STUDIES OF XIST A N D MACROH2A1.2 INDEPENDENT OF G E N E S I L E N C I N G : A UNIQUE H Y B R I D C E L L M O D E L S Y S T E M By E D M U N D K W O K B.Sc. University of British Columbia, 2000 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF M A S T E R OF SCIENCE In T H E F A C U L T Y OF G R A D U A T E STUDIES D E P A R T M E N T OF M E D I C A L G E N E T I C S We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA August 2002 © Edmund Kwok, 2002 In p resent ing this thesis in part ial fu l f i lment of t h e requ i remen ts fo r an advanced d e g r e e at the Univers i ty o f Brit ish C o l u m b i a , I agree that t h e Library shall make it f ree ly available fo r re ference and study. I fu r ther agree that permiss ion f o r extensive c o p y i n g o f th is thesis f o r scholar ly pu rposes may b e g ran ted by the h e a d o f my d e p a r t m e n t or by his o r her representat ives. It is u n d e r s t o o d that c o p y i n g or pub l i ca t i on o f th is thesis f o r f inancial gain shall n o t be a l l o w e d w i t h o u t m y w r i t t e n permiss ion . D e p a r t m e n t o f The Univers i ty o f Brit ish C o l u m b i a Vancouver , Canada DE-6 (2/88) ABSTRACT X chromosome inactivation results in the transcriptional silencing of one of the two X chromosomes present in mammalian female cells. XIST expression from the inactive X (Xi), and subsequent coating of the Xi by the XIST RNA, marks the initial major observable characteristic of X-inactivation. The Xi then begins to exhibit features of heterochromatin such as gene silencing, late replication, histone modification, hypermethylation, and the recruitment of core histone variants. However, the exact mechanistic events between XIST expression and the final heterochromatic state of the Xi remain unclear. To better understand the relationship between XIST, heterochromatin, and silenced gene expression, the first goal of this study was to develop and characterize a model system suitable for examination of those features of heterochromatin which can be induced by XIST independent of gene silencing. Previous work using rodent/human hybrids (containing active human and mouse X chromosomes) that had been demethylated to reactivate XIST/Xist expression showed that while human XIST RNA fails to localize properly to the human Xi, the mouse Xist localizes normally. In both cases no silencing of X-linked genes was observed. To verify this hybrid cell system as a model for studying the effects of XIST/Xist independent of inactivation, I examined the DNA methylation, H3 acetylation, and gene expression statuses of multiple human and mouse X-linked genes in these hybrids. The results confirm both i)the reactivation of XIST/Xist due to demethylation, and ii)the lack of gene silencing despite the presence of XIST/Xist. This study also examined the core histone variant macroH2A1.2 as the first candidate for study using this hybrid system. Past evidence has shown tight macroH2Al .2 localization to the Xi, and that this macrochromatin body (MCB) formation is abolished if the Xist gene was deleted. I created a macroH2A-GFP construct that was expressed within the hybrid system, and the results suggest that XIST RNA localization is required for proper human macroH2A1.2 localization, probably via a ribonucleoprotein complex. Furthermore, the evidence suggests that there may be species-specific differences in macroH2A1.2 that affects its interaction with XIST or Xist RNA. TABLE OF CONTENTS Page Number ABSTRACT ii TABLE OF CONTENTS — iv LIST OF TABLES vii LIST OF FIGURES viii LIST OF ABBREVIATIONS ix ACKNOWLEDGEMENTS xi 1 INTRODUCTION — — 1 1.1 X chromosome inactivation 1 1.1.1 Background 1 1.1.2 X chromosome inactivation center 2 1.1.3 Mechanism of X-inactivation 3 1.2 Features of Xi 6 1.2.1 XIST — 6 1.2.1.1 Gene structure 6 1.2.1.2 Gene expression - 6 1.2.1.2.1 Tsix 7 1.2.1.3 Role in X-inactivation - 8 1.2.1.3.1 Recruitment of chromatin proteins 9 1.2.1.4 Functional Domains — 10 1.2.2 DNA methylation - — —- 11 1.2.3 Histone modifications — - 15 1.2.3.1 Acetylation - 15 1.2.3.2 Methylation 18 1.2.3.3 Phosphorylation 20 1.2.3.4 Histone code 20 1.2.4 Replication timing - 22 1.2.5 Histone variants — 23 1.2.5.1 macroH2A 24 1.2.5.1.1 macroH2Al - 24 1.2.5.1.2 macroH2A2 26 1.2.5.1.3 macroH2A-Bbd 26 1.2.5.2 Role in X-inactivation 26 1.2.5.2.1 Macrochromatin bodies 26 1.2.5.2.2 Sequence Analysis - — - 28 1.2.5.2.3 Functional domains - 29 1.2.5.2.4 Involvement in heterochromatin 32 1.2.5.2.5 Possible mechanism of interaction — 33 1.3 Rodent/Human hybrid cells — 34 iv 1.3.1 XIST localization 37 1.3.2 Ideal model----- 40 1.4 Thesis goals 41 2 MATERIALS & METHODS 42 2.1 Tissue culture methods — 42 2.1.1 Cell lines —- —- 42 2.1.2 Freezing cells 42 2.2 Examining gene methylation status - 43 2.2.1 DNA extraction 43 2.2.2 Methylation-sensitive restriction digests 44 2.2.3 Polymerase chain reaction (PCR) 46 2.3 Examining gene expression status — 46 2.3.1 RNA extraction 47 2.3.2 DNase treatment of RNA —- 47 2.3.3 Reverse Transcription (RT) 48 2.3.4 PCR using cDNA 49 2.4 Cloning 52 2.4.1 Plasmid preparations from bacteria 52 2.4.2 Bacterial transformation 53 2.4.3 Cloning inserts into vectors 54 2.4.3.1 Inserts 54 2.4.3.2 Vector insertions 57 2.4.3.3 Construct verification 58 2.5 Lipofection 58 2.6 Chromatin immunoprecipitation (ChIP) assay 59 2.6.1 PCR ChIP results 61 2.7 Microscopy 63 2.7.1 Cremer preparations (fixing cells) — 63 3 CHARACTERIZATION OF HYBRID MODEL SYSTEM 64 3.1 Introduction 64 3.2 Results 67 3.2.1 Methylation status results 67 3.2.2 Expression status results 69 3.2.3 Acetylation status results 72 3.3 Discussion — 74 4MACROH2A1.2 —-- - 77 4.1 Introduction 77 4.2 Results 79 4.2.1 Missing human factor? — 79 4.2.2 macroH2A1.2 - GFP constructs 83 4.2.3 Localization 83 4.3 Discussion 91 v 5 DISCUSSION AND FUTURE DIRECTIONS — - 96 REFERENCES 100 APPENDIX 110 vi LIST OF TABLES Page Number Table 1-1: Human/Mouse Hybrid Cell Lines 36 Table 2-1: Primers for Determining Methylation Status — 45 Table 2-2: Primers for Determining (Human) Expression Status 50 Table 2-3: Primers for Determining (Mouse) Expression Status 51 Table 2-4: Primers for Amplifying macroH2A1.2 Inserts 55 Table 2-5: Primers for ChIP Analysis 62 Table 4-1: Hybrid Cell Lines Containing Human Xi and Other Chromosomes — 81 LIST OF FIGURES Page Number Figure 1-1: Schematic Diagram of the X-inactivation Process 5 Figure 1-2: mH2A Variants 25 Figure 1-3: Functional Domains of mH2Al 31 Figure 1-4: Reactivated XIST/Xist 39 Figure 2-1: Primers for Amplifying macroH2A1.2 Inserts 56 Figure 3-1: Location of Human X-linked Genes 66 Figure 3-2: Methylation Status of X-linked Genes 68 Figure 3-3: Expression Status of Human X-linked Genes 70 Figure 3-4: Expression Status of Mouse X-linked Genes 71 Figure 3-5: Acetylation of H3 Status of X-linked Genes 73 Figure 4-1: Expression of human macroH2Al Variants in Specific Hybrids 82 Figure 4-2: BOSC 23 Transfections 87 Figure 4-3: AHA-1 laBl Transfections 88 Figure 4-4: AHA-1 laBl-A52b Transfections - - 89 Figure 4-5: AHA-1 laBl-A52b-4Cl Transfections 90 vm LIST OF ABBREVIATIONS *NOTE: Throughout this thesis, gene names with all capitalized letters refer to human genes, whereas names with only the first letter capitalized refer to murine genes. 5-azaC - 5-azacytidine 5-azadC - 5-aza-2'-deoxycytidine Ada- encodes protein for part of transcription co activator complex w/Gcn5 AG-TG - azaguanine-thioguanine AR - androgen receptor ATF - transcription factor binding DNA-site ChIP - chromatin immunoprecipitation CREB - cyclic AMP response element-binding protein Dnmt - DNA cytosine methyltransferse ES - embryonic stem EST - expressed sequence tag G6pd - glucose-6-phosphate dehydrogenase, X-linked Gcn5 - HAT homolog in yeast, encodes protein for part of transcriptional co activator complex w/Ada FMR1 - fragile X mental retardation 1, X-linked GFP - green fluorescent protein HAT - histone acetyltransferase HDAC - histone deacetylase HMT - histone methyltransferase HP1 - heterochromatin protein 1 HPRT- hypoxanthineguanine-phosphoribosyl transferase, X-linked LINE - long interspersed nuclear element MBD - methyl CpG binding protein MCB - macrochromatin body MCD -macrochromatin domain MDF - microtubule depolymerizing factor MeCPl/2 - MBD proteins, X-linked mH2A - macroH2A (refers to variants 1.1, 1.2, and 2) MLE - maleless, a helicase protein MMLV-RT - Moloney murine leukemia virus reverse transcriptase MOF - males absent on the first, a histone acetyltransferase MSL-1, -2, -3 - male-specific lethal chromatin proteins mXist - mouse Xist PDHAl/Pdhal - pyruvate dehydrogenase (lipoamide) alpha 1, X-linked PGKl/Pgkl - phosphoglycerate kinase 1, X-linked PHKAl/Phkal - phosphorylase kinase, alpha 1, X-linked PNA - peptide nucleic acid roXl/roX2 - genes for non-coding RNA critical for X-linked dosage compensation in melanogaster RT - reverse transcription RT-PCR - reverse-transcription PCR SAGA - part of transcription complex in yeast Su(var) - suppressor of variegation TAFII250 - transcriptional regulator TIMP1 - tissue inhibitor of metalloproteinase 1, X-linked TSIX/Tsix - human/mouse antisense of XIST/Xist Ubel - ubiquitin-activating enzyme E l , X-linked Xa - active X chromosome Xce - X-controlling element (mouse) Xi - inactive X chromosome XlC/Xic - human/mouse X-inactivation center XIST/Xist - human/mouse inactive X transcript XPCT- X-linked PEST-containing transporter Zfx - zinc finger protein, X-linked ACKNOWLEDGEMENTS I have been extremely fortunate to have Dr. Carolyn Brown as my graduate studies supervisor; she has provided me with unparalleled guidance and enthusiasm, and her continual support is greatly appreciated. I would also like to extend my thanks to members of the Brown lab, in particular Cathy Anderson for her amazing patience, Sarah Baldry for her technical expertise, and Jennifer Chow for her wonderful insights and encouragement. Finally, I would like to acknowledge Tracy Tucker for being there every step of the way, and my family for their undying support. x i CHAPTER 1: Introduction 1.1 X chromosome inactivation 1.1.1 Background Spawned from a hypothesis initially proposed just over forty years ago by Mary Lyon, the concept of X chromosome inactivation refers to the transcriptional silencing of one of the two X chromosomes present in each mammalian female cell 1. Since each male cell contains only one X chromosome, X-inactivation compensates for dosage differences in X-linked genes between the sexes. In eutherians, X-inactivation occurs in a random fashion, meaning that during early development in the soma either one of the two X chromosomes can be silenced 2,3. In contrast, marsupials utilize a type of imprinted X-inactivation, with the paternal X being preferentially silenced 4>5 This latter mechanism has been thought to be a more ancestral form of X-inactivation 6, and this thesis will concentrate more on the random X-inactivation. Regardless which of these X chromosomes is inactivated, the silenced state is then faithfully inherited in all clonal descendants, and in the case of the random X-inactivation, results in a mosaic pattern of cells in mammalian females 2>3. Over the past forty years however, the precise details of X-inactivation have become more and more difficult to define. Indeed, although we think of X-inactivation as a chromosome-wide mechanism of gene silencing, the inactive X chromosome (Xi) is not completely inactivated. Roughly 10-15% of genes on the human X chromosome escape X-inactivation 7. Some of these genes have homologs residing on the Y 1 chromosome, most of which are located at the pseudoautosomal region, and thus are required to remain active for dosage compensation 8. 1.1.2 X chromosome inactivation center Several years after Lyon's original hypothesis in 1961, there had already been speculations about the existence of a genetic locus that might be responsible for X-inactivation. Rastan noted that in mouse X-autosomal translocations, specific regions of the X chromosome seemed to always be present in order for "spreading" of inactivation to autosomal regions 9. It wasn't until the nineties, however, that transgenic analysis in murine embryonic stem (ES) cells provided solid evidence of a X-inactivation center (XIC). Experimental evidence showed that mouse autosomes could be inactivated when certain X chromosome sequences (~450kb) from the mouse Xic were present in cis 1U\ The human XIC has been narrowed to a region located at Xql3, and is currently defined as a cis-acting locus that is essential for the X chromosome to be inactivated early in female embryogenesis. Contained within the XlC/Xic is a gene encoding an "inactive X transcript", or XIST/Xist, which is exclusively expressed from the inactive X chromosome in interphase nuclei 11-13. XIST is a functional RNA that appears to "coat" the Xi completely, and has thus been theorized to be critical in the mechanism of X-inactivation. The importance of this gene is discussed in greater detail in a following section. 2 1.1.3 Mechanism of X-inactivation It is generally accepted that X-inactivation occurs via a multi-step process: counting, choice, initiation, propagation, and maintenance. Early in embryogenesis, a mammalian female cell somehow measures the number of X chromosomes it has per haploid autosome set. Any number of X chromosomes above one per diploid set will be subject to silencing 14. Several studies have shown this counting step to be under the control of the Xic, specifically sequences 3' of the Xist gene, although the exact mechanism is still unclear 15,16 Next5 one of the X chromosomes is then randomly chosen to remain active. Once the choice of the inactive X chromosome is determined, the next series of steps happen in a less-well understood flurry of events. The first sign of the initiation of X-inactivation is decreased expression of the XIST antisense TSIX on the inactive X chromosome, followed by increased expression of XTSTRNA and its subsequent localization to the chosen Xi. Soon thereafter, the Xi exhibits many features of transcriptionally silent heterochromatin: hypermethylation, hypoacetylation of core histones, chromatin condensation and nuclease insensitivity. Recently, a histone variant called macroH2A has also been shown to localize preferentially to the inactive X, although only after initiation 17. All these features of X-inactivation appear to "spread" outwards from the XIC and along the whole X chromosome 18. This ability to induce heterochromatic features onto whole chromosomes seems to mainly depend on the physical presence of the XIC; in somatic cells containing X;4q translocations that included the XIC, most of the genes on the involved autosomal regions were inactivated 19. The actual distance of the affected genes' loci and the XIC did not seem to correlate with gene inactivation, as some genes 3 that remained active were located interstitially throughout the autosomal region. The fact that some genes appeared to be able to "resist" inactivation by the XIC suggests that there may be some unique feature of the X chromosome that allows for greater responsiveness to XlC-induced inactivation. This is not an absolute characteristic of the X chromosome, as there are genes that do escape inactivation 7. Early translocation studies suggested that perhaps the X chromosome contains a series of way stations throughout its chromatin, specific spots upon which XIST RNA can bind more readily 3,20 A recent proposal indicated long interspersed nuclear elements (LINEs) as being such way stations, since the X chromosome contains a significantly higher density of LINES (comparable to that found in pericentromeric regions) than the rest of the genome ^ . Once the X chromosome becomes transcriptionally silent, it remains so throughout the cell cycle, and the inactive state is maintained in all clonal descendants. One theory to explain such observations is thus: XZSTRNA recruits certain factors that silence genes and condense chromatin structure on the Xi (see Figure 1-1). Features such as DNA methylation, histone deacetylation, and macroFOA localization may act in maintaining this silenced state. It is very difficult to experimentally prove/disprove such a theory, however, since these heterochromatic features do not normally accumulate independent of each other; a step-wise breakdown of X-inactivation mechanism is thus poorly defined. 4 O C T | ~ XIC XIC Xist RNA coating iz? era Late Replication Gene Silencing Histone Media t ions? MacroH2A • DNA Methylation Figure 1-1: Schematic Diagram of the X-inactivation Process XIST/Xist expression from the XIC marks the first sign of X-inactivation, followed by the rapid accumulation of a succession of heterochromatic features. 5 1.2 Features of Xi 1.2.1 XIST First discovered in 1991H, the XIST gene (and its transcribed product's intimate relationship with the inactive X) remains as the primary candidate responsible for initiating X-inactivation. 1.2.1.1 Gene Structure One of the genes that is located within the XIC, the XIST gene encodes a relatively long RNA product: the human XIST is roughly 19.3kb, and the mouse Xist measures 17.8kb 21,22. The RNA itself is an alternatively spliced transcript, and subsequent sequence analysis of the XIST gene product using human/mouse comparisons failed to reveal any long conserved open reading frames 23,24 Combined with the observation that the XIST RNA, when examined using fluorescence in situ hybridization (FISH), maintains a nuclear localization, it is most likely a functional RNA 23,25 1.2.1.2 Gene Expression The XIST/Xist gene on the active X chromosome is silent, accompanied by methylation of its promoter region 26,27 Mouse embryos engineered with defective DNA methyltransferases demonstrate Xist expression in both males and females, presumably due to hypomethylation of the gene's promoter 2. On the inactive X chromosome, the XIST gene is transcribed, but the RNA itself remains untranslated and nuclearly localized 25. i n mouse, prior to X-inactivation both X chromosomes express 6 low levels of Xist. During the initiation of the whole inactivation process, the X chromosome destined to become the inactive X upregulates its Xist expression. Cytogenetically, one can observe the XIST RNA physically coating the inactive X, and colocalizing with the Barr body 25. Most agree that upregulation of XIST expression is the primary (observable) step in silencing the inactive X, and is undoubtedly an essential step for proper X-inactivation 28,29 j n a study using a modified Xist gene where the minimal promoter sequences were retained, it was shown that the presence of Xist RNA and not the actual expression of the gene is the crucial factor for inactivation 28, How XIST expression is actually regulated, however, remains unclear. In mice the X controlling element (Xce) locus, residing 3' to the Xist gene, acts as a strong modifier of choosing an inactive X chromosome; an X chromosome containing a "strong" allele for the Xce gene has a greater chance of being chosen as the active X chromosome 30. Although no orthologous locus has been identified in humans, recently an antisense locus 3' to and partly overlapping the Xist gene (and thus named Tsix) has been uncovered and is suspected of being the primary candidate for regulating Xist expression 16. A human ortholog TSIX was subsequently identified a couple years later 31. 1.2.1.2.1 Tsix Tsix encodes a transcript that starts roughly 40kb downstream of Xist and in its unprocessed form overlaps the entire Xist locus 16. A processed version of the primary transcript has been recently observed in murine ES cells, although its significance is unknown 32. Much like the Xist RNA, Tsix RNA has no recognizable open reading frame and resides in the nucleus; but unlike Xist, Tsix is eventually up regulated on the 7 active X chromosome, and down regulated on the inactive X chromosome. In fact, studies in mice with a deleted Tsix gene demonstrated that Tsix might act antagonistically to Xist 33. in the placental tissues of these mice, the paternal X chromosome is preferentially silenced (and thus Xist is expressed); but in the absence of the Tsix gene, the maternal allele of Xist in these cells also remains expressed 32,33 Lik;e Xist, Tsix is initially expressed from both X chromosomes prior to X-inactivation. Using RNA-FISH, Tsix RNA been observed to diminish from the future inactive X immediately prior to Xist RNA accumulation 34. 1.2.1.3 Role in X-inactivation Studies in developing cells have shown that XIST expression is an early step in X-inactivation 28,29 As mentioned, the XZSTRNA physically coats the inactive X chromosome in cis; this association is lost during mitosis, but continued transcription of the RNA and its localization is regained in the daughter cells 10,25 Some experiments have provided evidence that the actual stability of the XIST transcript may be the direct trigger of X-inactivation 29. Although substantial disagreement exists in explaining how this increase in transcription/transcript stability is achieved 35 s there is strong evidence that this trigger for X-inactivation must occur within a specific time window early in development. Deletion of the XIST gene in somatic cells that have already undergone inactivation generally fails to induce any kind of reactivation of the inactive X 36. i n fact, even with induced Xist expression and correct coating of Xist RNA to the X chromosome in adult somatic cells, X-inactivation cannot be achieved 37. Only when a Xist transgene is induced in undifferentiated ES cells can X-inactivation be induced; the 8 same Xist transgene failed to inactivate X chromosomes in cells that have already undergone differentiation. However, a recent study using human fibrosarcoma cells demonstrated that ectopic XIST expression was able to induce inactivation of autosomal regions, suggesting that human X-inactivation may be possible within a less stringent time frame 38. This i s the only study to date that was able to experimentally show XIST-induced inactivation in differentiated cells, and much further work is still required to confirm whether this is applicable to not only autosomal regions but also the human X chromosome. 1.2.1.3.1 Recruitment of Chromatin Proteins One likely role of XIST RNA is the recruitment of certain proteins to the inactive X chromosome, either for the initiation and/or maintenance of a heterochromatic state. This hypothesis is partly based on studies done in D. melanogaster, where X-linked dosage compensation also occurs albeit in a slightly different manner. In the fruit fly, the single X chromosome present in the male becomes hypertranscribed in order to match the regular transcription of genes on the 2 X chromosomes present in females. Much like mammalian X-inactivation, dosage compensation in Drosophila depends critically on functional RNAs, namely roXl and roX2 39. These non-coding RNA molecules exert their actions through the recruitment of 5 male-specific lethal proteins: MSL-1, -2, -3, MLE, and MOF 39. These proteins form a complex and are localized along the male X chromosome, presumably changing chromatin features and facilitating the upregulation of X-linked genes; the MOF protein for example, contains an innate histone acetyltransferase activity 39. 9 Similarly, human XIST RNA may also be responsible for interacting with and recruiting chromatin proteins to the inactive X chromosome. Chromatin immunoprecipitation experiments targeting hypoacetylated H3 and H4 core histones have found an association with Xist RNA, supporting the theory that the functional RNA might be recruiting certain protein factors capable of altering chromatin features 40 The most convincing evidence of a relationship between XIST RNA and a chromatin protein is the discovery of the core histone variant macroH2A 41, a topic that will be discussed in further detail in following sections of this thesis. 1.2.1.4 Functional Domains While the ultimate consequences of XIST/Xist expression are quite clear, how the actual functional RNA molecule exerts its actions remain largely unknown. Some light has recently been shed on the subject by Wutz et al 42; they performed a deletion analysis of the mouse Xist RNA to identify specific functional domains in the molecule. Utilizing a Cre-loxP based transgene approach, the researchers integrated different inducible modified Xist sequences into a specific site on the X chromosome of murine ES cells. Since silencing of the single X chromosome in the male cells results in cell death, the silencing ability of the different transgenic Xist sequences can be measured by degree of cell death. Their published results showed that a 0.9kb region of highly conserved poly-A repeats at the 5' end of Xist was crucial for proper silencing function; large deletions in either the middle or 3' end of Xist had little/no effect on Xist function. RNA FISH revealed that Xist molecules lacking the 0.9kb region were still able to localize to the Barr body, indicating that additional domains are required for localizing Xist to the X 10 chromosome. By fusing the 5' region minimally required for Xist function to different parts of Xist, they identified 3 domains that may contribute to the localization of Xist. To analyze the nature of how the 900bp critical region functions, the researchers inserted a Xist transgene that had the 900bp region placed at the 3' end of Xist. They observed no abolishment of silencing function, suggesting that actual location of this critical region in Xist is not important. Computational analysis demonstrated that this core sequence can form secondary structures of 2 stem loops, and that this 5' repeat region is comprised of 7.5 repeat units of a combination of these two stem loops. Replacement of bases with different bases, while maintaining proper base-pairing required for secondary structures to be conserved, failed to affect Xist's silencing function. Thus, Xist does appear to physically interact somehow with something in a sequence specific manner to initiate X-inactivation. Interestingly, deletions in the middle of Xist, as well as a 3' truncation of Xist, resulted in the loss of macrochromatin body (MCB) formation (see Section on macroH2A). One study used PNA (peptide nucleic acid) interference to functionally eliminate certain specific regions of the Xist RNA, and discovered that MCBs were lost when the C-terminal region of Xist was blocked 43. Although this confirms an intimate relationship between Xist RNA and macroH2A, whether this relationship is direct or indirect remains unclear. 1.2.2 DNA Methylation One well recognized feature of heterochromatin in mammals is the methylation of DNA. Enzymes known as DNA cytosine methyltransferases (DNMTs) add methyl-groups to the 5-position of specific cytosine residues 44,45 At the molecular level it has 11 been shown that the addition of a methyl-group to these sites can physically change the appearance of the major groove in a DNA helix, a chromatin alteration that may affect target sites for many DNA binding proteins, including transcriptional machinery 46,47 DNA methylation can thus serve as an epigenetic marker for modifying expression levels of affected genes, and in mammals methylation is associated with the reduction in transcriptional initiation. This particular epigenetic marker can be copied after DNA synthesis from cell cycle to cell cycle, resulting in a heritable feature of heterochromatin. It is important to note that 5-methylcytosine can spontaneously undergo deamination and be converted to a thymine residue 48 Whereas a non-methylated cytosine deamination (resulting in a uracil base) can easily be recognized and replaced by repair enzymes such as uracil DNA glycosylase, a base mismatch resulting from a C -> T conversion is more likely to be overlooked. Throughout evolution, CpG content in the human genome has slowly but consistently been depleted with the exception of certain regions, especially those found near the 5'-promoter sequences of genes; these are found to have the expected frequency of CpG dinucleotides 49. Clusters of CpG dinucleotides are thus, in general, underrepresented in the whole human genome. As a result, regions of relatively high CG content known as CpG islands (usually l-2kb in length) have restriction sites for certain rare-cutting enzymes. CpG islands in mammals are usually found near the 5' end of genes and, as a general rule, are unmethylated with a couple of exceptions. Imprinted genes, as well as inactivated genes in the Xi, are both found to be methylated. The first DNMT to be cloned in mammals was DNMT1, a ~200kDa protein with key functional domains in both its N-terminal (which contains regulator domains 12 responsible for targeting specificity) and C-terminal domains (which shows homology to bacterial cytosine-5 methylases) 50 There has been experimental evidence that the N-terminal domain actually targets replication foci of a replicating DNA strand, a function that has been hypothesized to contribute to the inheritance of methylation by the genomes of daughter cells 50 From observations that embryonic stem cells deficient for Dnmtl (Dnmtl-/-) can still methylate DNA de novo 51, however, it is apparent that other enzymes are capable of methylating DNA. Another group of DNMTs was recently found to have DNA methylating abilities, and these were labeled DNMT3a and DNMT3b. By observing expression levels of these genes in differentiated and ES murine cells, it was found that there was a correlation between increased tissue-specific gene expression and decreased Dnmt3a/b expression 45. it is now believed that while Dnmtl methylates hemimethylated DNA (and thus is involved in the maintenance and inheritance of methylation patterns), Dnmt3a and Dnmt3b are responsible for de novo methyltransferases activities in the cell 52, The actual association of DNA methylation to levels of gene expression was initially observed over two decades ago in experiments using 5-azacytidine. Treatment of mouse embryonic cell lines with 5-azacytidine resulted in a drastic change in phenotype, with cells becoming differentiated 53. Even more interestingly, mouse embryonic cells can be forced to differentiate when transfected with DNA taken from 5-azacytidine treated cells (i.e. methylation marks being copied onto newly synthesized DNA strands), demonstrating the heritable nature of this phenomenon 54. Originally 5-azacytidine and its deoxy- version 5-aza-2'-deoxycytidine were thought to be strong demethylating agents, with their mode of action being the replacement of cytosines with analog residues 13 resistant to methylation in the cellular DNA after DNA replication. Later it was suggested that 5-azacytidine and 5-aza-2'-deoxycytidine can covalently "trap" DNA methyltransferases to the DNA, resulting in a consistent loss of methylation with each successive round of cell division 55. Regardless of the how 5-azacytidine works, numerous other studies have supported the observations from the initial experiments in mouse embryonic cell lines: DNA methylation in promoter regions of genes appears to act as a powerful suppressor of gene expression 56 As mentioned before, methylated cytosines most likely affect transcription by modifying binding affinities of certain factors/proteins. For example, sequence-specific transcription factors such as ATF/CREB have target sequences that contain CpG dinucleotides, and methylation at these target sites may inhibit efficient binding 57. Additionally, there is a certain class of DNA binding proteins that targets methylated CpG specifically; MeCPl and MeCP2 are both members of this MBD family of proteins, and they may affect gene expression by competing with transcription factors for DNA binding, or by altering local chromatin structure to resemble transcriptionally silent heterochromatin 58. One aspect of this "chromatin change" is modification of core histones, a topic to be discussed in detail in a later section of this thesis. Generally, acetylation of core histone tails appears to correlate closely with a local "open" chromatin structure and high gene expression, whereas regions with hypoacetylated histones seem to repress transcription activity 59. Recently, it has been shown that MeCP2 can recruit a histone deacetylase (HDAC1) to methylated DNA, resulting in a closed chromatin conformation that is transcriptionally repressed 56,60 14 The exact reason why mammals have evolved to be so dependent on DNA methylation remains a mystery. Invertebrates such as C. elegans and D. melanogaster do not have extensive CpG methylation systems, and some researchers believe that only larger and more complex organisms require DNA methylation as a control mechanism for large volumes of transcriptional noise. Alternatively, some have argued that DNA methylation serves as a method of silencing expression from foreign viral DNA sequences; these theories are based on the fact that a large percentage of hypermethylated CpGs are found near/within retrotransposons. Regardless of the reason, methylation is essential for proper mammalian development. While mouse undifferentiated ES cells are not dependent on DNA methylation, proper cellular differentiation requires a concise, specific series of demethylation and methylation steps 61,62 Generally, parental methylation marks are removed during gametogenesis and a new set of methylation marks are established upon implantation and throughout gastrulation. 1.2.3 Histone Modifications 1.2.3.1 Acetylation In eukaryotes, naked DNA is constantly associated with many different DNA-binding proteins, and together this chromatin goes through a series of packaging and unwinding steps during a cell's life cycle. The most basic unit of chromatin compaction exists as the nucleosome, which is a complex of eight core histones (two of each of histones H2A, H2B, H3, and H4) around which is coiled 146 bp of double-stranded DNA helix 63. The N-terminal tails of core histones are rich in lysine residues, which are 15 positively charged under normal cellular conditions. This overall positive charge on histone tails has been proposed to allow interaction with the DNA directly, resulting in a higher level of chromatin packaging. However, more recent studies on the crystal structure of the nucleosomes revealed that the tails of the core histones actually pass through the gyres of the DNA double helix, and are free in solution instead of interacting with the DNA 64. More likely, these histone tails are candidates for interactions between nucleosomes, responsible for bringing the basic units of DNA compaction together to form higher order structures. In both cases, acetylation of the histone tails has a net effect of neutralizing the lysine charges, resulting in decreased histone-DNA and/or histone-histone interactions. Whereas transcription by RNA polymerase is prohibited in regions of DNA associated with deacetylated core histones, acetylation of these histone tails facilitates the previously silent transcription process, suggesting a loss of higher order compaction of chromatin 65. Molecularly, a large family of enzymes known as histone acetyltransferases (HAT) is responsible for adding acetyl groups to specific lysine residues on the core histone tails 66. Each HAT subfamily has its own specificity in regards to individual lysine residues on different core histones; more interestingly, in vivo HATs are usually complexed in multisubunit complexes that can modify their specificity 67. F o r example Gcn5, a HAT homologue in yeast, on its own has been shown to preferentially acetylate lysine 14 of histone H3 68,69 when complexed in the Ada or SAGA multiprotein units, however, the specificity expands to lysines 9, 14, and 18, as well as gaining a new preference for histone H2B 70. The actions of HATs in a cell are balanced by the effects of histone deacetylases (HDAC). At least eight HDACs have been cloned so far 71; and 16 there is increasing evidence that these also participate in forming multiprotein complexes resulting in a wide range of specificity 72. The acetylation status of core histones is thus a tempting candidate for a regulatory mechanism of chromatin compaction, and consequently accessibility for transcriptional factors. This is supported by not only the complexity and specificity of HATs and HDACs, but also experimental observations. As early as the 1960's, it had been noted that there is an association between the acetylation state of core histones and gene expression status 73,74 Follow-up studies revealed that regions of transcriptionally active chromatin are more likely to contain hyperacetylated core histones 75,76 More strikingly, many co activators of transcription were found to have acetyltransferase activity, one these being Gcn5 in yeast. In D. melanogaster, the presence of hypoacetylated H4 has been observed in the heterochromatic regions of metaphase chromosomes 77,78 Further supporting the idea that gene expression is affected by acetylation status, others have demonstrated that core histone acetylation preferentially occurs at promoters of subsequently activated genes 68,79 Similarly, deacetylation at promoters has been correlated with gene silencing 80. In the context of the inactive X chromosome, it is similar to heterochromatin in that overall it stains poorly with antibodies against acetylated core histones 81,82 -pne hypoacetylation of H4 has been shown in both eutherians and marsupials, demonstrating that this property is highly conserved 83. i n experiments where mouse embryonic stem (ES) cells were induced to differentiate (and thus initiate X-inactivation), it was observed that H4 acetylation at promoters of X-linked genes (Day 4) followed closely soon after Xist expression (Day 2) 84. 17 1.2.3.2 Methylation First observed in the 1960's, post-translational addition of methyl groups to histone tails had eluded functional characterization for nearly four decades 85. Histone methylation is facilitated by enzymes known as histone methyltransferases (HMTs), and the main targets of this modification are lysine residues on the amino-tails of H3 and H4 core histones 86. A varying range of the number of methyl groups at a single residue resulting in mono-, di-, or tri-methylated residues has been observed, although the significance of this remains unclear 87. What is clear is that, much like histone acetylation, there are specific lysine residues that are preferentially methylated 86. On both H3 and H4 tails, unique lysine residues are target for either acetylation or methylation; only one lysine (Lys9) in the H3 tail has been experimentally shown to be a target for both types of modification, although many investigators suspect that it is highly possible for other acetylated lysines to be targeted for methylation 87. Not unlike histone acetylation, histone methylation was initially thought to simply alter physical interactions between the histone tails and DNA. This is deduced from the fact that, although methylation of histone tails does not affect overall charge (unlike histone acetylation), increased histone tail methylation appears to strengthen its affinity for anionic substances such as DNA, as reflected by increased resistance to trypsin digestion 88,89 More recently it has become clear that methylation might also be a mark for recruiting chromatin-associated proteins to facilitate particular transcriptional responses 90. Either way, histone tail methylation has been shown to have an effect on the expression status of nearby genes. For example, DNA prepared from chromatin immunoprecipitations using hyperacetylated H4 (from transcriptionally active regions) 18 has been shown to be a target for H3 methylation y 1 . How these different modifications cooperate to upregulate transcription still remains unclear. Recent work in D. melanogaster and 5*. pombe, however, have provided strong evidence that HMTs play significant roles in regional silencing, or position effect variegation (PEV). PEV is an epigenetic phenomenon where if a transcriptionally active gene is removed from its euchromatic region and inserted into a heterochromatic position in the genome, the gene becomes silenced 92-94 This epigenetic effect is heritable through mitosis and meiosis 95,96 Genetic screens have been used extensively in D. melanogaster and S. pombe to identify genes responsible for PEV; these genes fall into two categories, enhancers and suppressors of variegation. To date, there are about 40 loci that have been identified as suppressors of variegation (Su(var)), responsible for chromatin changes resulting in regional silencing 87,92 These genes encode for proteins that are responsible for a variety of functions, including HDACs, protein phosphatases, and heterochromatin-associated proteins 97,98 One particular suppressor in D. melanogaster, Su(var)3-9, has been found to localize to heterochromatic regions, and seems to regulate the architecture of the condensed chromatin 99 i n a recent experiment, it was demonstrated that the human homolog of Su(var)3-9, SUV39H1, is actually a HMT targeting Lys9 of the core histone H3 1 0 ° . In the same study, researchers were able to show that overexpression of SUV39H1 resulted in ectopic heterochromatin formation, demonstrating this HMT's effect on chromatin architecture. Mouse knockouts for this gene resulted in divisional defects in mitosis with an overall decrease in heterochromatin formation, further supporting HMTs' importance in gene regulation 100 Qne of the prime candidates assumed to be partly responsible for 19 establishing a heterochromatin state at H3 Lys-9 is heterochromatin proteinl (HP1); it contains a well-conserved chromo-domain which has been shown to bind to methylated H3 Lys-9 in vitro 101,102 1.2.3.3 Phosphorylation Phosphorylation of H3 tails also appears to be important for chromosome condensation 103. Paralleling the observations seen for other core histone tail modifications, phosphorylation of specific amino acid residues is required for significant and reproducible effects. For example, Ser 10 of H3 appears to be a unique target for phosphorylation, and is associated with transcriptional activation along with Lys-9 acetylation 104-106 1.2.3.4 Histone Code It is thus clear that although histones throughout the genome have relatively generic structures, different modifications to their tails can result in a staggering array of variation. The identification of a rapidly growing number of accompanying enzymes, with multiple specificities for different histone modifications, is providing solid support for the existence of a "histone code" 107. This code maintained at the histone level provides a further extension of information beyond what is encoded by the DNA base pairs themselves. In their review, Jenuwein & Allis 107 identify what key components are necessary for a functional "histone code". First of all, the covalent histone modifications described previously must provide a target for effector molecules/enzymes to bind to. It 20 has been experimentally proven that many enzymes contain unique domains that target histone tail modifications; for example, the bromodomain on many transcriptional regulators has been shown to preferentially interact with specific acetylated lysine residues on histone tails 108-110 Chromodomains, like the one present in the HP1 protein, target methylated histone tails specifically 102. These effector molecules in turn presumably facilitate changes from euchromatin to heterochromatin, and vice versa. Indeed, many transcriptional activators contain bromodomains, whereas in the case of HP1 the presence of a "chromoshadow" domain suggests self-dimerizing abilities leading to inter-nucleosomal condensation 111,112. A second prediction of the histone code theory dictates that there is a further level of complexity beyond individual covalent modifications. It seems that some of these modifications (whether on the same or different histone tails) may be interdependent, and may in fact act in some combinatorial manner to signal final chromatin effects. For example, the transcriptional regulator TAFII250 contains tandem copies of the bromodomains and binds preferentially to diacetylated histone amino acid residues that are correctly spaced 113. In addition to the fact that acetylation, methylation, and phosphorylation all occur in close proximity on core histone tails, all three modifications can be found in both active euchromatin and inactive heterochromatin 107. i n f a ct 5 certain modifications have been shown to act synergistically/antagonistically with each other; Lys9 of H3 for example, when methylated prevents phosphorylation of SerlO 100. Likewise, phosphorylation of SerlO antagonizes Lys9 methylation, but at the same time couples synergistically with Lys9/Lysl4 acetylation, resulting in transcriptional activation 104-106 other modifications of histone tails such as ubiquitination have also 21 been observed, although they are less well understood. They do however, seem to fit in with the complexity of a histone code; for example, it was recently discovered that in the yeast, ubiquitination of Lysl23 on H2B is a prerequisite for methylation of Lys4 on H3, with the end effect of transcriptional silencing 114 Taken together, these observations further suggest that the histone code is more complex than simple associations between acetylated histones and active chromatin, etc.. 1.2.4 Replication Timing It has been known for decades that not all regions of the eukaryotic genome replicate at the same time during S phase 115,116 n w a s a j s o c i e a r those specific regions that were found to replicate earlier than others tend to exhibit this property through cell generation after generation 11?. In humans, these "early" replicating regions correspond to R-bands, which are GC rich and contain a high proportion of transcriptionally active genes. "Late" replicating regions thus correspond to G-bands, which are characterized by a low GC content and a transcriptionally silent heterochromatic state. Centromeres and telomeres fall within this latter category 11^ . The inactive X chromosome, although containing no less GC content than the active X, still exhibits late replication. The initial explanation for this difference in replication timing suggested a simple consequence of how chromatin was physically packaged. Euchromatin containing transcriptionally active genes was more open and thus more accessible for replication origin-binding proteins and other replication machinery. Euchromatin may also require less physical "unwinding" of DNA for replication to occur. 22 While the exact details at the molecular level are still being worked out in the human, it is clear that replication timing corresponds nicely to transcriptional activity. In females, asynchronous replication patterns between the active and inactive X chromosomes during S phase have been observed for over 40 years 119. i n f a ct ; this correlation can be observed at the level of individual gene loci on the inactive X chromosome 120; u s i n g single dot/double dot FISH technique first described by Selig et al. 121 s replication timing of individual genes that were subject to inactivation were compared to that of genes that escaped X-inactivation. Their observations supported the theory that transcriptionally active genes were early replicating while transcriptionally silent ones were late replicating. More interestingly, studies showed that late replication timing can be "spread" in X;autosome translocations 122. i n other words, autosomal regions that normally replicate early in S phase can become late replicating when translocated onto the inactive X chromosome. Evidence for correlation between this spread of late-replication and gene inactivation has been limited however, with only 4 reported cases presently 122-125 j n a r e c e n t study, however, a clinical case was presented where long-range inactivation of autosomal genes on a derivative chromosome resulting from a X;10 translocation was found, but without any spread of late-replication 126 1.2.5 Histone Variants Besides XIST, all the features of X-inactivation discussed thus far are general characteristics found in all heterochromatin and not exclusively on the inactive X chromosome. Expectedly, great interest was shown over the past 10 years on another 23 feature of X-inactivation that was exclusive to the inactive X; the recruitment of macroH2A to the Barr body. Initially isolated from rat liver nucleosomes 41, and subsequently from human liver tissue 127, the macroH2A (mH2A) protein is a class of histone-like variants that preferentially colocalizes to the inactive X chromosome. 1.2.5.1 macroH2A The gene for macroH2A encodes a protein three times the size (roughly 42kDa) of the core histone H2A. The N-terminal third of macroH2A is similar to H2A, although the C-terminal of macroH2A contains a large non-histone domain extension of unknown function 41. 1.2.5.1.1 macroH2Al To date there are several variants of macroH2A, one of which is macroH2Al. MacroH2Al has two isoforms that arise from alternative splicing called macroH2Al.l and macroH2Al .2, which differs only by a short stretch of amino acids within the non-histone domain corresponding to the 6th exon 127 ( s e e F ig U r e 1-2). Tissue-specific expression analysis showed that in the tissues examined, levels of spliced RNA for macroH2Al.2 were much more abundant than those for macroH2Al.l 128. More specifically, tissues that contained mostly few- or non-dividing cells such as the liver, kidney and brain demonstrated significant levels of expression of both variants, whereas tissues such as the thymus and testis showed detectable expression of only the macroH2A1.2 variant 129. 24 mH2Al A T G 1.2 S T O P • I • • • i n m rv v >v i *vn vm rx vi 1.1 mII2A2 A T G S T O P n m rv v vi vn vm Exon Intron Figure 1-2: mH2A Variants The two mH2Al variants are transcribed from the same gene but alternatively spliced. The mH2A2 gene resides on a completely different chromosome, but its sequence structure is highly conserved with the mH2Al variants. (Exons/introns not drawn to scale.) 25 1.2.5.1.2 macroH2A2 In 2001, Pehrson's group published a paper describing the discovery of a new macroH2A subtype, macroH2A2, using BLAST searches of expressed tagged sequences (ESTs) 130 A search through the human genome sequence revealed that the human gene for macroH2A2 is located on chromosome 10, as opposed to macroH2Al, which is located on chromosome 5. The study's analysis showed that macroH2A2 is, overall, 68% identical to macroH2A1.2 in amino acid sequence. 1.2.5.1.3 macroH2A-Bbd Another group of researchers using BLAST searches to identify novel H2A-related proteins, identified another core histone variant that appears to be nuclear but excluded from the Barr body and the Xi 131. Termed H2A-Bbd (Barr body deficient), the researchers observed a uniform nuclear distribution in male cells, whereas in female cells they found a number of "exclusion zones" to be one less than the total number of X chromosomes in different cell lines containing different number of X chromosomes. 1.2.5.2 Role in X-inactivation 1.2.5.2.1 Macrochromatin Bodies Early immunoflouresecence studies on macroH2A1.2 concentrated on determining its distribution pattern within the nucleus. It was shown that macroH2Al .2 localized into dense regions termed "macrochromatin bodies" (MCB). This was in stark contrast to macroH2A1.2 staining on autosomes, which although present, was quite faint 26 in comparison. These MCBs were mostly found in females, and more interestingly they colocalized with the Barr body. Further evidence of macroH2Al .2's involvement in X-inactivation was provided by studies in which the Xist gene was deleted in mouse embryonic fibroblasts in vitro; in these cells, MCB formation was disrupted, and the inactive X chromosome showed no greater colocalization of macroH2A1.2 than autosomal chromosomes 132 More convincingly, a recent study demonstrated that MCB formation can be induced by Xist expression alone 17. Utilizing ES cell lines that allow inducible expression of Xist RNA from Xist genes inserted at autosomal sites, these researchers were able to nucleate the formation of ectopic MCBs on autosomes expressing Xist RNA. Further more, studies have looked at the XY-body (or "sex vesicle") for clues to how X-inactivation occurs in the meiotic prophase of the male germline. This XY-body reflects the inactive, condensed state of the sex chromosomes during meiotic division, a mechanism that's been suggested to help prevent recombination between the X and Y chromosomes 133 Some experimental evidence suggests the possibility of Xist involvement in this male germline X-inactivation by demonstrating that Xist is transcribed only in these cells and not in other male tissues 134-136 Subsequent FISH studies revealed that macroH2A1.2 is also concentrated in the XY-body of early pachytene spermatocytes, further supporting the idea of a XIST RNA-macroH2A1.2 interaction 137. More convincingly, chromatin immunoprecipitation experiments targeting macroH2A were able to pull down Xist RNA 40. 27 1.2.5.2.2 Sequence Analysis Core histones in general are extremely well conserved between most eukaryotes. This is in agreement with the idea that histones play a crucial universal role in packaging genomic material in these organisms. Sequence comparisons reveal that macroH2A variants are also well conserved between species that have diverged a long time ago in evolution. For example, chickens and rats split in evolution roughly 300 million years ago, yet they still show tremendous conservation in their mH2Al.l and mH2A1.2 amino acid sequence: an overall 95% in mH2Al.l and 96% in mH2A1.2 138. The above comparison does not provide much clues about potential unique macroH2A functional domains, given that core histone sequences are well conserved as a rule, and the mH2A variants all share an H2A-similar histone N-terminal. A comparison of the non-histone tail sequence with other organisms will thus provide more useful information. This non-histone region of mH2A1.2 shows significant homology to a gene in some bacteria, roughly 34% with Alcaligenes eutrophus 139 and 30% with E. coli 138. The actual function of the bacteria gene remains to be identified. More intriguing is the homology found between the non-histone tail and some positive-strand RNA viruses: 24% identity to the sindbis virus and 25% to the rubella virus 140,141 j n e region of homology found in the sindbis virus is actually a part of a protein that associates with viral RNA and is required for synthesizing the negative-strand of the RNA virus 142,143 This then lends support for the theory that macroH2A variants can bind RNA species within the cell, a possible candidate being the XIST RNA. 28 1.2.5.2.3 Functional Domains A recent analysis of the macroH2Al and macroH2A2 sequences, with the availability of a continually changing human genome database, showed that the two histone variants share an overall 80% identity in their amino acid sequences 144 j n a n attempt to dissect which region(s) of macroH2A was crucial for MCB formation (and thus candidates for domains interacting with XIST), Chadwick et al. fused green fluorescent protein (GFP) to different truncated macroH2Al protein regions and observed the fusion proteins' ability to form MCBs 144 T n e histone-similar tail (i.e. the N-terminal third of macroH2A), marked by a GFP fusion, demonstrated MCB formation when transfected into human fibroblast cells. The non-histone two-thirds of the protein failed to show localization of any kind however, and was distributed throughout the nucleus and cytoplasm. To further identify a specific region responsible for MCB formation, the researchers fused the histone-similar tail of macroH2A that had targeted disruptions to GFP to be used in further transfections. The group identified a stretch of 19 amino acid residues that seemed to be essential for MCB formation; this amino acid sequence was conserved between mH2Al and mH2A2, but different from that of the core histone H2A. This domain was in turn termed the macrochromatin domain (MCD) 144 (see Figure 1-3). Interestingly, the non-histone tail also seemed to harbor a MCD of its own. In the same study, Chadwick et al. found that a fusion protein of the core histone H2B, non-histone tail of macroH2Al, and GFP formed a MCB when transfected into human fibroblast cells 144. This demonstrated that when localized to the nucleosome, the non-histone tail of macoH2Al could induce MCB formation. In addition, the use of a H2B 29 fusion instead of a H2A fusion suggests that the position of where this non-histone tail lies in the nucleosome structure does not seem to affect MCB formation efficiency. Instead, MCB formation efficiency appeared to be dependent upon the non-histone tail variant used: H2B fused to the mH2Al.l non-histone tail had much lower localization efficiency than a H2B-mH2Al .2 non-histone tail fusion protein 144. This difference seems to be attributable to a short region between amino acid position 120-175, which showed only 35% identity between mH2Al.l and mH2A1.2 as compared to 68% identity for the rest of the two variants. 30 H2A Nonhistone 19 A A 55 A A M C B region Localization region Figure 1-3: Functional Domains of m H 2 A l Both macroH2Al.l and macroH2A1.2 variants share similar important domains for MCB formation. The H2A tail is essential for proper localization and incorporation into the nucleosome, while two independent domains (one in each H2A and non-histone tails) are capable of MCB formation. 31 1.2.5.2.4 Involvement in Heterochromatin Many experiments have provided evidence of macroH2A's participation in heterochromatic structures. When it was first identified, macroH2Al .2 had been shown to replace the core H2A in some nucleosomes; in the rat it was estimated that there exists roughly one macroH2A molecule for every 30 nucleosomes throughout the genome 41. This was confirmed by experiments using macroH2A1.2 to artificially reconstitute chromatin. Reconstituted nucleosomes made from either H2A or macroH2A were both able to protect roughly 146bp of DNA from micrococcal nuclease digestion, showing that both demonstrated similar nucleosome structures, and that both H2A molecules in a nucleosome can be replaced with macroH2A1.2 145. There were however, some subtle but interesting differences between the two types of reconstituted nucleosomes. When digested with DNase I, the digestion pattern revealed that a particular region in the macroH2A nucleosomes was more resistant than a corresponding region in the H2A nucleosomes 145 This location in the protein is known as the "superhelix location zero" (SHLO) when referring to its crystallographic structure, and corresponds exactly to the non-histone tail of macroH2Al 64 A second difference between the two types of reconstituted nucleosomes was observed in sedimentation studies, in which macroH2A mononucleosomes preferentially associated in pairs 145. This supports earlier evidence that the non-histone tails of macroH2Al.l can form dimers 146. Taken together, the above observations suggest that although macroH2A does generally play a role in heterochromatin, it also contains unique properties that allows for a tighter and more specific DNA packaging role(s). One example of a special case of 32 heterochromatin formation in which macroH2A1.2 seems to play a role is meiotic sex chromosome inactivation, during which the sex chromosomes condense in order for proper pairing and segregation 137,147 j n addition to preferential localization to centrometric heterochromatin, macroH2A1.2 was found to colocalize to the pseudoautosomal region in male cells along with M31, a mammalian member of the conserved HP 1 protein family 148. 1.2.5.2.5 Possible Mechanism of Interaction The precise function of MCBs in the inactivation of the X chromosome remains unclear; however, studies using immunofluoresence have revealed some interesting clues. To better understand the timing of macroH2A1.2 recruitment, one study utilized differentiating murine ES cells to study MCB formation over time 149 j n m o u s e undifferentiated female ES cells, both X chromosomes are initially active, with random X-inactivation occurring during differentiation. By double-labeling macroH2A1.2 and Xist RNA, the study demonstrated that MCB does not colocalize with the Xist RNA until after the initiation and establishment of X-inactivation has been completed 149. l n a follow-up study, the same research group provided evidence that a "pool" of macrH2Al .2 is continually contained at the centrosome (in both males and females) prior to and throughout the mouse ES cell differentiation, before it becomes colocalized to the inactive X chromosome in the female cells (in the male cells, the macroH2A1.2 becomes dispersed) 1^ 0. These experiments suggest that macroH2A1.2 recruitment is a relatively late event in the sequential steps of X-inactivation, and is probably more important in the maintenance of the inactive state than in initiation and establishment of the Xi. 33 Nevertheless, there is no disputing the fact that experimental evidence also suggests an intimate relationship between Xist RNA and macroH2Al .2. Chromatin immunoprecipitation experiments targeting macroH2Al pulled down Xist RNA 40 . Using the Cre-loxP system, Csankovszki et al. observed female mouse fibroblasts in the absence of the Xist gene 132 They observed that although the inactive X chromosome still retained features of heterochromatin, such as silenced genes, late replication, and hypoacetylated H4, the macroH2A1.2 localization was lost. This suggests that macroH2A1.2 may not be crucial for maintenance of heterochromatin, but may be more important in mediating actions between XIST and the establishment of X-inactivation. Xist RNA has previously been shown to interact with the nuclear matrix; even after DNA and histones were extracted Xist RNA remained nuclearly localized 25. MacroH2Al .2, as described above, is tightly associated with nucleosomes. Thus, an interaction between Xist RNA and macroH2Al .2 would provide a convenient link between the inactive X chromatin and the nuclear matrix. However, whether XIST RNA physically interacts with and recruits the histone variant to the Xi remains to be proven. 1.3 Rodent/Human Hybrid Cells Rodent/human somatic cell hybrids were initially created for mapping purposes, as they retain few human chromosomes after fusion. Hybrids containing either the active or inactive human X chromosome can be selected for in culture 151,152 The hybrid AHA-1 laBl, was a subclone of a monochromosomal hybrid line AHA 11a (GMO10324) originally obtained from Dr. H. F. Willard (Case Western Reserve University). These somatic cell hybrids containing an active human X chromsome in a mouse background 34 were repeatedly treated with a demethylation agent such as 5-azadeoxycytidine (5-azadC), which inhibits methyltransferase activity, in order to reactivate XIST expression 153 Because most genes on the human active X chromosome lack methyl groups at their 5' promoter regions, 5-azadC will have little effect on their expression status. The XIST gene, however, is transcriptionally silent on the active X 153,154 Digestion of genomic DNA using methylation-sensitive restriction enzymes has revealed multiple methylation marks at the 5' end of the XIST gene 155. Initial expression analysis of X-linked genes in this cell line (AHA-11 aB 1 -A52b) revealed that only the XIST gene was affected by the demethylation treatment. Further rounds of treatment with 5-azadC resulted in hybrids that expressed both human XIST and mouse Xist37 (see Table 1-1). 35 Table 1-1: Human Mouse Hybrid Cell Lines Cell Line XIST Expression Localized toXi Xist Expression Localized toXi AHA-llaBl Human Xa + Mouse Xa NO - NO -AHA-llaBl-A52b Demethylated YES NO NO -AHA-1 laBl-A52b-4Cl Further demethylated YES NO YES YES 36 1.3.1 XIST Localization A handful of experiments have been done involving such (demethylated) hybrids, and their unexpected results identified some intriguing questions concerning XIST/Xist RNA localization and the RNA's ultimate function in X-inactivation. Generation of the AHA-1 laBl-A52b mouse/human hybrid cells was initially reported by Tinker & Brown 153 j n a n experiment where they demethylated mouse/human cells that contained an active human chromosome. The XIST gene was found to be reactivated, and the XIST RNA molecules were confirmed to have a nuclear localization. Using RT-PCR, they examined whether the expression status of 8 X-linked genes were affected by the reactivation of the XIST gene; despite previous evidence of XISTs importance in X-inactivation, none of the genes examined were silenced after XIST reactivation. This was confirmed with an azaguanine-thioguanine (AG-TG) growth assay, in which the media was supplemented with AG-TG and only cells that had lost expression (presumably via induced X-inactivation) of the X-linked gene hypoxanthineguanine-phosphoribosyl transferase (HPRT) would be unable to incorporate the toxic 8-azaguanine and 6-thioguanine. A similar study in the same year was performed using hamster-human somatic hybrid cells 154. The cells were put through the same demethylation process, and once again XIST RNA was found only in nuclear extracts of these cells. RT-PCR was also used to examine gene expression changes, and the two genes they looked at appeared to show no signs of induced silencing at all. More interestingly, when the researchers applied RNA FISH analysis on the induced XIST RNA they observed a general diffuse 37 localization, unlike normal female mammalian cells in which the XIST RNA tightly coats and "paints" the inactive X chromosome. Clemson et al. confirmed this unexpected result3 7 shortly after in mouse/human somatic cell hybrids. In fact, not only did the XIST RNA from a demethylated human Xa-containing hybrid demonstrate aberrant derealization, the XIST RNA from a normally inactive human X chromosome in Xi-containing hybrids also showed a similar deficiency in proper localization! Thus, despite recent identification of domain regions on the XIST RNA crucial for its localization to the Xi, a tight association with the X chromosome does not appear to be a pure innate feature of the XIST RNA itself. Clemson et al. 37 suggested that perhaps certain species-specific factors were missing in the somatic cells hybrids that interacts with XIST RNA and assists in proper localization. This theory is supported by their further observations that mouse Xist RNA appeared to localize properly in their demethylated mouse/human hybrids 37 ( s ee Figure 1-4). Much like previous observations however, the X-linked genes on the mouse X chromosome showed no signs of silencing despite proper Xist RNA localization. 38 Figure 1-4: Reactivated XIST/Xist RNA FISH analysis of AHA-1 laBl-A52b-4Cl hybrids. The green signal represents XIST RNA, which is drifting off the X chromosome. The red signal represents Xist RNA, which is localized correctly to the X chromosome. {Clemson et al 1998)37 1.3.2 Ideal Model As a result of the experiments described above, we have now at our disposal a unique hybrid model system in which to study X-inactivation. The key features of this model system are as follows: i) Active X chromosomes (human and mouse) are present, containing genes that have not been subjected to X-inactivation. ii) These hybrid cells have been demethylated to reactivate the XIST gene; the resulting XIST RNA is delocalized. iii) Some hybrid cells have been further demethylated to reactivate the Xist gene; the resulting Xist RNA is localized. iv) None of the X-linked genes examined thus far demonstrate any signs of silencing as an effect of induced XIST/Xist expression. In short, having all these features in a single cell line provides a unique system where one can begin to dissect the role of XIST RNA independent of gene silencing. As mentioned before, X-inactivation occurs in a cascade of events, with many features of X-inactivation appearing together in a brief window of time after XIST expression during development. This model system now allows us to investigate more rigorously how XIST RNA may function independent of complete X-inactivation, as well as how individual features of X-inactivation are affected by induced XIST expression. 40 1.4 Thesis Goals In an effort to utilize a unique somatic hybrid cell model to dissect and investigate individual features of X-inactivation, I have attempted the following: i) Verify the demethylated somatic hybrid cell lines as an useful model system for studying the role of XIST in X-inactivation independent of gene silencing; ii) Investigate the role of macroH2A in XIST localization, and X-inactivation in general, using the model system; iii) Characterize the histone acetylation status, another epigenetic feature of X-inactivation, in this model system. 41 CHAPTER 2: Materials & Methods 2.1 Tissue Culture Methods 2.1.1 Cell Lines The human/rodent hybrid cell lines were originally either obtained from H.F. Willard (Case Western Reserve University) and S. Hansen (University of Washington), or they were generated by S. Baldry using somatic cell fusion techniques described elsewhere 152,156,157 Once thawed from our lab stocks, cells were grown at 37C in alpha-Minimal Essential Medium (alpha-MEM) purchased from Gibco/BRL. To collect cell pellets for DNA and/or RNA extractions, the growth medium was first removed and the cells rinsed once with phosphate saline buffer (PBS). 0.25% trypsin-EDTA was used to release cells from the bottom of the culture flask, and these were resuspended in lOmL of growth medium. This cell suspension was spun down, and the supernatant aspirated leaving the cell pellets. 2.1.2 Freezing Cells The protocol for freezing cell lines in liquid nitrogen is simply a slight modification of the above protocol for obtaining cell pellets. Once the trypsin and normal growth medium has been removed, the cell pellet was resuspended in alpha-MEM enriched with 15% fetal calf serum (FCS) and 10% dimethyl sulfoxide (DMSO). This was then transferred to appropriate cryotubes, and placed at -70C in an isopropanol-filled container to allow for gradual cooling before long-term storage in liquid nitrogen. 42 2.2 Examining Gene Methylation Status In order to determine whether genes were methylated at their promoters, genomic DNA was first extracted from the cells of interest. This DNA was then digested with appropriate enzymes, before being used as a template for polymerase chain reactions using suitable primers. 2.2.1 DNA Extraction To obtain DNA samples, cell pellets were resuspended in 10 mL of TE (lOmM Tris pH7.5-8.0; ImM EDTA) and 1/20 volume of 20% sodium dodecyl sulfate (SDS). A pipette-tip-full of proteinase K (6mg) was added and the whole solution was incubated at room temperature overnight. The SDS (a detergent) and proteinase K (a proteolytic enzyme) works to rupture cell membranes and digest proteins, allowing the release of genomic DNA into solution. The following day, 0.8mL of 5M NaCl was added and incubated at 37C for 2 hours until in solution, after which 3.3mL more of 5M NaCl was added. The whole solution was shaken vigorously for 30 seconds, then centrifuged at 2500 rpm for 15 minutes. The supernatant was transferred to a 50mL tube, followed by the addition of 0.3mL of 20% SDS and 3.3mL 5M NaCl. The solution was shaken vigorously for 30 seconds before it was centrifuged at 2500rpm for 15 minutes. The supernatant was transferred to a fresh 50mL tube, and 2x volume of ethanol was added to precipitate the DNA. The DNA was gently spooled and transferred to a fresh tube, where it was resuspended in double distilled water (ddH20) and stored at 4C. 43 2.2.2 Methylation-sensitive Restriction Digests Genomic DNA from cells of interest was first pre-digested using EcoRI in order to cut the DNA strands into smaller fragments, followed by phenol/chloroform extraction and ethanol precipitation. Roughly 20-200ng of the pre-digested genomic DNA was incubated at 37C overnight in a reaction mix (total volume 20uL) with luL of a methylation-sensitive restriction enzyme (10 Units) and luL of lOx of the appropriate buffer. Methylation sensitive enzymes Hpall (NEB) and Hhal (NEB) were used; these enzymes will not cut if there is methylation of the C in CpG within the recognition sites. In the case of Hhal digestion, the reaction mix was supplemented with lOOug/mL bovine serum albumin (BSA). luL (lOOng) of the digestion product was then used as a template for subsequent PCR reactions. Species-specific primers were designed to flank the enzymes' recognition sites without flanking EcoRI sites (see Table 2-1). 44 OX) g '8 a s-•** S Q « «S a ixi © .S PH t> « o =J CD "5 N 2 bo - a W- .3 ^ PH bo J3 cu =tfc " O so Of! °<5 o C i f J eu a* „ 00 H CU B "3 "C OH PH C L O ro £ E 1 • H - N I I I T f T f CN Os >/-> t~-2 E < U E— U U u a u u H U H U H U a < O u < < a u u H < H a a < u a < u < H C/5 >< < os CN a. x T f os CN CO a. PC cd X PC o ro c c e £ £ £ — - H (N I I I cn oo CN os >n r~ s E o < O H U < u a H U H H a a o u u u u o a < < a < a < o H U o < < u u CM ft CQ in H in 00 ft ro ft OJ >/o >n X) rep ro OS X rep CN CN CS ft PC c c e £ £ 1 — — CN I I I VI O N C\ m r-U a H a u a < a < u u a H u < < a < u u H < U H U U a H U a o a < < o t-o. H U O co ft PC cs X! PC c c e £ £ £ — —i CN I I I T f T f CN Os >n f~ E o < U U H U U H U a < u u < u u < u u E -H U < u a u u u u o < < u H U u a E-U a a a a E -o H U a i x LH ft UU ft X CN os ro ft PC CS X PC o ro c c e E £ £ — — CN I I I T f CN as m r~ s E < u u a u ta S w ta OH S 3 w CS H o o H o u o < < o o H E— o H CJ H < O o < O o o < o H H o o E-u o CJ E-o o o CJ < o < < < u o c j G CJ O < CJ CJ E-< O < o o CJ < < o CJ CJ E-o CJ o E— o H U CJ CJ CJ H o o o E-< < CJ G S-H < O H H G H O < O H o G G t— CJ CJ G O H CJ H CJ E-G H O o G G G E-CJ E-< b-G H CJ O G G o < CJ H G G G < H H E-CJ H < CJ H G G < o G ft. ft. s H CJ Q ft. X o-B °3 "G rt rt (50 < 8 « o o rt i t 3 C 1 e O-fN .S3 ^ T3 rx T3 tU C < fN X s C .o u J? if-a u H r»j U O PP r i 4 3 03 O O T3 C 03 T3 rt GO d o3 a o o T3 Cft > 3 i_ T3 JO £ O o CJ a a > a « P-1 0 - a E 8 a ° 1 3 •§ H ^ cd 0 0 "2 •a £ °-o 5 U a c o o 43 tfi CJ .3 cn -rt a e « cd •o is OJ *-i Cft ,OJ 13 8 4 - 1 o c is ffi < < 2 3 DG i s 2 3 e > co ^ -A s & & -a 03 X) S £ .9 « f i CN < U < Q •o CJ c T 3 fl fe a GO c O SJ S4—• Of) C 2 H fN < fl 1 3 CO cu rv t j E s - i 00 i2 ca ta 2 -fl M ' S cj cu S ° CD 1 s .S is -fl oo U SS 3 £ •>-* oo •I-* — CU cj 15 cu 0= "C X •— C J T 3 fl 3 cu o X fN < a 00 C * -It) & < fN g 5 2 > o "O -rt « n < 00 CH GO "c3 s - O CU 3 O X 89 oo s © • •H *-> CJ .33 3 S •~ H y « N • CQ cs i < X < • s-3 DC • PM 5 _Cu g I . CD u 5 •4-» I—H °< ^ < e Q a 0 3 « S OH £ PH - O O £ c+-s ° tn _0 o 3 3 CJ 0 3 *-» a 2 ccj >H +H C CJa c o CJ fl cSfe O DC •S > !fe o O & a -a 3 cu j - ° -T 3 O CU >-H-> cu CU 00 cn C 3 ,5 CS t+a ° O "io 73 O 3 es U T3 Cu 3 S « 3 a co 1—1 o y £TJ" 00 3 0 3 cj cs CJ O 4.5 Discussion To immediately tackle the question of whether macroH2A is the missing human factor responsible for proper XIST RNA localization, the expression status of different macroH2A variants was examined in various hybrids that had delocalized XIST RNA. The variable expression of macroH2Al.l/macroH2A1.2 (determined by RT-PCR) seen in the different hybrids failed to provide definite answers. While some hybrids showed the presence of macroH2Al.l/macroH2A1.2 mRNA, others didn't. The hybrids expressing macroH2Al .2 proved that macroH2Al .2 by itself is not the lone factor responsible for XIST RNA localization; otherwise it would have restored proper XIST RNA localization. However, there are still a couple of possible explanations. MacroH2Al.l/macroH2A1.2 may not be important at all for XIST A localization; while past experiments have shown that XIST expression is necessary for MCB formation, it does not mean that macroH2A expression is required for XIST RNA localization. If indeed macroH2Al. l/macroH2Al .2 does have a role in XIST localization, it is likely neither of the histone variants is solely sufficient for XIST coating of the X chromosome. However, it cannot be ruled out that macroH2Al .2 may still play a crucial part in a multi-factor process. The localization studies involving the macroH2Al.2-GFP constructs provided some insights towards the relationship between XIST and the histone variant, while at the same time prompted a number of intriguing mysteries. As mentioned before, previous work in the literature has demonstrated that prevention of Xist expression resulted in lack of MCB formations. However, in the demethylated AHA-1 laBl-A52b hybrid cells examined here, MCBs were not formed despite having XIST expression. This leads to the 91 conclusion that XIST gene expression alone is not sufficient for MCB formation. Most likely, the physical association of XIST to the X chromosome is required for proper macroH2Al .2 localization, as reflected in the normal MCB formation seen in the transfected BOSC 23 cells. Alternatively, macroH2A1.2 localization may be dependent on i)XIST expression, and ii)some other species-specific factor that directly brings macroH2Al .2 to the X chromosome, which is absent in the hybrid cells. It is also possible that this same species-specific factor may be responsible for proper XIST RNA localization. In either case, the possibilities all point towards the existence of some ribonucleoprotein complex involving X75T RNA, macroH2A1.2, and other factor(s). Upon closer comparison of the images taken from transfected AHA-1 laBl and AHA-1 laBl-A52b hybrids, the distribution of GFP signals in the two cell lines appeared to be slightly different. As expected, the macroH2A1.2-GFP fusion proteins demonstrated a diffuse pattern of distribution within the nuclei of non-demethylated hybrids (i.e. no XIST expression). In the X/ST-reactivated AHA-11 aB 1 -A52b hybrids, however, there seems to exist small regions of slightly higher density of GFP signals. Although no true MCBs were formed it is possible that, according to the ribonucleoprotein complex theory, the macroH2A1.2-GFP proteins are somehow interacting with XIST RNA. Since in these hybrids XIST RNA has been shown to be delocalized, and to drift off into a more dispersed region away from the X chromosome, the GFP signal patterns observed here could be reflecting the drifting off of macroH2A1.2 associated with XIST RNA. The observations strongly support the theory that XIST expression alone is not sufficient for MCB formation, and that macroH2A1.2 localization to the X chromosome requires XIST RNA colocalization as well. 92 The most startling result came from the AHA-1 laBl-A52b-4C1 transfections. The main difference between these hybrids and the AHA-1 laBl-A52b hybrids is the expression of Xist, and the proper localization of the Xist RNA, in the 4C1 cells. Yet the macroH2A1.2-GFP fusion proteins demonstrated a drastic change in its localizing behavior. Initially, it was hypothesized that the introduction of localized Xist RNA would provide a different result than the presence of delocalized XIST. Combined with the fact that human and mouse macroH2A1.2 share a high sequence similarity, and that histones and histone variants are well conserved between the two species, it was expected that perhaps the localized Xist RNA will bring about proper MCB formation. The large signal seen at the periphery of the cell nucleus, which is 2-3 times the size of normal MCBs seen in BOSC 23 cells, raises several possibilities. Firstly, the dispersed "pseudo-MCB" could reflect an incomplete formation of a proper MCB. For some reason, the human macroH2Al .2-GFP cannot fully respond to the Xist RNA that is coating the X chromosome. Although it is assumed that the macroH2Al .2 is highly conserved between the two species, it is still possible for small species-specific differences within the non-histone tail to have a slight interfering effect on proper interaction with other proteins and/or RNA molecules. It is worth noting that the X chromosomes in these hybrids are still active; thus, unlike the inactive X chromosomes in normal cells, the active X chromatin may not have the same degree of compact packaging. So even with proper Xist localization, the more "loose" chromatin of the active X chromosome might also contribute to the larger MCBs observed. However, further investigation using mouse embryonic fibroblasts suggests that there are very likely subtle species-specific factors involved in MCB formation. 93 Essentially, a female MEF transfected with the same macroH2A1.2-GFP construct demonstrated a diffused GFP signal identical to the ones seen in the AHA-1 laBl-A5 2b-4C1 hybrid transfections. In this case, human macroH2A1.2 seems to exhibit this unexpected localization despite being in a cell that has proper Xist expression, Xist localization, and X-inactivation. The MEF transfections thus provide greater support for the theory that slight species-specific differences in macroH2A1.2 do in fact affect the effectiveness of MCB formations. Another factor for consideration is the possible role of centrosomes in these hybrids. As has been previously demonstrated, macroH2Al .2 proteins are initially "stored" at centrosomes just prior to the beginning of X-inactivation in a cell, before its eventual migration to the inactive X chromosome within the nucleus. It is possible that XIST/Xist expression triggers some signal to direct an upregulation of expression, and concentration of macroH2A at the centrosome in anticipation of X-inactivation 160 j^g dispersed MCBs observed in the MEFs and AHA-1 laBl-A52b-4C1 hybrids may be reflecting a concentrating of macroH2A1.2 at the centrosome near the periphery of the nucleus, presumably persisting there due to some incompatibility between human macroH2Al .2 and mouse Xist RNA (or some other factor) resulting in an inability to migrate to the inactive X chromosome. Alternatively, the large GFP signal could also simply be a reflection of heterochromatic "junk" accumulating in unhealthy and/or dying cells. The lipofection procedure is quite straining on cells, and it possible that after transfection many of the cells observed are critically damaged. However, since all cell types were subjected to the same procedures, one would expect that such "heterochromatic globs" be present at a 94 certain frequency in all the hybrids examined. This is obviously not the case, as the dispersed MCBs were only observed in the AHA-1 laBl-A52b-4Cl hybrids. Taking these results together, it is clear that: i)XIST expression alone is not sufficient for MCB formation, and that other factor(s) such as XIST RNA localization is (are) important; ii)A7STRNA probably has some physical interaction with macroH2A1.2 either directly or indirectly via a ribonucleoprotein complex; iii)species-specific differences exist within the sequence of human and mouse macroH2Al .2 that affects proper MCB formations. Any species-specific differences probably acts either by interfering with direct interactions between Xist RNA and human macroH2A1.2, or by affecting a third-party protein/factor. 95 CHAPTER 5: Discussion and Future Directions The initial goal for this study was to characterize in greater detail the hybrid model system proposed for the study of X-inactivation. Previous work has established the human/mouse hybrids containing an active X chromosome as unique cell systems where XIST/Xist can be reactivated, and where the XIST RNA demonstrates abnormal localization 37. The DNA methylation, H3 acetylation, and gene expression experiments in this study confirmed that despite having reactivated XIST/Xist, no silencing of any X-linked genes was observed as a consequence of the reactivation. This hybrid model system thus provides several advantages and disadvantages for the study of the role of XIST/Xist independent of gene silencing. The primary advantage to this system is of course the presence of XIST/Xist expression without a final inactivated state of the X chromosome. In essence, the primary major observable step in the X-inactivation process has been "teased" out of and separated from some of the tightly linked cascade of events leading to heterochromatin formation. We can thus begin to investigate more rigorously what immediate effects XIST/Xist RNA have by identifying what these functional RNAs interact with. The fact that the human XIST RNA is delocalized in this system provides a further advantage in studying XIST function; some factor(s) important in physically localizing the RNA to the Xi must be deficient in these cells, and thus these cells provide perfect vessels for introducing and testing candidates for such "missing factors". However, the delocalized XIST in these cells is also indicative of a defective X-inactivation system. Because the human Xi is in a mouse cell background, many other factors encoded on autosomal regions will be missing in these cells. Therefore, there will always be an uncertainty 96 about whether an abnormal behavior of XIST (and inactivation in general) in these cells is due to some unknown factor. Nevertheless, this study has demonstrated that these hybrid cells are useful tools in studying certain aspects of X-inactivation. More specifically, factor(s) that may have a direct physical interaction with XIST can be investigated in these cells. Proteins considered to be candidates for being participants in the ribonucleoprotein complex with XIST RNA can be introduced into this hybrid system, and any change in XIST localization can be examined. Moreover, any such factors identified will likely be species-specific, since Xist RNA localizes normally in these cells. A more exciting possibility for this hybrid system is the study of factors involved in the potential for X-inactivation beyond the "developmental window" 38, Once again, one can complement these cells with human specific candidate factors, and look for signs of XIST localization and other features of X-inactivation, such as DNA methylation and, histone acetylation. This study investigated macroH2A1.2 as the first candidate for possible interaction with XIST RNA. The results, although not completely conclusive, confirmed and further defined the nature of the relationship between the histone variant and XIST RNA. XIST localization appears to be required for proper MCB formation, which suggests a physical relationship with macroH2Al .2. The aberrant localization of macroH2Al .2 in response to Xist expression identifies that there may be species-specific factors involved in this relationship. Immediate follow-up experiments to this study include further ChlPs targeting mH2Al.l and mH2A1.2 to confirm if there is a direct physical interaction with XIST/Xist RNA. More sophisticated cytogenetic techniques must also be used to clarify the images presented in this study, such as FISH labeling of 97 the X chromosome in order to pinpoint more accurately where the GFP signals are distributed, as well as RNA-FISFf to visualize XIST RNA localization simultaneously. One consideration for future investigation into macroH2A1.2's functional role in mammalian silencing of the X chromosome is the imprinted version of X-inactivation. In some mammals, cells undergoing early development are subject to imprinted X-inactivation, in that the paternally inherited X chromosome is preferentially inactivated, and these cells eventually give rise to the extra-embryonic tissues 161. Past experimental evidence has shown that macroH2A1.2 colocalization on the inactive X chromosome occurs much earlier in the imprinted form of X-inactivation 162 The significance of this difference between the two types of X-inactivation remains unknown, although it does suggests that perhaps the role of macroH2A1.2 is more important in one than the other. Could it be that macroH2A1.2 is important in imprinted inactivation of the paternal X chromosome, thus is recruited immediately after XIST RNA coating, and that MCB formation in random X-inactivation is simply a non-crucial "residual" effect? Indeed, the fact that the loss of macroH2A1.2 on the inactive X does not lead to reactivation supports the hypothesis that macroH2A1.2's role may not lie in the actual maintenance of the inactive state 163 There has also been skepticism about the actual uniqueness of macroH2A1.2 localization to the inactive X chromosome. Perche et al. recently showed that many different histones, and not just macroH2A, were found to be much more concentrated on the inactive X chromosome than on autosomes 164. Their staining technique revealed higher signal intensities for H2A, H2B, and H3, and suggested that macroH2A only appears to be tightly associated with the inactive X chromosome because of the highly 98 heterochromatic packaging of the chromatin. This claim was soon challenged by experiments using epitope-tagged H2A and H2B showing no significant increase in intensely staining on the inactive X chromosome 165. Another factor to consider is that it is very probably that macroH2Al .2's role in the mammalian cell extends beyond X-inactivation. The very fact that abundant expression of macroH2Al has been observed in both sexes is indicative of functions that are not involved in X-inactivation 128 The main problem thus remains: most experiments on macroH2A have been involved in visually characterizing localization patterns. Although these studies are all important in identifying the possible existence of an intimate role in X-inactivation, the identification of interacting factors (possibly from the hypothetical ribonucleoprotein complex) will be needed in order to fully dissect the true relationship between XIST RNA and macroH2A. 99 REFERENCES 1. Lyon, M.F. Gene action in the X-chromosome of the mouse (Mus musculus L.). Nature 190, 372-373 (1961). 2. Heard, E., Clerc, P. & Avner, P. X-chromosome inactivation in mammals. Annu. Rev. Genet. 31, 571-610 (1997). 3. Gartler, S.M. & Riggs, A.D. Mammalian X-chromosome inactivation. Ann. Rev. Genet. 17, 155-190(1983). 4. Sharman, G.B. Late DNA replication in the paternally derived X chromosome of female kangaroos. Nature 230, 231-232 (1971). 5. Cooper, D.W., VandeBerg, J.L., Sharman, G.B. & Poole, W.E. Phosphoglycerate kinase polymorphism in kangaroos provides further evidence for paternal X inactivation. Nature New Biol. 230, 155-157 (1971). 6. Graves, J.A.M. Mammals that break the rules: genetics of marsupials and monotremes. Annu. Rev. Genet. 30, 233-260 (1996). 7. Carrel, L., Cottle, A.A., Goglin, K.C. & Willard, H.F. A first-generation X-inactivation profile of the human X chromosome. Proc. Natl. Acad. Sci., USA 96, 14440-14444(1999). 8. Disteche, C M . Escape from X inactivation in human and mouse. Trends in Genet. 11, 17-22(1995). 9. Rastan, S. Non-random X-chromosome inactivation in mouse X-autosome translocation embryos-location of the inactivation centre. J. Embryol. Exp. Morph. 78, 1-22 (1983). 10. Lee, J.T. & Jaenisch, R. Long-range cis effects of ectopic X-inactivation centres on a mouse autosome. Nature 386, 275-279 (1997). 11. Brown, C.J. et al. A gene from the region of the human X inactivation centre is expressed exclusively from the inactive X chromosome. Nature 349, 38-44 (1991). 12. Borsani, G. et al. Characterisation of a murine gene expressed from the inactive X chromosome. Nature 351, 325-329 (1991). 13. Brockdorff, N. et al. Conservation of position and exclusive expression of mouse Xist from the inactive X chromosome. Nature 351, 329-331 (1991). 14. Brown, C.J. & Willard, H.F. Molecular and genetic studies of human X chromosome inactivation. Adv. Dev. Biol. 2, 37-72 (1993). 15. Clerc, P. & Avner, P. Role of the region 3' to Xist exon 6 in the counting process of X-chromosome inactivation. Nat. Genet. 19, 249-253 (1998). 16. Lee, J.T., Davidow, L.S. & Warshawsky, D. Tsix, a gene antisense to Xist at the X-inactivation center. Nat. Genet. 21, 400-404 (1999). 17. Rasmussen, T.P., Wutz, A., Pehrson, J. & Jaenisch, R. Expression of Xist RNA is sufficient to initiate macrochromatin body formation. Chromosoma 110, 411-420 (2001). 18. Lyon, M.F. X-chromosome inactivation: a repeat hypothesis. Cytogenet. Cell Genet. 80,133-137 (1998). 19. White, W.M., Willard, H.F., Van Dyke, D.L. & Wolff, D.J. The spreading of X inactivation into autosomal material of an X;autosome translocation: evidence for 100 a difference between autosomal and X-chromosomal DNA. Am. J. Hum. Genet. 63, 20-28 (1998). 20. Riggs, A.D. DNA methylation and late replication probably aid cell memory, and type I DNA reeling could aid chromosome folding and enhancer function. Phil. Trans. R. Soc. Lond. 326, 285-297 (1990). 21. Hong, Y.K., Ontieros, S.M., Chen, C , and Strauss, W.M. A new structure for the murine Xist gene and its relationship to chromosome choice/counting during X-chromosome inactivation. Proceedings of the National Academy of Science USA 96, 6829-6834 (1999). 22. Hong, Y.K., Ontiveros, S.D., and Strauss W.M. A revision of the human XIST gene organization and structural comparison with mouse Xist. Mammalian Genome 11, 220-224 (2000). 23. Brown, C.J. et al. The human XIST gene: analysis of a 17 kb inactive X-specific RNA that contains conserved repeats and is highly localized within the nucleus. Cell 71, 527-542(1992). 24. Brockdorff, N. et al. The product of the mouse Xist gene is a 15kb inactive X-specific transcript containing no conserved ORF and located in the nucleus. Cell 71,515-526 (1992). 25. Clemson, C M . , McNeil, J.A., Willard, H.F. & Lawrence, J.B. XIST RNA paints the inactive X chromosome at interphase: evidence for a novel RNA involved in nuclear/chromosome structure. J. Cell Biol. 132, 259-275 (1996). 26. Beard, C , Li, E. & Jaenisch, R. Loss of methylation activates Xist in somatic but not in embryonic cells. Genes & Dev. 9, 2325-2334 (1995). 27. Hendrich, B.D. & Willard, H.F. Epigenetic [^ regulation of gene expression: the effect of altered chromatin structure from yeast to mammals. Hum. Mol. Genet. 4, 1765-1777(1995). 28. Penny, G.D., Kay, G.F., Sheardown, S.A., Rastan, S. & Brockdorff, N. Requirement for Xist in X chromosome inactivation. Nature 379, 131-137 (1996). 29. Avner, P. & Heard, E. X-chromosome inactivation: counting, choice and initiation. Nature Reviews: Genetics 2, 59-67 (2001). 30. Cattanach, B.M. & Papworth, D. Controlling elements in the mouse V. Linkage tests with X-linked genes. Genet. Res. 38, 57-70 (1981). 31. Migeon, B.R., Chowdhury, A.K., Dunston, J.A. & Mcintosh, I. Identification of TSIX, encoding an RNA antisense to human XIST, reveals differences from its murine counterpart: implications for X inactivation. Am. J. Hum. Genet. 69, 951-960 (2001). 32. Sado, T., Wang, Z., Sasaki, H. & Li, E. Regulation of imprinted X-chromosome inactivation in mice by Tsix. Development 128, 1275-1286 (2001). 33. Lee, J.T. Disruption of imprinted X inactivation by parent-of-origin effects at Tsix. Ce//103, 17-27 (2000). 34. Lee, J.T. & Lu, N. Targeted mutagenesis of Tsix leads to nonrandom X inactivation. Cell 99, 47-57 (1999). 35. Johnston, C M . et al. Developmentally regulated Xist promoter switch mediates initiation of X inactivation. Cell 94, 809-817 (1998). 36. Brown, CJ. & Willard, H.F. The human X inactivation center is not required for maintenance of X inactivation. Nature 368, 154-156 (1994). 101 37. Clemson, C M . , Chow, J.C, Brown, CJ. & Lawrence, J.B. Stabilization and localization of Xist RNA are controlled by separate mechanisms and are not sufficient for X inactivation. J. Cell Biol. 142,13-23 (1998). 38. Hall, L.L. et al. An ectopic human XIST gene can induce chromosome inactivation in postdifferentiation human HT-1080 cells. Proc. Natl. Acad. Sci. USA 99, 8677-8682 (2002). 39. Stuckenholz, C , Kageyama, Y. & Kuroda, M.I. Guilt by association: non-coding RNAs, chromosome-specific proteins and dosage compensation in Drosophila. Trends Genet. 15, 454-458 (1999). 40. Gilbert, S.L., Pehrson, J. & Sharp, P.A. XIST RNA associates with specific regions of the inactive X chromatin. J. Biol. Chem. 275, 36491-36494 (2000). 41. Pehrson, J.R. & Fried, V.A. MacroH2A, a core histone containing a large nonhistone region. Science 257, 1398-1400 (1992). 42. Wutz, A., Rasmussen, T.P. & Jaenisch, R. Chromosomal silencing and localization are mediated by different domains of Xist RNA. Nat. Genet. 30, 1-8 (2002). 43. Beletskii, A., Hong, Y.-K., Pehrson, J., Egholm, M. & Strauss, W.M. PNA interference mapping demonstrates functional domains in the noncoding RNA Xist. Proc. Natl. Acad. Sci., USA 98, 9215-9220 (2001). 44. Bestor, T., Laudano, A., Mattaliano, R., and Ingram V. Cloning and sequencing of a cDNA encoding DNA methyltransferase of mouse cells. Journal of Molecular Biology 203, 971-983 (1988). 45. Okano, M., Xie, S., and Li, E. Cloning and characterization of a family of novel mammalian DNA (cytosine-5) methyltransferases. Nature Genetics 19, 219-220 (1998). 46. Razin, A., Riggs, A.D. DNA methylation and gene function. Science 210, 604-610(1980). 47. Jones, P. A., and Takai, D. The Role of DNA Methylation in Mammalian Epigenetics. Science 293, 1068-1070 (2001). 48. Bird, A. The essentials of DNA methylation. Cell 70, 5-8 (1992). 49. Cooper, D.N., and Krawczak, M. Cytosine methylation and the fate of CpG dinucleotides in vertebrate genomes. Human Genetics 83,181-188 (1989). 50. Leonhardt, H., Page, A.W., Weier, H., and Bestor, T.H. A targeting sequence directs DNA methyltransferase to sites of DNA replication in mammalian nuclei. CW/71, 865-873 (1992). 51. Lei, H., Oh, S.P., Okano, M., Juttermann, R., Goss, K.A., Jaenisch, R.,and Li, E. De novo DNA cytosine methyltransferase activities in mouse embryonic stem cells. Development 122, 3195-3205 (1996). 52. Bestor, T.H. The DNA methyltransferases of mammals. Human Molecular Genetics 9, 2395-2402 (2000). 53. Taylor, S.M., and Jones, P.A. Multiple new phenotypes induced in 10T1/2 and 3T3 cells treated with 5-azacytidine. Cell 17, 771-779 (1979). 54. Jones, P.A. Altering gene expression with 5-azacytidine. Cell 40,485-486 (1985). 55. Jutterman, R., Li, E. & Jaenisch, R. Toxicity of 5-aza-2'deoxycytidine to mammalian cells is mediated primarily by covalent trapping of DNA 102 methyltransferase rather than DNA demethylation. Proc. Natl. Acad. Sci. 91, 11797-11801 (1994). 56. Jones, PA. & Laird, P.W. Cancer epigenetics comes of age. Nat. Genet. 21, 163-167(1999). 57. Tate, P.H., and Bird, A.P. Effects of DNA methylation on DNA-binding proteins and gene expression. Current Opinion in Genetics and Development 3, 226-231 (1993). 58. Boyes, J. & Bird, A. Repression of genes by DNA methylation depends upon CpG density and promoter strength: evidence for involvement of a methyl-CpG binding protein. EMBO J. 11, 327-333 (1992). 59. Wolffe, A.P., and Pruss, D. Targeting chromatin disruption: Transcriptional regulators that acetylate histones. Cell 84, 817-819 (1996). 60. Nan, X., Ng, H.H., Johnson, C.A., Laherty, CD., Turner, B.M., Eisenman, R.N., and Bird, A. Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex. Nature 393, 386-389 (1998). 61. Li, E., Bestor, T.H. & Jaenisch, R. Targeted mutation of the DNA methyltransferase gene results in embryonic lethality. Cell 69, 915-926. (1992). 62. Okano, M., Bell, D.W., Haber, D.A. & Li, E. DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development. Ce//99, 247-257(1999). 63. Strachan, T.a.R., A.P. Human Molecular Genetics, (BIOS Scientific Publishers Limited, New York, 1996). 64. Luger, K., Mader, A.W., Richmond, R.K., Sargent, D.F., and Richmond, T.J. Crystal structure of the nucleosome core particle at 2.8 A resolution. Nature 389, 251-260 (1997). 65. Tse, C , Sera, T., Wolffe, A.P. & J.C, H. Disruption of higher-order folding by core histone acetylation dramatically enhances transcription of nucleosomal arrays by RNA polymerase III. Mol. Cell. Biol. 18, 4629-4638 (1998). 66. Wolffe, A.P. Chromatin: Structure Function, (Academic Press, London, 1992). 67. Marmorstein, R. Structure and function of histone acetyltransferases. CMLS 58, 693-70 (2001). 68. Kuo, M.H. et al. Transcription-linked acetylation by Gcn5p of histones H4 and H4 at specific lysines. Nature 383, 269-272 (1996). 69. Wang, L., Liu, L. & Berger, S.L. Critical residues for histone acetylation by Gcn5, functioning in Ada and SAGA complexes, are also required for transcriptional function in vivo. Genes Dev. 12, 640-653 (1998). 70. Grant, PA. et al. Expanded lysine acetylation specificit of Gcn5 in native complexes. J. Biol. Chem. 274, 5895-5900 (1999). 71. Timmermann, S., Lehrmann, A., Polesskaya, A. & Harel-Bellan, A. Histone acetylation and disease. CMLS 58, 728-736 (2001). 72. Cress, W.D. & Seto, E. histone deacetylases, transcriptional control, and cancer. J. Cell Physiol. 184, 1-16(2000). 73. Allfrey, V.G., Faulkner, R. & Mirsky, A.E. Acetylation and methylation of histones and their possible role in the regulation of RNA synthesis. Proc. Natl. Acad. Sci. USA 51, 786-794 (1964). 103 74. Pogo, B.G.T., Allfrey, V.G. & Mirsky, A.E. RNA synthesis and histone acetylation during the course of gene activation in lymphocytes. Proc. Natl. Acad. Sci. USA 55, 6212-6222 (1966). 75. Hebbes, T.R., Thorne, A.W. & Crane-Robinson, C. A direct link between core histone acetylation and transcriptionally active chromatin. EMBO J. 7, 1395-1403 (1988). 76. Vettese-Dadney, M. et al. Acetylation of histone H4 plays a primary role in enhancing transcription factor binding to nucleosomal DNA in vitro. EMBO J. 15, 2508-2518 (1996). 77. Turner, B., Birley, A. & Lavender, J. Histone H4 isoforms acetylated at specific lysine residues define individual chromosomes and chromatin domains in Drosophila polytene nuclei. Cell 69, 375-384 (1992). 78. Lavender, J.S., Birley, A.J., Palmer, M.J., Kuroda, M.I. & Turner, B.M. Histone H4 acetylated at lysine 16 and other components of the Drosophila dosage compensation pathway colocalize on the male X chromosome through mitosis. Chromos. Res. 2, 398-404 (1994). 79. Parekh, B.S. & Maniatis, T. Virus infection leads to localized hyperacetylation of histones H3 and H4 at the IFN-beta promoter. Mol. Cell 3, 125-129 (1999). 80. Kadosh, D. & Struhl, K. Histone deacetylase activity of Rpd3 is important for transcriptional repression in vivo. Genes Dev. 12, 797-805 (1998). 81. Jeppeson, P. & Turner, B. The inactive X chromosome in female mammals is distinguished by a lack of histone H4 acetylation, a cytogenetic marker for gene expression. Cell 74, 281-189 (1993). 82. Belyaev, N.D., Keohane, A.M. & Turner, B.M. Differential underacetylation of histones H2A, H3 and H4 on the inactive X chromosome in human female cells. Hum. Genet. 97, 573-578 (1996). 83. Wakefield, M.J., Keohane, A.M., Turner, B.M. & Graves, J.A.M. Histone underacetylation is an ancient component of mammalian X chromosome inactivation. Proc. Natl. Acad. Sci. 94, 9665-9668 (1997). 84. Turner, B.M. Histone acetylation as an epigenetic determinant of long-term transcriptional competence. Cell Mol. Life Sci. 54, 21-31 (1998). 85. Murray, K. The occurrence of epsilon-N-methyl lysine in histones. Biochemistry 3(1964). 86. van Holde, K.E. Chromatin, (Springer-Verlag, New York, 1988). 87. Rice, J.C. & Allis, CD. Histone methylation versus histone acetylation: new insights into epigenetic regulation. Curr. Opin. Cel. Biol. 13, 263-273 (2001). 88. Byvoet, P., Shepherd, G.R., Hardin, J.M. & Noland, B.J. The distribution and turnover of labeled methyl groups in histone fractions of cultured mammalian cells. Arch. Biochem. Biophys. 148, 558-567 (1972). 89. Baxter, CS. & Byvoey, P. Intercalating agents as probes of the spatial relationship betweenchromatin components. Biochem. Biophys. Res. Commun. 63, 286-291 (1975). 90. Strahl, B.D. & Allis, CD. The language of covalent histone modifications. Nature 403, 41-45 (2000). 104 91. Annunziato, A.T., Eason, M.B. & Perry, C A . Relationship between methylation and acetylation of arginine-rich histones in cycling and arrested HeLa cells. Biochemistry 34, 2916-2924 (1995). 92. Reuter, G. & Spierer, P. Position effect variegation and chromatin proteins. Bioessays 14, 605-612 (1992). 93. Karpen, G.H. Position-effect variegation and the new biology of heterochromatin. Curr. Opin. Genet. Dev. 4, 281-291 (1994). 94. Wakimoto, B.T. Beyond the nucleosome: epigenetic aspects of position-effect variegation in Drosophila. Cell 93, 321-324 (1998). 95. Grewal, S.I. & Klar, A.J. Chromosomal inheritance of epigenetic states in fission yeast during mitosis and meiosis. Cell 86, 95-101 (1996). 96. Cavalli, G. & Paro, R. The Drosophila Fab-7 chromosomal element conveys epigenetic inheritance during mitosis and meiosis. Cell 93, 505-518 (1998). 97. Grewal, S.I., Bonaduce, M.J. & Klar, A.J. Histone deacetylase homologs regulate epigenetic inheritance of transcriptional silencing and chromosome segregation in fission yeast. Genetics 150, 563-576 (1998). 98. Eissenberg, J.C. & Elgin, S.C. The HP1 protein family: getting a grip on chromatin. Curr. Opin. Genet. Dev. 10, 204-210 (2000). 99. Jenuwein, T., Laible, G., Dorn, R. & Reuter, G. SET domain proteins modulate chromatin domains in eu- and heterochromatin. Cell. Mol. Life. Sci. 54, 80-93 (1998). 100. Rea, S. et al. Regulation of chromatin structure by site-specific histone H3 methyltransferases. Jature 406, 593-599 (2000). 101. Lachner, M., O'Carroll, D., Rea, S., Merchtier, K. & Jenuwein, T. Methylation of histone H3 lysine 9 creates a binding site for HP1 proteins. Nature 401, 116-120 (2001). 102. Bannister, A.J. et al. Selective recognition of methylated lysine 9 on histone H3 by the HP1 chromo domain. Nature 410, 120-124 (2001). 103. Wei, Y., Yu, L., Bowen, J., Gorovsky, M.A. & Allis, CD. Phosphorylation of histone H3 is required for proper chromosome condensation and segregation. Cell 97, 99-109 (1999). 104. Cheung, P. et al. Synergistic coupling of histone H3 phosphorylation and acetylation in response to epidermal growth factor stimulation. Mol. Cel. 5, 905-915 (2000). 105. Lo, W.S. et al. Phosphorylation of serine 10 in histone H3 is functionally linked in vitro and in vivo to Gcn5-mediated acetylation at lysine 14. Mol. Cell. 5, 917-926 (2000). 106. Clayton, A.L., Rose, S., Barratt, M.J. & Mahadevan, L . C Phosphoacetylation of histone H3 on c-fos- and c-jun-associated nucleosomes upon gene activation. EMBOJ. 19,3714-3726(2000). 107. Jenuwein, T. & Allis, CD. Translating the histone code. Science 293, 1074-1080 (2001). 108. Dhalluin, C. et al. Structure and ligand of a histone acetyltransferase bromodomain. Nature 399,491-496 (1999). 109. Winston, F. & Allis, CD. The bromodomain: a chromatin-targeting module? Nat, Struct. Biol. 6, 601-604 (1999). 105 110. Owen, D.J. et al. The structural basis for the recognition of acetylated histone H4 by the bromodomain of histone acetyltransferase gcn5p. EMBO J. 19, 6141-6149 (2000). 111. Brasher, S.V. et al. The structure of mouse HP1 suggests a unique mode of single peptide recognition by the shadow chromo domain dimer. EMBO J19, 1587-1597 (2000). 112. Cowieson, N.P., Partridge, J.F., Allshire, R.C. & McLaughlin, P.J. Dimerisation of a chromo shadow domain and distinctions from the chromodomain as revealed by structural analysis. Curr Biol 10, 517-525 (2000). 113. Jacobson, R.H., Ladurner, A.G., King, D.S. & Tjian, R. Structure and function of a human TAFII250 double bromodomain module. Science 288, 1422-1425 (2000). 114. Sun, Z.W. & Allis, CD. Ubiquitination of histone H2B regulates H3 methylation and gene silencing in yest. Nature 418, 104-108 (2002). 115. Taylor, J.H. Rates of chain growth and units of replication in DNA of mammalian chromosomes. J. Mol. Biol. 31, 579-594 (1968). 116. Huberman, J.A. & Riggs, A.D. On the mechanism of DNA replication in mammalian chromosomes. J. Mol. Biol. 32, 327-341 (1968). 117. Jackson, D.A. & Pombo, A. Replicon clusters are stable units of chromosome structure: evidence that nuclear organization contributes to the efficient activation and propagation of S phase in human cells. J. Cell Biol. 140, 1285-1295 (1998). 118. Ten Hagen, K.G., Gilbert, D.M., Willard, H.F. & Cohen, S.N. Replication timing of DNA sequences associated with human centromeres and telomeres. Mol. Cell. Biol. 10, 6348-6355 (1990). 119. Taylor, J.H. Asynchronous duplication of chromosomes in cultured cells of Chinese hamster. Biophysic. Biochem. Cytol. 7, 455-463 (1960). 120. Boggs, B.A. & Chinault, A.C. Analysis of replication timing properties of human X-chromosomal loci by fluorescence in situ hybridization. Proc. Natl. Acad. Sci. 91,6083-6087 (1994). 121. Selig, S., Okumura, K., Ward, D.C. & Cedar, H. Delineation of DNA replication time zones by fluorescence in situ hybridization. EMBO J. 11, 1217-1225 (1992). 122. Mohandas, T., Crandall, B.F., Sparkes, R.S., Passage, M.B. & Sparkes, M.C. Late replication studies in a human X/13 translocation: correlation with autosomal gene expression. Cytogenet. Cell Genet. 29, 215-220 (1981). 123. Couturier, J. et al. Evidence for a correlation between late replication and autosomal gene inactivation in a familial translocation t(X;21). Hum. Genet. 49, 319-326(1979). 124. Mohandas, T., Sparkes, R.S. & Shapiro, L.J. Genetic evidence for the inactivation of a human autosomal locus attached to an inactive X chromosome. Am. J. Hum. Genet. 34, 811-817 (1982). 125. Taysi, K., Sparker, R.S., O'Brien, T.J. & Dengler, D.R. Down's syndrome phenotype and autosomal gene inactivation in a child with presumed (X;21) de novo translocation. J. Med. Genet. 19, 144-148 (1982). 126. Sharp, A., Robinson, D.O. & Jacobs, P.A. Absence of correlation between late-replication and spreading of X inactivation in an X;autosome translocation. Hum. Genet. 109,295-302(2001). 106 127. Lee, Y. et al. Isolation of cDNA clones encoding human histone macroH2Al subtypes. Biochimica et Biophysica Acta 1399, 73-77 (1998). 128. Rasmussen, T.P. et al. Messenger RNAs encoding mouse histone macroH2Al isoforms are expressed at similar levels in male and female cells and result from alternative splicing. Nucleic Acids Re. 27, 3685-3689 (1999). 129. Pehrson, J.R., Costanzi, C. & Dharia, C. Developmental and tissue expression patterns of histone macroH2Al subtypes. J. Cell Biochem. 65, 107-113 (1997). 130. Costanzi, C. & Pehrson, J.R. MacroH2A2, a new member of the MacroH2A core histone family. J. Biol. Chem. 276, 21776-21784 (2001). 131. Chadwick, B.P. & Willard, H.F. A novel chromatin protein, distantly related to histone H2A, is largely excluded from the inactive X chromosome. J. Cell Biol. 152, 375-384 (2001). 132. Csankovski, G., Panning, B., Bates, B., Pehrson, J.R. & Jaenisch, R. Conditional deletion of Xist disrupts histone macroH2A localization but not maintenance of X inactivation. Nat. Genet. 22, 323-324 (1999). 133. McKee, B.D. & Handel, M.A. Sex chromosomes, recombination, and chromatin conformation. Chromosoma 102, 71-80 (1993). 134. McCarrey, J.R. & Dilworth, D.D. Expression of Xist in mouse germ cells correlates with X-chromosome inactivation. Nature Genetics 2, 200-203 (1992). 135. Richler, C , Soreq, H. & Wahrman, J. X inactivation in mammalian testis is correlated with inactive X-specific transcription. Nature Genetics 2, 192-195 (1992). 136. Salido, E.C., Yen, P.H., Mohandas, T.K. & Shapiro, L.J. Expression of the X-inactivation-associated gene XIST during spermatogenesis. Nature Genetics 2, 196-199(1992). 137. Hoyer-Fender, S., Costanzi, C. & Pehrson, J.R. Histone macroH2Al .2 is concentrated in the XY-body by the early pachytene stage of spermatogenesis. Exp. Cell Res. 258, 254-260 (2000). 138. Pehrson, J.R. & Fuji, R.N. Evolutionary conservation of histone macroH2A subtypes and domains. Nucleic Acids Res. 26, 2837-2842 (1998). 139. Valentin, H.E., Zwingmann, G., Schonebaum, A. & Steinbuchel, A. Metabolic pathway for biosynthesis of poly(3-hydroxybutyrate-co-4-hydroxybutyrate) from 4-hydroxybutyrate by Alcaligenes eutrophus. Eur. J. Biochem. 227, 43-60 (1995). 140. Dominguez, G., Wang, C Y . & Frey, T.K. Sequence of the genome RNA of rubella virus: evidence for genetic rearrangement during togavirus evolution. Virology 177, 225-238 (1990). 141. Gorbalenya, A.E., Koonin, E.V. & Lai, M.M. Putative papain-related thiol proteases of positive-strand RNA viruses. Identification of rubi- and aphthovirus proteases and delineation of a novel conserved domain associated with proteases of rubi-, alpha- and coronaviruses. FEBS Lett. 288, 201-205 (1991). 142. LaStarza, M.W., Lemm, JA. & Rice, C M . Genetic analysis of the nsP3 region of Sindbis virus: evidence for roles in minus-strand and subgenomic RNA synthesis. J. Virol. 68, 5781-5791 (1994). 143. Wang, Y.F., Sawicki, S.G. & Sawicki, D.L. Alphavirus nsP3 functions to form replication complexes transcribing negative-strand RNA. J. Virol. 68, 6466-6475 (1994). 107 144. Chadwick, B.P., Valley, C M . & Willard, H.F. Histone variant macroH2A contains two distinct macrochromatin domains capable of directing macroH2A to the inactive X chromosome. Nucl. Acids Res. 29, 2699-2705 (2001). 145. Changolkar, L.N. & Pehrson, J.R. Reconstitution of nucleosomes with histone macroH2A1.2. Biochem. 41, 179-184 (2002). 146. Vijay-Kumar, S., Chandra, N., Dharia, C. & Pehrson, J.R. Crystallization and preliminary X-ray crystallographic studies of nonhistone region of macroH2A. 1. Proteins 22, 290-292 (1995). 147. Richler, C , Dhara, S.K. & Wahrman, J. Histone macroH2A1.2 is concentrated in the XY compartment of mammalian male meiotic nuclei. Cytogenet. Cell Genet. 89,118-120(2000). 148. Turner, J.M.A., Burgoyne, P.S. & Singh, P.B. M31 and macroH2A1.2 colocalise at the pseudoautosomal region during mouse meiosis. J. Cell Sci. 114, 3367-3375 (2001). 149. Mermoud, J.E., Costanzi, C , Pehrson, J.R. & Brockdorff, N. Histone macroH2Al .2 relocates to the inactive X chromosome after initiation and propagation of X-inactivation. J . Cell Biol. 147, 1399-1408 (1999). 150. Mermoud, J.E., Tassin, A.M., Pehrson, J.R. & Brockdorff, N. Centrosomal association of histone macroH2A1.2 in embryonic stem cells and somatic cells. Exp. Cell Res. 268, 245-251 (2001). 151. Littlefield, J.W. Selection of hybrids from matings of fibroblasts in vitro and their presumed recombinants. Genomics 29, 353-363 (1964). 152. Willard, H.F., Brown, C.J., Carrel, L., Hendrich, B. & Miller, A.P. Epigenetic and chromosomal control of gene expression: molecular and genetic analysis of X chromosome inactivation. Cold Spring Harbor Quant. Symp. Biol. 58, 315-322 (1993). 153. Tinker, A.V. & Brown, CJ. Induction of XIST expression from the human active X chromosome in mouse/human somatic cell hybrids by DNA demethylation. Nucl. Acids Res. 26, 2935-2940 (1998). 154. Hansen, R.S., Canfield, T.K., Stanek, A.M., Keitges, E.A. & Gartler, S.M. Reactivation of XIST in normal fibroblasts and a somatic cell hybrid: abnormal localization of XIST RNA in hybrid cells. Proc. Natl. Acad. Sci. USA 95, 5133-5138 (1998). 155. Hendrich, B.D., Brown, CJ. & Willard, H.F. Evolutionary conservation of possible functional domains of the human and murine XIST genes. Hum. Mol. Genet. 2, 663-672 (1993). 156. Hansen, R.S., Ellis, N.A. & Gartler, S.M. Demethylation of specific sites in the 5' region of the inactive X-linked human phosphoglycerate kinase gene correlates with the appearance of nuclease sensitivity and gene expression. Mol. Cell. Biol. 8, 4692-4699 (1988). 157. Brown, CJ. & Willard, H.F. Noninactivation of a selectable human X-linked gene that complements a murine temperature-sensitive cell cycle defect. A m . J. Hum. Genet. 45, 592-598 (1989). 158. Henke, W., Herdel, K., Jung, K., Schnorr, D. & Leoning, S.A. Betaine improves the PCR amplificaiton of GC-rich DNA sequences. Nucleic Acids Res. 25, 3957-3958 (1997). 108 159. Carrel, L. et al. X inactivation analysis and DNA methylation studies of the ubiquitin activating enzyme El and PCTAIRE-1 genes in human and mouse. Hum. Mol. Genet. 5, 391-402 (1996). 160. Rasmussen, T.P., Mastrangelo, M.A., Eden, A., Pehrson, J.R. & Jaenisch, R. Dynamic relocalization of histone macroH2Al from centrosomes to inactive Xchromosomes during X inactivation. J. Cell Biol. 150, 1189-1198 (2000). 161. Takagi, N. & Sasaki, M. Preferential inactivation of the paternally derived X chromosome in the extraembryonic membranes of the mouse. Nature 256, 640-642 (1975). 162. Costanzi, C , Stein, P., Worrad, D.M., Schultz, R.M. & Pehrson, J.R. Histone macroH2Al is concentrated in the inactive X chromosome of female preimplantation mouse embryoes. Development 127, 2283-2289 (2000). 163. Boumil, R.M. & Lee, J.T. Forty years of decoding the silence in X-chromosome inactivation. Hum. Mol. Genet. 10, 2225-2232 (2001). 164. Perche, P.-Y. et al. Higher concentrations of histone macroH2A in the Barr body are correlated with higher nucleosome density. Curr. Biol. 10, 1531-1534 (2000). 165. Chadwick, B.P. & Willard, H.F. Histone H2A variants and the inactive X chromosome: identification of a second macroH2A variant. Hum. Mol. Genet. 10, 1101-1113 (2001). 109 APPENDIX Primer: L I G G F P 5 ' - T C T T G T T A G T T G C C G T C G T C C T T - 3 ' primer mH2A1.2 306bp EGFPN1 I 617bp Primer: C1SEQ1 5 ' - A T G G A C G A G C T G T A C A A G T C C G G A - 3 ' primer EGFPC1 mH2A1.2 1 15bp 872bp Figure A - l : Diagram of Sequencing Results A unique primer was designed for each of the two macroH2A1.2-GFP constructs to be used in sequencing reactions. Both produced sequence data that overlapped the corresponding ligation junction and confirmed that the two protein fragments were in frame. 110 < u < u o < < o H U a 3 < H o a a o f -o a E -a < u H E -U o H U O t— a o E -a a < < a i -H u < a o o 3 E -< y a 3 a u H H U y < f— a u < a • J H U CJ H U a u o H a o < < a < a f— a a fc 8 O O < o u o R « § 3 8 « E- < R o u R a a ^ 9 ° 8 fc < 3 fc O $ < R < R u b S 2 £ R s < O U H O U u y a u o u a H h h fc ° < ^ O R ^ < O H r j O P < o R u R a o < y H o fc u y 2 t 5 H E 8 « S I I I I ^ < fc ° o R R a 2 2 R S fc O u 8 R < 2 E U C U u R U u a u a g o R <; u H u E- u E- u o a u u o u o < a u f- U E-u —' 71 W W o o 111