recruit PcG complexes to the reporter and thus does not silence the reporter (Poux et al, 2001a). One explanation for these results and for the dual role of G A F as an activator and repressor is that the function of GAF is dependent on the chromatin context of where it is bound. The chromatin context converts the activator into a repressor, or vice versa, by the adjacent binding of other factors. Examples of such scenarios are widespread. In Drosophila, Dorsal is a transcriptional activator but is converted to a repressor by the proximity of other D N A binding factors and by the interaction with Groucho, a protein that interacts with the histone deacetylase Rpd3 (Dubnicoff et al, 1997). The mammalian transcription factor Y Y 1 , which is the homologue of the Drosophila pho gene, is involved in repressing and activating a diverse number of promoters (Thomas and Seto, 1999). Depending upon chromatin context, Y Y 1 interacts with a number of key regulatory proteins (e.g. TBP, TFIIB, TAFII55, Spl , and E l A) and these interactions are important for determining which particular function of Y Y 1 is displayed at a specific promoter. In this particular case, GAF binding, hence GAF recruitment of a PcG complex, may follow the recruitment of an initial silencing complex that has predisposed the element for silencing. Alternatively, G A F or GAF-PcG complexes could be recruited to the element by other PcG complexes. The combined action of these multiple complexes may stabilize PRE activity. The mutation analysis of the d(GA)3 binding sites within MHS-70 is consistent with such a hypothesis. Mutation of these elements did not cause as severe a loss of PRE activity as did deletion of the entire element, arguing that other sequences contribute to the function of the MHS-70 element. In light of these results and those previously published, I propose that trl (gene encoding GAF) be reclassified as belonging to the ETP group of genes (Enhancer of trithorax and Polycomb) that act both in activation and silencing of target loci (Gildea et al, 2000). Pipsqueak (Psq) belongs to a family of proteins defined by the BTB (POZ) domain that has been implicated in protein-protein interactions (Horowitz and Berg, 1996). Like other members of this family, Psq is an important developmental regulator in Drosophila, having pleiotropic functions during oogenesis, embryonic pattern formation (Siegel et al, 1993), and adult development (Weber et al, 1995). The role of Psq in homeotic gene regulation or in PcG-70 mediated silencing has never been documented prior to our investigation. Like GAF, Psq may be involved in recruiting PcG complexes to the PRE and may have dual function depending on chromatin context. It would be interesting to determine i f Psq is a member of other PcG complexes and, i f so, does it directly interact with any PcG proteins. F. Role of POU homeodomain proteins in fexrf-mediated silencing The POU (Pit-1, Oct-1, unc-86) family of transcription factors are involved in the transcriptional regulation of a variety of genes related to cell cycle regulation, development, and hormonal signals (Phillips and Luisi, 2000). It has been shown that Oct-1 acts not only as a transcriptional activator but also as a transcriptional repressor for certain genes. The mechanism of the repressive function of Oct-1 is not well understood. However, the POU domain of Oct-1 directly interacts with a silencing mediator for retinoid and thyroid hormone receptors (SMRT) (Kakizawa et al, 2001). The POU genes of Drosophila, pdm-1 and pdm-2, are expressed at high levels during early embryogenesis and at lower levels throughout the rest of development. Tests of functional interaction revealed that the ability of the bxd5.\ PRE to silence the Ubx/lacZ reporter was not affected in pdm-1 and pdm-2 mutant embryos (data not shown). However, like trl and psq mutations, certain pdm-1 and pdm-2 alleles enhanced the sex combs phenotype of ph, implicating these genes in PcG-mediated silencing (Hodgson and Brock, unpublished data). In all these cases it is believed that the maternal contribution suffices to rescue the zygotic loss of these gene products. Perhaps the time has come to use R N A interference to inactivate both the maternal and zygotic transcripts and look for changes in focci-mediated silencing. 71 Chapter 3 Modular structure of the bithoraxoid 1.5 Polycomb Response element I. Introduction Since the initial discovery of eukaryotic enhancers 21 years ago (Banerji et al, 1981), several additional classes of cis-regulatory sequences have been identified: transcriptional silencer elements (Brand et al., 1985), locus control regions (Dillon and Grosveld, 1991) and chromatin boundary elements (Kellum and Schedl, 1991). Silencer elements were first described in yeast and implicated in the regulation of the silent mating type loci. Since then, silencer elements have been identified and characterized in many organisms and found to affect many processes. The Polycomb response elements (PREs), first identified by Simon et al. (1990), are required to maintain the transcriptionally silent state of the homeotic genes in a number of higher eukaryotes, and are the focus of this thesis. PREs are thought to function like the yeast silencers to propagate long-range changes in chromatin structure at target loci (reviewed in Pirrotta, 1997; 1998). This chapter deals with the functional dissection of the bithoraxoid 1.5kb Polycomb response element, bxdl.5, one of the best-characterized PREs in Drosophila. A. Properties of the bxdl.5 PRE During embryogenesis, the Ubx promoter is regulated by a number of parasegmental enhancers. These parasegmental enhancers are targets of the early segmentation genes and are either activated or repressed to establish the correct Ubx expression domain. Later in development, as the products of the segmentation genes are only transiently available, the off or on state of Ubx transcription is maintained through subsequent cell divisions by the PcG and trxG proteins, respectively. The bxd\.5 PRE is the target of these proteins and is the minimal cis-regulatory maintenance element responsible for ensuring that Ubx remains off or on in the appropriate parasegments. In transgenic analyses, the bxd\.5 PRE can simultaneously silence the flanking SI and S2 Ubx parasegmental enhancers during embryogenesis (Chan et al, 1994; Hodgson et al, 2001; Chapter 2). In the absence of the bxdl.5 PRE, these two enhancer elements initially drive the expression of a reporter in the appropriate parasegments. Later in development, when the initial repressors are absent, the SI and S2 enhancer elements drive expression of a reporter gene 72 beyond their prescribed boundaries (Chan et al, 1994; Horard et al., 2000). When the bxd\.5 PRE is included with these enhancers in the same transposon, the parasegmental enhancers are silenced in the appropriate parasegments and remain active in the Ubx domain for the remainder of embryogenesis. The Ubx expression pattern established in the embryo is remembered and propagated into the larval stages by the bxdl.5 PRE thereby establishing the correct parasegment-specific expression pattern of the Ubx imaginal disc enhancers (Chan et al., 1994; Fritsch et al, 1999). The ability to maintain the silent state is dependent on PcG function while the ability to maintain the active state is dependent on trxG function. A n important question that arises from these observations is whether the bxdl.5 PRE controls transcription by inactivating the function of the enhancers, or does it directly interfere with basal promoter activity? Insight into the mechanism of bxdl.5 PRE function comes from two key observations. First, transgenic analyses using the pCaSpeR Ubx-lacZ transformation vector revealed that, when tested alone, the bxdl.5 PRE did not express LacZ (Chan et al, 1994). The lack of expression was explained by stating that this element does not contain enhancer elements. An alternate explanation, which will be tested in this chapter, is that the bxdl.5 PRE directly represses the Ubx promoter found in this construct. Second, constructs containing the bxdl.5 PRE can repress the mini-white reporter gene present in the same transposon resulting in its variegated expression pattern in the adult eye (Chan et al, 1994). This repression is enhanced when the flies are homozygous for the construct, a phenomenon known as pairing-sensitive repression. Pairing-sensitive repression of the mini-white gene indicates that the repressive effects of the bxdl.5 PRE are twofold. First, the bxdl.5 PRE can act on heterologous promoters and not just the Ubx promoter. Second, it can act long-range to simultaneously repress the Ubx promoter and the mini-white promoter contained in the same transposon. B . Structural organization of the bxdl.5 P R E The bxdl.5 PRE has been subdivided into three regions: a central 661bp region known as the pairing-sensitive region (PSR) (Sigrist and Pirrotta, 1997), flanked upstream by a 412bp fragment termed UPS (Upstream of pairing-sensitive region) and downstream by a 489bp fragment termed DPS (Downstream of pairing-sensitive region) (Hodgson et al, 2001) (Figure 3-1). The subdivision of the bxdl.5 PRE into the UPS, PSR and DPS subfragments is based on two key findings. First, it was shown via transgenic analysis that the region corresponding to the PSR fragment is the minimal fragment necessary and sufficient for pairing-sensitive repression of the mini-white reporter gene (Sigrist and Pirrotta, 1997). Second, biochemical studies on the bxdl.5 73 Figure 3-1. Summary of the biochemical and genetic analyses of the bxdl.5 PRE. Fragments in green were identified and described in Hodgson et al, 2001, Chapter 2 of this thesis and Hodgson et al., in preparation. Fragments in blue were characterized by Horard et al., 2000. Fragments in red were characterized in Tillib et al., 1999. Fragments in gold were characterized in Fritsch et al., 1999. Asterisks denote G A F binding sites and ovals denote PHO binding sites. See text for more details. Figure drawn to scale. 74 a O oj u y S fin «N >) OH 5? 5 a? 1 5 X u fi-fe X < a. O OH u PH X < a, a cr if) OH ft, O3 u OH U OH ON 7> u OH ^ u Xii OH X OH U OH 75 PRE revealed the presence of 2 restriction fragments within the DPS fragment and 1 within UPS that supported the binding of PH-containing protein complexes (Hodgson et al, 2001; Hodgson et al, in preparation). No PH-containing complexes formed on restriction fragments from the PSR subfragment. The importance of this biochemical finding on the role of the UPS and DPS subfragments to bxdl .5 PRE function was underscored by the findings that deletions of any of the PH-binding sites, in the context of bxd5.l fragment, disrupted the PRE's ability to maintain embryonic silencing (see Chapter 2). Surprisingly, deletion of the PH-binding sites did not effect pairing-sensitive repression of the mini-white gene suggesting that at the bxdl.5 PRE, embryonic silencing and pairing-sensitive repression may be separable functions and conferred by different sequences. Deletion analysis of the PH binding sites within the bxdl .5 PRE revealed that the UPS and DPS subfragments are not functionally equivalent (Chapter 2). Although deletions within UPS and DPS abrogated PRE activity, the misexpression phenotypes were not similar. Deletion of MHN-90 located within UPS showed a metameric misexpression pattern that was clearly distinct from the misexpression pattern of the DPS deletions. Thus, the picture emerging from the transgenic and biochemical studies is that the bxdl.5 PRE is comprised of several regions or modules defined by UPS, PSR and DPS that make distinct contributions to the overall function of this element. Indeed, three independent laboratories have provided further experimental evidence for this assertion. In the proceeding sections I will review and discuss the recent advances made in understanding the structure and function of the bxdl.5 PRE (summarized in Figure 3-1). i. PREs and TREs within the PSR fragment of bxdl.5 consist of closely situated but separable sequences The early findings that PcG and trxG proteins colocalize at some PREs suggested that PcG and trxG proteins act as molecular switches through common chromosomal elements, to either direct the adjacent chromatin into an active or inactive state (Chang et al, 1995; Strutt et al, 1997). In order to investigate this hypothesis, Tillib et al. (1999) conducted an impressive analysis of the bxdl.5 PRE. In vitro immunoprecipitation and polytene localization experiments revealed that T R X binds to a fragment, termed fragment C, that is located entirely within the PSR subfragment of bxdl.5 PRE (Tillib et al, 1999). This discovery led to the hypothesis that within bxdl.5, PREs and Trithorax response elements (TREs) are intermingled. To test this idea, the authors conducted a germline transformation assay similar to that performed in Chapter 2 of this thesis. Deletion of fragment C (AC), in the context of a 13kb construct containing the entire 76 bxd regulatory region fused to a Ubx/lacZ reporter, resulted in misexpression of lacZ mRNA in PS 1 to 5 during germ band extension (4.5 hours A E L ) and resulted in reduced levels of expression of lacZ mRNA in PS6-13 compared to the wildtype construct (Tillib et al., 1999). These results indicate that fragment C, in addition to being recognized by TRX, contains PRE activity and is required for silencing. Furthermore, the decrease in expression in PS6-13 is reminiscent of decreased Ubx expression observed in embryos homozygous for trx null alleles (Breen and Harte, 1993; Sedkov et al., 1994) suggesting that AC gives an expression pattern similar to that given when trx function is removed. Further reduction of expression is not observed in AC transgenic lines in a trx mutant background suggesting that fragment C contains stimulating activity and is an essential TRE. The finding that PREs and TREs are closely situated suggested that these elements may be the same or overlapping D N A sequences. To test this hypothesis, the authors conducted a similar germ line transformation assay and determined that fragment C can be subdivided into 3 subfragments. Subfragment C l contained a functional TRE, while subfragments C2 and C3 were required for PRE activity. The results of this study show that the primary PRE and TRE activities within the C element act independently and require distinct D N A sequences. These findings provide further support for the composite nature of the bxdl.5 PRE. ii. bxdl.5 PRE-mediated silencing in imaginal discs requires the PSR subfragment Tests for silencing function of bxdl.5 subfragments revealed that the PREv subfragment, corresponding to PSR and a small portion of DPS (see Figure 3-1) was capable of completely silencing a Ubx imaginal disc enhancer (IDE) anterior to PS6 in imaginal discs (Fritsch et al., 1999). PREc, corresponding to DPS, was partially capable of silencing, and PRE A, corresponding to UPS, did not silence. The ability of the PREu subfragment to silence the IDE in the imaginal discs was abrogated in Pc and pho mutant backgrounds demonstrating a late requirement for PC and PHO. Indeed, mutation of all six PHO binding sites within PRED abolished imaginal disc silencing indicating that these sites are essential for PRE® activity. The requirement of the PcG on PREc activity was not tested. These bxdl .5 PRE subfragments were unable to silence the P B X embryonic enhancer located in the same transposon. The data presented by these authors provides strong evidence that the bxdl.5 PRE can be subdivided into smaller modules with differing functions. In addition, this study establishes a direct link between PHO and the PSR module and demonstrates that PHO binding is essential for PSR-mediated silencing of the IDE in imaginal discs. 77 iii. The bxdl.5 PRE is a compound element composed of sequences with different PRE-Iike properties that interact in vitro with different PcG complexes To determine the contribution of bxdl.5 subfragments to PRE activity a transgenic analysis was performed (Horard et al, 2000). Subfragments of bxdl.5, ranging from 77 to 306bp (Figure 3-1), were multimerized (3 to 6 times) and subcloned downstream of the bithoraxoid S2 enhancer in the Ubx/lacZ transformation vector. The S2 enhancer does not display PRE activity itself, drives expression of lacZ in all parasegments, and is a native target of 6xJ1.5-mediated silencing. Although all of the fragments were capable of silencing the mini-white reporter in the adult eye assay in a PcG-dependent manner, the BP-191 (6X) fragment of the PSR was the only fragment capable of silencing the S2 enhancer in PS 1-5 in the embryo. Silencing mediated by BP-191 (6X) was lost in embryos homozygous for a Pc mutation confirming the dependence on PC. The ability of these bxdl.5 PRE subfragments to bind PcG proteins was tested by immunoprecipitation experiments using embryonic nuclear extracts (Horard et al, 2000). Fragments BP, HH2 and HS, all which contain d(GA) n sequences, are more efficiently precipitated by anti-PC antibodies, while HH1, H A and A B , which lack d(GA) n sequences, are more efficiently precipitated by anti-PSC antibodies. As expected from the presence of d(GA) n sequences, the BP fragment also formed complexes that were immunoprecipitated by anti-GAF antibodies indicating that G A F is a component of a subset of PcG complexes or may participate in the assembly of PcG complexes at the bxdl.5 PRE. In conclusion, this group was able to show that different fragments of bxdl.5 PRE have differing silencing activities and that PcG silencing may not be equivalent in embryos and larva. Furthermore, the different fragments of bxdl.5 PRE were able to differentially interact in vitro with PcG complexes present in nuclear extracts suggesting the presence of multiple PcG complexes. Taken together with the findings of Chapter 2, the three studies discussed above strongly support the hypothesis that the bxdl .5 PRE is built up of multiple interaction sites with differing functions that recruit different PcG-containing DNA-binding activities and contribute to the overall function of the PRE. In each of the studies discussed above, the roles of the bxdl.5 PRE subfragments were defined by their effect on the activity of Ubx parasegmental enhancers or by their effect on the mini-white reporter gene. This was accomplished by using very large bxd fragments that contain known parasegmental enhancers (Tillib et al, 1999; Chapter 2) or by fusing the element of interest to known embryonic or imaginal disc enhancers (Horard et al, 2000; Fritsch et al, 2000). In addition, creating high-affinity binding sites through 78 multimerization may stimulate binding in a cooperative manner, a phenomenon that may or may not occur on this fragment, resulting in the stabilization of transient or weak binding that normally occurs at this site. These studies have provided support for the model of bxdl .5 PRE function which states that PREs act at long-range to hinder the access of transcriptional activators to enhancer elements or to prevent enhancer-promoter interactions necessary for transcription (see model in Figure 3-2). A n alternative model, but not necessarily mutually exclusive, is that the bxdl.5 PRE may directly interfere with basal Ubx promoter activity. Thus, an important question that arises is: Does the bxdl.5 PRE, or subfragments thereof, directly repress the Ubx promoter? The work presented in this chapter attempts to rectify some of these problems and contribute to the knowledge of how the bxdl.5 PRE works. C. Chromatin Boundary Elements An inherent problem with germline transformation assays is the phenotypic variability observed among different transgenic lines harbouring the same construct. These effects are presumably due to chromosomal position-effects, the influence on transgenic promoters by chromosomal enhancer or silencer elements located in the vicinity of the transgene's site of insertion. As the chromosomal environment interacting with the transgene varies from one insertion site to the next, independently generated lines often display different expression patterns. Thus, the behaviour of transgenes in Drosophila most likely reflects the proximity of enhancers or silencers, a fact that is exploited in enhancer trap screens (O'Kane and Gehring, 1987). One solution to this problem is to place chromatin boundary elements in the transformation vector to shield the transgene from external influences. Enhancers or silencers are capable of exerting their influence over long distances in an orientation-independent manner to orchestrate the complex gene expression patterns required for embryonic development. The effects of enhancers or silencers must be confined to the genes they regulate. Chromatin domain boundaries have been identified and serve to isolate chromosomal regions, confining the regulatory effects of the enhancer or silencer to their respective targets. It is believed that chromatin boundaries organize the chromosomes into a series of discrete and topologically independent higher-order domains. Evidence for this is the specialized chromatin structures (scs and scs') found flanking the coding sequence of the Drosophila hsp70 gene (Kellum and Schedl, 1991). These sequences are associated with specific chromatin structures and serve as boundaries that can prevent activation by enhancer elements. The Drosophila gypsy insulator functions similarly conferring position-independent 79 Model 1 H P R E Prom Model 2 P R E Enh Prom Model 3 Prom Figure 3-2. Model of the regulation of Ubx expression by the bxdl.5 PRE. In the anterior of the embryo the Ubx promoter (Prom) is off. Model 1 posits that the PRE directly interferes with the activity of the parasegmental enhancers (Enh). The model tested in this chapter, Model 2, predicts that the bxdl.5 PRE directly interferes with basal Ubx promoter activity. Model 3 predicts that PRE silencing is achieved through interaction with both enhancers and prompter. 80 transcription to genes and preventing activation or silencing of promoters by enhancers or silencers, respectively, separated from proximal promoters by insulator elements (Roseman, Pirrotta and Geyer, 1993). The Suppressor of Hairy-wing {Su(Hw)} gene encodes a zinc finger protein which binds to the gypsy insulator and is both necessary and sufficient to disrupt the communication between a large number of enhancers/silencers and promoters and to protect transgenes from chromosomal position effects (Gdula and Corces, 1997; Sigrist and Pirrotta, 1997; M a i l i n g a/., 1998). For the next series of experiments, a new P-element transformation vector was engineered in which chromatin boundary elements {Su(Hw) binding sites} were strategically inserted to shield the transgenes from chromosomal position-effects (see Materials & Methods and Figure 5-5). D. Overa l l Rat ionale Although the studies discussed above have provided valuable insight into the structure and function of the bxdl.5 PRE, several fundamental issues remain to be addressed. In the absence of parasegmental enhancers, do single copies of each bxdl.5 PRE module directly repress the Ubx promoter? If so, do each of the modules contribute equally to PRE-mediated repression or do they make distinct contributions at distinct development stages? Are embryonic and pairing-sensitive repression independent features of the bxdl.5 PRE? Do different PcG genes interact differentially with each of the PRE modules and are they required specifically for embryonic or pairing-sensitive repression or for both processes? Determining the function of each of bxdl.5 PRE modules as a means to understanding how these modules cooperate to impart wildtype bxdl.5 is the major goal of this chapter. A germline transformation assay, similar to Chapter 2, was used to address these fundamental issues regarding bxdl.5 PRE-mediated repression. The results presented in this chapter show that the bxdl.5 PRE has modular organization and is composed of at least three modules that have different PRE activities. The UPS and DPS modules directly repress the Ubx promoter in select parasegments during different stages of development. The DPS module is capable of pairing-sensitive repression whereas the UPS module is not. Although the PSR module is capable of pairing-sensitive repression of the mini-white gene, it is unable to completely repress the Ubx promoter as seen by the reduced levels of LacZ throughout the embryo. The results also show that the modules interact preferentially with some PcG and trxG genes and that a subset of those tested are required for either embryonic or 81 pairing-sensitive repression or for both processes. The implications of these results are discussed. II. Results The subdivision of the bxdl.5 PRE into 3 modules was based on the ability of the PSR fragment to confer pairing-sensitive repression on the mini-white reporter gene and the identification of PH binding sites within UPS and DPS that were required for embryonic silencing. To determine if the bxdl.5 PRE modules retain PRE activity a germline transformation assay was performed to determine the contribution of each of these modules to PRE function. A diagram illustrating the important features of the transformation vector and the constructs tested are shown in Figure 3-3. Two crucial control constructs were generated and transformed into Drosophila embryos. The first control was the transformation vector containing the basal Ubx promoter in the absence of any PRE fragments. This important construct allowed comparison of the basal Ubx expression pattern to the expression pattern obtained from the test constructs, and provided a baseline to determine the overall silencing capacity of each PRE module. The second control was the intact bxdl.5 PRE fused to the basal Ubx promoter. Since I have taken a reductionist approach to understanding the mechanisms of bxdl.5 PRE-mediated silencing, it was important to determine the overall function of the wildtype bxdl.5 PRE and use this phenotype as a point of reference for the smaller modules. A. What is the Basal Ubx Expression Pattern? The Ubx basal construct contains a 1.65kb Stul fragment that includes 680bp of 5' promoter and promoter proximal elements and the entire 968bp untranslated leader plus the first 7 codons of Ubx fused in-frame to lacZ (Qian et al, 1991)(Figure 3-4A). The term Ubx promoter will be used henceforth to refer to the transcriptional control elements contained in this 1.65kb Ubx fragment. This transformation vector also contains the mini-white eye reporter and the yellow marker. Transgenic embryos harbouring this basal construct were used as a control for transformation of constructs containing the bxdl.5 PRE subfragments for two reasons. First, the basal construct is active throughout embryogenesis in all parasegments thus providing the ideal reporter for a silencing assay. Second, the effects of the bxdl.5 subfragments on the basal Ubx expression pattern and on pairing-sensitive repression of the mini-white gene could be determined. Figures 3-4B and C show the typical basal Ubx expression pattern at two different 82 Figure 3-3. Diagrammatic representation of the p{y+, Ubx/lacZ, w+} transformation vector and the constructs tested. The Ubx/lacZ reporter construct contains the yellow and mini-white transformation ' markers in addition to the chromatin boundary elements labelled Su(Hw). The name used to describe each of the test constructs is given below each line drawing and its length in base pairs. The PH binding sites within the UPS and DPS modules are indicated with black boxes. PHO binding sites within bxdl.5 are indicated by gray diamonds and the d(GA)3 repeat elements are indicated with white triangles. 83 yellow Ubx promoter lacZ mini-white 3 'P Su(Hw) Su(Hw) Su(Hw) 5 ' P EcoRl M H N - 9 0 Ndel UPS M P A - 1 6 8 M H S - 7 0 Pstl Sty\ PSR DPS EcoRl UPS+PSR (1.07kb) Pstl Ndel PSR+DPS (1.15kb) EcoRl Ndel UPS (412bp) Ndel PSR (661bp) Pstl H Pstl DPS (489bp) EcoRl Ndel Pstl tet» UTD (1.56kb) Figure 3-4. The basal Ubx/lacZ construct does not have PRE activity. (A) Schematic representation of the basal Ubx/lacZ construct as contained in the p{y+Ubx/lacZw+} transformation vector. The bxdl.5 PRE (shown below vector) and its modules were subcloned upstream of Ubx/lacZ fusion gene. The PH binding sites within the UPS and DPS modules of the bxdl .5 PRE are indicated by black boxes. PHO binding sites are indicated by gray diamonds and d(GA)3 repeat elements are indicated by white triangles. (B-E) A l l embryos are mounted with anterior to the left and dorsal side up. LacZ expression in embryos was detected immunohistochemically using the VectaStain Kit (Vector Laboratories). (B) A germ band extended embryo showing the basal Ubx expression pattern. Expression is detected in the anterior compartment of every parasegment in the ectoderm. (C) At germ band retraction expression of LacZ is still detected in the anterior compartment of each parasegment. (D) The bxdl.5 PRE partially silences the Ubx promoter during germ band extension in every parasegment. Some head staining is detected. (E) Silencing persists into germ band retraction. (F) The basal construct is incapable of conferring pairing-sensitive repression of the mini-white transformation marker. Adults heterozygous for the transgene are shown on the left (labelled HET) and adults homozygous for the transgene are shown on the right (labelled HOM). (G) Transgenic heterozygous adults carrying the bxdl.5 PRE exhibit variegated expression of the mini-white gene. Homozygous adults have less pigment in their eyes indicating that this construct is capable of pairing-sensitive repression. 85 A Ubx Basal Ubx construct bxdl.5 PRE 86 stages of development. At germ band extension (4.5 hrs of embryogenesis), LacZ is detected in lateral epidermal patches in the anterior portions of every parasegment. The epidermis, derived from ectoderm, is the outer epithelial layer of the embryo, larva and adult; it secretes cuticle, the exoskeleton of the fly. The ectoderm is the outer germ layer of the embryo, and is easily distinguishable from the mesoderm and endoderm. Significant LacZ staining is detected in the head region. The head staining is routinely observed in transformants that contain D N A from the Ubx promoter and it is believed to be a background pattern derived from the basal Ubx construct (Bienz et al., 1988; Simon et al., 1990). By germ band retraction (10 hrs of embryogenesis), the spatial expression pattern in the anterior portion of each parasegment persists but the intensity of staining fades equally in all parasegments. This expression pattern was observed in 7 of 7 independent transformed lines and is identical to that described previously for Ubx/lacZ fusion genes containing greater than 680 bp of 5' flanking Ubx D N A (Bienz et al., 1988; Simon et al, 1990). The basal Ubx expression pattern is clearly distinct from the endogenous Ubx expression pattern as it lacks an anterior boundary of expression and LacZ is detected in all parasegments. This difference confirms that the endogenous Ubx expression pattern is dependent on other Ubx regulatory elements to maintain its unique domain of expression (PS5-13). The colour of the adult eyes in heterozygous adults containing the basal Ubx construct was uniform, non-variegated, and ranged from yellow to bright orange, depending on the insertion site. Homozygous adult eyes were always uniform, darker than the heterozygotes and ranged from dark orange to red (Figure 3-4F). This is consistent with an increased dose of the white* product. Thus, the basal Ubx construct was unable to silence lacZ in the appropriate parasegments and was incapable of conferring pairing-sensitive repression of the mmi-white gene located in the same transposon. B. What is the overall function of the bxdl.S PRE? In order to determine the contribution of the bxdl.5 subfragments to PRE activity, it was necessary to know the overall function of the intact element. Therefore, an important objective was to determine if the bxdl.5 PRE can directly repress the -680 to +986 Ubx promoter used in the transformation vector. Transgenic embryos carrying the bxdl.5 PRE directly repress the Ubx promoter in every parasegment at germ band extension (Figure 3-4D). By germ band retraction, LacZ is detected in the head and tail regions and nearly complete silencing is observed in PS 1-13 (Figure 3-4E). This result was observed in 4 of 4 transformed lines obtained, except that one of 87 the transgenic lines obtained exhibited a stochastic loss of repression in a small number of cells within the ectoderm and mesoderm. This feature is totally random as loss of repression does not follow a parasegment-specific pattern and varies from embryo to embryo. A l l transgenic embryos carrying the bxdl .5 PRE exhibited head and tail staining similar to the control embryos carrying the basal Ubx promoter construct. A l l four of the bxdl.5 PRE transgenic lines displayed some level of variegated eye colour in heterozygous adults. The size and location of the coloured spots, which varied from yellow to light orange, varied from eye to eye (data not shown). This phenotype suggests that loss of mini-white silencing could have occurred early or late in development and the activated state was subsequently inherited through multiple cell divisions. Not surprisingly, all four lines exhibited strong pairing-sensitive repression of the mini-white gene resulting in the complete or almost complete loss of eye pigment (Figure 3-4G). These results clearly show that the bxdl.5 PRE can repress the basal Ubx promoter in every parasegment of the developing embryo, and is capable of pairing-sensitive repression of the mini-white gene. C. Do the fctrfl.5PRE modules retain P R E activity? In the next series of experiments, the function of the bxdl.5 PRE modules were determined by inserting them upstream of the Ubx/lacZ fusion gene in the ~?{y+ Ubx/lacZ w+J transformation vector and transforming them into Drosophila embryos (see Figure 3-3). The first modules tested included every adjacent double subfragment combination, UPS+PSR and PSR+DPS. I reasoned that i f these larger fragments didn't produce a phenotype then it would be unlikely that the smaller ones would. The subsequent modules tested included the three separate subfragments UPS, PSR and DPS. Transgenic embryos harbouring these constructs were stained for LacZ expression and heterozygous and homozygous adults were compared for pairing-sensitive repression of the mini-white reporter gene. i . UPS+PSR and PSR+DPS: The larger contiguous fragments, UPS+PSR and PSR+DPS, are capable of repressing the basal Ubx promoter (Figure 3-5B and C). These transgenic lines showed a variable degree of repression, but did not completely repress in every parasegment in all stages of development. Loss of repression, especially in the posterior parasegments 7-13, was detected during germ band 88 Figure 3-5. Embryonic silencing activities of the bxdl.5 PRE modules at germ band extension. A schematic representation of the basal Ubx/lacZ construct as contained in the p{y+Ubx/lacZw+} transformation vector and the important features of the bxdl.5 PRE are shown on top. The PH binding sites within the UPS and DPS modules are indicated by black boxes. PHO binding sites are indicated by gray diamonds and d(GA)3 repeat elements are indicated by white triangles. The module being tested is drawn under each embryo. (A) The bxdl.5 PRE silences the Ubx promoter in most cells with slight derepression observed. Head staining is detected at all stages. (B and C) The UPS+PSR and PSR+DPS fragments partially silence the Ubx promoter, the latter being the stronger silencer. Note the misexpression in the mid-region of these embryos, PS7-10. (D) The PSR module reduces the expression level of the Ubx promoter in every parasegment but is not able to completely silence it. (E) The UPS module represses the Ubx promoter in a parasegment-specific manner. It completely represses LacZ expression in the posterior parasegments but does not affect expression in the anterior PS. (F) The DPS module represses the Ubx promoter in a variegated fashion; a punctate pattern of LacZ misexpression is observed and varies from embryo to embryo. 89 extension and persisted into the later stages. Misexpression was generally confined to the epidermis but some mesodermal misexpression was observed. In general, the four PSR+DPS transgenic lines obtained showed stronger repression than the three UPS+PSR lines obtained. Since both of these larger fragments contain the PSR module it was not surprising that both exhibited very strong pairing-sensitive repression of the mini-white gene. Because the function of the bxdl.5 PRE was partly retained in these constructs I proceeded to test the function of the smaller units. i i . UPS: Gel mobility shift analyses using embryo nuclear extract and restriction fragments spanning the bxdl.5 PRE revealed a 90bp region, MHN-90, within the UPS module that promoted the formation of a Polyhomeotic (PH)-containing nucleoprotein complex (Hodgson et al., in preparation). This 90bp region was necessary for bxd5.l PRE-mediated silencing as a deletion of this 90bp partially abrogated PRE activity. Of particular interest, deletion of M H N -90, in the context of bxd5.l, resulted in a metameric misexpression pattern suggesting that segmentation gene products may recognize this site to promote PcG silencing. This being the case, I predicted that the UPS module would show a metameric parasegment-specific repression pattern. Indeed, all four transformed lines carrying the UPS module displayed a parasegment-specific repression pattern. However, the expression pattern observed was completely unexpected. At germ band extension, the UPS module repressed the basal Ubx promoter in the posterior parasegments but could not repress in the anterior parasegments (Figure 3-5E). By germ band retraction, the UPS module could not repress the basal Ubx promoter in any parasegment indicating that this module's contribution to PRE activity is early in development (data not shown). The eyes of heterozygous adult flies carrying the UPS transgene were uniform orange to light red and never displayed a variegated eye phenotype. Homozygous animals never exhibited pairing-sensitive repression of the mini-white gene. i i i . PSR: It has been previously reported that a 661bp bxdl.5 PRE fragment, corresponding to the PSR module, generally causes variegated expression of the mini-white reporter gene located in the same transposon and this variegation is dependent on PcG genes (Chan et al, 1994; Sigrist and Pirrotta, 1997). The same construct, when homozygous, exhibits very strong pairing-91 sensitive repression of the mini-white gene. Another property of the PSR fragment is that it contains the only T R X binding site within the bxdl.5 PRE (Tillib et al, 1999). The embryonic activity of the PSR module is not known. Five transgenic lines harbouring the PSR module were obtained. Because of the strong adult effects of this construct and because the BP-171 (6X) construct can silence the S2 enhancer in a parasegment-specific pattern (Horard et al, 2001), I predicted that this construct, like the intact bxdl.S PRE, would strongly repress the basal Ubx promoter in every parasegment throughout the entire course of development. This was not the outcome. Blind studies, comparing transgenic embryos showing the basal Ubx expression pattern to embryos exhibiting the PSR expression pattern, revealed that the PSR transgenes showed reduced levels of LacZ expression during all stages of development. To confirm these observations, the head staining, which is independent of upstream sequences, was used as a reference for quantifying the levels of expression from the body of the embryos. Under high magnification, the basal Ubx construct showed an approximate equal intensity in head to epidermal staining whereas in the PSR lines the head staining was much stronger than the epidermal staining. Thus, the PSR module reduces the expression of the basal Ubx promoter in every parasegment of the embryo but is unable to completely repress it (Figure 3-5D). As expected, all five PSR lines caused variegated expression of the mini-white gene and exhibited very strong pairing-sensitive repression. The strength of pairing-sensitive repression, as measured by the distribution and the extent of eye colouration, was equal to the bxdl.S transgenic adults. iv. DPS: Gel mobility shift analyses using embryo nuclear extract and restriction fragments spanning the bxdl.5 PRE revealed a 70bp region, MHS-70, within the DPS module that promoted the formation of a PH-containing nucleoprotein complex. Competition analysis identified an additional fragment within DPS that was recognized by a PH-containing DNA-binding activity, MPA-168. In vivo deletion analysis revealed that these sites constitute one functional unit of PRE activity and are required for bxd5.\ PRE activity (Hodgson et al, 2001; Chapter 2). Because of this, I predicted that the DPS module would be a very strong repressor. Indeed, the DPS module repressed the basal Ubx promoter. In 3 of 3 lines tested, a punctate pattern of LacZ misexpression was observed at germ band extension that varied from embryo to embryo (Figure 3-5F and 3-6A). The level of misexpression within the parasegments was sporadic but limited to the ectoderm, and showed no parasegmental-selectivity. The DPS 92 Figure 3-6. The DPS module silences the Ubx promoter in a metameric pattern in germ band retracted embryos. A schematic representation of the basal Ubx/lacZ construct and the DPS module are shown on top. The PH binding sites within the DPS module are indicated by black boxes. PHO binding sites are indicated by gray diamonds and d(GA)3 repeat elements are indicated by white triangles. (A) At germ band extension the DPS module silences the Ubx promoter in a stochastic fashion. (B) By germ band retraction, LacZ is expressed in alternate PS, the odd-numbered PS (indicated by arrowheads), showing that the DPS module operates in a parasegment- and stage-specific manner. 93 Ubx 1 MPA-168 MHS-70 94 expression pattern differs from the PSR expression pattern as the levels of LacZ in the expressing cells are not reduced but equal to the level of LacZ detected in the cells of the head. By germ band retraction, a metameric misexpression pattern is observed. The DPS module is capable of repressing the basal Ubx promoter in the even-numbered parasegments (from PS2 to 14) and does not affect LacZ expression in the odd-numbered parasegments (Figure 3-6B). These results suggest that the DPS module is required for both early and late Ubx repression, and that its function changes during development. A l l three DPS lines caused variegated expression of the mini-white gene and exhibited very strong pairing-sensitive repression. This result indicates that the ability to confer pairing-sensitive repression is not exclusive to the PSR module. In retrospect, this is not a surprising result as both the PSR and DPS modules each contain three PHO binding sites that are necessary but not sufficient for pairing-sensitive repression of the mini-white reporter (Tillib et al., 1999; Shimmel et al., 2000; Mishra et al., 2001). These two modules also share in common d(GA) 3 repeat elements that are also involved in pairing-sensitive repression (Poux et al., 2002). D. Are embryonic repression and pairing-sensitive repression separable functions of the bxdl.5 PRE? The theme emerging from the previous experiments is that the bxdl.5 PRE is a mosaic of multiple subfragments with different functions that work together to impart full PRE activity. In keeping with this idea it was important to determine if the modules conferring embryonic repression of the basal Ubx promoter could be separated from those responsible for pairing-sensitive repression. The experiments described above provide partial support that these two activities are separable functions of the bxdl.5 PRE. To solidify this hypothesis one additional construct was tested and compared to the smaller modules. The fragments tested and compared were UTD, which consists of UPS, a 660bp spacer D N A and DPS, versus the PSR fragment (Figure 3-7). The spacer D N A was taken from the bacterial tetracycline gene and it was used to preserve the normal spacing between UPS and DPS. Competition analysis has shown that the tetracycline gene is not recognized by any bxdl.5 PRE binding activities (Hodgson and Brock, unpublished). As mentioned previously, PSR reduces the expression level of the Ubx promoter in every parasegment but is not able to completely repress it (Figure 3-7A). Furthermore, PSR is capable of mini-white variegation and pairing-sensitive repression (Figure 3-7C). Eight of eight UTD transgenic lines selectively repress the Ubx promoter in anterior PS 1-5 and in posterior PS11-14 95 Figure 3-7. Embryonic silencing and pairing-sensitive repression are separable functions of the bxdl.5 PRE. A schematic representation of the basal Ubx/lacZ construct is shown on top. The left panel (A and C) shows the embryonic and adult activities of the PSR module while the right panel (A, D and E) shows the embryonic and adult activities of the UTD module. (A) The PSR module reduces the level of LacZ expression in all parasegments at germ band extension. (B) The UTD module is capable of silencing the Ubx promoter in PS 1-5 and PS 10-14 at germ band extension. (C) Transgenic adults carrying the PSR module exhibit pairing-sensitive repression. The heterozygous flies, labelled Ffet, have more pigment that the homozygotes, labelled Horn. (D) Half of the UTD transgenic lines exhibited pairing-sensitive repression while the other half did not (E). 96 97 (Figure 3-7B). Half of the transgenic UTD lines are capable of mini-white variegation and pairing-sensitive repression whereas the other half are not (Figure 3-7D and E). This latter observation implies that the sequences flanking the PSR module also contribute to pairing-sensitive repression and indeed this is the case as DPS is capable of this activity. These results also suggest that the UPS module may interfere with the ability of DPS to exhibit pairing-sensitive repression as only one half are capable of this activity compared to DPS alone. Although these transgenes are shielded by boundary elements, this ambiguity may also be explained by chromosomal position-effects. Collectively, the results of these experiments clearly show that the UPS and DPS modules confer the embryonic repression capacity of the bxdl.5 PRE and that the PSR and DPS modules confer pairing-sensitive repression. Thus, the different bxdl.5 PRE modules have distinct roles in embryonic repression and in pairing-sensitive repression. E. What are the functional components that contribute to the activity of the different bxdl.5 PRE modules? Despite our knowledge that PREs are physical targets of PcG complexes in Drosophila, we know very little about the site-specificity of PcG action or the contribution of individual PcG proteins to silencing. Predictions about the identity and function of putative interacting factors can be tested in vivo by examining the effects of mutations in genes encoding the putative factors on embryonic repression or pairing-sensitive repression by the bxd5.l PRE. In an attempt to do so, I tested the responsiveness of the basal Ubx promoter, the bxdl.5 PRE and its modules in several PcG and trxG mutant backgrounds includingph 2,ph 4 0 9, Pc4, Psc\pho\ 6 2 T 6 2 T^ F 13 E 1 3 E(zr, tr?1 ', psq and trx . The impetus for selecting these particular alleles is that most are believed to be loss-of-function mutations (see Table 5-1 for description and references) and are representative of the known PcG complexes or trxG complexes (in the case of trx). All alleles were selected to determine the contribution of these genes to bxdl.5 PRE function in the embryonic repression and adult pairing-sensitive repression assays. Since most of the alleles being tested are homozygous lethal, dying at late embryonic or larval stages, the pairing-sensitive assay was conducted in PcG/trxG heterozygous backgrounds. Two independent transgenic lines of each construct were tested to control for variability caused by position-effects. Transgenic flies homozygous for the basal Ubx construct were crossed to flies heterozygous for ph2,phm, Pc4, Psc\pho\ E(zf2, trl62 and trxnu to determine if the basal Ubx expression pattern is dependent on PcG or trxG function respectively. Embryos were stained for 98 LacZ expression and PcG/trxG mutants were compared to sibling controls. As expected, the expression pattern of the Ubx/lacZ fusion was not altered in these mutant backgrounds (data not shown) confirming the absence of a PRE or TRE in the Ubx promoter. Consistent with this finding, Chiang et al. (1995) have shown that the basal Ubx construct does not create a novel PcG binding site in polytene chromosomes. Transgenic flies homozygous for the intact bxdl.5 PRE were crossed to embryos carrying aph 2 ,ph 4 0 9 , Pc4, Pscl,pho\ E(z)62, trl62 or a trxEl3 mutation. No misexpression of LacZ was detected in any of these transheterozygous embryos. One possible reason for undetected LacZ misexpression pattern in the transheterozyogotes is that one copy of the transgene in a PcG mutant background may not cause sufficient misexpression to be detected by this assay system. Another explanation is that smaller PRE units could be less sensitive to the loss of half the dose of PcG proteins than larger PRE fragments. Alternatively, all PcG genes are expressed in the Drosophila female germline and maternally deposited wild-type protein often rescues homozygous mutant embryos to a considerable extent (Struhl, 1981; Breen and Duncan, 1986; Soto et al, 1995). For these reasons, two generation crosses were performed to obtain embryos homozygous for the transgene and homozygous for the PcG/trxG mutation. From these crosses, approximately 12.5% of the embryos were homozygous for the transgene and homozygous or hemizygous for the X-linked ph allele and approximately 6% of embryos were homozygous for the transgene and homozygous for the autosomal mutation. Indeed, when a given transgene was responsive to a specific mutant allele, the number of embryos exhibiting the response was quite low, as expected. The results of the genetic analyses performed with the intact bxdl.5 PRE fall into three classes. The first class consists of those embryos that showed a stochastic loss of repression during germ band extension (4.5 hrs) in a variable number of cells throughout the embryo (Figure 3-8B). This result was obtained in the ph2 and ph409 homozygous embryos. The second class consists of those embryos that showed a loss of repression during germ band extension resulting in the partial restoration of the basal Ubx promoter expression pattern (Figure 3-8C). This result was observed in the Pc4, Psc1 and E(zf2 mutant backgrounds with Pc4 showing the strongest loss of repression. The third class consists of those embryos that showed no response and included the pho\ trf1 and trxBU mutant backgrounds (data not shown). It is important to note that the subdivision of the misexpression patterns into three classes only pertains to the intact bxdl.5 PRE. In all other transgene/mutant combinations (see below), I did not see the 99 Figure 3-8. The bxdl.5 PRE responds differently to ph mutations than all other mutations tested. (A) Germ band extended embryo homozygous for the bxdl.5 PRE transgene. (B) Embryos homozygous or hemizygous for ph2 or ph409 and homozygous for the bxdl.5 PRE transgene showed a stochastic loss of silencing during germ band extension in a variable number of cells throughout the embryo. This type of response was categorized as Class I. (C) The second class consists of those embryos that showed a loss of silencing during germ band extension resulting in the partial restoration of the basal Ubx promoter expression pattern. This was observed in ph2, ph409, Pc4 (shown here), Psc and E(zf2 homozygous backgrounds. 100 + .P{bxdl.5-UbxlacZ\ + 'P{bxd\.5-UbxlacZ} B ^ 2 ,P{bxd\.5-UbxlacZ) ph2 9P{bxdl.5-UbxlacZ} Class I Class II P{bxd\.5-UbxlacZ\ pc* P{bxd\ .5-UbxlacZ]5 ~ stochastic misexpression pattern characteristic of the ph-bxdl.5 PRE combinations, thus falling into the second and third classes. To test the contribution of individual PcG or trxG proteins to the activity of each bxdl.5 PRE module, I generated embryos homozygous for each module and homozygous for the PcG mutations mentioned above. A summary of the effects of PcG/trxG mutations on the embryonic PRE activity of the bxdl.5 PRE and its modules is shown in Table 3-1. A genetic response was classified as strong (++) when the basal Ubx expression pattern was fully restored. A response was classified as weak or moderate (+) when the basal Ubx expression pattern was partially restored. Consistent with the in vitro experiments of Hodgson et al. (2001) and Hodgson and Brock (unpublished data), the UPS and DPS modules showed a strong genetic requirement for ph. The UPS module loses the ability to repress the basal Ubx promoter in the posterior parasegments. The basal Ubx expression pattern is almost restored in the ph409 background and partially restored in the ph2 background (data not shown). This difference could reflect a strict requirement for the proximal form of ph (ph409 affects the proximal unit of ph, PHP) and suggests that the PH fusion product generated by the ph2 mutant could retain some PHP function. The punctate expression pattern observed in the DPS transgenic lines is lost in ph2 and ph409 mutants resulting in a more uniform expression pattern similar to the basal Ubx expression pattern. In addition, DPS-mediated repression in the even-numbered parasegments at germ band retraction is lost (Figure 3-9B). Thus, DPS-mediated repression is/?/z-dependent at both early and later stages of development. The level of LacZ expression detected in the PSR lines was not altered in ph2 and ph409 mutant backgrounds. Based on the immunoprecipitation experiments of Horard et al. (2000), a strong requirement for Pc and Psc on the repressive function of all the modules was predicted. Indeed, a strong Pc and Psc requirement was observed for all modules even for the weak activity of the PSR module (Figure 3-9D). However, the UPS module responded weakly to Psc (Figure 3-9F). Inconsistent with Horard et al. (2000) were the findings that the PSR module was unresponsive to E(zf2 and that the DPS module did so weakly. The expression patterns of all modules were unaffected in the pho] mutant background indicating that pho is dispensable for bxd PRE-mediated embryonic repression. The PSR module displayed reduced levels of LacZ in the trx™ mutant background indicating that trx functions through this module to stimulate expression. The severity of misexpression detected in this experiment ranged from very little to severe probably reflecting the degree of requirement for these gene products or the severity of the mutation. 102 Table 3-1. Effects of homozygous PcG and trxG mutations on the embryonic PRE activity of the bxdl.5 PRE and its modules. Genetic background 3 Construct" ph2 ph409 Pc4 Psc1 pho1 E(zf trf2 trx™ psq] Basal bxdl.5 + ++ ++ ++ + • UPS + ++ + + - NT PSR - - ++ ++ - +d NT DPS + ++ ++ ++ + a From the crosses, a small percentage of the embryos were of the genotype that yielded detectable responses; homozygous for the transgene and homozygous for the mutant allele. b Two independent transgenic lines harbouring each construct were tested. 0 Loss of silencing is the response. The lack of response is indicated by a minus (-) sign. A small or moderate response is indicated by a single +, while a strong response is indicated by ++. d This indicates a decrease in expression NT- Not tested 103 Figure 3-9. Effects of homozygous PcG mutations on the embryonic PRE activity of the bxdl.5 PRE modules. The parental genotypes of the final cross prior to collection and staining are shown above the embryos. The genotypes of interest are shown below each embryo. The left panel of embryos (A, C and E) represents embryos heterozygous or homozygous for the transgene and may be heterozygous for the mutant allele being tested. The right panel of embryos (B, D and F) represents the minority of embryos that exhibited a response and thus must be homozygous for the transgene and homozygous or hemizygous for the mutant allele. The percentages in the upper right corner represent the expected number of progeny with the desired genotype. (A) The DPS module silences the Ubx promoter in a metameric pattern in germ band retracted embryos. (B) The metameric expression pattern is lost in a homozygous or hemizygous ph409 mutant background. This embryo is indistinguishable from embryos carrying the basal Ubx construct (compare Figure 3-3C to 3-8B), thus it was classified as a strong response (++). (C) The PSR module reduces the level of expression from the Ubx/lacZ transgene. (D) A germ band extended embryo presumably homozygous for the PSR transgene and the Pc4 mutation. A very strong response was reported. (E) The UPS module has a unique expression pattern in germ band extended embryos. LacZ is detected in the anterior PS but not in the posterior. (F) In a Psc] homozygous background, a weak response is observed as UPS partially loses its ability to silence in the posterior PS. 104 phmw- . P{y\ DPS-UbxrtacZ, w+} m # ph409W- P{y+, BVS-UbxAacZ, w+} B —10% >• • P{V+, DPS-Ubx/lacZ, w+} ph409H>- DPS-t/ZutfacZ, w+} pk*"w ' P{y+, BPS-Ubx/lacZ, w+} yw\Pb>+,VSR-Ubx/lacZ,w+},Pc4 Y / M r . P { y + , PSR-Ubx/lacZ, H>+} Pc4 £ - 7 5 % J) - 6 % | m^L- * -id P{y+,VSR-Ubx/lacZ, w+} P{y+,PSR-Ubx/lacZ, p c 4 P{y + ,PSR - [7^acZ, pc4 yw\ p{y+, VPS-Ubx/lacZ, w+]. Psc1 V P (F + , UPS-Ubx/lacZ, w+}_ Psc1 k ^-w' + » + E ~75% M Y1 ^ 'cJJBk ..if - 6 % P[rM)PS-r?/>.v>7«fZ, i r : P{y+,UPS-£/fextfacZ, P s c l P{y+,VVS-Ubx/lacZ, w+}' pscl The d(GA)3 repeat elements within the MHS-70 site of the DPS fragment were required for the full silencing activity of the bxd5.\ PRE (Chapter 2). If binding of G A F or Psq to these elements is critical for PRE activity, then the repressive activity of the DPS module should depend on the trl and psq genes. In fact, the DPS-mediated repression was unaltered in these mutant backgrounds (data not shown). These confounding results may be due to rescue of the zygotic loss of GAF and Psq by maternally deposited products. A summary of the effects of PcG/trxG mutations on the pairing-sensitive repression activity of the bxdl.5 PRE and the PSR and DPS modules is shown in Table 3-2. The ability of the bxdl.5 PRE lines to show pairing-sensitive repression was lost i n P c 4 and Psc1 heterozygous backgrounds (data not shown) and pho] homozygous backgrounds (Figure 3-1 OA). The ability of the PSR and DPS modules to confer pairing-sensitive repression on the mini-white gene was also lost in Pc4 (Figure 3-10B), Psc'andphox mutant backgrounds (data not shown). These results confirm previous data showing the necessity of Pc (Fritsch et al., 1999) and pho or PHO binding sites (Tillib et al, 1999; Fritsch et al, 1999; Shimell et al, 2000; Mishra et al, 2001) to pairing-sensitive repression and establish a role for Psc in this activity. A recent study has shown that the ability to confer pairing-sensitive repression by the BP-191 (X6) subfragment of the PSR module is dependent on both PHO and GAF (Poux et al, 2002). Dependence on trl function for this activity was not observed in the bxdl.5, PSR or DPS transgenic lines. The previous studies have shown that the bxdl .5 PRE modules retain residual PRE activity and that two of these modules, UPS and DPS, which contain PH binding sites, lose their ability to repress in ph mutant embryos. The PSR module, which does not contain PH binding sites, was unresponsive to ph mutations. Based on these data, and the correlation with the immunoprecipitation experiments of Horard et al. (2000), the genetic interaction assay is an ideal system to confirm existing biochemical data about the functional components of each module and, by extension, can be used to determine what the functional components are in the absence of biochemical data. III. Discussion A germline transformation assay was undertaken to functionally dissect the bxdl.5 PRE with the aim of understanding how it represses gene expression. The term repression, instead of silencing, was used to describe the inhibitory effects of the bxdl.5 PRE modules on the Ubx promoter for one key reason. The classical definition of silencing states that silencing can be 106 Table 3-2. Effects of heterozygous PcG and trxG mutations on the pairing-sensitive repression activity of the bxdl.5 PRE and the PSR and DPS modules. Genetic backgroundb Construct3 ph2 ph409 Pc4 Psc1 pho1 E(zf trl62 trx*13 psq] bxdl.5 + + + -PSR + + + + DPS + + + _ a Two independent transgenic lines harbouring each construct were tested. b A three-way comparison was made between heterozygous adults, homozygous adults and homozygous adults in a mutant background. Due to the homozygous lethality of the mutations, the mutant backgrounds were in fact heterozygous backgrounds. c The loss of pairing-sensitive repression is indicated by a + while the lack of a response is indicated by a minus sign (-)• 107 Figure 3-10. The ability of the bxdl.5 PRE and the PSR module to exhibit pairing-sensitive repression is abolished in pho and Pc mutant backgrounds. (A) Transgenic flies containing the bxdl.5 PRE exhibit pairing-sensitive repression of the mini-white gene located in the same transposon. Compare the heterozygote, bxdl.5/+, to the homozygote, bxdl.5/bxdl.5 (upper left and upper right). Pairing-sensitive repression is abolished in apho]/pho] mutant background. The expression of the mini-white gene in this fly (bottom) is almost identical to the heterozygote (upper left). (B) Transgenic flies containing the PSR module also exhibit pairing-sensitive repression of the mini-white gene. The heterozygote, PSR/+, has more eye pigmentation than the homozygote, PSR/PSR (compare upper left and upper right). Pairing-sensitive repression is abolished in a Pc4 heterozygous background. The expression of the mini-white gene in this sensitized background (bottom) is almost identical to the heterozygote (upper left). 108 W 1 . 5 + b x d l . 5 b x d l . 5 if bxdl.5.pho{ bxdl.5\phox B P S R separated into three distinct phases; initiation, maintenance and inheritance (Loo and Rine, 1995). The Ubx basal promoter, which is expressed ubiquitously, lacks those sequences required for the initiation phase, as does the bxdl.5 PRE. Thus, the inhibitory effects of the bxdl.5 PRE and of its modules on the basal Ubx expression pattern are best described as repressive. The results, summarized in Table 3-3, indicate that the bxdl.5 PRE is a complex element built up of at least three modules with differing functions that contribute to the overall function of the PRE. The UPS module represses the Ubx promoter early in embryogenesis in the posterior parasegments and is incapable of pairing-sensitive repression. The DPS module represses the Ubx promoter in a variegated manner early in development and then selectively in the even-numbered parasegments during germ band retraction. The DPS module, like the PSR module, is capable of pairing-sensitive repression indicating that multiple pairing-sensitive regions exist within the bxdl.5 PRE. Unlike the other modules, the PSR module does not completely repress the Ubx promoter but does reduce its level of expression in every parasegment. These results clearly show that embryonic repression and pairing-sensitive repression are separable functions of the bxdl.5 PRE. Genetic analyses indicate that the embryonic activity of the UPS and DPS modules is dependent on the function of the PcG genes ph, Pc, and Psc. The DPS module showed an additional requirement for E(z) indicating that different PcG complexes are recruited to different regions of the PRE which may account for the differing activities of each module. Consistent with the finding that the bxdl.5 PRE also contains a TRE, the PSR module showed further reduction in LacZ levels in a trx background. Genetic analyses also revealed that Pc, Psc andpho are essential for PSR and DPS-mediated pairing-sensitive repression of mxm-white. A . PRE-mediated gene silencing and basal transcription This study is the first to show that smaller fragments of the bxdl.5 PRE, the DPS, PSR and UPS modules, are able to directly repress the basal Ubx promoter in a PcG-dependent manner, and in the absence of parasegmental enhancers. This is a key finding as one of the more accepted models of PcG function posits that silencing is achieved by PRE interference with enhancer-promoter interactions. D N A looping has been implicated in mediating these long-range interactions. The results presented in this chapter do not rule out this hypothesis but do indicate that several mechanisms of PcG-mediated silencing exist and that one of these must be at the level of the basal Ubx promoter. Several key findings support the model that the PcG silences transcription through interference with the basal transcription machinery. In transgenic analyses, the bxd PRE works 110 Table 3-3. Summary of the PRE activities associated with the bxdl.5 PRE and its modules. Construct Number of transformed lines mini-white variegation3 Repress Ubx promoter at germ band extension" Repress Ubx promoter at germ band retraction Pairing-sensitive repression' Basal 7 0/7 - - 0/7 bxdl.5 4 4/4 PS1-14 PS1-14 4/4 UPS+PSR 4 4/4 PS1-5 PS1-14 4/4 PSR+DPS 3 3/3 PS1-7 PS1-14 3/3 UPS 4 0/4 PS7-14 - 0/4 PSR 5 5/5 - - 5/5 DPS 3 3/3 PSl-14 d PS2,4,6,8,10, 12,14 3/3 UTD 8 3/8 PS1-5/10-14 - 4/8 Ratios in this column represent the number of lines with variegated white expression in heterozygous animals to the total number of lines tested b Data in this column represent the parasegments (PS) that the Ubx promoter//acZ fusion gene is silenced. A minus symbol (-) indicates that the construct was unable to silence the Ubx promoter. c Ratios in this column represent the number of lines with decreased white expression in homozygotes versus heterozygotes to the total number of lines tested. Variegated misexpression phenotype varying from embryo to embryo. I l l more efficiently when assayed together with the Ubx promoter (Chan et al., 1994; Muller, 1995). ChIP experiments have shown that PC associates with Ubx promoter sequences in embryos only after the initiation of gene silencing has occurred (Orlando et al., 1998), yet the basal Ubx construct does not respond to PcG mutations nor does it recruit PcG proteins to its insertion site on polytene chromosomes (Chiang et al., 1995; this work). These conflicting results suggest that PcG binding to the Ubx promoter depends on interaction with the PRE, a hypothesis that could easily be tested using ChIP on my transgenic lines. Furthermore, mutations within the Ubx promoter reduced the ability of the abx PRE to maintain gene silencing in embryos (Laney and Biggin, 1992). Similarly, Zeste and G A F binding sites at the Ubx proximal promoter were required to maintain, but not to initiate, the silent state of a Ubx transgene containing bxd PRE sequences (Hur et al., 2002). Lastly, the biochemical purification of the PcG complex PRC1 revealed that, in addition to PcG proteins, it also contains general transcription factors (GTFs), namely TBP (TATA-binding protein) and TAFs (TBP-associated factors), that play a critical role in transcriptional activation (Shao et al., 1999; Saurin et al., 2001). How does the bxdl.5 PRE directly repress the basal Ubx promoter? One possibility is that the Ubx promoter is permissive to both transcription factor binding and assembly of the pre-initiation transcription complex (PIC), but, in the presence of a functional bxd PRE, promoter clearance or transcription elongation is prevented. The idea for this hypothesis stems from the finding that silenced chromatin at the HMRal promoter in yeast is permissive to activator binding and PIC recruitment (Sekinger and Gross, 2001) and from the finding that TBP is a component of both activating and repressing complexes (reviewed in Pugh, 2000). The mechanisms for repressing TBP transcriptional activity are diverse and include keeping TBP dissociated from D N A or inhibiting the incorporation of other GTFs into the assembling TBP-containing complex. Thus, PcG complexes at PREs may sequester TBP and retain it in its inactive form or sterically hinder access of select GTFs to the PIC. In vitro transcription assays using well defined PRE and promoter templates could help determine some of the functions of PcG-containing complexes. B. Embryonic function of the bxdl.5 P R E modules The variegated misexpression pattern observed in DPS transgenic embryos at germ band extension indicates that PcG-mediated repression of the Ubx promoter via this module is conducted on a cell-by-cell basis. Thus, the PcG complexes formed on this element cannot discriminate between the different regions of the embryo as they do with larger bxd PRE 112 fragments. Such difficulties in initiating repression in the appropriate parasegments could indicate that this module is impartial to the function of the early segmentation genes involved in establishing the Ubx expression pattern. This is consistent with the finding that PcG-containing D N A binding activities were not detected on any DPS fragment in 0-3 hr embryo extracts (Hodgson and Brock, unpublished data). The fact that DPS confers some degree of PcG-mediated repression indicates that this module has some level of affinity for PcG complexes. The most likely interpretation for the variegated misexpression pattern is that the PcG complexes formed on the DPS module are unstable or transient and dissociate in different cells during germ band extension. The transition from variegated repression in germ band extended embryos to a metameric misexpression pattern in germ band retracted embryos most likely reflects different mechanisms of PcG complex recruitment at later stages of development. Indeed, biochemical differences were observed between the PcG-containing complexes formed on DPS (the MHS-70 site) throughout embryogenesis (Hodgson and Brock, unpublished data). These late repressive complexes are more stable and fine tuned to be active in the even-numbered parasegments. Even though the products of the segmentation genes have since decayed, their downstream effects may be interpreted by the DPS module and acted upon accordingly. A clearer understanding of DPS function awaits the biochemical purification of the PcG complexes formed on this module at different developmental stages. In the meantime, predictions about the identity of the putative factors can be tested in vivo by examining the effects of mutations in genes encoding these factors on DPS-mediated repression. The UPS module represses the Ubx promoter in the posterior parasegments (PS7-14) at germ band extension. The UPS module (site MHN-90) is recognized by a PcG-containing complex early in development (0-3 hrs) thus implicating a role for this module in the initiation of silencing (Hodgson and Brock, unpublished data). This was verified by the early misexpression phenotype observed when site MHN-90 was deleted from the bxdSA PRE (Chapter 2). The biochemical and genetic data cannot explain the UPS expression pattern as it conflicts with the expected role of the bxd PRE which is to maintain Ubx silencing in PS 1-5. As proposed for the DPS module, this result most likely reflects the nature of the D N A binding factors that recruit repressive complexes. By germ band retraction the UPS module is inactive and the basal Ubx expression pattern is observed. The repressive capacity of this module is lost probably owing to the decreased stability of PcG complexes formed at this site or the absence of PcG complex recruiting factors. Tracking the biochemical composition of the PcG complexes formed on this module would provide insight into how parasegment-selective silencing is achieved. 113 The PSR module was unable to completely repress the Ubx promoter at any stage of development but was able to reduce the level of Ubx transcription during germ band extension. This result appears to be in contrast to the findings of Horard et al. (2000) who have shown that when the BP-191 subfragment of PSR is multimerized six times it is able to silence expression of the S2 enhancer in PS 1-5. These differences could be explained i f after multimerization, a high-affinity PcG complex binding site has been generated that is otherwise not present in the single copy. The inhibitory effects of the PSR module were lost in several PcG mutants indicating that this fragment is responsive to the PcG and is thus a PRE, albeit a weak one. Consistently, biochemical experiments have shown that single copies of PSR fragments effectively immunoprecipate PcG-containing complexes (Horard et al., 2000). The fact that the PSR module is incapable of completely silencing the Ubx promoter could be explained by the finding that this module also contains a TRE (Tillib et al, 1999; this work) and the decision to become a PRE or a TRE may depend on the activity of the flanking modules. The question that arises is how does the function of the PSR module, or any of the modules for that matter, correspond to the overall function of the bxdl.5 PRE? A clearer understanding of the function of the bxdl.5 PRE modules awaits the biochemical purification of the PcG complexes formed on these elements at different developmental stages. In the meantime, predictions about the identity of the putative interacting factors can be tested in vivo by examining the effects of mutations in genes encoding these factors on UPS, DPS and PSR-mediated repression. C. Mechanisms of bxdl.5 PRE-mediated embryonic repression A prediction made about the overall function of the bxdl.5 PRE was that it would be the sum of the activities of each module. The results obtained in this chapter argue that this is not the case. Each module possessed distinct functions that, when combined additively, do not match the function of the intact element. For example, at germ band retraction the DPS module represses the Ubx promoter in every other parasegment, while the UPS and PSR modules have no effect on the Ubx promoter at this stage. A sum of the parts would be a metameric misexpression pattern during germ band retraction. In fact, the intact element completely represses the Ubx promoter in every parasegment at this developmental stage. Therefore, functional interactions between the modules must be an integral property of the bxdl .5 PRE. Evidence to support this comes from the phenotypes of the larger contiguous PRE fragments, UPS+PSR, PSR+DPS and UTD, which behave more like the intact element as they repress the 114 Ubx promoter very effectively and much stronger than the individual modules. It appears that the different modules cooperate to achieve much stronger repression to a degree that is not attained by the single modules alone. These interactions are not observed when the modules are placed in a context that is quite different from the context in which they normally function. The increase in strength of repression may be due to increased interactions between PcG proteins and the cooperative assembly of PcG complexes, facilitated by the proximity of the adjacent PRE modules. Functional interactions between different D N A elements are a recurring theme within the bxd PRE and in other silencing systems. Competition analyses (Hodgson et al, 2001; Hodgson et ah, in preparation), in combination with the deletion analysis presented in Chapter 2, revealed that separate fragments within the bxd5.\ PRE bound related complexes and that two of these fragments, MHN-90 of the UPS module and S1HB-90 of the SI enhancer, act synergistically to promote PRE activity. Functional interactions based on competition analysis were not detected between fragments of the three modules but were detected within the DPS module, sites MHS-70 and MPA-168. Increasing the sensitivity of the competition assay by using highly purified PcG-containing D N A binding activities might reveal biochemical interactions between the bxd5.l PRE modules. These interactions can then be tested in vivo to determine their functional relevance. D N A elements that have little or no silencing activity on their own can cooperate over distances of several kilobases to establish silencing. In yeast, binding sites for the repressor activator protein 1 (Rapl), the autonomous replicating sequence (ARS) binding factor 1 (Abfl) and the origin recognition complex (ORC) cooperate to initiate nucleation of a SIR-containing silencing complex at the silent mating type locus (Boscheron et al., 1996). A single binding site for any of these proteins can enhance the action of a distant silencer without acting as a silencer on its own. The same type of mechanism might be envisaged to explain the function of the bxdl.5 PRE. It has been shown that the early transcription state of a promoter at the time of initiation of PcG-mediated silencing is decisive in conferring subsequent repressive or activating function to PREs/TREs in a particular cell (Poux et al., 1996; Cavalli and Paro; 1998; Cavalli and Paro, 1999). Therefore, function of the UPS and DPS modules may be to decipher the early transcriptional state established by the segmentation genes, and to recruit PcG complexes onto these modules in the correct cells. Once the decision is made to become a silencing element, PcG complexes on the DPS and UPS modules promote the formation of PcG complexes on the PSR module, converting the PSR module into a strong PRE thus abrogating any PSR-associated 115 Figure 3-11. Model of bxdl .5 PRE-mediated repression of Ubx promoter. Early in embryogenesis (upper panels) the Ubx promoter is repressed (left) or activated (right) in the appropriate parasegments by the action of the segmentation genes. The function of the UPS and DPS modules may be to decipher the early transcriptional state and to recruit PcG complexes onto these modules only in those cells where Ubx is off. The PSR module, which contains a TRE, is responsible for maintaining the active state in those cells where Ubx is on. As embryogenesis proceeds, PcG complexes on the DPS and UPS modules promote the formation of PcG complexes on the PSR module, converting the PSR module into a strong PRE. The three modules cooperate to directly silence the Ubx promoter. Late trans-interactions between the complexes formed on the PSR and DPS modules (double-headed arrows) stabilize and strengthen PcG-mediated silencing and result in pairing-sensitive repression (bottom left). In the absence of the signals that initiate Ubx silencing, the PSR module recruits stimulating trxG complexes thus inhibiting access of PcG complexes and counteracting PcG-mediated silencing. 116 TRE activity. In the absence of the signals that initiate Ubx silencing, the PSR module recruits stimulating trxG complexes thus inhibiting access of PcG complexes and counteracting PcG-mediated silencing (Figure 3-11). Real understanding of the mechanism of bxdl.5 PRE function awaits the development of order of recruitment experiments on embryonic cells sorted on the basis of Ubx transcription states. The model proposed above for bxdl.5 PRE activity is supported by several findings which suggest that PcG and trxG proteins act as molecular switches through common chromosomal elements, PREs/TREs, to either direct the adjacent chromatin into a transcriptionally active or inactive state. Several trxG proteins are components of the SWI/SNF chromatin-remodelling complex (Crosby et al., 1999; Kal et al., 2000). In vitro chromatin remodelling assays using the PcG-containing complex PRC1 and a SWI/SNF complex suggest that these two complexes compete with each other for interaction with the nucleosomal template (Shao et al., 1999). Molecular and genetic analyses on the Fab-7 PRE of Abd-B have demonstrated the ability to switch from an inactive to active state of transcription during embryogenesis and this switching is mediated by the antagonistic function of trxG and PcG genes (Cavalli and Paro, 1998, 1999). D. What is pairing-»sensitive repression? In 1954, 38 years after the observation that homologous chromosomes are paired in the somatic cells of Drosophila (Metz, 1916), E.B. Lewis showed that homologue pairing does influence gene expression. He termed this form of regulation transvection and illustrated it by showing how complementation between two Ubx alleles was antagonized by disruptions in somatic pairing (Lewis, 1954). Since then transvection has been demonstrated for a number of loci in many different species (reviewed in Wu and Morris, 1999). Pairing-sensitive repression is a form of transvection that has only been described in transgenic analyses. Pairing-sensitive repression has not been demonstrated in embryos indicating that this is strictly a larval or adult phenomenon (Kassis et al, 1991; Muller et al, 1999, Horard et al., 2000; this work). In these studies, silencing conferred by known pairing-sensitive elements was never enhanced in homozygous embryos as compared to the heterozygotes indicating that the regulatory proteins involved in pairing-sensitive repression are inactive or not present in the embryo. It has been proposed that cooperative interactions between PcG-containing complexes located on separate chromosomes facilitate pairing-sensitive repression (Pirrotta, 1998). Furthermore, the short cell 118 cycles during embryogenesis may prevent pairing-sensitive repression in embryos while the prolonged interphases in larval phases may promote this form of silencing. The pairing-sensitive capacity of the bxdl.5 PRE and of the PSR and DPS modules is abrogated in Pc, Psc and pho mutant backgrounds implicating a role for these genes in pairing-sensitive repression. Pc and Psc mutations also interfered with the embryonic silencing capacity of the bxdl.5 PRE and DPS module whereas pho mutations had no effect. That PHO is required for pairing-sensitive repression is exemplified by the fact that all known pairing-sensitive elements lose their ability to induce pairing-sensitive repression in pho mutants or when the PHO binding sites are mutated (Brown et al, 1998; Muller et al, 1999; Cavalli and Paro, 1999; Shimell et al, 2000; Mishra et al, 2001; Americo et al, 2002). This differential requirement for the PcG suggests that PcG silencing may not be equivalent in embryos and larva. It has been suggested that PHO may be a component of a PcG-containing complex that is only required or functional in the larval or adult stages. Analysis of pho mutant embryos and of PRE constructs whose PHO-binding sites are mutated suggests that, while PHO is important for silencing in imaginal discs, it is not necessary for embryonic PcG silencing (Poux et al., 2001a). Although PHO binding sites are necessary for pairing-sensitive repression, they are not sufficient (Brown et al., 1998). Americo et al. (2002) have examined the sequences important for the pairing-sensitive silencing activity of a 181bp fragment from the engrailed gene. They have shown that at least five and as many as eight distinct protein-binding sites are required for this activity. Mutational analysis revealed that loss of any one of these sites abrogated pairing-sensitive repression. One of the required sites was the PHO binding site and another was the sequence G A G A G , which binds GAF and Psq and is important for the embryonic function of the bxd PRE (Hodgson et al, 2001; Chapter 2). The identity of the other proteins is unknown. The PSR and DPS modules contain multiple PHO and GAF/Psq binding sites and require PHO as they lose their pairing-sensitive repression activity in a pho homozygous background. The genetic requirement of trl (encoding GAF) or psq for pairing-sensitive repression could not be discerned using this genetic assay as the trl and psq mutations used were homozygous lethal. Hence, all adults scored were heterozygous for the mutant allele and the wildtype product in these flies may have been sufficient for PSR and DPS-mediated pairing-sensitive repression. However, in other studies, mutation of all GAF/Psq binding sites in the iab-7 PRE (Mishra et al., 2001) and in the PSR subfragment BP-191 (6X) resulted in the inability of these constructs to induce pairing-sensitive repression (Poux et al, 2002). These data suggest a surprising degree of complexity in the DNA-binding proteins required for pairing-sensitive repression. 119 Pairing-sensitive repression and other forms of transvection may not be essential for viability. At the endogenous Ubx locus, chromosome rearrangements, including deletions, that disrupt somatic pairing, generally do not result in lethality (Micol et al., 1990). In addition, an analysis of haploid mosaics in Drosophila revealed that haploid patches of tissue survive to adulthood (Santamaria, 1983). Pairing-sensitive repression may exist because it uses factors involved in a mechanism that is essential and also involves gene pairing. Indeed, PcG function is essential for viability but does it require gene pairing? A l l genes of all organisms homologously pair at least once per cell cycle; as they are passing through the replication fork. Homologue pairing events of this kind mediate replication control (Keeney and Kleckner, 1996). Is it possible that some factors of the replication machinery serve dual roles as modulators of pairing-sensitive repression? Evidence that D N A replication and the PcG share common regulatory pathways is that the PcG gene cramped (crm) genetically interacts with mus209, the gene encoding Proliferating Cell Nuclear Antigen (PCNA), a protein involved in D N A replication (Yamamoto et al., 1997). Furthermore, the products of these two genes co-localize at sites of D N A replication in S-phase nuclei. The importance of the interplay between D N A replication and silencing processes is underscored by the finding in yeast. The silencer elements flanking the silent mating loci have the ability to allow autonomous replication of plasmids. In addition, proteins of origin of recognition complex (ORC) are required for both D N A replication and gene silencing (reviewed in Loo and Rine, 1995). I propose that the co-localization of silencing complexes with D N A replication machinery plays a role in establishing the epigenetic mark required to propagate the silent state. Furthermore, pairing-sensitive elements are origins of replication and their ability to confer pairing-sensitive repression is a secondary feature. The role of crm in bxd PRE-mediated pairing-sensitive repression remains unknown. 120 Chapter 4 General Discussion I. Main conclusions Overall, the main conclusions of my thesis are: 1) The bxd5.l PRE contains four, non-overlapping PH binding sites that are essential for embryonic silencing. 2) Functional interactions between at least two of these PH binding sites promote PRE activity. 3) Mutation of the d(GA) 3 sequences within the multipartite MHS-70 site disrupts PRE activity demonstrating a sequence-specific requirement for a PcG complex. 4) The bxdl.5 core PRE is composed of at least three modules that repress transcription through direct interaction with the Ubx promoter. 5) Embryonic silencing and pairing-sensitive repression are independent features of the bxdl.5 PRE. II. Modularity in the bxd P R E The notion that cis-regulatory elements are composed of multiple genetic elements, or modules, was first described for viral promoters and included the HSV thymidine kinase (tk) and SV40 early transcription units (reviewed by McKnight and Tjian, 1986). These studies have shown that promoters are composed of discrete functional modules, ranging from 7-20bp, contain recognition sites for transcription factors and retain some functional activity. The best example of this is the T A T A box that functions to position the start site for transcription. Since then, the term modular has been used to describe the structure of most enhancer elements and, more recently, silencer elements. A combination of genetic, molecular and biochemical approaches led to the observation that the bithoraxoid PRE has modular structure. Modular organization of cis-regulatory elements allows transcription to be regulated in response to diverse intracellular signals. The differences observed in the activities of the different bxdl.5 PRE modules most likely reflect a qualitative difference of the PcG complexes formed on each of these modules. This is supported by the differential genetic response of the modules in the different PcG mutant backgrounds (this work) and by the biochemical differences of the PH-containing complexes formed at the UPS and DPS modules (Hodgson and Brock, unpublished). That PcG proteins form different multimeric protein complexes is clearly established in the literature (Shao et al, 1999; Tie et al, 2001; Hodgson et ah, 2001). The time-of-action and function of these PcG-containing complexes at their target sites remains a mystery. 121 The results obtained in this thesis indicate that the modular nature of the bxd PRE is responsible for the orchestrated recruitment of different PcG complexes, those involved in the initiation followed by those involved in the maintenance of silencing. Preliminary biochemical evidence of this has been obtained for the bxd PRE (Orlando et al., 1998; Hodgson and Brock, unpublished). Detailed understanding of DPS, PSR and UPS function awaits the biochemical purification and identification of the PcG complexes that bind to these modules at different developmental stages as a way of determining how spatial- and temporal-selective silencing is established. In the interim, predictions about the identity of putative DNA-recognition factors can be tested in vivo by examining the effects of mutations in genes encoding the putative factors on embryonic silencing or pairing-sensitive repression by the bxd5.\ PRE. Because PcG proteins directly interact with each other, developmentally regulated interactions between the PcG complexes formed on the different modules may be an integral feature of bxdl.5 PRE-mediated silencing. Indeed, functional interactions between the PH binding sites in UPS and SI enhancer were observed. These interactions can also be inferred from the ability of PcG proteins to bind cooperatively. The spatial arrangement and the proximity of the modules may facilitate these interactions. The importance of the spatial distribution and proximity of the modules relative to one another remains to be seen and could be easily tested. Rearranging the order of the modules or altering the spacing between them may provide valuable insight into how the bxd PRE is built. Although the bxdl.5 P R E has been subdivided into three modules, it is highly probable that these modules could be further subdivided and still retain PRE activity. Indeed, even within the multipartite MHS-70 PH binding site, there are discrete sequences for at least two distinct binding activities. I predict that the DPS region encompassing the PH binding sites MHS-70 and MPA-168, which constitutes one functional unit of PRE activity, wil l behave similarly to DPS. It would also be interesting to test the effects of each single PH binding site on Ubx expression. To explore the mechanisms involved in pairing-sensitive repression it will be important to find minimal elements conferring this activity, and this activity alone. I believe the distal portion of MPA-168, which contains the PHO and G A F binding sites, will behave this way. Identification of modules within modules will be an important step in determining the intricacies of bxdl.5 PRE design. Complex cis-regulatory elements presumably evolve from simpler precursors. Hence, an approach to identify important regulatory sequence motifs is to compare corresponding sequences from related but distant species. The reasoning is that since the 122 regulatory proteins remain highly conserved, their target sequences will also be conserved. Dellino et al. (2002) have taken this approach and have found extensive conservation of bxd PRE sequences from several Drosophila species after a divergence of 60 million years. To date, no function has been ascribed to the region with the highest conservation (proximal end of the DPS module). It would be interesting to determine if this fragment retains PRE activity in transgenic assays. It is not yet clear to what extent modular organization can be extended to other PREs. However, it is clear that embryonic silencing and pairing-sensitive repression are separable functions at many PREs. The Mcp PRE core fragment is sufficient to silence an embryonic reporter but requires additional flanking sequences to confer pairing-sensitive repression in adults (Muller et al., 1999). The iab-2 (1.7kb) PRE is capable of pairing-sensitive repression but it is unable to act as a PRE in embryos (Shimmel et al., 2000). It will be interesting to learn if the modularity observed at the bxd PRE, with regards to embryonic activity, is an intrinsic feature of PRE design. III. PRE-Promoter Interactions One of the more appealing models of how PREs silence the expression of their target genes is the " D N A looping model" (Pirrotta and Rastelli, 1994). This model is based on cooperative binding and postulates that weak PcG binding sites are distributed throughout the genome at strategic points along transcriptional domains. PREs consist of multiple PcG binding sites and these binding sites serve as nucleation centers. Thus, the formation of a large, fully functional PcG complex at PREs is stabilized by protein-protein interactions with PcG proteins bound at the weaker PcG binding sites. These interactions result in a stable looping configuration of the D N A that could,prevent enhancers from interacting with their respective promoters thus creating a structure unfavourable to transcription. Until recently, this has been the favoured model for PcG function. The results of Chapter 2, in combination with the gel-mobility shift assay, clearly show that there are multiple PH binding sites along the bxd5.l PRE. Furthermore, the biochemical properties and the spatial arrangement of these sites along this PRE strongly indicate that D N A looping may be one mechanism used by this element to silence gene expression. Pending direct visualization of these D N A loops via electron microscopy, this remains a testable hypothesis. Although there is no evidence to contradict the D N A looping model, the results of Chapter 3 suggest that PcG complexes on PREs directly interfere with basal transcription at the 123 level of the Ubx promoter. A common characteristic of PREs is that they are sensitive to the state of activity of their target genes in such a way that PcG silencing is established only at transcriptionally silent genes and trxG activation is established at transcriptionally active genes. This indicates that the PRE must sense and respond to promoter activity accordingly. This is most likely achieved through changes in chromatin structure at the promoter that are established by the early regulatory cascade of transcription factors that dictate the initial transcriptional state. The repressed state is recognized by the PcG and forces the PRE to adopt a structure that will maintain the off state at the promoter. Conversely, an active promoter will be recognized by the trxG and direct the formation of trxG complexes at the PRE, converting it into a TRE. Like PC, TRX has also been shown to bind to Ubx promoter sequences during the maintenance phases (Orlando et al., 1998). Insight into the nature of PREs/TREs has been provided by experiments that switch a silenced PRE to an active state. Using a transgene carrying the Fab7 PRE flanked by reporter genes under the control of a GAL4 transcriptional activator, the silent state of the PRE was switched into an active one by a pulse of G A L 4 expression (Cavalli and Paro, 1998; 1999). Hyperacetylation of histone H4 was found to be associated with Fab7 after activation, suggesting that changes in chromatin structure are the cause or consequence of the switch. Does the bxd PRE silence transcription through interference with enhancer-promoter interactions or does it directly interfere with basal transcription? The results presented in this thesis indicate the latter but do not rule out the former. Since the PRE activity of each of the modules was assessed in a context that is quite different from the context in which they normally function, it is difficult to conclude which mechanism is used by the endogenous PRE. One possible experiment that I can envision to help distinguish between these different modes of silencing involves the use of excisable cassettes via the yeast FRT-FLP recombinase system (Broach and Hicks, 1980). By removing adjacent enhancer elements at different times during development, we would be able to conclude that the bxd PRE directly silences the promoter if the off state persists. Until these experiments are developed and the hypotheses tested, it is probably safe to conclude that the bxdl.5 PRE uses both mechanisms to silence Ubx gene expression. 124 IV. Final Thoughts A recurring theme in the analysis of complex cis-regulatory elements is the idea of functional interactions between the different modules that make up the element. One of the best-characterized cis-regulatory elements belongs to the endol6 gene of the sea urchin Stronglyocentrotus purpuratos (Yuh et al., 2001). Early in development the endol6 gene participates in the specification events that define the endomesoderm; later in development it functions as a gut-specific differentiation gene. Extensive research of this gene's cis-regulatory element has led to the generation of complex computational logic models that were used to explain this switch. Logic considerations predicted that developmentally controlled functional interactions between two modules, named Module A and B, mediate this switch in endol6 function. Indeed, this prediction was confirmed experimentally and a distinct set of functional interactions between the modules that mediate the switch function was demonstrated. The endol6 computational model now provides a detailed explanation of the information processing functions executed by the cis-regulatory system of this gene throughout embryogenesis. One of the greatest undertakings in PcG biology will be to reach the stage where we can devise computational models to explain how PREs decipher the regulatory inputs and generate novel regulatory outputs. Useful models for PRE function must be capable of explaining the temporal and spatial dynamics of cis-regulatory response as input and output data become available. The functional dissection of the bxd PRE presented in this thesis forms a foundation upon which we can build on to achieve this goal. 125 Chapter 5 Materials and Methods Preparation of D N A , restriction enzyme digestion, end-repair, D N A ligation, bacterial transformation, agarose gel electrophoresis and Southern blotting were performed according to standard procedures (Sambrook et al., 1989). Enzymes were purchased from New England Biolabs or Boehringer Mannheim. Gel elution (Qiaex gel extraction kit, Qiagen) and large scale D N A preparations (Maxi-prep kit, Qiagen) were performed according to the manufacturer's directions. A. Plasmids and P-element transformation vectors The Bithorax complex genomic subclone 3101 containing the 11 kb BamHl bithoraxoid fragment of phage L2212 (Bender et al., 1983) was digested with BamHl and Hindlll to isolate the bxd5.l fragment. The bxd5.l fragment was subcloned into pBluescript K S + (pBS(KS+)} to generate pBS-3103HB. Deletion derivatives of pBS-3101HB or restriction fragments isolated from pBS-3103HB were end-repaired and subcloned into the appropriate transformation vectors. The transformation vector used for the Chapter 2 mutational analysis was a modified Cahsneo-ZacZ, a Carnegie 4-derived plasmid (pCaSpeR4). It was modified by subcloning the Ultrabithorax basal promoter upstream of the bacterial fi-galactosidase gene to generate a Ubx-lacZ fusion gene (Qian et al., 1991) and subsequently named pCaSpeR4-Ubx/lacZ (Figure 5-1). The Ubx basal promoter is a 1.65kb Stul fragment including 680 bp of 5' upstream sequence and the entire 968bp untranslated leader plus the first 7 codons. Translation of the Ubx/lacZ gene begins with the first 7 codons of Ubx fused in-frame to the lacZ coding sequence. The transformation marker present on this vector is the mim-white gene which contains the minimal eye enhancer and promoter fused to the white cDNA (Pirrotta et al., 1985). The m\m-white gene encodes a membrane protein related to ATP-dependent transport proteins and is necessary for deposition of pigment in the eye (Dreesen et al., 1988). The eye pigment produced by the mini-white gene ranges from pale yellow to strong red, is affected by chromosomal insertion site and is sensitive to the long-range effects of Polycomb Response elements within the same transposon (reviewed by Pirrotta and Rastelli, 1994). 126 Ubx Figure 5-1. Schematic representation of the transformation plasmid pCaSper4-Ubx/lacZ. Wildtype bxd5.\ and all deletion derivatives thereof were subcloned upstream of the Ubx promoter//acZ fusion gene in the Xhol site. The Drosophila mini-white gene is used as a transformation marker and as an indicator for pairing-sensitive repression. The P-element terminal ends required for transposition are labelled 5' P and 3' P. The direction of transcription from the Ubx promoter//acZ fusion gene and from the mini-white gene is indicated by the arrow. The transcription start site of these genes is indicated by +1. 127 The transformation vector containing the gypsy chromatin boundary elements used for Chapter 3 was obtained from John Burr (O'Connor Laboratory) and was subsequently modified. The original vector was named yellowlSu(Hw) bxdl RR83 and is pUC derived. The role of chromatin boundary elements has been discussed in Chapter 3. The multiple cloning site (MCS) from pCaSpeR4 was subcloned into the unique Spel site that separates the yellow and mini-white genes. A 5.8kb fragment containing the entire Ubx/lacZ fusion gene was also isolated from pCaSpeR4 and subcloned into the repaired Xhol site of the MCS generating P{Su(Hw) yellow* Su(Hw) Ubx/lacZ white* Su(Hw)} or P{y+ Ubx/lacZ w+} for short (Figure 5-2). The,yellow gene (y) included in this vector is involved in pattern-specific melanin pigmentation of the cuticle of the Drosophila adult (Biessmann, 1985) and is used as an additional transformation marker. This yellow gene is 5.2kb in length and contains the body colour enhancer element that directs yellow gene expression in the cuticle of the adult fly (Geyer and Corces, 1987). B . Polymerase Chain Reaction (PCR) For all PCR reactions the following standard conditions were used: lu,g of plasmid template, 0.5ul lOmg/ml acetylated BSA, 5ul 10X Thermopol buffer (NEB), 0.7ul 2.5mM dNTPs, l u l lOOmM MgS04, l u M of each primer, l u l (2 units) Vent polymerase (NEB) and H 2 0 to make 50ul final volume, overlayed with 50uf mineral oil. The typical temperature cycles were: 5 minutes at 96°C, followed by 7X (1 minute at 96°C; 1 minute at 52°C; 1 minute at 72°C). A l l PCR reactions were performed on a D N A Thermal Cycler 480 (Perkin-Elmer). A low number of cycles with a large amount of template and an error-correcting D N A polymerase (Taq polymerase) were used to minimize the chance of PCR-induced mutagenesis. The PCR products were gel-eluted and subsequently purified by phenol/CHCl3 extraction followed by ethanol precipitation. After subcloning, all PCR generated fragments were sequenced on an automated sequencer (NAPS unit, UBC) using fluorescent dye termination. C. D N A manipulations The fragment deleted in each of the constructs described below was the smallest fragment shown to be sufficient for complex formation in vitro. 128 Spel Notl Figure 5-2. Schematic representation of the transformation plasmid P{y+ Ubx/lacZ w }. Fragments of the M-element were subcloned in the Notl site upstream of the Ubx promoter//acZ fusion gene. The mini-white gene is used as a transformation marker and as an indicator for pairing-sensitive repression. The yellow gene is used as an additional transformation marker. The relative positions of the Su(Hw) binding sites are also shown. 129 i . Generation of fca/5.1ASl-HB90 A flowchart for the generation of bxd5.l AS 1-HB90 is shown in Figure 5-3. The PCR template used to generate fcraf5.1ASl-HB90 was pBS-3103HB. The PCR primers used were: AS 1 -HB90-p 1,5' - C C G G A A A C G T A A T A C C A T G G G A A G C A A G C T ASl-HB90-p2, 5 ' - C C G G G A A T T C C A A A A T T G C A A T G A A A A A G C ASl-HB90-p3, 5 ' - C C G G T A A A A T T C G T C G C A G C T T T T C A T T T G G AS 1 -HB90-p4, 5 ' - C C G G G G A T C C C A C G A G C C C G T G A C T A A C T T T The 80bp amplified product from primer pair ASl-HB90-pl and ASl-HB90-p2 was isolated after gel electrophoresis, digested with Kpnl and EcoRl (bolded sequence) and subcloned into pBS, generating pBS-KE. The 600bp isolated amplified product from primer pair ASl-HB90-p3 and ASl-HB90-p4 was digested with EcoRl and BamHl (underlined sequence) and subcloned into pBS-KE, generating pBS-KB. A 639bp KpnllStyl fragment containing AS1-HB90 was excised from pBS-KB and subcloned into pBS-3101HB which had been liberated of the KpnllStyl fragment. The final construct pBS 6xJ5.1ASl-HB90 was sequenced to verify the fidelity of the PCR reactions. This deletion construct was excised from pBS with Hindlll and BamHl, end-repaired and subcloned into the repaired Xhol site upstream of the Ubx/lacZ fusion gene. i i . Generation of bxd5.1 AMHS-70 A flowchart for the generation of bxd5.lAMHS-70 is shown in Figure 5-4. The PCR template used to generate bxd5.1 AMHS-70 was pBS-3103HB. The primers used were: AMHS70-p 1, 5 ' -CCGGCTCGAGCCTGTTGCCTTGGCGGCTCT; AMHS70-p2, 5 ' - C C G G | G A A T T C | G C T A G C C A T A C G C A C G G C T G T T A G A A ; AMHS70-p3, 5' - C C G G G C T A G C C A A G C G A G AGCTTTTC A T A G ; AMHS70-p4, 5 ' - C C G G G C G G C C G C G A A G C C A T A A C G G C A G A A C C The isolated 200bp amplified product from primer pair AMHS70-pl and AMHS70-p2 was digested with Xhol (underlined sequence) and EcoRl (boxed sequence) and subcloned into pBS, generating pBS-XR. The isolated 450bp amplified product from primer pair AMHS70-p3 and AMHS70-p4 was digested with Nhel (bolded sequence) and Notl (double underlined sequence) and subcloned into pBS-XR, generating pBS-XN. A 600 bp Bgll fragment containing AMHS-70 recovered from pBS-XN was subcloned into a pBS-3103HB vector with the 571bp Bgll 130 \ Figure 5-3. Construction of pBS -W5 .1ASlHB-90. The S1HB-90 fragment within bxd5.1 is shown as a dark box (top). See text for details. 131 Figure 5-4. Construction of pBS-fcw/5.1 AMHS-70. The MHS-70 fragment within bxd5.l is shown as a dark box (top). See text for details. 132 fragment removed. The final construct, pBS-focc?5.1AMHS70 was sequenced to confirm orientation and the fidelity of the PCR reactions and subsequently cloned into the pCaSpeR4 Ubx/lacZ transformation vector. i i i . Generation of 6xrf5.1AMPA-168 A flowchart for the generation of focci5.1AMPA-168 is shown in Figure 5-5. A 571bp BgU fragment from the M region was subcloned into pBS, generating pBS-BB. pBS-BB was digested with Pstl and Asel, releasing MPA-168, end-repaired using Klenow enzyme (NEB) and re-ligated. The mutated pBS-BB, named pBS-BBmut, was then digested with BgK releasing a 403bp fragment that was subcloned into a pBS-3101HB vector with the 571bp BgU fragment removed. iv. Generation of fctrf5.1AMHN-90 A flowchart showing the generation of 6xc75.1AMHN-90 is shown in Figure 5-6. A 412bp Ndel/EcoRl fragment spanning MFIN-90 was excised from pBS-3103HB generating pBS-3101-NE. The PCR template used to generate W5 .1AMHN-90 was pBS-3103BH and the primers used were; AMHN90-pl , 5' - C C G G C A T A T G G C T A G C C A G T T C T T T T T T T A T C T T A A - 3 ' and AMHN90-p2, 5' -CCGGTTCGGTTTTATGCTGCCCGC-3 ' Primer 1 contains a Ndel restriction site (underlined sequence) used for subcloning, and a Nhel restriction site (bolded sequence) which was used as an insertion site for the in vivo substitution mutant analysis. The 350bp amplified product was isolated, digested with Ndel and iscoRl and subsequently subcloned into pBS-3101-NE generating pBS-6jc^5.1AMHN-90. The final construct was sequenced to confirm the fidelity of the PCR reaction. The wild-type bxd5A and mutated derivatives thereof were subcloned after end-repair into the filled in Xhol site upstream of the Ubx/lacZ fusion gene in pCaSpeR4 (Pirrotta, 1988). v. Generation of 6xrf5.1ASl-HB90+AMHN-90 A 2.2kb Pstl/Hindlll fragment was released from pBS-focd5.1ASl-HB90 and subcloned into pBS, generating pBS-PH. A 2.8kb BamHl/Pstl fragment released from pBS-tW5.1AMHN-90 was subcloned into pBS-PH, generating pBS-fotc?5.1 AS1-HB90+ A M H N -90. 133 Bgll Styl EcoRl 2. Isolate Bgll fragment and subclone into pBS Pstl A s e l Hindll BamH 1. Partial Bgll digestion 3. Partial Digestion with Pstl & Asel. End-repair and re-ligate EcoRl 4. Recover Bgll fragment and ligate Hindll lBamU Figure 5-5. Construction of pBS-fera?5.1AMPA-168. The MPA-168 fragment within bxd5.1 is shown as a dark box (top). See text for details. 134 PI Figure 5-6. Construction of pBS-W5.1AMHN-90. The MHN-90 fragment within bxdSA is shown as a dark box (top). See text for details. v i . Generat ion of fctrf5.1AMHS-70+AMPA-168 pBS-XN, an intermediate product from the generation of bxd5.lAMHS-70, was digested with Pstl and Asel, releasing MPA-168, end-repaired and re-ligated. The double mutated Bgll fragment was subcloned into a pBS-3103HB vector with the 571bp Bgll fragment removed. v i i . Generat ion of bxdS.\-LS\l9 The mutant oligomer MHS70-LS1/9 was end-repaired and subcloned into the newly generated Nhel restriction site of pBS-foc^5.1 AMHS-70. The construct was sequenced to confirm the orientation of the mutant oligomer. v i i . Generat ion of U T D The 354bp EcoRI/Ndel fragment corresponding to UPS was end-repaired and subcloned into the Hincll site of pBS generating pBS-UPS. A 620bp fragment isolated from the coding region of the bacterial tetracycline gene was end-repaired and subcloned into the EcoRN site of pBS-UPS to generate pBS-UT. The 485bp PstllStyl fragment corresponding to DPS was end-repaired and subcloned into the Smal site of pBS-UT, generating pBS-UTD. UTD was released from this vector and subsequently subcloned into the P{y+ Ubx/lacZ w+} transformation vector. E. Drosophi la E m b r y o Transformat ion Germline transformation was performed essentially as described by Rubin and Spradling (1982). Flies deficient for the allele(s) that marked the transformants (whitelu8 or yellow1 whitexm double mutants depending on the transformation vector used) were the host flies for P-element transformation. Egg laying plates were 2% technical agar, 5% table sugar and 1% balsamic vinegar. The injection solution was prepared with a 3:1 molar ratio of transformation vector to p:t25.7wc helper plasmid (Karess and Rubin, 1984) with a total concentration of 0.6 mg/ml. Plasmid D N A used for injections was prepared by Qiagen Maxi large scale preparations according to the manufacturer's directions. After mixing the D N A in its proper amounts, it was ethanol precipitated, spun and washed with cold 70% ethanol. The D N A was subsequently dessicated and dissolved in injection buffer solution, I X PBS (see Sambrook et al, 1989), 1% glycerol in ddH20 to a final volume of 20ul This solution was centrifuged for an additional 10 minutes and the supernatant was recovered into a fresh tube. The microinjection needles were made from borosilicate glass capillary tubes (World Precision Instruments, Inc.) with an outer diameter of 1mm and an inner diameter of 0.65mm. 136 The capillary tubes were pulled with a Model P-80 Brown-Flaming micropipette puller (Sutter Instrument Company) to make needles optimal for Drosophila transformations. Prior to injection, the blunt tip of the injection needle was broken by touching the tip of the needle to the edge of a slide under the microscope to produce a sharp point. In order to introduce the D N A before cellularization, the entire process from embryo collection to injection spanned 30 minutes. After collection, the chorion was removed by immersing the embryos in a freshly made 3% sodium hypochlorite solution (Chlorox bleach) for 1 minute. The embryos were collected in a strainer with a fine mesh, and washed extensively in ddF^O. The embryos were delicately transferred from the mesh onto a clean glass slide. Using a dissecting needle, the embryos were transferred from the glass slide onto a coverslip that had a piece of non-toxic double-stick tape (1/4" 3M, type 415 double-stick tape) along the entire length of one side. The embryos were dessicated with a standard hair blow-dryer for 5 to 10 minutes to facilitate microinjection. Embryo injections were performed at 18°C using the Eppendorf microinjector 5242 apparatus attached to a Leitz Laborlux 12 microscope. After injection, the embryos were covered in a thin layer of Halocarbon Oil Series 700 (Ffalocarbon Products Corp.) to prevent them from desiccating during development. Injected embryos were placed in a humidified environment at 18°C. Larvae began hatching after two days and were collected at least twice a day using a dissection needle. Approximately 50 larvae were placed in each vial of fly food. The following crossing scheme was used to screen for transformants: (Go) Surviving adult X ^ o r 1 Screen for rescue of w 1 1 1 8 or y ' w " 1 8 (Fi) After hatching, Go flies were crossed with the original marker stock (w 1 1 1 8 or ylwlus). A single Go was placed in a vial with four marker flies of the opposite sex. Depending on the transformation vector used, transformants were detected by rescue of the white eye colour phenotype or the rescue of both white' and the yellow' body colour phenotype. Transformants were subsequently amplified and crossed to the appropriate stocks to generate balanced stocks and to determine the chromosomal location of each transgenic line. The transformed lines were 137 examined by Southern blot hybridization to ensure transgene integrity and to determine transgene copy number. In a number of transgenic lines, the insertion site was determined by in situ hybridization to larval salivary gland chromosomes. F. In situ localization of transposons on polytene chromosomes In situ localization of transposons on polytene chromosomes was performed as described in Zink and Paro (1995). Polytene chromosome spreads were prepared from salivary glands of third instar larvae which were grown at 18°C in non-crowded conditions. Salivary glands were dissected in 45% acetic acid. For fixation, a pair of salivary glands was transferred into a drop of solution containing 3 parts acetic acid: 2 water: 1 lactic acid on a siliconized coverslip (18mm2). The coverslip was picked up with a clean slide and tapped a few times with the lead tip of a pencil. To ensure proper spreading of the chromosomes the slide was turned over onto blotting paper and firmly pressed with the thumb. The slides were incubated at 4°C overnight. The coverslip was removed with a razor blade after freezing in liquid nitrogen. The slides were immediately immersed in 95% ethanol for 10 minutes to dehydrate the chromosomes and air-dried. The bacterial lacZ gene, which is contained in all the transposons, was used as a probe for the in situ hybridization. The probe was labelled using the D N A Random Priming kit (Boehringer Mannheim) using ImM digoxygenin (DIG) labelled-dUTP and 500ng of D N A in a 20u.l reaction. After 1 hour, the probe D N A was ethanol precipitated, washed and desiccated. The probe was resuspended in hybridization buffer (50% deionized formamide, 2X SSC, 10% dextran sulfate, 100u,g/ml sheared salmon sperm DNA). The slides were heat treated in 2X SSC at 60°C for 30 minutes. Slides were then dipped in 2X SSC at room temperature for 2 minutes. Chromosomes were acetylated for 10 minutes in 0.1M triethanolamine in 0.125% acetic anhydride. Slides were washed 2X for 5 minutes each in 2X SSC and dehydrated 2X for 5 minutes each in 70% ethanol and 5 minutes in 95% ethanol, and air-dried. Chromosomes were denatured in 0.07N NaOH at room temperature for 3 minutes. The probe was denatured by heating to 75°C for 3 minutes and 15ul of this solution was applied to each chromosome preparation. Hybridization was performed overnight at 37°C with coverslips sealed with rubber cement. After hybridization, coverslips were removed and slides were washed for 3X for 20 minutes each in 2X SSC at 53°C followed by I X for 10 minutes in 2X SSC at room temperature. Slides were placed at room temperature I X for 10 minutes in Buffer 1 (0.1M Tris pH7.5; 0.1M NaCI; 2mM M g C l 2 ; 0.05 Triton X-100) followed by I X for 20 minutes in pre-warmed (42°C) 138 Buffer 2 (identical to Buffer 1 except made 2% in BSA). Slides were incubated with l u l of a-DIG in 1ml of Buffer 1 for 15 minutes at room temperature. Slides were washed 3X for 3 minutes each in Buffer 1 and 3X for 3 minutes each in Buffer 3 (0.1M Tris pH9; 0.1M NaCI; 50mM MgCh). The slides were incubated with the substrate solution containing 4.4u.l NBT and 3.3uf BCIP (DNA Random Priming kit) in 1ml Buffer 3 for 10 minutes. The reaction was stopped by rinsing the slides in ddFI^O and air-dried. The chromosomes were stained for 30 seconds with Giemsa (1:20 dilution in lOmM sodium phosphate buffer at pH6.8) and washed for 30 seconds in ddHaO and air-dried. Preparations were mounted in a drop of water and examined using phase contrast. G. P-element mobilization Where low numbers of transformed lines were obtained, additional lines were generated by P-element mobilization (Robertson et al., 1988). The source of transposase was provided by the stockyw; kiA2-3/TM3Sb (obtained from W. Gehring laboratory). The following crossing scheme was used: (Q\ rf w . P{W, bxd5.1 -Ubx/lacZ} y O O y w~ • kiA2-3 + A + + W'TMlSb 1 (p 2) ^ 1 yw P {w~, bxd5.1 - Ubx/lacZ} Select F 2 which have different eye colour than Go and have lost Ki In this example the original P-element has inserted on the second chromosome. Identical strategies were used for X and third chromosome insertions. Single males of each transgenic line were crossed withjw: KiA2-3/TM3Sb virgin females generating flies trans-heterozygous for the transgene and the transposase source (Fi). Transposition of the P-element occurs in the 139 germline of the Fi males. The transposase is out-crossed and the F2 male progeny are screened for differences in eye colour as compared to the Go males. H . Immunohistochemical Staining Developmentally staged embryos were dechorionated in a 3% sodium hypochlorite solution (Chlorox bleach) for 1 minute and fixed for 20 minutes at room temperature in 5 ml of heptane, 0.5 ml of 37% formaldehyde and 4.5 ml of 35mM sodium phosphate, 120mM sodium chloride (PBS), pH 7.4. The embryos were removed from the fix solution and devitellinized in a 1:1 heptane:methanol solution. The heptane methanol solution was removed and the embryos were washed 2X with methanol. The embryos were either stored in methanol at -20°C or immunostained directly. Immunohistochemical staining of Drosophila embryos was performed essentially as described in Patel (1994). Embryos were rehydrated by washing 3X for 20 minutes each in I X PBS, 0.1% Tween and 0.2% bovine serum albumin (BSA). Rehydrated embryos were blocked with 5% heat inactivated normal goat serum (Jackson Laboratories) in I X PBS, 0.1% Tween for 30 minutes and then incubated with either a polyclonal mouse anti-|3-galactosidase antibody (Jackson Laboratories) or a polyclonal mouse anti-UBX antibody at a dilution of 1:1000 and 1:100 respectively. Incubation with the primary antibodies was performed for 1 hour at room temperature followed by overnight at 4°C on a rotating shaker. After extensive washing in I X PBS, 0.1% Tween, 3X for 20 minutes each, the embryos were incubated for 90 minutes at room temperature with a rabbit anti-mouse biotinylated secondary antibody (Vector Laboratories) at a dilution of 1:1000. After extensive washing, 3X for 20 minutes each in I X PBS, 0.1% Tween, the biotinylated antibody complex was detected with avidin coupled to biotinylated horseradish peroxidase (ABC reaction; Vector Laboratories). Avidin and biotinylated horseradish peroxidase were prepared in a 1:1 ratio and diluted 1:100 in I X PBS, 0.1% Tween. Incubation with the A B C reagents was carried out at room temperature for 30 minutes. The embryos were washed 3X for 10 minutes each. The colour reaction was developed with 0.1 mg/ml diaminobenzidine tetrahydrochloride (DAB) in 50mM Tris, pH 7.2 and 0.01% hydrogen peroxide and was allowed to proceed for 30 minutes. The reaction was stopped by washing 3X for 5 minute each in I X PBS, 0.1% Tween. The embryos were mounted in 80% glycerol in 20mM Tris pH 8.0 for microscopy. 140 I. Scoring misexpression of Ubx/lacZ reporter in embryos Transgenic flies were identified on the basis of their white" or yellow+white+ phenotypes and established as homozygous or as balanced heterozygous stocks. For antibody staining, overnight embryo collections were used in order to ensure that all developmental stages were accounted for in the preparations. Typically, all developmental stages were represented in equal proportions. Embryos were staged according to the parameters described by Campos-Ortega and Ffartenstein (1985). Different stages were looked at to chart the progression from initial activation of the bxd-Ubx/lacZ transgene (2.5 to 3 hours after egg laying) to initiation of silencing (3-4 hours A E L ) and finally to maintenance of silencing (4-18 hours) or lack thereof. Homozygous control and experimental genotypes were simultaneously fixed and immunostained. This was to limit the variability often observed between homozygous and heterozygous embryos of the same genotype and to ensure that embryos of different genotypes were handled equivalently. Detection of LacZ in >80% of the examined embryos (n=150) in any parasegment anterior to PS 6 between 4 to 6 hours A E L constituted early misexpression or early loss of maintenance while detection in the anterior PS beyond 6 hours A E L constituted late misexpression. Slides were often scored blind to control for observer bias. To test the effects of PcG and trxG mutations on the expression from the bxd5.\-Ubx/lacZ fusion gene construct (Chapter 2), homozygous transgenic lines, which had the P-element insert on a different chromosome than the PcG/trxG gene being tested, were crossed to flies carrying the PcG/trxG allele. The majority of the alleles tested in this genetic assay altered the expression pattern in the trans-heterozygous state. This indicated that the bxd5A-Ubx/lacZ transgene is sensitive to the partial loss of the PcG and trxG proteins. Conversely, heterozygous mutations in PcG and trxG genes were found to have no effect on the LacZ expression pattern of heterozygous transgenic flies carrying minimal bxd fragments fused to Ubx/lacZ. Therefore, to obtain embryos homozygous for a PcG or a TrxG mutation and homozygous for a bxd fragment-Ubx/lacZ transposon (Chapter 3), the mutant strain was first crossed with transgenic flies carrying the transposon on a different chromosome and their non-balancer progeny were mated together to produce embryos suitable for antibody staining. An example using the ph allele phm is shown below: 141 fyr^ y'W _ P{y+, bxd- Ubx/lacZ, w+} v r " ) 0 phmw\ + ~^ ' P{y+, bxd-Ubx/lacZ, w+} X + + FM7C ' T 1 p/i 4 0V. P{y + , bxd-Ubx/lacZ, w+} ^ Q Q P ^ ^ " . P{y+, bxd-Ubx/lacZ, w+} ^ ' + • > yw' ' + 1 Collect and stain embryos Two problems exist with this crossing scheme. The first problem is that the homozygous mutants constitute 37% of the embryos generated (50% for phm/Y or ph409/phm X 75% for at least one copy of the reporter) and approximately 12.5% of the embryos are homozygous for transgene and homozygous or hemizygous for the ph allele (50% for ph409/Yov ph409/ph409 X 25% for two copies of the transgene). It is the latter class of embryos that exhibit derepression. This was resolved by examining many embryos, approximately 150 per slide, and noting misexpression phenotypes exhibiting Mendelian segregation. The second problem with using such a crossing scheme is that these homozygous PcG/trxG embryos lack any zygotic contribution but retain maternal product. For this reason, I cannot conclusively refute a role for those PcG/trxG alleles that did not alter a specific transgene's LacZ expression pattern. J. Scoring pairing-sensitive repression of the mini-white gene Flies were positive for pairing-sensitive repression when the eye colour of a fly homozygous for a particular transgenic line was either lighter or indistinguishable from the heterozygote (see Introduction for details). Many of the transgenic lines obtained displayed sectored or mottled pigmentation patterns. These lines were scored positive for pairing-sensitive repression i f expression of mini-white changed appropriately. The level of expression from the mini-white gene dictates the amount of pigmentation deposited in the fly eye. Fly eye colour also depends on the age and sex of the fly (Qian and Pirrotta, 1995) and the temperature at which they were grown. In addition, the genetic background of the fly, more specifically, the type of balancer chromosome used, affects the eye colour of flies containing bxd PRE/mini-w/nte transgenes. For these reasons, when scoring for 142 pairing-sensitive repression of the mini-white gene, the eye colour of flies with similar age and sex in the absence of any balancer chromosomes, where possible, were compared and scored. Virgins of the same sex, reared at the same temperature (25°C), were collected in the morning and aged for 48 hours. Adult flies to be compared were killed by freezing at -20°C for 1 hour, glued to a glass slide using Krazy Glue and photographed. K . Microscopy and Photography A l l embryos were photographed with a Zeiss Axiophot photomicroscope using differential interference contrast optics on Kodak Ektachrome 160T Tungsten film 160 ASA. The 35mm slides were scanned on a Kodak Professional RFS 2035 Plus film scanner at a resolution of 600 dpi. Adult eyes were photographed with a Spot digital camera (Diagnostic Instruments Inc.) attached to a Wild Heerbrugg dissecting microscope. A l l figures were assembled on Adobe Photoshop® Version 6.0. L . Drosophila culture Fly stocks were maintained at 25°C on standard yeast-cornmeal medium containing Tegosept as a mold inhibitor. A l l crosses were performed at 25°C. See Table 5-1 for stock details. A l l mutant alleles used are described in more detail in Lindsley and Zimm (1992). 143 Table 5-1. List and description of Drosophila strains used in this thesis. Gene Allele Description Reference polyhomeotic 409 ph phw ph2 Hypomorphic alleles affecting p h p r m i m M Temperature-sensitive hypomorph Durae/a/., 1987 Dura etal., 1987 Dura et al., 1985 Polycomb Pc Pc* XT 109 Null Null Jurgens, 1985 Franked a/., 1995 Posterior sex combs Psc1 Strong hypomorph Adlerefa/., 1989 Enhancer of zeste E(z)6 Null Carrington and Jones, 1996 pleiohomeotic pho Strong hypomorph Browne^/., 1998 Additional sex combs Asx Gain of function Sinclair et al, 1992 Sex combs on midleg Scm Null Breen and Duncan, 1986 trithorax trx trx Hypomorph Strong, recessive lethal with dominant phenotypes Gindhart and Kaufman, 1995 Kennison and Tamkun, 1988 trithorax-like 13C trl trf2 Hypomorph Null Farkase/a/., 1994 Farkas etal, 1994 pipsqueak psq psq2' „RF13 Null Null Horowitz and Berg, 1995 Horowitz and Berg, 1996 144 References Adler, P.N., Charlton, J. and Brunk, B. (1989). Genetic interactions of the Suppressor 2 of zeste region genes. Dev Genet 10, 249-260. Adler, P.N., Martin, E., Charlton, J. and Jones, K. (1991). Phenotypic consequences and genetic interactions of a null mutation in the Drosophila Posterior Sex Combs gene. Dev Genet 12, 349-361. Akam, M.E. (1983). The location of Ultrabithorax transcripts in Drosophila tissue sections. EMBO J 2, 2075-2084. Akam, M.E. , Martinez-Arias, A. , Weinzierl, R. and Wilde, C D . (1985). Function and expression of Ultrabithorax in the Drosophila embryo. Cold Spring Harb Symp Quant Biol 50, 195-200. Akam, M . (1987). The molecular basis for metameric pattern in the Drosophila embryo. Develop 101, 1-22. Akasaka, T., Kanno, M . , Balling, R., Mieza, M.A. , Taniguchi, M . and Koseki, H. (1996). A role for mel-18, a Polycomb group-related vertebrate gene, during the anteroposterior specification of the axial skeleton. Develop 122, 1513-1522. Alkema, M.J. , Wiegant, J., Raap, A . K . , Berns, A . and van Lohuizen, M . (1993). Characterization and chromosomal localization of the human proto-oncogene BMI-1. Hum Mol Genet 2, 1597-1603. Alkema, M.J. , Bronk, M . , Verhoeven, E., Otte, A. , van't Veer, L.J . , Berns, A. and van Lohuizen, M . (1997a). Identification of Bmil-interacting proteins as constituents of a multimeric mammalian Polycomb complex. Genes Dev 11, 226-240. Alkema, M.J. , Jacobs, J., Voncken, J.W., Jenkins, N.A. , Copeland, N.G. , Satijn, D.P., Otte, A.P., Berns, A. and van Lohuizen, M . (1997b). MPc2, a new murine homolog of the Drosophila polycomb protein is a member of the mouse polycomb transcriptional repressor complex. J Mol Biol 14, 993-1003. Americo, J., Whiteley, M . , Brown, J.L., Fujioka, M . , Jaynes, J.B. and Kassis, J.A. (2002). A complex array of DNA-binding proteins required for pairing-sensitive silencing by a Polycomb group response element from the Drosophila engrailed gene. Genetics 160, 1561-1571. Austen, M . , Luscher, B. and Luscher-Firzlaff, J .M. (1997). Characterization of the transcriptional regulator Y Y 1 . The bipartite transactivation domain is independent of interaction with the T A T A box-binding protein, transcription factor IIB, TAFII55, or cAMP-responsive element-binding protein (CBP)-binding protein. J Biol Chem 272, 1709-1717. Bajusz, I., Sipos, L. , Gyorgypal, Z., Carrington, E.A., Jones, R.S., Gausz, J. and Gyurkovics, H . (2001). The Trithorax-mimic allele of Enhancer of zeste renders active domains of target genes 145 accessible to Polycomb-group-dependent silencing in Drosophila melanogaster. Genetics 159, 1135-1150. Banerji, J., Rusconi, S. and Schaffher, W. (1981). Expression of a beta-globin gene is enhanced by remote SV40 D N A sequences. Cell 27, 299-308. Beachy, P.A., Helfand, S.L. and Hogness, D.S. (1985). Segmental distribution of bithorax complex proteins during Drosophila development. Nature 313, 545-551. Beachy, P.A. (1990). A molecular view of the Ultrabithorax homeotic gene of Drosophila. Trends Genet 6, 46-51. Bel, S., Core, N . , Djabali, M . , Kieboom, K., Van der Lugt, N . , Alkema, M.J. and Van Lohuizen, M . (1998). Genetic interactions and dosage effects of Polycomb group genes in mice. Develop 125,3543-3551. Bender, W., Akam, M . , Karch, F., Beachy, P.A., Peifer, M . , Spierer, P., Lewis, E.B. and Hogness, D.S. (1983). Molecular genetics of the Bithorax complex in Drosophila melanogaster. Science 221, 23-29. Bender, W., Weiffenbach, B., Karch, F. and Peifer, M . (1985). Domains of cis-interaction in the bithorax complex. Cold Spring Harb Symp Quant Biol 50, 173-180. Bienz, M . , Saari, G., Tremml, G., Muller, J., Zust, B. and Lawrence, P.A. (1988). Differential regulation of Ultrabithorax in two germ layers of Drosophila. Cell 53, 567-576. Bienz, M . and Tremml G. (1988). Domain of Ultrabithorax expression in Drosophila visceral mesoderm from autoregulation and exclusion. Nature 333, 576-578. Bienz, M . (1992). Molecular mechanisms of determination in Drosophila. Curr Opin Cell Biol 4, 955-961. Biggin, M.D. and Tjian, R. (1988). Transcription factors that activate the Ultrabithorax promoter in developmental^ staged extracts. Cell 53, 699-711. Birve, A. , Sengupta, A . K . , Beuchle, D., Larsson, J., Kennison, J.A., Rasmuson-Lestander, A . and Muller, J. (2001). Su(z)12, a novel Drosophila Polycomb group gene that is conserved in vertebrates and plants. Develop 128, 3371-3379. Boivin, A. , Fauvarque, M.O. and Dura, J.M. (1999). One-to-one correspondence between the two genetic units and the tandemly duplicated transcriptional units of the polyhomeotic locus of Drosophila. Mol Gen Genet 261, 196-200. Boivin, A . and Dura, J.M. (1998). In vivo chromatin accessibility correlates with gene silencing in Drosophila. Genetics 150, 1539-1549. Bornemann, D., Miller, E. and Simon J. (1996). The Drosophila Polycomb group gene Sex comb on midleg (Scm) encodes a zinc finger protein with similarity to Polyhomeotic protein. Develop 122, 1621-1630. 146 Boscheron, C , Maillet, L. , Marcand, S., Tsai-Pflugfelder, M . , Gasser, S.M. and Gilson, E. (1996). Cooperation at a distance between silencers and proto-silencers at the yeast H M L locus. EMBO J 15,2184-2195. Boumil, R . M . and Lee, J.T. (2001). Forty years of decoding the silence in X-chromosome inactivation. Hum Mol Genet 10, 2225-2232. Bownes, M . (1990). Preferential insertion of P elements into genes expressed in the germ-line of Drosophila melanogaster. Mol Gen Genet 222, 457-460. Brand, A . H . , Breeden, L. , Abraham, J., Sternglanz, R. and Nasmyth, K . (1985). Characterization of a "silencer" in yeast: a D N A sequence with properties opposite to those of a transcriptional enhancer. Cell 41, 41-48. Brannan, C.I. and Bartolomei, M.S. (1999). Mechanisms of genomic imprinting. Curr Opin Genet Dev 9, 164-170. Breen, T.R., and Duncan, I. (1986). Maternal expression of genes that regulate the Bithorax complex of Drosophila melanogaster. Dev Biol 118, 442-456. Breen, T.R. and Harte, P.J. (1991). Molecular characterization of the trithorax gene, a positive regulator of homeotic gene expression in Drosophila. Mech Dev 35. 113-127. Breiling, A. , Bonte, E., Ferrari, S., Becker, P.B., and Paro, R. (1999). The Drosophila Polycomb protein interacts with nucleosomal core particles in vitro via its repression domain. Mol Cell Biol 19, 8451-8460. Breiling, A. , Turner, B . M . , Bianchi, M.E. and Orlando, V . (2001). General transcription factors bind promoters repressed by Polycomb group proteins. Nature 412, 651-655. Broach, J.R. and Hicks, J.B. (1980). Replication and recombination functions associated with the yeast plasmid, 2 mu circle. Cell 21, 501-508. Brock, H.W., and van Lohuizen, M . (2001). The Polycomb group—no longer an exclusive club? Curr Opin Genet Dev 11, 175-181. Brown, J.L., Mucci, D., Whitely, M . , Dirksen, M. -L . , and Kassis, J.A. (1998). The Drosophila Polycomb group gene pleiohomeotic encodes a sequence-specific D N A binding protein with homology to the multifunctional mammalian transcription factor Y Y 1 . Mol Cell 1, 1057-1064. Brunk, B.P., Martin, E.C., and Adler, P. (1991). Drosophila genes Posterior Sex Combs and Suppressor two of zeste encode proteins with homology to the murine bmi-1 oncogene. Nature 353,351-353. Buchenau, P., Hodgson, J., Strutt, H . and Arndt-Jovin, D.J. (1998). The distribution of Polycomb-group proteins during cell division and development in Drosophila embryos: Impacts on models for silencing. J Cell Biol 141, 469-481. 147 Bunker, C A . and Kingston, R.E. (1994). Transcriptional repression by Drosophila and mammalian Polycomb group proteins in transfected mammalian cells. Mol Cell Biol 14, 1721-1732. Busturia, A. and Bienz, M . (1993). Silencers in Abdominal-B, a homeotic Drosophila gene. EMBO J 12, 1415-1425. Busturia, A. , Wightman, C.D. and Sakonju, S. (1997). A silencer is required for maintenance of transcriptional repression throughout Drosophila development. Develop 124, 4343-4350. Busturia, A. , Lloyd, A. , Bejarano, F., Zavortink, M . , Xin, H . and Sakonju, S. (2001). The MCP silencer of the Drosophila Abd-B gene requires both Pleiohomeotic and G A G A factor for the maintenance of repression. Develop 128, 2163-2173. Campbell, R.B., Sinclair, D.A., Couling, M . , and Brock, H.W. (1995). Genetic interactions and dosage effects of Polycomb group genes of Drosophila. Mol Gen Genet 246, 291-300. Campos-Ortega, J.A., and Hartenstein, V . (1985). The embryonic development of Drosophila melanogaster (Springer Verlag, Heidelberg). Capdevila, M.P. and Garcia-Bellido, A. (1981). Genes involved in the activation of the bithorax complex of Drosophila. Wilhelm Roux's Arch Dev Biol 190, 339-350. Carrington, E.A., and Jones, R.S. (1996). The Drosophila Enhancer of zeste gene encodes a chromosomal protein: examination of wild type and mutant protein distribution. Develop 122, 4073-4083. Carroll, S.B., Grenier, J.K., and Weatherbee, S.D. (2001). From D N A to diversity: Molecular genetics and the evolution of animal design (Blackwell Scientific, Oxford). Casanova, J., Sanchez-Herrero, E. and Morata, G. (1985). Prothoracic transformation and functional structure of the Ultrabithorax gene of Drosophila. Cell 42, 663-669. Casares, F. and Sanchez-Herrero, E. (1995). Regulation of the infraabdominal regions of the bithorax complex of Drosophila by gap genes. Develop 121, 1855-1866. Cavalli, G. and Paro, R. (1998). The Drosophila Fab-7 chromosomal element conveys epigenetic inheritance during mitosis and meiosis. Cell 93, 505-518. Cavalli, G. and Paro, R. (1999). Epigenetic inheritance of active chromatin after removal of the main transactivator. Science 286, 955-958. Chan, C , Rastelli, L . and Pirrotta, V . (1994). A Polycomb response element in the Ubx gene that determines an epigenetically inherited state of repression. EMBO J 13, 2553-2564. Chang, Y . L , Peng, Y . H . , Pan, I.C., Sun, D.S., King, B., Huang, D.H. (2001). Essential role of Drosophila Hdacl in homeotic gene silencing. PNAS 98, 9730-9735. 148 Chen, H., Rossier, C. and Antonarakis, S.E. (1996). Cloning of a human homolog of the Drosophila Enhancer of zeste gene (EZH2) that maps to chromosome 21q22.2. Genomics 15, 30-37. Cheng, N . N . , Sinclair, D.A. , Campbell, R.B. and Brock, H.W. (1994). Interactions of polyhomeotic with Polycomb group genes of Drosophila melanogaster. Genetics 138, 1151-1162. Cherbas, L. , Moss, R. and Cherbas, P. (1994). Transformation techniques for Drosophila cell lines. Meth Cell Biol 44, 161-179. Chiang, A. , O'Connor, M.A. , Paro, R., Simon, J. and Bender, W. (1995). Discrete Polycomb-binding sites in each parasegmental domain of the Bithorax complex. Develop 121, 1681-1689. Core, N . , Bel, S., Gaunt, S.J., Aurrand-Lions, M . , Pearce, J., Fisher, A. , and Djabali, J. Altered cellular proliferation and mesoderm patterning in Polycomb-M33-deficient mice. Develop 124, 721-729. Cote, J., Quinn, J., Workman, J.L. and Peterson, C L . (1994). Stimulation of G A L 4 derivative binding to nucleosomal D N A by the yeast SWI/SNF complex. Science 265, 53-60. Crosby, M.A. , Miller, C , Alon, T., Watson, K . L . , Verrijzer, C P . , Goldman-Levi, R. and Zak, N.B. (1999). The trithorax group gene moira encodes a Brahma-associated putative chromatin-remodeling factor in Drosophila melanogaster. Mol Cell Biol 19, 1159-1170. Deatrick, J., Daly, M . , Randsholt, N.B. and Brock HW. (1991). The complex genetic locus polyhomeotic in Drosophila melanogaster potentially encodes two homologous zinc-finger proteins. Gene 15, 185-195. DeCamillis, M . , Cheng, N.S., Pierre, D. and Brock, H.W. (1992). The polyhomeotic gene of Drosophila encodes a chromatin protein that shares polytene chromosome-binding sites with Polycomb. Genes Dev 6, 223-232. Dellino, G.I., Tatout, C. and Pirrotta, V . (2002). Extensive conservation of sequences and chromatin structures in the bxd Polycomb response element among Drosophilid species. Int J Dev Biol 46, 133-141. Dingwall, A . K . , Beek, S.J., McCallum, C M . , Tamkun, J.W., Kalpana, G.V., Goff, S.P., and Scott, M.P. (1995). The Drosophila snrl and brm proteins are related to yeast SWI/SNF proteins and are components of a large protein complex. Mol Biol Cell 6, 777-791. Dreesen, T.D., Johnson, D.H., Henikoff, S. (1988). The brown protein of Drosophila melanogaster is similar to the white protein and to components of active transport complexes. Mol Biol Cell 8, 5206-5215. Duncan, I. (1982) Poly comblike: a gene that appears to be required for the normal expression of the Bithorax and Antennapedia complexes oi Drosophila melanogaster. Genetics 102, 49-70. Duncan, I. (1987). The Bithorax complex. Ann Rev Genet 21, 285-319. 149 Dura, J.-M., Brock, H.W., and Santamaria, P. (1985). Polyhomeotic: A gene of Drosophila melanogaster required for correct expression of segmental identity. Mol Gen Genet 198, 213-220. Dura, J.M., Randsholt, N.B. , Deatrick, J., Erk, I., Santamaria, P., Freeman, J.D., Freeman, S.J., Weddell, D. and Brock, H.W. (1987). A complex genetic locus, polyhomeotic, is required for segmental specification and epidermal development mD. melanogaster. Cell 51, 829-839. Dura, J.M. and Ingham, P. (1988). Tissue- and stage-specific control of homeotic and segmentation gene expression in Drosophila embryos by the polyhomeotic gene. Develop 103, 733-741. Espinas, M.L . , Canudas, S., Fanti, L. , Pimpinelli, S., Casanova, J. and Azorin, F. (2000). The G A G A factor of Drosophila interacts with SAP18, a Sin3-associated polypeptide. EMBO Rep 1,253-259. Farkas, G., Gausz, J., Galloni, M . , Reuter, G., Gyurkovics, H . and Karch, F. (1994). The Trithorax-like gene encodes the Drosophila G A G A factor. Nature 371, 806-808. Fauvarque, M.-O. and Dura, J .-M. (1993). polyhomeotic regulatory sequences induce developmental regulator-dependent variegation and targeted P-element insertions in Drosophila. Genes Dev 7, 1508-1520. Fauvarque, M.O., Zuber, V . and Dura, J .M. (1995). Regulation of polyhomeotic transcription may involve local changes in chromatin activity in Drosophila. Mech Dev 52, 343-355. Fitzgerald, D.P. and Bender, W. (2001). Polycomb group repression reduces D N A accessibility. Mol Cell Biol 21, 6585-6597. Francis, N.J. and Kingston, R.E. (2001). Mechanisms of transcriptional memory. Nat Rev Mol Cell Biol 2, 409-421. Franke, A. , DeCamillis, M . , Zink, D., Cheng, N . , Brock, H.W. and Paro, R. (1992). Polycomb and Polyhomeotic are constituents of a multimeric protein complex in chromatin of Drosophila melanogaster. EMBO J 11, 2941-2950. Franke, A. , Messmer, S. and Paro, R. (1995). Mapping functional domains of the Polycomb protein of Drosophila melanogaster. Chromosome Res 3, 351-360. Frei, E., Bopp, D., Burri, M . , Baumgartner, S., Edstrom, J.E. and Noll , M . (1985). Isolation and structural analysis of the extra sex combs gene of Drosophila. Cold Spring Harb Symp Quant Biol 50, 127-134. Fritsch, C , Brown, J.L., Kassis, J.A. and Muller, J. (1999). The DNA-binding Polycomb group protein Pleiohomeotic mediates silencing of a Drosophila homeotic gene. Develop 126, 3905-3913. 150 Gasser, S.M. and Cockell, M . M . (2001). The molecular biology of the SIR proteins. Gene 279, 1-16. Gdula, D.A. and Corces, V . G . (1997). Characterization of functional domains of the su(Hw) protein that mediate the silencing effect of mod(mdg4) mutations. Genetics 145, 153-161. Gecz, J., Gaunt, S.J., Passage, E., Burton, R.D., Cudrey, C , Pearce, J.J. and Fontes, M . (1995). Assignment of a Polycomb-like chromobox gene (CBX2) to human chromosome 17q25. Genomics 26, 130-133. Gildea, J.J., Lopez, R. and Shearn, A. (2000). A screen for new trithorax group genes identified little imaginal discs, the Drosophila melanogaster homologue of human retinoblastoma binding protein 2. Genetics 156, 645-663. Gindhart, J.G. and Kaufman, T.C. (1995). Identification of Polycomb and trithorax group responsive elements in the regulatory region of the Drosophila homeotic gene Sex combs reduced. Genetics 139, 797-814. Girton, J.R. and Jeon, S.H. (1994). Novel embryonic and adult homeotic phenotypes are produced by pleiohomeotic mutations in Drosophila. Dev Biol 161, 393-407. Goodrich, J., Puangsomlee, P., Martin, M . , Long, D., Meyerowitz, E . M . and Coupland, G.A. (1997). Polycomb-group gene regulates homeotic gene expression in Arabidopsis. Nature 386, 44-51. Grewal, S.I.S. and Elgin, S.C.R. (2002). Heterochromatin: new possibilities for the inheritance of structure. Curr Opin Genet Dev 12, 178-187. Grossniklaus, U . , Vielle-Calzada, J..P, Hoeppner, M.A.and Gagliano, W.B. (1998). Maternal control of embryogenesis by M E D E A , a Polycomb group gene in Arabidopsis. Science 280, 446-450. Grunstein, M . , Hecht, A. , Fisher-Adams, G., Wan, J., Mann, R.K., Strahl-Bolsinger, S., Laroche, T. and Gasser, S. (1995). The regulation of euchromatin and heterochromatin by histones in yeast. J Cell Sci 19,29-36. Gunster, M.J. , Satijn, D.P., Hamer, K . M . , den Blaauwen, J.L., de Bruijn, D., Alkema, M.J., van Lohuizen, M . , van Driel, R. and Otte, A.P. (1997). Identification and characterization of interactions between the vertebrate Polycomb-group protein BMI1 and human homologs of Polyhomeotic. Mol Cell Biol 17, 2326-2335. Hagstrom, K. , Muller, M . and Schedl, P. (1997). A Polycomb and G A G A dependent silencer adjoins the Fab-7 boundary in the Drosophila Bithorax complex. Genetics 146, 1365-1380. Hashimoto, N . , Brock, H.W., Nomura, M . , Kyba, M . , Hodgson, J., Fujita, Y . , Takihara, Y. , Shimada, K. and Higashinakagawa, T. (1998). RAE28, BMI1 and M33 are members of heterogeneous multimeric mammalian Polycomb group complexes. B B R C 245, 356-365. 151 Haupt, Y . , Alexander, W.S., Barri, G., Klinken, S.P. and Adams, J .M. (1991). Title. Cell 65, 753-760. Hecht, A. , Laroche, T., Strahl-Bolsinger, S., Gasser, S.M. and Grunstein, M . (1995). Histone H3 and H4 N-terminal interact with SIR3 and SIR4 proteins: a molecular model for the formation of heterochromatin in yeast. Cell 80, 583-592. Hemenway, C.S., Halligan, B.W. and Levy, L.S. (1998). The Bmi-1 oncoprotein interacts with dinG and MPh2: the role of RING finger domains. Oncogene 16, 2541-2547. Henikoff, S. (2000). Heterochromatin formation in complex genomes. Biochim Biophys Acta 1470, 1-8. Hobert, O., Sures, I., Ciossek, T., Fuchs, M . and Ullrich, A . (1996). Isolation and developmental expression analysis of Enx-1, a novel mouse Polycomb group gene. Mech Dev 55, 171-184. Hodgson, J.W., Cheng, N.N. , Sinclair, D.A., Kyba, M . , Randsholt, N .B. and Brock, H.W. (1997). The polyhomeotic locus of Drosophila melanogaster is transcriptionally and post-transcriptionally regulated during embryogenesis. Mech Dev 66, 69-81. Hodgson, J.W., Argiropoulos, B. and Brock, H.W. (2001). Site-specific recognition of a 70-base-pair element containing d(GA)(n) repeats mediates bithoraxoid Polycomb group response element-dependent silencing. Mol Cell Biol 21, 4528-4543. Hogness, D.S., Lipshitz, H.D., Beachy, P.A., Peattie, D.A., Saint, R.B., Goldschmidt-Clermont, M . , Harte, P.J., Gavis, E.R. and Helfand, S.L. (1985). Regulation and products of the Ubx domain of the bithorax complex. Cold Spring Harb Symp Quant Biol 50, 181-194. Holdeman, R., Nehrt, S.and Strome, S. (1998). MES-2, a maternal protein essential for viability of the germline in Caenorhabditis elegans, is homologous to a Drosophila Polycomb group protein. Develop 125, 2457-2467. Horard, B., Tatout, C , Poux, S. and Pirrotta, V . (2000). Structure of a Polycomb response element and in vitro binding of Polycomb group complexes containing G A G A factor. Mol Cell Biol 20,3187-3197. Horowitz, H. and Berg, C A . (1996). The Drosophila pipsqueak gene encodes a nuclear BTB-domain-containing protein required early in oogenesis. Develop 122, 1859-1871. Hur, M.W., Laney, J.D., Jeon, S.H., A l i , J. and Biggin, M.D. (2002). Zeste maintains repression of Ubx transgenes: support for a new model of Polycomb repression. Develop 129, 1339-1343. Ingham, P.W. (1984). A gene that regulates the Bithorax complex differently in larval and adult cells of Drosophila. Cell 37, 815-823. Ingham, P.W. and Martinez-Arias, A . (1986). The correct activation of Antennapedia and bithorax complex genes requires the fushi tarazu gene. Nature 324, 592-597. 152 Jones, R.S., and Gelbart, W . M . (1990). Genetic analysis of the Enhancer of zeste locus and its role in gene regulation in Drosophila melanogaster. Genetics 126, 185-199. Jones, C.A., Ng, J., Peterson, A.J . , Morgan, K. , Simon, J., and Jones, R.S. (1998). The Drosophila esc and E(z) proteins are direct partners in Polycomb group-mediated repression. Mol Cell Biol 18, 2825-2834. Jorgenson, R.A. (1995). Cosuppression, flower color patterns, and metastable gene expression states. Science 268, 686-691. Jurgens, G. (1985). A group of genes controlling the spatial expression of the bithorax complex in Drosophila. Nature 316, 153-155. Kal, A.J. , Mahmoudi, T., Zak, N.B. , and Verrijzer, C P . (2000). The Drosophila Brahma complex is an essential coactivator for the trithorax group protein Zeste. Genes Dev 14, 1058-1071. Kapoun, A . M . , and Kaufman, T.C. (1995). Regulatory regions of the homeotic gene proboscipedia are sensitive to chromosomal pairing. Genetics 140, 643-658. Kassis, J.A., VanSickle, E.P. and Sensabaugh, S.M. (1991). A fragment of engrailed regulatory D N A can mediate transvection of the white gene in Drosophila. Genetics 128, 751-761. Kassis, J. (1994). Unusual properties of regulatory D N A from the Drosophila engrailed gene: Three "pairing-sensitive" sites within a 1.6kb region. Genetics 136, 1025-1038. Kassis, J. (2002). Pairing-sensitive silencing, Polycomb group response elements, and transposon homing in Drosophila. Adv Genet 46, 421-438. Kaufman, T . C , Seeger, M.A. , and Olsen, G. (1990). Molecular and genetic organization of the Antennapedia gene complex of Drosophila melanogaster. Adv Genet 27, 309-362. Keeney, S. and Kleckner, N . (1996). Communication between homologous chromosomes: genetic alterations at a nuclease-hypersensitive site can alter mitotic chromatin structure at that site both in cis and in trans. Genes Cells 1, 475-489. Kehle, J., Beuchle, D., Treuheit, S., Christen, B., Kennison, J.A., Bienz, M . and Muller, J. (1998). dMi-2, a hunchback-interacting protein that functions in Polycomb repression. Science 282, 1897-1900. Kellum, R. and Schedl, P. (1991). A position-effect assay for boundaries of higher order chromosomal domains. Cell 64, 941-950 Kennison, J.A., and Tamkun, J.W. (1988). Dosage-dependent modifiers of Polycomb and Antennapedia mutations in Drosophila. PNAS 85, 8136-8140. Kennison, J.A. and Southworth, J.W. (2002). Transvection in Drosophila. Adv Genet 46, 399-420. 153 Kim, C.A., Gingery, M . , Pilpa, R . M . and Bowie, J.U. (2002). The S A M domain of Polyhomeotic forms a helical polymer. Nat Struct Biol Published online: 6 May 2002. Krumlauf, R. (1994). Hox genes in vertebrate development. Cell 78, 191-201. Kyba, M . and Brock, H.W. (1998a). The S A M domain of Polyhomeotic, Rae28, and Scm mediates specific interactions through conserved residues. Dev Genet 22, 74-84. Kyba, M . and Brock, H.W. (1998b). The Drosophila Polycomb group protein Psc contacts Ph and Pc through specific conserved domains. M o l Cell Biol 18, 2712-2720. Landecker, H.L., Sinclair, D.A. and Brock, H.W. (1994). Screen for enhancers of Polycomb and Polycomblike in Drosophila melanogaster. Dev Genet 15, 425-434. Laney, J.D. and Biggin, M.D. (1997). Zeste-mediated activation by an enhancer is independent of cooperative D N A binding in vivo. P N A S 94, 3602-3604. Lehmann, M . , Siegmund, T., Lintermann, K . G . and Korge G. (1998). The Pipsqueak protein of Drosophila melanogaster binds to G A G A sequences through a novel DNA-binding domain. J Biol Chem 273, 28504-28509. Lewis, E. B. (1978). A gene complex controlling segmentation in Drosophila. Nature 276', 565-570. Lindsley, D. and Zimm, G. (1992). The genome of Drosophila melanogaster. Academic Press, San Diego, CA. Lloyd, A. and Sakonju, S. (1991). Characterization of two Drosophila POU domain genes, related to oct-1 and oct-2, and the regulation of their expression patterns. Mech Dev 36, 87-102. Lobell, R.B., and Schleif, R.F. (1990). D N A looping and unlooping by AraC protein. Science 250, 528-532. Locke, J., Kotarski, M.A . , Tartof, K .D. (1988). Dosage-dependent modifiers of position effect variegation in Drosophila and a mass action model that explains their effect. Genetics 120, 181-198. Lonie, A. , D'Andrea, R., Paro, R. and Saint R. (1994). Molecular characterisation of the Polycomblike gene of Drosophila melanogaster, a trans-acting negative regulator of homeotic gene expression. Develop 120, 2629-2636. Loo, S. and Rine, J. (1995). Silencing and heritable domains of gene expression. A n n u Rev Cell Dev Biol 11, 519-548. Lupo, R., Breiling, A. , Bianchi, M.E. and Orlando, V . (2001). Drosophila chromosome condensation proteins Topoisomerase II and Barren colocalize with Polycomb and maintain Fab-7 PRE silencing. M o l Cel l 7, 127-136. 154 Mallin, D.R., Myung, J.S., Patton, J.S. and Geyer, P.K. (1998). Polycomb group repression is blocked by the Drosophila suppressor of Hairy-wing [su(Hw)] insulator. Genetics 148, 331-339. Martin, E.C. and Adler, P.N. (1993). The Polycomb group gene Posterior sex combs encodes a chromosomal protein. Develop 117,641-655. Martinez-Arias, A . and Lawrence, P. (1985). Parasegments and compartments in the Drosophila embryo. Nature 313, 639-642. McCall, K. and Bender, W. (1996). Probes of chromatin accessibility in the Drosophila bithorax complex respond differently to Polycomb-mediated repression. EMBO J 15, 569-580. McKeon, J. and Brock, H.W. (1991). Interactions of Polycomb group of genes with homeotic loci of Drosophila. Roux's Arch Dev Biol 199, 387-399. McKeon, J., Slade, E., Sinclair, D.A., Cheng, N . , Couling, M . and Brock, H.W. (1994). Mutations in some Polycomb group genes of Drosophila interfere with regulation of segmentation genes. Mol Gen Genet 244, 474-483. McKnight, S. and Tjian, R. (1986). Transcriptional selectivity of viral genes in mammalian cells. Cell 46, 795-805. Micol, J.L., Castelli-Gair, J.E. and Garcia-Bellido, A . (1990). Genetic analysis of transvection effects involving cis-regulatory elements of the Drosophila Ultrabithorax gene. Genetics 126, 365-373. Mihaly, J., Mishra, R.K. and Karch, F. (1998). A conserved sequence motif in Polycomb-response elements. Mol Cell 1, 1065-1066. Mishra, R.K., Mihaly, J., Barges, S., Spierer, A . , Karch, F., Hagstrom, K. , Schweinsberg, S.E. and Schedl, P. (2001). The iab-7 Polycomb response element maps to a nucleosome-free region of chromatin and requires both G A G A and Pleiohomeotic for silencing activity. Mol Cell Biol 21, 1311-1318. Moazed, D. and O'Farrell, P.H. (1992). Maintenance of the engrailed expression pattern by Polycomb group genes in Drosophila. Develop 116, 805-810. Morata, G. and Kerridge, S. (1981). Sequential functions of the bithorax complex of Drosophila. Nature 290, 778-781. Moretti, P. and Shore, D. (2001). Multiple interactions in Sir protein recruitment by Raplp at silencers and telomeres in yeast. Mol Cell Biol 21, 8082-8094. Muller, J. and Bienz, M . (1991). Long range repression conferring boundaries of Ultrabithorax expression in the Drosophila embryo. EMBO J 10, 3147-3155. Muller, J., Gaunt, S. and Lawrence, P.A. (1995). Function of the Polycomb protein is conserved in mice and flies. Develop 121, 2847-2852. 155 Muller, J. (1995). Transcriptional silencing by the Polycomb protein in Drosophila embryos. E M B O J 14, 1209-1220. Muller, M . , Hagstrom, K. , Gyurkovics, H. , Pirrotta, V . and Schedl, P. (1999). The mcp element from the Drosophila melanogaster bithorax complex mediates long-distance regulatory interactions. Genetics, 153, 1333-1356. Ng, J., Hart, C .M. , Morgan, K. , Simon, J.A. (2000). A Drosophila ESC-E(Z) protein complex is distinct from other Polycomb group complexes and contains covalently modified ESC. M o l C e l l B i o l 20, 3069-3078. Nomura, M . , Takihara, Y . and Shimada, K. (1994). Isolation and characterization of retinoic acid-inducible cDNA clones in F9 cells: one of the early inducible clones encodes a novel protein sharing several highly homologous regions with a Drosophila Polyhomeotic protein. Differentiation 57, 39-50. Nusslein-Volhard, C. and Wieshaus, E. (1980). Mutations affecting segment number and polarity in Drosophila. Nature 287, 795-801. Nusslein-Volhard, C , Frohnhofer, H.G. and Lehmann, R. (1987). Determination of anteroposterior polarity in Drosophila. Science 238, 1675-1681. O'Connor, M.B. , Binari, R., Perkins, L . A . and Bender, W. (1988). Alternative RNA products from the Ultrabithorax domain of the bithorax complex. E M B O J 7, 435-445. O'Kane, C.J. and Gehring, W.J. (1987). Detection in situ of genomic regulatory elements in Drosophila. P N A S 84, 9123-9127. Ohad, N . , Yadegari, R., Margossian, L. , Harmon, M . , Michaeli, D., Harada, J.J., Goldberg, R.B. and Fischer, R.L. (1999). Mutations in FIE, a WD Polycomb group gene, allow endosperm development without fertilization. Plant C e l l 11, 407-416. Okada, M . and Hirose, S. (1998). Chromatin remodeling mediated by Drosophila G A G A factor and ISWI activates fushi tarazu gene transcription in vitro. M o l C e l l B i o l 18, 2455-2461. Orlando, V . and Paro, R. (1993). Mapping Polycomb-repressed domains in the Bithorax complex using in vivo formaldehyde cross-linked chromatin. C e l l 75, 1187-1198. Orlando, V . and Paro, R. (1995). Chromatin multiprotein complexes involved in the maintenance of transcription patterns. C u r r O p i n Genet Dev 5, 174-179. Orlando, V. , Jane, E.P., Chinwalla, V . , Harte, P.J. and Paro, R. (1998). Binding of Trithorax and Polycomb proteins to the bithorax complex: dynamic changes during early Drosophila embryogenesis. E M B O J 17, 5141-5150. Pal-Bhadra, M . , Bhadra, U . , and Birchler, J.A. (2002). R N A i related mechanisms affect both transcriptional and posttranscriptional transgene silencing in Drosophila. M o l C e l l 9, 315-327. 156 Palladino, F., Laroche, T., Gilson, E., Axelrod, A. , Pillus, L. , and Gasser, S.M. (1993). SIR3 and SIR4 proteins are required for the positioning and integrity of yeast telomeres. Cell 75, 543-555. Papoulas, O., Beek, S.J., Moseley, S.L., McCallum, C M . , Sarte, M . , Shearn, A. , and Tamkun, J.W. (1998). The Drosophila trithorax group proteins B R M , ASH1 and ASH2 are subunits of distinct protein complexes. Develop 125,3955-3966. Paro, R., and Hogness, D.S. (1991). The Polycomb protein shares a homologous domain with the heterochromatin-associated protein in Drosophila. PNAS 88, 263-267. Paro, R. (1993). Mechanisms of heritable gene expression during development of Drosophila. Curr Opin Cell Biol 5, 999-1005. Patel, N . H . (1994). Imaging neuronal subsets and other cell types in whole-mount Drosophila embryos and larvae using antibody probes. Methods Cell Biol 44, 445-487. Pearce, J.J.H., Singh, P.B., and Gaunt, S.J. (1992). The mouse has a Polycomb-like chromobox gene. Develop 114, 921-929. Peifer, M . and Bender, W. (1986). The anterobithorax and bithorax mutations of the bithorax complex. EMBO J 5, 2293-2303. Pelegri, F. and Lehmann, R.A. (1994). A role of Polycomb group genes in the regulation of gap gene expression in Drosophila. Genetics 136, 1341-1353. Peterson, A.J. , Kyba, M . , Bornemann, D., Morgan, K. , Brock, H.W. and Simon, J. (1997). A domain shared by the Polycomb group proteins Scm and Ph mediates heterotypic and homotypic interactions. Mol Cell Biol 17, 6683-6692. Petruk, S., Sedkov, Y . , Smith, S., Tillib, S., Kraevski, V. , Nakamura, T., Canaani, E., Croce, C M . and Mazo, A. (2001). Trithorax and dCBP acting in a complex to maintain expression of a homeotic gene. Science 294, 1331-1334. Phillips, M.D. and Shearn, A. (1990). Mutations inpolycombeotic, a Drosophila Polycomb-group gene, cause a wide range of maternal and zygotic phenotypes. Genetics 125, 91-101. Phillips, K. and Luisi, B. (2000). The virtuoso of versatility: POU proteins that flex to fit. J Mol Biol 302, 1023-1039. Pirrotta, V . , Steller, H . and Bozzetti, M.P. (1985). Multiple upstream regulatory elements control the expression of the Drosophila white gene. EMBO J 16, 3501-3508. Pirrotta, V . and Rastelli, L . (1994). White gene expression, repressive chromatin domains and homeotic gene regulation in Drosophila. Bioessays 16, 549-556 Pirrotta, V . (1995). Chromatin complexes regulating gene expression in Drosophila. Curr Opin Gen Dev 5, 466-472. 157 Pirrotta, V . (1997). PcG complexes and chromatin silencing. Curr Opin Genet Dev 7, 249-258. Pirrotta, V . (1998). Polycombing the genome: PcG, trxG, and chromatin silencing. Cell 93, 333-336. Pirrotta, V . (1999). Transvection and chromosomal trans-interaction effects. Biochim Biophys Acta 1424, M l - 8 . Poux, S., Kostic, C. and Pirrotta, V. (1996). Hunchback-independent silencing of late Ubx enhancers by a Polycomb group response elements. EMBO J 15, 4713-4722. Poux, S., McCabe, D. and Pirrotta, V . (2001a). Recruitment of components of Polycomb Group chromatin complexes in Drosophila. Develop 128, 75-85. Poux, S., Melfi, R. and Pirrotta, V . (2001b). Establishment of Polycomb silencing requires a transient interaction between PC and ESC. Genes Dev 15, 2509-2514. Poux, S., Horard, B. , Sigrist, C.J.A. and Pirrotta, V . (2002). The Drosophila Trithorax protein is a coactivator required to prevent re-establishment of Polycomb silencing. Develop 129, 2483-2493. Pugh, F.B. (2000). Control of gene expression through regulation of the TATA-binding protein. Gene 255, 1-14. Qian, S., Capovilla, M . and Pirrotta, V . (1991). The bx region enhancer, a distant cis-control element of the Drosophila Ubx gene and its regulation by hunchback and other segmentation genes. EMBO J 10, 1415-1425. Qian S, Capovilla M , and Pirrotta V. (1993). Molecular mechanisms of pattern formation by the BRE enhancer of the Ubx gene. EMBO J 12, 3865-3877. Qian, S. and Pirrotta, V . (1995). Dosage compensation of the Drosophila white gene requires both the X chromosome environment and multiple intragenic elements. Genetics 139, 733-744. Randsholt, N.B. , Maschat, F. and Santamaria, P. (2000). polyhomeotic controls engrailed expression and the hedgehog signaling pathway in imaginal discs. Mech Dev 95, 89-99. Rastelli, L. , Chan, C S . and Pirrotta, V. (1993). Related chromosome binding sites for zeste, Suppressor of Zeste, and Polycomb group protein and their dependence on Enhancer of zeste function. EMBO J 12, 1513-1522. Robertson, H . M . , Preston, C R . , Phillis, R.W., Johnson-Schlitz, D .M. , Benz, W.K. and Engels, W.R. (1988). A stable genomic source of P element transposase in Drosophila melanogaster. Genetics 118,461-470. Roseman, R.R., Morgan, K. , Mallin, D.R., Roberson, R., Parnell, T.J., Bornemann, D.J., Simon, J.A., Geyer, P.K. (2001). Long-range repression by multiple Polycomb group (PcG) proteins targeted by fusion to a defined DNA-binding domain in Drosophila. Genetics 158, 291-307. 158 Rubin, G .M. and Spradling, A .C . (1982). Genetic transposition of Drosophila with transposable element vectors. Science 218, 348-353. Sabbattini, P., Lundgren, M . , Georgiou, A. , Chow, C , Warnes, G. and Dillon, N . (2001). Binding of Ikaros to the lambda5 promoter silences transcription through a mechanism that does not require heterochromatin formation. EMBO J 20, 2812-2822. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989). Molecular cloning: A laboratory manual. Cold Spring Harbor Laboratory Press, USA. Sanchez-Herrero, E., Vernos, I., Marco, R. and Morata, G. (1985). Genetic organization of Drosophila bithorax complex. Nature 313, 108-113. Santamaria, P. (1983). Analysis of haploid mosaics in Drosophila. Dev Biol 96, 285-295. Santamaria, P. and Randsholt, N .B . (1995). Characterization of a region of the X chromosome of Drosophila including multi sex combs (mxc), a Polycomb group gene which also functions as a tumour suppressor. Mol Gen Genet 246, 282-290. Satijn, D.P., Olson, D.J., van der Vlag, J., Hamer, K . M . , Lambrechts, C , Masselink, H. , Gunster, M.J., Sewalt, R.G., van Driel, R. and Otte, A.P. (1997). Interference with the expression of a novel human polycomb protein, hPc2, results in cellular transformation and apoptosis. Mol Cell Biol 17, 6076-6086. Satijn, D.P. and Otte, A.P. (1999). Polycomb group protein complexes: do different complexes regulate distinct target genes? Biochim Biophys Acta 1447, 1-16. Sato, T., Russel, M . A . and Denell, R.E. (1983). Homeosis in Drosophila: a new enhancer of Polycomb and related homeotic mutations. Genetics 105, 357-370. Saurin, A.J. , Shiels, C , Williamson, J., Satijn, D.P.E., Otte, A.P. , Sheer, D. and Freemont, P.S. (1998). The human Polycomb group complex associates with pericentromeric heterochromatin to form a novel nuclear domain. J Cell Biol 142, 887-898. Saurin, A.J. , Shao, Z., Erdjument-Bromage, H., Tempst, P. and Kingston, R.E. (2001). A Drosophila Polycomb group complex includes Zeste and dTAFII proteins. Nature 412, 655-660. Schlopherr, J., Eggert, H. , Paro, R., Cremer, S. and Jack, R.S. (1994). Gene inactivation in Drosophila mediated by the Polycomb gene product or by position-effect variegation does not involve major changes in the accessibility of the chromatin fibre. Mol Gen Genet 243, 453-462. Schumacher, A. , Faust, C. and Magnuson, T. (1996). Positional cloning of a global regulator of anterior-posterior patterning in mice. Nature 383, 250-253. Schumacher, A. , Lichtarge, O., Schwartz, S. and Magnuson, T. (1998). The murine Polycomb-group gene eed and its human orthologue: functional implications of evolutionary conservation. Genomics 54, 79-88. 159 Sedkov, Y . , Tillib, S., Mizrokhi, L . and Mazo, A. (1994). The bithorax complex is regulated by trithorax earlier during Drosophila embryogenesis than is the Antennapedia complex, correlating with a bithorax-like expression pattern of distinct early trithorax transcripts. Develop 120, 1907-1917. Sekinger, E.A. and Gross, D.S. (2001). Silenced chromatin is permissive to activator binding and PIC recruitment. Cell 105, 403-414. Shao, Z., Raible, F., Mollaaghababa, R., Guyon, J.R., Wu, C.T., Bender, W. and Kingston, R.E. (1999). Stabilization of chromatin structure by PRC1, a Polycomb complex. Cell 98, 37-46. Shearn, A. , Hersperger, E. and Ffersperger, G. (1987). Genetic studies of mutations at two loci of Drosophila melanogaster which cause a wide variety of homeotic transformations. Roux's Arch Dev Biol 196, 231-242. Shearn A. (1989). The ash-1, ash-2 and trithorax genes of Drosophila melanogaster are functionally related. Genetics 121,517-525. Shi, Y . , Seto, E., Chang, L.S. and Shenk, T. (1991). Transcriptional repression by Y Y 1 , a human GLI-Kruppel-related protein, and relief of repression by adenovirus E1A protein. Cell 67, 377-388. Shimell, M.J. , Peterson, A.J . , Burr, J., Simon, J.A. and O'Connor, M . B . (2000). Functional analysis of repressor binding sites in the iab-2 regulatory region of the abdominal-A homeotic gene. Dev Biol 218, 38-52. Siegel, V. , Jongens, T.A., Jan, L . Y . and Jan, Y . N . (1993). pipsqueak, an early acting member of the posterior group of genes, affects vasa level and germ cell-somatic cell interaction in the developing egg chamber. Develop 119, 1187-1202. Sigrist, C.J. and Pirrotta, V . (1997). Chromatin insulator elements block the silencing of a target gene by the Drosophila Polycomb response element (PRE) but allow trans interactions between PREs on different chromosomes. Genetics 147, 209-221. Simon, J., Peifer, M . , Bender, W. and O'Connor, M . (1990). Regulatory elements of the bithorax complex that control expression along the anterior-posterior axis. EMBO J 9, 3945-3956. Simon, J., Chiang, A . and Bender, W. (1992). Ten different Polycomb group genes are required for spatial control of the abdA mdAbdB homeotic products. Develop 114, 493-505. Simon, J., Chiang, A. , Bender, W., Shimell, M.J. and O'Connor, M . (1993). Elements of the Drosophila bithorax complex that mediate repression by Polycomb group products. Dev Biol 158, 131-144. Simon, J. (1995). Locking in stable states of gene expression: transcriptional control during Drosophila development. Curr Opin Cell Biol 7, 376-385. 160 Simon, J., Bornemann, D., Lunde, K. and Schwartz, C. (1995). The extra sex combs product contains WD40 repeats and its time of action implies a role distinct from other Polycomb group products. Mech Dev 53, 197-208. Simon, J.A. and Tamkun, J.W. (2002). Programming off and on states in chromatin: mechanisms of Polycomb and trithorax group complexes. Curr Opin Genet Dev 12, 210-218. Sinclair, D.A.R., Campbell, R.B., Nicholls, F., Slade, E. and Brock, H.W. (1992). Genetic analysis of the Additional sex combs locus of Drosophila melanogaster. Genetics 130, 817-825. Sinclair, D.A., Clegg, N.J., Antonchuk, J., Milne, T.A., Stankunas, K. , Ruse, C , Grigliatti, T.A., Kassis, J.A. and Brock, H.W. (1998a). Enhancer of Polycomb is a suppressor of position-effect variegation in Drosophila melanogaster. Genetics 148, 211-220. Sinclair, D.A.R., Milne, T.A., Hodgson, J.W., Shellard, J., Salinas, C.A., Kyba, M . , Randazzo, F. and Brock, H.W. (1998b). The Additional sex combs gene of Drosophila encodes a chromatin protein that binds to shared and unique Polycomb group sites on polytene chromosomes. Develop 125, 1207-1216. Smouse, D. and Perrimon, N . (1990). Genetic dissection of a complex neurological mutant, polyhomeotic, in Drosophila. Dev Biol 139, 169-185. Smouse, D., Goodman, C , Mahowald, A . and Perrimon, N . (1988). polyhomeotic: a gene required for the embryonic development of axon pathways in the central nervous system of Drosophila. Genes Dev 2, 830-842. Soeller, W.C., Oh, C.E. and Kornberg, T.B. (1993). Isolation of cDNAs encoding the Drosophila G A G A transcription factor. Mol Cell Biol 13, 7961-7970. Soto, M.C. , Chou, T., and Bender, W. (1995). Comparison of germline mosaics of genes in the Polycomb group of Drosophila melanogaster. Genetics 140, 231-243. Stankunas, K. , Berger, J., Ruse, C , Sinclair, D.A., Randazzo, F. and Brock, H.W. (1998). The Enhancer of Polycomb gene of Drosophila encodes a chromatin protein conserved in yeast and mammals. Develop 125, 4055-4066. Stanojevic D, Hoey T, Levine M . (1989). Sequence-specific DNA-binding activities of the gap proteins encoded by hunchback and Kruppel in Drosophila. Nature 341, 331-335. Struhl, G. (1981). A gene product required for correct initiation of segmental determination in Drosophila. Nature 293, 36-41. Struhl, G. and Akam, M.E. (1985). Altered distribution of Ultrabithorax in extra sex combs mutant embryos of Drosophila. EMBO J 4, 3259-3264. Struhl, G. and White, R.A. (1985). Regulation of the Ultrabithorax gene of Drosophila by other bithorax complex genes. Cell 43, 507-519. 161 Strutt, H. , Cavalli, G. and Paro, R. (1997). Co-localization of Polycomb protein and G A G A factor on regulatory elements responsible for the maintenance of homeotic gene expression. EMBO J 16, 3621-3632. Strutt, H . and Paro, R. (1997). The Polycomb group protein complex of Drosophila melanogaster has different compositions at different target sites. Mol Cell Biol 17, 6773-6783. Tagawa, M . , Sakamoto, T., Shigemoto, K. , Matsubara, H., Tamura, Y . , Ito, T., Nakamura, I., Okitsu, A. , Imai, K . and Taniguchi, M . (1990). Expression of novel DNA-binding protein with zinc finger structure in various tumor cells. J Biol Chem 265, 20021-20026. Taillebourg, E. and Dura, J.M. (1999). A novel mechanism for P element homing in Drosophila. PNAS 96, 6856-6861. Takihara, Y . , Tomotsune, D., Shirai, M . , Katoh-Fukui, Y . , Nishii, K. , Motaleb, M.A. , Nomura, M . , Tsuchiya, R., Fujita, Y . , Shibata, Y . , Higashinakagawa, T. and Shimada, K . (1997). Targeted disruption of the mouse homologue of the Drosophila polyhomeotic gene leads to altered anteroposterior patterning and neural crest defects. Develop 124, 3673-3682. Takihara, Y . and Hara, J. (2000). Polycomb-group genes and hematopoiesis. Int J Hematol 72, 165-72. Tamkun, J.W., Deuring, R., Scott, M.P., Kissinger, M . , Pattatucci, A . M . , Kaufman, T . C , and Kennison, J.A. (1992) brahma: a regulator of Drosophila homeotic genes structurally related to the yeast transcriptional activator SNF2/SWI2. Cell 68, 561-572. Takihara, Y . , Tomotsune, D., Shirai, M . , Katoh-Fukui, Y . , Nishii, K. , Motaleb, M . A . , Nomura, M . , Tsuchiya, R., Fujita, Y . , Shibata, Y . , Higashinakagawa, T. and Shimada, K. (1997). Targeted disruption of the mouse homologue of the Drosophila polyhomeotic gene leads to anteroposterior patterning and neural crest defects. Develop 124, 3673-3682. Tie, F., Furuyama, T., Prasad-Sinha, J., Jane, E. and Harte, P.J. (2001). The Drosophila Polycomb Group proteins ESC and E(Z) are present in a complex containing the histone-binding protein p55 and the histone deacetylase RPD3. Develop 128, 275-286. Tillib, S., Petruk, S., Sedkov, Y. , Kuzin, A. , Fujioka, M . , Goto, T. and Mazo, A . (1999). Trithoraz and Polycomb group response elements within a Ultrabithorax transcription maintenance unit consist of closely situated but seperable sequences. Mol Cell Biol 19, 5189-5202. Tsukiyama, T., Becker, P. and Wu, C. (1994). ATP-dependent nucleosome disruption at a heat-shock promoter mediated by binding of G A G A transcription factor. Nature 367, 525-532. Tsukiyama, T. and Wu, C. (1995). Purification and properties of an ATP-dependent nucleosome remodeling factor. Cell 83, 1011-1020. van de Vosse, E., Walpole, S.M., Nicolaou, A. , van der Bent, P., Cahn, A. , Vaudin, M . , Ross, M.T., Durham, J., Pavitt, R., Wilkinson, J., Grafham, D., Bergen, A . A . , van Ommen, G.J., Yates, 162 J.R., den Dunnen, J.T. and Trump, D. (1998). Characterization of SCML1, a new gene in Xp22, with homology to developmental Polycomb genes. Genomics 49, 96-102. van der Lugt, N.M.T. , Alkema, M.J. , Berns, A. and Deschamps, J. (1996). The Polycomb-group homologue Bmi-1 is a regulator of murine Hox expression. Mech Dev 58, 153-164. van der Vlag, J. and Otte, A.P. (1999). Transcriptional repression mediated by the human polycomb-group protein EED involves histone deacetylation. Nat Genet 23, 474-478. van Lohuizen, M . , Tijms, M . , Willem Boncken, J., Schumacher, A. , Magnuson, T., and Wientjens, E. (1998). Interaction of mouse Polycomb-group (Pc-G) proteins Enxl and Enx2 with Eed: Indication for separate Pc-G complexes. Mol Cell Biol 18, 3572-3579. Varga-Weisz, P.D., Wilm, M . , Bonte, E., Dumas, K. , Mann, M . and Becker, P.B. (1997). Chromatin-remodelling factor C H R A C contains the ATPases ISWI and topoisomerase II. Nature 388, 598-602. Wade, P.A., Jones, P.L., Vermaak, D., and Wolffe, A.P. (1998). A multiple subunit Mi-2 histone deacetylase from Xenopus laevis cofractionates with an associated Snf2 superfamily ATPase. Curr Biol 8, 843-846. Wakimoto, B.T. (1998). Beyond the nucleosome: epigenetic aspects of position-effect variegation in Drosophila. Cell 93,321-324. Weber, U. , Siegel, V . and Mlodzik, M . (1995). pipsqueak encodes a novel nuclear protein required downstream of seven-up for the development of photoreceptors R3 and R4. EMBO J 14, 6247-6257. White, R.A. and Wilcox, M . (1984). Protein products of the bithorax complex in Drosophila. Cell, 163-171. Wieschaus, E., Nusslein-Volhard, C. and Jurgens, G. (1984). Mutations affecting the pattern of the larval cuticle in Drosophila melanogaster. Zygotic loci on the X chromosome and fourth chromosome. Roux's Archive of Dev Biol 193, 296-307. Wu, C T . and Morris, J.R. (1999). Transvection and other homology effects. Curr Opin Genet Dev 9, 237-246. Yamaguchi, K. , Hidema, S. and Mizuno, S. (1998). Chicken chromobox proteins: cDNA cloning of CHCB1, -2, -3 and their relation to W-heterochromatin. Exp Cell Res 242, 303-314. Yamamoto, Y . , Girard, F., Bello, B., Affolter, M . Gehring, W.J. (1997). The cramped gene of Drosophila is a member of the Polycomb-group, and interacts with mus209, the gene encoding Proliferating Cell Nuclear Antigen. Develop 124, 3385-3394. Yuh, C.H., Bolouri, H . and Davidson, E.H. (2001). Cis-regulatory logic in the endol6 gene: switching from a specification to a differentiation mode of control. Develop 128, 617-629. 163 Zhang, C.C., Muller, J., Hoch, M . , Jackie, H. and, Bienz, M . (1991). Target sequences for hunchback in a control region conferring Ultrabithorax expression boundaries. Develop 113, 1171-1179. Zhang, C , and Bienz, M . (1992). Segmental determination in Drosophila conferred by hunchback (hb), a repressor of the homeotic gene Ultrabithorax (Ubx). PNAS 89, 7511-7515. Zhang, Y . , LeRoy, G., Seelig, H.-P., Lane, W.S. and Reinberg, D. (1998). The dermatomyositis-specific autoantigen Mi2 is a component of a complex containing histone deacetylase and nucleosome remodelling activities. Cell 95, 279-289. Zink, B., and Paro, R. (1989). In vivo binding pattern of a trans-regulator of the homeotic genes in Drosophila melanogaster. Nature 337, 468-471. Zink, B., Engstrom, Y . , Gehring, W.J. and Paro, R. (1991). Direct interaction of the Polycomb protein with Antennapedia regulatory sequences in polytene chromosomes of Drosophila melanogaster. E M B O J 10, 153-162. Zink, B., and Paro, R. (1995). Drosophila Polycomb-group regulated chromatin inhibits the accessibility of a trans-activator to its target DNA. E M B O J 14, 5660-5671. 164 Abbreviations AEL After egg laying Ant-C Antennapedia complex BTB brie a brae, tramtrack, broad complex domain BX-C Bithorax complex bxd bithoraxoid DPS downstream of pairing-sensitive region module ChIP chromatin immunoprecipitation GAF G A G A Factor GTFs general transcription factors ETP Enhancer of trithorax and Polycomb FAB frontal-abdominal HAT histone acetyltransferase HB Hunchback protein HDAC histone deacetylase iab infra-abdominal Mcp miscadastral pigmentation MCS multiple cloning site NTF-1/GRH Nuclear transcription factor-1/Grainyhead protein PBS phosphate buffered saline Pc Polycomb gene PcG Polycomb group PCR polymerase chain reaction ph polyhomeotic gene PH Polyhomeotic protein P H O Pleiohomeotic protein PIC preinitiation complex P O U Pit-1, Oct-1, unc-86 domain PS parasegment psq pipsqueak gene PRC-1 Polycomb repressive complex-1 P R E Polycomb response element PSR pairing-sensitive region module T R E trithorax response element trxG trithorax group Ubx Ultrabithorax gene U P S upstream of pairing-sensitive region module U T D UPS - tetracycline spacer - DPS 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0099654/manifest

Comment

Related Items