UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mechanisms for XIST RNA cis-localisation Kelsey, Angela 2013

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


24-ubc_2013_fall_kelsey_angela.pdf [ 2.07MB ]
JSON: 24-1.0074225.json
JSON-LD: 24-1.0074225-ld.json
RDF/XML (Pretty): 24-1.0074225-rdf.xml
RDF/JSON: 24-1.0074225-rdf.json
Turtle: 24-1.0074225-turtle.txt
N-Triples: 24-1.0074225-rdf-ntriples.txt
Original Record: 24-1.0074225-source.json
Full Text

Full Text

   Mechanisms for XIST RNA cis-localisation  by  Angela Kelsey B.Sc., The University of Manchester, 2010 (Honours)  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in The Faculty of Graduate and Postdoctoral Studies (Medical Genetics) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   August 2013  ? Angela Kelsey, 2013    ii Abstract X inactivation is the process of silencing one of the two X chromosomes in mammalian female cells in order to equalise the dosage of X-linked genes with males. The process is initiated by the long noncoding RNA XIST, which is transcribed from the future inactive X and localises to it in cis. How XIST RNA is able to localise to the X chromosome is not well defined. The aim of the current study was to deduce mechanisms of XIST RNA localisation. This was addressed in various ways, including 1) testing the ability of an XIST transgene integrated into a variety of autosomes to localise to those autosomes, as opposed to the X chromosome; 2) assessing the ability of XIST transgenes with different regions deleted to localise, in order to identify sequences required for localisation; and 3) knocking down various proteins implicated in X inactivation in order to assess any effect on the ability of XIST to localise. We find that the XIST transgene is able to localise to a wide variety of different autosomes and furthermore, is able to direct the enrichment of the histone variant macroH2A on an autosome and the deposition of a repressive histone modification, H3K27me3, onto an autosome. We also find that a region of XIST encompassing repeats B and C, and sequences downstream of exon 1 are involved in localising XIST RNA, and that they do so in a redundant fashion. Lastly, we show that the knockdown of five proteins - YY1, hnRNP-U, SPOP, CUL3 and ASH2L - prevent the formation of an intact XIST focus. The results presented here add to the limited knowledge of how XIST RNA is able to localise, an essential step in the process of X chromosome inactivation.  iii Preface The XIST transgene integrations and the XIST deletion constructs were made previously in the lab. Two of the deletions, delta (PflMI and del 3? XIST), were described in a previous paper (Chow et al. 2007). The knockdowns were performed by Jakub Minks (at the time a PhD student in the laboratory). Images were taken by ADK and Irene Qi (at the time an undergraduate student in the laboratory).   iv Table of Contents Abstract ..................................................................................................................................... ii Preface ...................................................................................................................................... iii Table of Contents ..................................................................................................................... iv List of Tables ............................................................................................................................ vi List of Figures ......................................................................................................................... vii List of Abbreviations ............................................................................................................. viii List of Gene Names.................................................................................................................. ix Acknowledgements .................................................................................................................. x 1 Introduction ........................................................................................................................ 1 1.1 Thesis overview ................................................................................................... 2 1.2 Dosage compensation and the X-inactivation centre ............................................ 2 1.3 Features of the Xi ................................................................................................. 3 1.3.1 Histone methylation ................................................................................ 3 1.3.2 DNA methylation ..................................................................................... 8 1.3.3 Cot-1 holes ............................................................................................. 9 1.3.4 Perinuclear/perinucleolar localisation ...................................................... 9 1.3.5 Other features of the Xi ......................................................................... 10 1.4 Sequences of XIST RNA .................................................................................... 10 1.5 Spreading of XIST RNA on autosomes .............................................................. 13 1.6 Proteins implicated in X inactivation ................................................................... 15 1.6.1 YY1 ...................................................................................................... 17 1.6.2 hnRNP-U/SAF-A ................................................................................... 18 1.6.3 SMCHD1 .............................................................................................. 18 1.6.4 ASH2L .................................................................................................. 18 1.6.5 CBX7 .................................................................................................... 18 1.6.6 CUL3 and SPOP .................................................................................. 19 1.7 Thesis objective ................................................................................................. 19 2 Methods ............................................................................................................................ 20 2.1 Cell culture ......................................................................................................... 21 2.2 Fixation of cells for FISH and immunofluorescence ............................................ 21 2.3 RNA FISH .......................................................................................................... 21 2.4 Fluorescence microscopy and image analysis .................................................... 21 2.5 RNA isolation, reverse transcription and Q-PCR ................................................ 22 2.6 Immunofluorescence combined with RNA FISH ................................................. 22 2.7 siRNA-mediated knockdown .............................................................................. 22 3 Results .............................................................................................................................. 25 3.1 The experimental system ................................................................................... 26 3.2 XIST RNA can function when expressed from different chromosomes ............... 26 3.2.1 XIST transgenes produce RNA that can localise to various autosomes ............................................................................................ 26 3.2.2 XIST expression increases perinucleolar localisation of autosomes ..... 27 3.2.3 XIST is able to form Cot-1 holes on a variety of autosomes .................. 31 3.2.4 XIST RNA forms Cot-1 holes to a similar degree on different autosomes ............................................................................................ 32 3.2.5 An XIST-expressing autosome recruits H3K27me3 and macroH2A ...... 35 3.3 XIST sequences act redundantly to localise XIST RNA ...................................... 37 v 3.4 Proteins that affect XIST localisation .................................................................. 38 4 Discussion ........................................................................................................................ 40 5 References ....................................................................................................................... 50    vi List of Tables Table 1.1 Histone modifications enriched and depleted on the inactive X .................................. 6 Table 1.2: Proteins implicated in XCI ....................................................................................... 16 Table 2.1 List of qPCR primers ................................................................................................ 23  vii List of Figures  Figure 1.1 Epigenetic pathways of facultative and constitutive heterochromatin......................... 5 Figure 1.2 PRC2 and PRC1 subunits and their functions ........................................................... 7 Figure 1.3 Structure of mutant Xist alleles ................................................................................ 12 Figure 4.1 XIST is able to localise to a variety of autosomes .................................................... 27 Figure 4.2 XIST expression increases nucleolar association of an XIST-expressing autosome 30 Figure 4.3 XIST is able to form Cot holes on a variety of autosomes ....................................... 32 Figure 4.4 XIST is equally able to form Cot holes on a variety of different autosomes.............. 34 Figure 4.5 H3K27me3 and mH2A are recruited to XIST transgene-bearing chromosomes ...... 36 Figure 4.6 XIST sequences act redundantly to localise XIST RNA ........................................... 38 Figure 4.7 Knockdowns that affect XIST focus formation ......................................................... 39 Figure 5.1 Model for compaction of nucleosomes containing either histone H1 or macroH2A .. 48   viii List of Abbreviations bp = base pairs BSA ? bovine serum albumin cDNA = complementary deoxyribonucleic acid ChIP = chromatin immunoprecipitation Cot = concentration over time Cot-1 = concentration over time fraction 1 DAPI = 4?-6-Diamidino-2-phenylindole DNA = deoxyribonucleic acid DNMT/Dnmt = DNA methyltransferase ES(C) = embryonic stem (cell) FISH = fluorescence in situ hybridization GFP = green fluorescence protein kb = kilobase pairs L1 = long interspersed nuclear element 1 LINE = long interspersed nuclear element LTR = long terminal repeat Mb = mega base pairs PBS = phosphate buffered saline PBT = PBS / BSA / Tween 20 PCR = polymerase chain reaction PRC2 = polycomb repressive complex 2 Q-PCR = quantitative-polymerase chain reaction RNA = ribonucleic acid RNA pol II = RNA polymerase II RT-PCR = reverse transcription polymerase chain reaction SINE = short interspersed elements SSC = sodium citrate, sodium chloride TSS = transcription start site Xa = active X chromosome Xi = inactive X chromosome XCI = X chromosome inactivation XIC/Xic = X inactivation centre   ix List of Gene Names Note that gene names in all capital letters refer to human genes; gene names with just the first letter capitalized refer to mouse genes. ASH2L = (absent, small or homeotic)-like CBX7 = chromobox homolog 7 CUL3 = cullin 3 DAXX = death-domain associated protein EED = embryonic ectoderm development EHMT1 = euchromatic histone-lysine N-methyltransferase 1 EZH2 = enhancer of zeste homolog 2 HBiX1 = HP1-binding protein enriched in the inactive X chromosome 1 HNRNP-U (AKA SAF-A) = heterogeneous ribonucleoprotein U  MH2A = macroH2A RNF2 (AKA RING1B) = ring finger protein 2 SAF-A (AKA HNRNP-U) = scaffold attachment factor A SMCHD1 = structural maintenance of chromosomes flexible hinge domain containing 1 SPOP = speckle-type POZ protein YY1 = Yin and Yang 1   x Acknowledgements I would like to thank everyone who has helped me throughout my thesis. In particular, I would like to thank my supervisor, Dr. Carolyn Brown, for being such a welcoming and inspiring person and for making my experience of research here at UBC such a good one. Also thanks to my committee members, Dr. Dixie Mager and Dr. Evica Rajcan-Separovic for their time and valuable thoughts on my project. I would also like to thank all members of the Brown lab, who have been a fun and friendly team to be a part of. And thanks also to my family, for their love and support.                    1  1 Introduction    2  1.1 Thesis overview The inactivation of one of the two X chromosomes in mammalian female cells is an example of epigenetic regulation. The inactive X chromosome is therefore a good model for the study of epigenetics, in particular of facultative heterochromatin. The localisation of XIST RNA to the future inactive X is necessary to the process of X inactivation. However, the mechanism of XIST RNA localisation remains largely unknown. In this study we use inducible XIST transgenes  integrated into the genome of HT1080 cells (a male fibrosarcoma cell line) to explore XIST?s ability to localise. Specifically, we explore XIST?s ability to localise: 1: when transcribed from a range of different autosomes, as opposed to the X chromosome; 2: upon deletion of different regions of XIST, and 3: upon knockdown of various proteins implicated in X inactivation. 1.2 Dosage compensation and the X-inactivation centre Mammalian females have two X chromosomes, whereas males have a single X chromosome (and a Y chromosome). To avoid a gene dosage imbalance between the sexes, in placental mammals one of the two X chromosomes in females in transcriptionally silenced in the process of X chromosome inactivation, hereafter referred to as XCI (Lyon 1961). XCI occurs early in development and is random, in that either the paternal or maternal X can become the inactive X. In each cell, whichever X is chosen to be silenced will remain silenced throughout all subsequent cell divisions, leading to clonal patches. These clonal patches are exemplified by the patches of different fur colours on female tortoiseshell cats (Lyon 1961). The X-inactivation centre (XIC) is the region of the X chromosome that controls the process of XCI, and is located at Xq13 (Brown et al. 1991). Within the XIC is the XIST gene (X-inactive specific transcript), which produces a ~17 kb spliced and polyadenylated long non-coding RNA (lncRNA) (Brown et al. 1992). Following its transcription, XIST RNA localises back to the chromosome from which it was transcribed and coats it in cis. The coating of the Xi (inactive X; the active X is referred to as Xa) by XIST RNA results in a substantial epigenetic transformation. During XCI, the chromosome loses epigenetic modifications associated with active chromatin and modifications associated with inactive chromatin (see section 1.3 for more details). The importance of XIST in XCI is demonstrated by the heterozygous deletion of the Xist gene in mice (note that the human gene is written XIST whereas the mouse gene is written Xist), which results in non-random XCI in which only the chromosome bearing the Xist gene becomes inactivated; the other chromosome, without Xist, is unable to initiate XCI (Penny et al. 1996). 3  The upregulation of Xist expression occurs immediately prior to the onset of XCI (Kay et al. 1993). However, once XCI has been established, XIST expression is no longer required for the silence of the chromosome to be maintained (Brown & Willard 1994; Csankovszki et al. 1999; Csankovszki et al. 2001); it is thought that at this stage the silent status has been ?locked in? by chromatin modifications.  1.3 Features of the Xi As mentioned previously, the coating of the future Xi by XIST RNA results in epigenetic changes on the X chromosome. These features are discussed below. 1.3.1 Histone methylation Histones are the proteins around which DNA is wrapped, forming chromatin. A section of DNA wrapped around a set of histones forms a nucleosome. There are four canonical core histones ? H2A, H2B, H3 and H4 ? and each nucleosome contains two copies of each of these. In addition, there is a linker histone protein, H1, which lies outside of the nucleosome. H1 is involved in chromatin compaction. The core histones have N-terminal ?tails? that protrude outwards from the nucleosome. These tails can undergo a variety of modifications that affect gene transcription levels (Lennartsson & Ekwall 2009).  The inactive X is heterochromatic. Heterochromatin can be classed as either constitutive or facultative. Constitutive heterochromatin, which includes centromeres and telomeres, is permanently in a heterochromatic state. On the other hand, facultative chromatin,  as represented by the inactivation of an X chromosome, has the potential to be either heterochromatic or euchromatic. In terms of epigenetics, facultative and constitutive heterochromatin are distinct. H3K9me2 (i.e. dimethylation of lysine 9 on histone H3), which is laid down by EHMT1 (GLP) and EHMT2 (G9a), is a mark associated with facultative heterochromatin, along with H3K27me3 and H4K20me1. Constitutive chromatin, on the other hand, is associated with H3K9me3, H3K27me1 and H4K20me3 (Peters et al. 2003; Schotta et al. 2004), and also HP1 (heterochromatin protein 1). HP1 binds di- and tri-methylated H3K9 via its chromodomain (Lachner et al. 2001), though with much greater affinity to the trimethylated form (Jacobs et al. 2001) and dimerises via its chromoshadowdomain (Cowieson et al. 2000). Two histone methyltransferases, SUV4-20H1 and SUV4-20H2, bind HP1 and lay down H4K20me3 (Schotta et al. 2004). Figure 1.1 shows pathways that are thought to occur in each type of heterochromatin. 4  The Xi is compartmentalised into two different types of heterochromatin, one associated with XIST, H3K27me3 and macroH2A, and the other associated with H3K9me3, HP1 and H4K20me3 (Chadwick & Willard 2004). This indicates that the Xi is composed of domains showing features of either facultative or constitutive heterochromatin.   HBiX1 binds HP1 and is enriched on the Xi (Nozawa et al. 2013). HBiX1 also binds SMCHD1. SMCHD1 is also enriched on the Xi and is required for DNA methylation of X-linked genes (Blewitt et al. 2008). HBiX1 and SMCHD1 form a bridge between the two different domains of the Xi (i.e. the domains bearing hallmarks of either constitutive or facultative heterochromatin). HBiX1 binds HP1 in the constitutive heterochromatin domain, and binds SMCHD1, which is closer to the facultative heterochromatin domain and presumably binds a protein in the facultative heterochromatin domain. Knockdown of either SMCHD1 or HBiX1 results in decompaction of the chromosome (Nozawa et al. 2013).  5   Figure 1.1 Epigenetic pathways in constitutive and facultative heterochromatin * PRC1 is also recruited by a PRC2-independent mechanism.  Compared with the Xa, the Xi is enriched for various epigenetic modifications that are associated with inactive chromatin and depleted of epigenetic modifications associated with active chromatin, summarised in Table 1.1, and discussed below.        6  Table 1.1 Histone modifications enriched and depleted on the inactive X Enriched Depleted H3K27me3 (Silva et al. 2003) H3K4me3 (Boggs et al. 2002; Goto et al. 2002) H2AK119u1 (De Napoles et al. 2004; Fang et al. 2004) H3K36me2 (Chadwick & Willard 2003) H3K9me2 (Heard et al. 2001; Mermoud et al. 2002; Peters et al. 2002) Histone acetylation (Belyaev et al. 1996; Jeppesen & Turner 1993; Yan & Boyd 2006) H3K9me3 (Chadwick &  Willard 2004)  H4K20me1 (Kohlmaier et al. 2004)   H3K27me3 is enriched on the Xi. This mark is laid down by the EZH2 subunit of PRC2 (polycomb repressive complex 2). A common model for the downstream effect of H3K27me3 is the recruitment of PRC1 (Czermin et al. 2002). PRC1 lays down H2AK119u1 (i.e. monoubiquitination of lysine 119 one histone H2A). PRC1 binds H3K27me3 via a chromodomain-containing CBX subunit (diagrams showing the subunits of PRC2 and PRC1 are shown in Figure 1.1 PRC2 and PRC1 subunits). However, PRC2 is not necessary for PRC1 recruitment; mouse ES cells lacking Eed (a subunit of PRC2) are unable to recruit the Mph1 subunit of PRC1 (=?HPH1? in Figure 1.1), but they do still recruit the catalytic subunit of PRC1, Ring1B (also known as RNF2) which proceeds as usual to lay down H2AK119u1 (Schoeftner et al. 2006). This indicates that PRC1 is recruited by both the PRC2 pathway and additional PRC2-independent mechanism.   7   Figure 1.1 PRC2 and PRC1 subunits  PRC1 ubiquitinates H2A using its RING1B/RNF2 subunit (Wang et al. 2004). uH2A (ubiquitinated H2A) has repressive functions mediated by different mechanisms, as follows.  The ubiquitinated lysine in uH2A, lysine 119, lies in the C-terminal tail of H2A, which can reach histone H1 (Luger et al. 1997). Mutating lysine 119 decreases the association between the C-terminal tail of H2A and H1 (Zhou et al. 2008). This suggests that the ubiquitination of H2A increases the association with H1. H1 can be phosphorylated, and the phosphorylated form of H1 is associated with open chromatin. Cells unable to deubiquitinate H2A exhibit increased uH2A and decreased H1 phosphorylation (and conversely knocking down Ring1b decreases ubiquitination of H2A and increases phosphorylation of H1). Therefore, uH2A increases the association of the C-terminal tail of H2A with H1 and prevents H1 phosphorylation, resulting in chromatin compaction. uH2A also inhibits promoter escape of RNA polymerase ll (Zhou et al. 2008).  PRC1 also has chromatin compaction capabilities independent of its ubiquitin ligase activity. Polycomb complexes are required to compact chromatin at Hox loci - absence of either PRC2 or PRC1 results in loss of compaction (Eskeland et al. 2010). Interestingly, PRC1?s capacity to compact chromatin and repress gene expression is not dependent on its ubiquitination of 8  histone H2A (Endoh et al. 2012; Eskeland et al. 2010). It would seem logical then that the RING1B/RNF2 subunit of PRC1 (the subunit that ubiquitinates H2A) might also not be required for compaction; however, absence of RING1B/RNF2 results in decompaction (Eskeland et al. 2010). This shows that RNF2/RING1B has compaction activity in addition to that associated with its ubiquitination of H2A.   1.3.2 DNA methylation  Another feature of the Xi is DNA methylation; specifically, the methylation of cytosine in the context of CpG dinucleotides. CpG dinucleotides are unevenly distributed throughout the genome; they are found in clusters known as CpG islands, which are found at 60% of gene promoters (Larsen et al. 1992). DNA methylation is laid down by the DNA methyltransferases (DNMT). DNMT3A and DNMT3B are de novo methyltransferases that methylate previously unmethylated DNA. Following DNA replication, only the parental DNA strand is methylated and the DNA is therefore hemimethylated; DNMT1 methylates the unmethylated strand so that methylation persists throughout cell divisions (Hermann et al. 2004). DNA methylation of a CpG island at a gene promoter is generally associated with transcriptional silencing of that gene (Stein et al. 1982). On the Xi, genes that are subject to inactivation have methylated CpG islands whereas genes that escape inactivation have unmethylated CpG islands (Hansen & Gartler 1990). Treatment of human cells with DNA methyltransferase inhibitors results in reactivation of genes on the Xi (Venolia et al. 1982), demonstrating the importance of DNA methylation in the maintenance of X-linked gene silencing. Mutations in both the de novo and the maintenance DNA methyltransferases result in XCI defects. Mutations in the de novo DNA methyltransferase DNMT3B in humans results in ICF syndrome (immunodeficiency, centromere instability and facial anomalies) (Hansen et al. 1999). People with ICF syndrome have undermethylated promoters of X-linked genes, resulting in escape from inactivation (Hansen et al. 2000). On the other hand, mutations in the maintenance methyltransferase Dnmt1 result in the ectopic expression of Xist in male mice and Xist expression from both Xs in female mice, leading to ectopic XCI (Beard et al. 1995; Panning & Jaenisch 1996). Therefore, various lines of evidence point towards a critical role of DNA methylation in XCI. 9  1.3.3 Cot-1 holes Another feature of the Xi is so-called ?Cot-1 holes?. The complexity of a genome can be measured using DNA reassociation kinetics, which is a technique that measures how quickly denatured DNA reanneals.  DNA reassociation kinetics is based on the principle that highly repetitive DNA sequences reanneal much faster than DNA sequences that are rare in the genome. This is because it is much more likely that a sequence that is abundant in the genome will encounter its complementary sequence than would a sequence that is only found once in the genome. The reassociation of a denatured genome can be shown on a Cot (concentration over time) curve. Eukaryotic genomes show Cot curves with three distinct regions, each one representing the reannealing of a different class of DNA sequences. The first class to reanneal is composed of highly repetitive regions of the genome (centromeres and telomeres). Single copy sequences reanneal last. Between these two groups are the middle repetitive sequences, which include, amongst other sequences, LINEs (long interspersed nuclear elements) and SINEs (short interspersed nuclear elements). This class of sequences has a Cot value of one and is therefore referred to as Cot-1 DNA. Cot-1 DNA can be used as a probe in RNA FISH (fluorescence in situ hybridisation) experiments to indicate the transcriptional status of these repetitive elements (Hall et al. 2002). Shortly after XIST RNA coats the future Xi, a so-called ?Cot-1 hole? appears. A cot-1 hole is a depletion of the Cot-1 RNA FISH signal coincident with the Xi, and it indicates transcriptional silencing of the Cot-1 DNA on the Xi. This Cot-1 hole coincides with the absence of RNA pol ll from the Xi (Chaumeil et al. 2006). Cot-1 holes were originally thought to represent overall silencing of the Xi. However, the formation of a Cot-1 hole is independent of gene silencing; genic sequences lie on the periphery of the Cot-1 domain. In other words, the Xi is composed of a core of repetitive sequences surrounded by genic sequences. A Cot-1 hole only indicates silencing of the repetitive elements and not necessarily of genes. Following the formation of this transcriptionally silent core, genes that are subject to XCI are pulled in towards the silent Cot-1 domain, whereas those genes that escape XCI remain more peripheral (Chaumeil et al. 2006).  1.3.4 Perinuclear/perinucleolar localisation Another feature of the Xi is its localisation within the nucleus. The Xi is often found either at the edge of the nucleus (i.e. a perinuclear position) or next to the nucleolus (i.e. a perinucleolar position) (Barr & Bertram 1949; Bourgeois et al. 1985; Borden & Manuelidis 1988). One hypothesis for the perinuclear localisation is that the perinuclear compartment is a general sink for heterochromatin (Bourgeois et al. 1985). However, the Xa is also attached to the nuclear 10  envelope (Eils et al. 1996). Perinucleolar localisation, however, is Xi-specific. Supporting the role of the perinucleolar compartment in XCI is the fact that both human and mouse autosomes harbouring XIST or XIC transgenes localise to the nucleolus (Hall et al. 2002; Zhang et al. 2007). Interestingly, the Xi is located next to the nucleolus specifically during S phase (Zhang et al. 2007), suggesting that perinucleolar positioning may be important for the replication of the Xi. This is supported by the fact that the perinucleolar region is enriched for Snf2h, the catalytic subunit of a complex required for progression of the replication fork through condensed chromatin (Collins et al. 2002). A role for the perinucleolar region in maintaining repression on the Xi is supported by the fact that another long noncoding RNA, kcnq1ot1, mediates transcriptional repression by targeting the chromosomal domain it regulates to the perinucleolar region during S phase (Mohammad et al. 2008).  1.3.5 Other features of the Xi Heterochromatin replicates later in S-phase than does euchromatin (Limadefa & Jaworska 1968). The Xi follows this rule and replicates later than the Xa (and the autosomes). The late replication of the Xi has often been used as a method of identifying it (e.g. Sharp et al. 2001). The Xi is enriched in the variant of histone H2A called macroH2A/mH2A (Costanzi & Pehrson 1998). The accumulation of mH2A on the Xi results in the formation of a macrochromatin body, or MCB. mH2A is composed of an N terminal region that is similar to that of canonical H2A and a C terminal region that consists of a large nonhistone macro domain. These domains are connected by a basic linker region. The linker region reduces access to linker DNA (linker DNA is the stretch of DNA between nucleosomes) (Chakravarthy et al. 2012) and has chromatin compaction activity (Muthurajan et al. 2011).  Due to its condensed nature, the Xi is distinctly visible upon DNA staining. The DNA-dense Xi is referred to as a Barr body (Barr and Bertram, 1949). The Xi is enriched for a variety of proteins (for more information on these proteins, see section 1.6).  1.4 Sequences of XIST RNA Comparative sequence analysis reveals relatively poor conservation of the primary sequence of XIST between different mammalian species. However, there are conserved regions of tandem repeats, termed A-F (Brockdorff et al. 1992; Nesterova et al. 2001). Human XIST has eight 11  exons. Exon 1 is the largest exon (11kb) and contains all the repeats except for repeat E, which is in exon 6. Exon 6 is the second largest exon (~4.6kb). The locations of the repeats are shown in Figure 3.7. Little is known about the functions of the different sequences within XIST, and in particular about which sequences of XIST are responsible for localising XIST RNA to the X. Repeat A is a ~500bp sequence consisting of 9 copies of a 26mer sequence at the 5? end of XIST. Deleting repeat A results in a lack of gene silencing (Wutz et al. 2002). Repeat A forms putative stem loop structures that interact with PRC2 in vitro (Zhao et al. 2008), and repeat A is thus thought to be involved in PRC2 recruitment to the Xi. However, it is not necessary for PRC2 recruitment, as mouse ES cells lacking repeat A are capable of recruiting H3K27me3, the mark laid down by PRC2 (Pullirsch et al. 2010). As mentioned in section 1.3.3, the Xi is composed of a core of transcriptionally silent repetitive elements surrounded by genic sequences, and during XCI the genes subject to inactivation are pulled in towards the silent core (Chaumeil et al. 2006). Importantly, the relocation of genes into the silent core is dependent on repeat A. It therefore seems likely that the requirement for repeat A in the silencing of X-linked genes lies in its involvement in the relocation of genes into the silent compartment. Various studies have aimed to identify regions of XIST required for its cis-localisation. Repeat C has been assessed in several studies for involvement in XIST RNA localisation. In mice there are 14 copies of the repeat C monomer, whereas in humans there is only one. PNA (peptide nucleic acid) interference of repeat C results in Xist RNA delocalisation (Beletskii et al. 2001), as does LNA (locked nucleic acid) interference (Sarma et al. 2010). Importantly, however, this is only the case for mouse cells; LNA interference of repeat C in human cells does not delocalise XIST RNA.  The molecular role of repeat C in XIST localisation is in the initial tethering of Xist RNA to the X chromosome (Jeon & Lee 2011). The transcription factor YY1 binds both Xist RNA and the Xist gene, and anchors the Xist RNA to the X chromosome at the Xist gene. Importantly, YY1 binds to Xist RNA at repeat C (and also repeat B).  A series of deletions throughout Xist and subsequent Xist RNA FISH show that, unlike the well-defined sequence requirement for silencing (i.e. repeat A), sequences that localise Xist RNA do so in a redundant fashion (Wutz et al. 2002). In contrast to the above studies implicating repeat C in Xist localisation, deleting repeat C in mouse ES cells does not disrupt localisation (Wutz et 12  al. 2002). Therefore, while repeat C does seem to have a role in Xist localisation through its interaction with YY1, it is not necessary.  The 211 nucleotide exon 4 of Xist is highly conserved, perhaps indicating importance in XCI. Furthermore, computational prediction indicates that it forms a stem-loop structure, perhaps indicating protein binding (Caparros et al. 2002). Deletion of exon 4 in mouse ES cells does not, however, disrupt Xist localisation. This allele and others described below are shown diagrammatically in Figure 1.2 Structure of mutant Xist alleles Another study used the above mutant allele of Xist (i.e. the deletion of exon 4) and in addition created an inversion that disrupts conserved sequences extending from 6kb into exon 1 to intron 5 (Senner et al. 2011).. This allele, XistINV, consists of the first 6kb of Xist (which goes as far as between repeat C and repeat D) followed by either 6kb or 14kb (two isoforms were found) of sequence antisense to Xist. RNA FISH on interphase cells shows much smaller Xist foci than wild type Xist, and Xist RNA FISH on metaphase chromosomes shows that XistINV does not localise to the X. In mice, XCI is imprinted in the placenta ? the paternal X is always inactivated. Paternal transmission of a mutant Xist allele unable to initiate silencing is therefore lethal, due to failure of imprinted XCI in the placenta. (In humans, XCI is random in the placenta (de Mello et al. 2010).) Paternal inheritance of XistINV results in embryonic lethality due to failure of imprinted XCI in the placenta, and maternal inheritance results in secondary skewed inactivation due to elimination of cells expressing the XistINV allele. This shows that large scale inversion of sequences in Xist affect Xist RNA localisation.  Figure 1.2 Structure of mutant Xist alleles 13  Figure shows various Xist alleles discussed in the text. Sequence deleted in XistEx4del is shown by dashed lines. Sequence inverted in XistINV is shown in grey. Location in XistIVS of exogenous 16bp is marked by the white arrow. Figure adapted, with permission, from Sado and Brockdorff, 2013.  Another Xist allele contains a 1kb intron exogenously introduced 0.9kb from the transcription start site (Hoki et al. 2011). When spliced out, the resulting Xist RNA is the same as the wild type except for a 16bp sequence left over from the introduction of the intron. This allele, XistIVS, produced RNA that is able to coat the X chromosome in TS (trophoblast stem) cells, leading to H3K27me3 accumulation and depletion of H4 acetylation. However, gene silencing is compromised and Xist localisation, along with H3K27me3, are reduced upon differentiation of these cells. Mice with a paternally inherited XistIVS allele never survive past midgestation due to a failure of imprinted XCI. Xist RNA levels decreased upon differentiation, hence the loss of Xist localisation. It is possible that splicing of the exogenous exon disrupted Xist RNA processing and resulted in an unstable transcript. In summary, not much is known about the sequences of XIST required for its localisation. Repeat A is required for silencing, and repeat C is involved in the initial tethering of Xist RNA to the Xi, though is not required. Knowing the sequences involved would narrow down the search for proteins that bind Xist and help in the development of a model of how XIST localises. Additionally, analysis of the sequences involved would mean that any sequence features identified could be searched for in other lncRNAs to help elucidate how other lncRNAs function.  1.5 Spreading of XIST RNA on autosomes X;autosome translocations were used to define the region of the X that is required for X inactivation to occur. i.e. the XIC (Russel 1963; Rastan 1983; Rastan & Robertson 1985; Lyon et al. 1986; Keer et al. 1990; Brown et al. 1991). In females that carry an unbalanced X;autosome translocation it is the derivative X that is normally inactivated, because the silencing of the X can spread to some extent into the autosomal segment and suppress transcription from these excess copies of genes from the translocated autosome. However, the spread of XIST RNA into the autosomal material is highly variable and for some X;A translocations, spreading of XIST RNA stops at the X;A boundary, not extending into the autosomal segment (Duthie et al. 1999; Keohane et al. 1999). It has been hypothesised that there are DNA elements, putative 14  ?way stations?, that help XIST RNA to spread along the X chromosome (Gartler & Riggs 1983). Such putative elements would be more abundant on the X chromosome than on autosomes, thereby explaining the limited spread of XIST RNA onto the autosomal segment of X;A translocations. There is much evidence, though, that Xist RNA can spread on autosomes. Such evidence comes from studies of autosomes harbouring Xist or Xic transgenes. Such studies have shown not only Xist localising to autosomes but also resulting in various features of an inactive X on autosomes. Mouse ES cells are a useful experimental system for the study of X inactivation. In the undifferentiated state, XCI has not yet taken place and female cells have two active Xs; upon differentiation one of the two Xs becomes inactivated. This allows the study of the earliest steps in the process of XCI. A 450kb transgene containing the mouse Xic integrated into three different autosomes in mouse ES cells produced Xist RNA that localises to the autosomes (Lee et al. 1996). Visualisation of the chromosomes in late prophase when the chromosomes are condensed supports that Xist RNA is coating the autosomes and not just localising around the chromosome. Furthermore, expression of a lacZ transgene cointegrated with the Xic transgene on the autosomes became silenced upon differentiation of the cells, showing that Xist RNA is capable of silencing a gene on an autosome. Fibroblasts from mice derived from one of these cell lines displayed a two-fold reduction in expression of endogenous genes on chromosome 12 by RT-PCR(chromosome 12 was identified as the Xic-bearing autosome). An additional two genes on chromosome 12 showed a shift from biallelic expression in control cells to monoallelic expression in transgenic cells, supporting the silencing of autosomal genes by Xist. The transgenic chromosome 12 also displayed chromosome-wide histone H4 hypoacetylation, a characteristic feature of the Xi (see section 1.3). The transgenic chromosome 12 also displayed late replication, another hallmark of the Xi. This demonstrates that Xist RNA can direct, at least partially, the heterochromatinisation of an autosome.  The above studies were done in mouse cells, and similar results have been obtained in human cells. Integration of human XIST into several autosomes of HT1080 cells (a human male fibrosarcoma cell line, and the cell line used in this thesis) resulted in XIST RNA localisation to the autosomes (Hall et al. 2002). XIST localisation resulted in histone H4 hypoacetylation and late replication. Similarly to the silencing of lacZ seen by Lee et al. (1996) in mouse cells, XIST localisation in HT1080s resulted in silencing of a neomycin resistance gene cointegrated with XIST. Cot-1 holes (see section 1.3 for definition of a Cot-1 hole) also formed on the autosomes, indicating long-range silencing. 15  In a study from the Brown laboratory (Chow et al. 2007), also performed in human cells, enrichment of macroH2A, H3K27me3 and H4K20me1 resulted from XIST expression from an autosome bearing an inducible XIST transgene in HEK293s. Assessment of chromatin changes at the promoter of a cointegrated EGFP gene by ChIP (chromatin immunoprecipitation) revealed a decrease in various ?active? epigenetic marks at the promoter, including H4 acetylation and H3K4 di- and tri-methylation, and an increase in the ?silent? epigenetic mark H4K20me1 as well as HP1. XIST expression also resulted in Cot-1 holes.  In addition, expression of an inducible Xist cDNA transgene integrated into chromosome 11 of mouse ES cells results in enrichment of H3K27me3 and H2AK119u1 (the mark laid down by PRC1) on chromosome 11 (Pullirsch et al. 2010). Two proteins enriched on the Xi, Ash2l and Hnrnp-u, were also recruited to the chromosome 11 ? see ?section 1.6 for details of these proteins.) Together, these transgene studies demonstrate that XIST RNA is able not only to spread along autosomal material but also result in a host of epigenetic modifications and other features normally observed on the Xi. While the above observations are not incompatible with the notion of way stations that aid in the spreading of XIST RNA, they do indicate that the lack of spreading of XIST onto autosomal segments of X;A translocations is not due to a lack of putative way stations on the autosomes. While the results described above indicate that XIST is capable of localising to autosomes, only a subset of autosomes have been assessed for their capacity to support XIST localisation. It remains possible that there is some sequence specificity to XIST RNA localisation and that further assessment of autosomes bearing XIST transgenes will identify such sequence specificities.   1.6 Proteins implicated in X inactivation Various proteins have been implicated in the process of XCI. The table below lists these proteins, some of which are discussed in more detail below.     16  Table 1.2: Proteins implicated in XCI Protein / complex Enriched on Xi? Required for Xist localisation? Required for gene silencing? YY1 No (Jeon & Lee 2011) Localisation ? knocking down YY1 results in loss of Xist foci by RNA FISH (Jeon & Lee 2011).  PRC2 Transiently, during initiation of XCI (Silva et al. 2003)  Silencing - Eed (a component of PRC2) mutant mice are unable to maintain XCI in trophoblast cells (Wang et al. 2001). PRC1 PRC1 components transiently enriched during ES cell differentiation (De Napoles et al. 2004; Bernstein et al. 2006). Also enriched in differentiated cells (MEFs and HEK293s) (Plath et al. 2004)  Has gene silencing activity but has not specifically been shown to be required for silencing on the Xi. hnRNP-U (SAF-A) Yes (Helbig & Fackelmayer 2003; Pullirsch et al. 2010) Localisation ? knockdown of hnRNP-U results in delocalised Xist RNA (Hasegawa et al. 2010) SMCHD1 Yes (Blewitt et al. 2008) Silencing ? knockout of Smchd1 in mice leads to partial upregulation of X-linked genes together with  abnormalities in DNA methylation (Blewitt et al. 2008). Also required for compaction of the Xi (Nozawa et al. 2013) SATB1 Not enriched on Xi in lymphocytes (Agrelo et al. 2009) Silencing ? Loss of SATB1 diminishes silencing by Xist (Agrelo et al. 2009).  17  Protein / complex Enriched on Xi? Required for Xist localisation? Required for gene silencing? DNMT3A/3B/1 DNMT3B depleted on Xi in human epithelial cells (Chadwick & Willard 2003) Silencing ? treatment of cells with DNA methyltransferase inhibitors results in Xi reactivation (Venolia et al. 1982) SPOP and CUL3 CUL3 not enriched; SPOP not tested (Hern?ndez-Mu?oz et al. 2005) Silencing ? knockdown of SPOP or CUL3 in mouse fibroblasts with a silent GFP transgene on the Xi leads to reactivation of GFP in the presence of inhibitors of DNA methylation and histone deacetylation (Hern?ndez-Mu?oz et al. 2005) ATRX Yes (Baumann & De La Fuente 2009) Unknown EHMT1 Unknown (but the mark that it lays down, H3K9me2, is enriched on the Xi) Unknown CBX7 Yes (Bernstein et al. 2006) Unknown ASH2L Yes (Pullirsch et al. 2010) Ash2l knockdown in mouse ES cells containing an Xist transgene on chromosome 11 did not result in lack of silencing of a cointegrated puromycin resistance gene (Pullirsch et al. 2010)  1.6.1 YY1 YY1 is a ubiquitous transcription factor that can activate and repress gene expression, and is involved in processes including proliferation, differentiation and apoptosis (reviewed in He & Casaccia-Bonnefil (2008)). Jeon and Lee (2011) used EMSA (electrophoretic mobility shift assay) to show that YY1 binds near repeat F of Xist DNA and used a pulldown assay to show that YY1 binds Xist RNA through repeat C, and possibly also repeat B. 18  1.6.2 hnRNP-U/SAF-A hnRNP-U contains both RNA and DNA binding domains (Fackelmayer et al. 1994) and is enriched on the Xi. Enrichment requires the RNA binding RGG domain of hnRNP-U,  (Helbig &  Fackelmayer 2003) and also the DNA binding SAP domain (Pullirsch et al. 2010). Hasegawa et al. (2010) used UV cross-linking followed by immunoprecipitation to show that hnRNP-U interacts with Xist RNA via its RNA-binding RGG domain, with the greatest enrichment in the centre of exon 1. Deletion of Xist results in loss of Xi enrichment of hnRNP-U on the Xi (Pullirsch et al. 2010). hnRNP-U enrichment is not perturbed by deletion of repeat A (Pullirsch et al. 2010). 1.6.3 SMCHD1 In a screen aimed at identifying genes involved in gene silencing, Blewitt et al. (2005) identified a mutation that when homozygous resulted in embryonic lethality in female but not male mice, which is indicative of a role in XCI. Indeed, the mutation is in Smchd1. Smchd1 produces a protein that is enriched on the Xi.  Mutations in Smchd1 result in hypomethylation of genes on the Xi, leading to loss of silencing of X-linked genes (Blewitt et al. 2008). SMCHD1 is also involved in compaction of the Xi ?SMCHD1 is required to bridge together the constitutive heterochromatin and facultative chromatin domains of the Xi (Nozawa et al. 2013). Depletion of SMCHD1 results in loss of the Barr body and the spreading apart of FISH signals showing the location of X-linked sequences, demonstrating that SMCHD1 is required for Xi compaction. 1.6.4 ASH2L ASH2L is a trithorax group protein involved in the deposition of H3K4me3, a mark found at the promoters of active genes. Surprisingly, Ash2l is enriched on the Xi of mouse 3T3 cells and differentiated mouse ES cells, even though H3K4me3 is depleted on the Xi (Pullirsch et al. 2010). Conditional deletion of Xist in the 3T3 cells results in loss of Ash2l enrichment on the Xi. Knockdown of Ash2l in mouse ES cells harbouring an Xist transgene on chromosome 11 did not result in loss of silencing of a gene co-integrated with Xist, indicating that Ash2l is not required for silencing. The molecular role of Ash2l on the Xi is unknown. 1.6.5 CBX7 CBX7 is a subunit of PRC1. The mammalian CBX proteins are homologs of the Drosophila Polycomb (Pc) protein that binds the H3K27me3 mark via its chromodomain (the five homologs are CBX2, 4, 6, 7 and 8). Each of these CBX proteins, with the exception of CBX4, are enriched 19  on the Xi in differentiating mouse ES cells (Bernstein et al. 2006). CBX4, howeveris enriched on the Xi in MEFs and HEK293s (Plath et al. 2004). Though Drosophila Pc binds solely to the H3K27me3 mark, the binding preferences of its mammalian homologs are more complex, with the chromodomains of the different CBX proteins having preferential binding affinity for H3K27me3 or H3K9me3. The chromodomain of Cbx7, however, has strong affinity for both H3K27me3 and H3K9me3 (Bernstein et al. 2006).  Cbx7?s chromodomain binds single-stranded RNA and removal of single-stranded RNA from ES cells results in loss Cbx7 enrichment on the Xi (Bernstein et al. 2006). These data might be indicative of a role of Xist RNA in recruiting Cbx7 to the Xi.  1.6.6 CUL3 and SPOP CUL3 and SPOP together form a ubiquitin ligase complex. The CUL3/SPOP complex ubiquitinates two proteins that are recruited to the Xi ? BMI1 (a subunit of PRC1) and mH2A (Hern?ndez-Mu?oz et al. 2005). Knockdown of either CUL3 or SPOP results in loss of mH2A from the Xi. Knockdown of either protein also results in reactivation of an X-linked GFP transgene in the presence of inhibitors of DNA methylation and histone deacetylation. The CUL3/SPOP complex may therefore be required for silencing of the Xi.  As shown above, various proteins are known to be involved in XCI. However, there is not yet an overarching model of how these proteins come together to silence an X chromosome. The identification of other proteins involved is a key step to understanding the inactivation process.   1.7 Thesis objective The aim of my project was to elucidate mechanisms of XIST RNA localisation. Specifically, the aims were 1) to assess XIST RNA?s ability to localise in a variety of different chromosomal contexts using HT1080s with an XIST transgene inserted into a range of different chromosomes; 2) to identify regions of XIST required for its localisation, using XIST deletion constructs; and 3) to identify proteins required for XIST RNA localisation, by assessing XIST localisation in cells that have had various proteins knocked down. Many studies on XCI have been done in mouse cells. Even though mouse ES cells are a convenient system for the study of X inactivation, there are multiple differences in how XCI occurs between mice and humans. Studies in human cells are therefore important. 20 2 Methods  21 2.1 Cell culture Cells were grown at 37?C with 5% CO2. HT1080 cells were grown in DMEM (Gibco) supplemented with penicillin/streptomycin (Gibco), non-essential amino acids (Gibco) and 10% V/V fetal bovine serum (PAA Laboratories Inc.). XIST expression was induced with the addition of 1?g/ml doxycycline to the culture medium. 2.2 Fixation of cells for FISH and immunofluorescence Cells were grown on glass coverslips. Upon removal from cell culture the coverslips were first rinsed in ice-cold CSK buffer (0.3M sucrose, 100?M NaCl, 10?M PIPES, 3?M MgCl2,), then permeabilised with 0.5% Triton-X 100 in CSK for eight minutes on ice and then fixed in 4% paraformaldehyde (Electron Microscopy Sciences) for eight minutes at room temperature. Coverslips were stored at 4?C in 70% ethanol. 2.3 RNA FISH Just prior to RNA FISH, the coverslips were immersed in 100% ethanol for five minutes and left to air dry. FISH was performed with two probes: an XIST probe and a Cot-1 probe. Cot-1 DNA from Invitrogen was used as for the Cot-1 probe. Both probes had been directly fluorescently labeled using the Nick Translation Reagent Kit (Abbott Molecular Inc) with Spectrum red-UTP (Vysis) for Cot-1 DNA probes and Spectrum green-UTP (Vysis) for XIST probes. ~150ng of each probe was mixed together along with 20?g salmon testes DNA. This probe mixture was air dried in a speed vacuum, resuspended in 10?l deionized formamide (Sigma), denatured at 80?C for ten minutes, and then mixed with 10?l hybridization buffer (20mg/ml BSA (Roche), 4XSSC (Invitrogen), 20% dextran sulphate). This was pipetted onto a small square of Parafilm and the coverslip was placed on top of the probe mixture. Another piece of Parafilm was then placed on top and the edges were sealed to prevent the drying out of the coverslip. Hybridisation took place overnight in a humidified chamber at 37?C. The next day the coverslips were rinsed as follows: 20 minutes in 50% formamide / 50% 4XSSC at 37?C, 20 minutes in 2X SSC at 37?C and 20 minutes in 1X SSC at room temperature. Coverslips were then stained with DAPI and mounted onto microscope slides with Vectashield (Vector Laboratories). 2.4 Fluorescence microscopy and image analysis Cells were observed on a Leica inverted microscope (DMI 6000B) at 100X magnification and images were obtained using a Retiga 4000R (Q-Imaging) camera and using Openlab software (PerkinElmer). Images were processed using Adobe Photoshop CS4 to reduce background and correct for variation in FISH efficiency between different images. ImageJ was used to define the area of the XIST signal and to measure the intensity of the Cot-1 staining under the XIST signal relative to the intensity in the nucleus.  22 2.5 RNA isolation, reverse transcription and Q-PCR RNA was isolated from cell pellets stored at -70?C using TRIZOL (Invitrogen) according to the manufacturer?s instructions and then treated with DNase1 (Roche). RNA concentration was determined by spectrophotometry and 0.5-2.5?g RNA was put into a reverse transcription reaction using M-MLV reverse transcriptase (Invitrogen). The resulting cDNA was then used for qPCR on a StepOnePlusTM Real-Time PCR System (Applied Biosystems, Darmstadt, Germany), using Maxima Hot Start Taq (Fermentas) and EvaGreen dye (Biotium). The following conditions were used: 95? for 5 min, followed by 40 cycles of [95? for 15 s, 60? for 30 s, 72? for 1 minute], and a melt curve stage of [95? for 15 s, 60? for 1 min, increase of 0.3? until 95?]. Gene of interest levels were normalised to actin or PGK1. Primer sequences are found in Table 2.2 List of qPCR primers. 2.6 Immunofluorescence combined with RNA FISH Coverslips, which had been stored at 4?C in 70% ethanol, were first rinsed in PBS. They were then placed onto a small amount of PBT (PBS with 1% BSA (Amresco) and 0.1% Tween 20) containing 0.4U/?l Ribolock (Thermo Scientific). Coverslips were sealed between two layers of Parafilm (Bemis) and left in the blocking buffer at room temperature for 20 minutes. The coverslips were then transferred from blocking buffer to PBT (1% BSA, 0.1% Tween 20 in PBS) containing 1:100 primary antibody and 0.4U/?l Ribolock, sealed between two layers of Parafilm and left at room temperature for four to six hours. Coverslips were then washed three times, for five minutes each time, at room temperature in PBS containing 0.1% Tween 20 (Fisher Scientific), then put onto a small amount of PBT containing 1:250 fluorescently labeled secondary antibody and 0.4U/?l Ribolock, sealed between two layers of Parafilm and left at room temperature in the dark for 45 minutes. Coverslips were then washed three times, for five minutes each time, at room temperature in the dark in PBS containing 0.1% Tween 20. Coverslips were then fixed in 4% PFA in PBS for 10 minutes at room temperature in the dark, and washed for five minutes in PBS before continuing on to RNA FISH, as described in 2.3, making sure that the coverslips remained in the dark throughout the RNA FISH procedure.  Antibodies used in immunofluorescence include: anti-ASH2L from Bethyl Laboratories (A300-107A); anti-H3K27me3 (07-449 from Millipore); anti-macroH2A (07-219 from Millipore). 2.7 siRNA-mediated knockdown siRNA-mediated knockdown was carried out according to the manufacturer?s protocol. Knockdowns were done in a 24-well plate, with 30,000 cells seeded per well. The day after seeding the cells, siRNA was added. 0.5-1.0?l DharmaFect 4 transfection reagent (Thermo Scientific) and 2.5-5.0?l of 5?M siGenome SMARTpool siRNA (Thermo Scientific) were used 23 per 500?l medium in each well. The day after transfection, the transfection medium was removed and replaced with medium containing 1?g/ml dox to induce XIST expression. After two days, coverslips were fixed ? see 2.2 Fixation of cells.  Table 2.1 List of siRNAs Target Gene Accession number Product number ASH2L NM_004674 M-019831-01 CUL3 NM_003590 M-010224-02 HNRPU NM_004501 M-013501-01 SPOP NM_001007228 M-017919-02 YY1 NM_003403  M-011796-02    Table 2.2 List of qPCR primers Primer name Sequence qACTB 1  TTGCCGACAGGATGCAGAA  qACTB 2  GCCGATCCACACGGAGTACTT  qPgk1 1F  GGCACTTGGCGCTACACAA  qPgk1 1R  CCTACCGGTGGATGTGGAAT  qCUL3_F  TCAGTCAGCCACACCAAAGTG  qCUL3_R  CACTGTGTTTGGCTAAGTAGAACCTT  24 Primer name Sequence qSPOP_F  TTCCAGGCTCACAAGGCTATC  qSPOP_R  TTGCTCTCCTCCATTTCATGTTC  qHNRPU_F  GCGAAATTTTATTCTGGATCAGACA  qHNRPU_R  GCTGGAAGCCTGCAAACAG  qYY1_F  ACCTGGCATTGACCTCTCAGA  qYY1_R  TTTTTCTTGGCTTCATTCTAGCAA  qASH2L_F  GGCTGACACATTTGGCATAGATAC  qASH2L_R  GATGGCAGACGTTGCAATGA       25 3 Results The knockdowns in part 3.7 were performed by Jakub Minks, at the time a PhD candidate in the lab. Subsequent RNA FISH, microscopy and image analysis was performed by ADK along with Irene Qi, at the time an undergraduate student doing her honours project in the lab. qRT-PCR to assess knockdown efficiency was performed by Jakub Minks. Integration sites experiments were replicated by Irene Qi.   26 3.1 The experimental system All the experiments described in this thesis were performed on HT1080 (human male fibrosarcoma) cells harbouring an XIST cDNA transgene. The following paragraphs give an overview of the experimental system.  A cDNA construct of the XIST gene had previously been integrated into the genome of HT1080 cells by Sarah Baldry in our laboratory using the Flp-In T-Rex system (Invitrogen). An advantage of using male cells is that there is no endogenous Xi and thus no need to distinguish between the endogenous and exogenous XIST RNA. To make the cells, a FRT site was randomly integrated into the genome followed by integration of the XIST cDNA construct at the FRT site (for more details on the generation of the HT1080 integration lines see Chow et al. 2007). Various HT1080 clones harbouring a FRT site had the XIST cDNA construct integrated, such that we have a number of HT1080 lines, each with XIST integrated into a different site in the genome. The integration sites include: 1p, 3q, 4q, 7p, 7q, 8p, 12q, 15q and Xq (note: the autosomal integrations were created in the Brown laboratory; the X integration was a kind gift of Dr. Chunhong Yan). The XIST construct is present as a single copy and is under the control of a doxycycline (DOX)-inducible CMV promoter. In the absence of DOX, Tet repressor proteins bind to Tet operator sequences in the CMV promoter and block transcription of XIST. When cells are cultured with DOX present in the media, the DOX prevents binding of Tet repressor proteins to the promoter, resulting in XIST transcription. The 3q integration site has previously been shown to produce XIST transcripts that localise to chromosome 3 (Chow et al., 2007). 3.2 XIST RNA can function when expressed from different chromosomes As stated, the first objective of my thesis was to assess XIST?s ability to localise to the chromosome from which it was transcribed, but when transcribed from autosomes instead of the X. This could shed light on the idea that there are specific DNA sequences (?way stations?) that are more enriched on the X chromosome than on autosomes that aid in XIST localisation. The identification of an integration site that is not compatible with XIST localisation would allow the analysis of various DNA sequence elements that might be depleted either on the chromosome as a whole or around the integration site. Such elements would be potential way stations. Sections 3.1.X of the results describe the experiments performed to address this objective. 3.2.1 XIST transgenes produce RNA that can localise to various autosomes In the present study we used the XIST integration cell lines described above to test whether XIST RNA can localise to the various chromosomes containing an XIST transgene. We used 27 RNA FISH to assess the localisation of XIST RNA to these chromosomes. Figure 3.1 shows that XIST RNA can localise to all the chromosomes tested. This indicates that XIST RNA can localise in a wide range of chromosomal contexts.   Figure 3.1 XIST RNA is able to localise to a variety of autosomes Shown is RNA FISH showing XIST (green) when an inducible XIST cDNA transgene is located in a variety of different autosomes (and the X chromosome) in HT1080 cells. In all cases XIST expression was induced for five days. Also shown is a control cell, AG (=AG14412), a female fibroblast. DNA is stained by DAPI (blue).  3.2.2 XIST expression increases perinucleolar localisation of autosomes As the Xi is known to preferentially reside in particular subnuclear locations, i.e. at the edge of the nucleus (perinuclear) and next to the nucleolus (perinucleolar), we next wished to determine if the expression of XIST from these chromosomes results in translocation of the chromosomes to these locations. This would provide evidence that XIST is not only localising to the chromosomes but has downstream effects on them. If XIST RNA could localise to these 28 autosomes but not have any downstream effects, this could indicate that XIST is not localising properly. The locations of the XIST-expressing chromosome were determined for each integration line. The gaps in the Cot-1 RNA FISH signal were used as a nucleolar marker. The XIST RNA FISH signal was classed as being either 1) ?perinuclear? (i.e. at the nuclear periphery); 2) ?perinucleolar? (i.e. adjacent to a nucleolus); 3) ?both? (i.e. both at the nuclear periphery and adjacent to a nucleolus; and 4) ?neither? (i.e. neither at the nuclear periphery nor adjacent to a nucleolus). These subnuclear positions are shown in Figure 3.2.  Figure 3.2 Subnuclear position categories of XIST RNA FISH signals Shown are examples of the four subnuclear positions used to categorise XIST RNA FISH signals. A=perinuclear, B=perinucleolar, C=both perinucleolar and perinuclear, D=neither perinucleolar nor perinuclear. XIST RNA FISH is shown in green. Cot-1 RNA FISH is shown in red (nucleoli were identified by Cot-1 ?ve regions). DAPI is shown in blue. Photos are of the 7q integration site following 5d in dox.  Figure 3.3A shows that when the XIST-bearing chromosomes are expressing XIST, they show substantial localisation in perinuclear and perinucleolar regions. This raises the possibility that XIST expression from these chromosomes is able to influence their subnuclear positioning to make the positioning similar to that of an inactive X. To confirm that the pattern of subnuclear 29 positioning is actually due to XIST, RNA FISH was used to assess the subnuclear positioning of one of the transgenic sites (the 3q integration site) prior to XIST integration (using a probe against the plasmid that was used to integrate the FRT site into the cells). If XIST expression is influencing the subnuclear positioning of the chromosomes, the chromosome would be expected to not show this positioning without XIST integrated. Figure 3.3B shows that, while the perinuclear localisation of chromosome 3 does not change upon XIST expression, the perinucleolar association increases ~three-fold upon XIST expression (p<0.01, chi-square test). This supports that XIST is able to influence the subnuclear positioning of an autosome, which in turn indicates that XIST RNA is able to function on these autosomes.   30   Figure 3.3 XIST expression increases nucleolar association of an XIST-expressing autosome A) Graph shows the distribution of subnuclear positions of XIST-expressing autosomes. AG14412 cells are normal female fibroblasts. N=60 for each integration site. B) Graph shows the distribution of an autosome (3q integration site) both prior to XIST integration and after XIST integration with five days of XIST expression. N=60. For A and right-hand part of B, chromosome was visualised by RNA FISH showing XIST; for left-hand part of B, chromosome was visualised by RNA FISH using a probe that binds the FRT site. Nucleoli were identified 31 using by lack of a Cot-1 RNA FISH signal. Perinucleolar association increases upon XIST expression (p<0.01, chi-square test).  3.2.3 XIST is able to form Cot-1 holes on a variety of autosomes We next determined whether XIST expression from these autosomes results in the formation of Cot-1 holes (see 1.3.3 for a description of a Cot-1 hole). Cot-1 holes are a characteristic feature of an inactive X (Hall et al. 2002). The presence of Cot-1 holes on the integration lines would provide further evidence that XIST RNA is able to confer features of an inactive X onto autosomes. To see if Cot-1 holes are formed in the transgene-bearing chromosomes, dual RNA FISH against both XIST and Cot-1 RNA was performed on each integration line. We found that XIST RNA is able to form Cot-1 holes when expressed from each integration site. Figure 3.4 shows example images showing Cot-1 holes in each integration site and a control female cell line. 32  Figure 3.4 XIST is able to form Cot-1 holes on a variety of autosomes Shown is dual RNA FISH showing XIST (green) and Cot-1 (red) for each XIST integration site, following five days of XIST expression. Also shown is a control female cell, AG (=AG14412). Note the decrease in the intensity of Cot-1 underneath the XIST signal.  3.2.4 XIST RNA forms Cot-1 holes to a similar degree on different autosomes We next wanted to assess whether XIST?s ability to form Cot-1 holes might differ between the different integration sites. This would show whether XIST RNA is sensitive to the sequence content of the chromosome to which it localises. If XIST RNA were better able to direct the formation of Cot-1 holes on one chromosome than another, this would enable us to look for differences in sequence composition between the two chromosomes. Such sequences may be 33 involved in XIST RNA function. To this end, we measured the intensity of the Cot-1 RNA FISH signal underneath the XIST RNA FISH signal to obtain an average intensity for each integration site. Figure 3.5 shows that the Cot-1 intensity varies significantly between the integration sites (p<0.05, 1 way ANOVA). In particular, the 3q and 4q integration sites showed lower Cot-1 intensity, which could indicate that XIST is better able to have an effect in these integration sites than the others. However, from looking at the FISH images it is evident that Cot-1 intensity is generally weaker at the nuclear periphery, and the 3q and 4q integration sites also showed high perinuclear positioning (see Figure 3.3). It is thus possible that the low Cot-1 intensities are a reflection of highly perinuclear positioning. To test this, we also measured the mean Cot-1 intensities under XIST for each integration site, but this time according to the different nuclear positions of the XIST signal in the different integration sites (see Figure 3.3). Figure 3.44B shows that, when taking the subnuclear position into account, there are not great differences between the integration sites in terms of how well XIST can form a Cot-1 hole. This shows that XIST is equally able to silence repetitive elements regardless of which chromosome it localise to.  34  Figure 3.5 XIST is equally able to form Cot-1 holes on a variety of different autosomes A) Graph shows the mean Cot-1 intensity under the XIST signal of ~60 cells for each integration site. Bars show 95% confidence intervals. The 3q and 4q integrations show significantly lower Cot-1 intensity than other integrations (p<0.05, 1-way ANOVA). B) Graph shows the same data as A) but for each integration site the values are shown according to subnuclear position of the XIST signal. AG14412 are normal female fibroblasts.  35 3.2.5 An XIST-expressing autosome recruits H3K27me3 and macroH2A To further explore XIST?s capability to function in the integration sites, we next looked for enrichment of the repressive histone modification H3K27me3, and also for enrichment of the histone variant macroH2A/mH2A. Both H3K27me3 and mH2A are enriched on the Xi. Seeing enrichment of either of these epigenetic features would lend further support to the notion that XIST RNA can function on autosomes. Additionally, if either of these epigenetic features were not enriched on the autosomes, it would indicate that the other features of an Xi that we have thus seen on the autosomes (i.e. nucleolar association and Cot-1 holes) do not depend on H3K27me3 or mH2A. This is important as deducing which features of the Xi depend on each other would aid in the development of an overall model of how the Xi is formed.  Figure 3.6 shows that when the XIST transgene is located on chromosome 3 or the X chromosome, it is able to recruit both H3K27me3 and mH2A. While this does not allow us to address whether the other features we have seen depend on these epigenetic modifications, it does support the notion that XIST is able to function normally on the XIST-bearing autosomes. This in turn supports a lack of sequence specificity of XIST. 36   Figure 3.6 H3K27me3 and mH2A are recruited to XIST transgene-bearing chromosomes A) H3K27me3 immunofluorescence (red) is shown with XIST RNA FISH (green) in cells bearing an XIST transgene either on the X chromosome (F55; top panel) or on chromosome 3 (3q; bottom panel). B) mH2A immunofluorescence (red) is shown with XIST RNA FISH (green) in cells bearing an XIST transgene either on the X chromosome (F55; top panel) or on chromosome 3 (3q; bottom panel).  37  3.3 XIST sequences act redundantly to localise XIST RNA The second objective of my thesis was to deduce which sequences of XIST are required for its localisation as little is known of this. To try to build upon the limited existing knowledge of sequences of XIST required for XIST RNA localisation, three XIST deletion constructs were analysed by RNA FISH for their ability to localise. One construct, del 3?XIST, consists of most of the first exon of XIST and includes all the repeats except for repeat E; delta PflMI lacks repeats B and C; and delta delta is a combination of the above two deletions, consisting of exon 1 of XIST minus the B and C repeats. Repeat C has previously been suggested as important for Xist localisation (see section 1.4). The deletions are shown diagrammatically in Figure 3.7 The XIST deletion constructs. Figure 3.8 XIST sequences act redundantly to localise XIST RNA shows that either of the single deletions alone does not disrupt XIST localisation, whereas the construct with both deletions is delocalised. This shows that sequences in the two deleted regions are involved in XIST localisation, and that they localise XIST RNA in a redundant fashion.    Figure 3.7 The XIST deletion constructs Shown is the structure of XIST cDNA (?Full length?), showing the locations of the tandem repeats (A-F). Below this are the deletion constructs, with black lines showing the regions deleted in each of the constructs.  38  Figure 3.8 XIST sequences act redundantly to localise XIST RNA XIST RNA FISH (green) was used on HT1080s with different XIST constructs integrated into chromosome 3. The full length XIST construct (A), along with the del 3?XIST construct (B) and the delta PflMI construct (C) produce RNA able to localise. A construct bearing both the deletions, however, produces RNA unable to localise (D). 46 of 68 XIST=expressing cells (68%) with the double deletion showed this delocalisation of XIST RNA. DAPI staining is shown in blue.  3.4 Proteins that affect XIST localisation The final objective was to assess whether various proteins known to be involved in XCI are required for XIST RNA localisation. siRNA was used to knock down several proteins implicated in XCI, followed by RNA FISH to assess XIST RNA localisation. HT1080s with full length XIST integrated into chromosome 3 were used for these experiments. Cells were treated with siRNA for 24 hours, at which point XIST expression was dox-induced for a further 48 hours (with siRNA in the media throughout). Cells were then fixed and RNA FISH performed to visualize XIST RNA.  Controls treated with transfection reagent but no siRNA showed a normal XIST signal, as did some of the knockdowns (data not shown). However, knockdown of YY1, hnRNP-U, SPOP, CUL3 and ASH2L prevent the formation of a normal XIST signal (Figure 3.8), perhaps reflecting a role of these proteins in XIST localisation.  39  Figure 3.9 Knockdowns that disrupt the XIST RNA signal Panels show XIST RNA FISH (green) in HT1080 cells with an XIST transgene integrated into chromosome 3. Cells have had the indicated protein knocked down by siRNA. Nuclei are stained by DAPI (blue). Bars on the right of each image show the proportion of 30 cells that showed a disrupted XIST signal. Red lettering shows the respective mRNA levels assessed by qRT-PCR.  As ASH2L has previously been reported to be enriched on the Xi (Pullirsch et al. 2010), we performed immunofluorescence combined with XIST RNA FISH to test for ASH2L enrichment in the HT1080s; however, no enrichment coincident with the XIST signal was observed (enrichment was also not seen with control female fibroblast cells, GM04626). hnRNP-U has also been seen to be enriched on the Xi (Helbig & Fackelmayer 2003; Pullirsch et al. 2010). However, we did not see enrichment in HT1080 cells or GM04626 control cells by immunofluorescence with XIST RNA FISH.       40 4 Discussion   41 Localising the XIST RNA to the X chromosome is a critical step in the process of XCI. However, little is known of how this occurs on a molecular level. In the present study we used HT1080 cells containing a doxycycline-inducible XIST cDNA transgene to investigate various facets of XIST RNA localisation. We analysed the ability of XIST to localise to a variety of autosomes, the sequences of XIST required for its cis-localisation, and proteins potentially involved in localising XIST RNA.   We used HT1080s with an XIST transgene located on eight different chromosomes (chromosomes 1, 3, 4, 7, 8, 12 and 15 and the X) to see how well XIST can function in such a variety of chromosomal contexts. The lack of localisation of XIST RNA often seen in the autosomal segment of X;autosome translocations suggests that XIST RNA may have some sequence specificity for the X chromosome. This was addressed in this study by assessing whether XIST RNA can localise to the various autosomes into which it was integrated. We found that XIST was able to localise in all the integration sites tested. This is in agreement with previous studies that have demonstrated XIST RNA localisation to autosomes when expressed from a transgene on an autosome. This autosomal localisation may reflect a lack of sequence specificity of XIST RNA for the X chromosome. However, while XIST was able to localise and form an XIST RNA focus by RNA FISH, the extent of spreading on the chromosome was not assessed in the present study and should be examined. One way to examine more precisely the extent of XIST RNA spreading would be to perform XIST RNA FISH on metaphase chromosomes. In humans, XIST RNA normally falls off the chromosomes at metaphase; however, it can be retained on the chromosome through the use of Aurora kinase B inhibitors (Hall et al. 2009). Another approach is RNA antisense purification (RAP). RAP allows the identification of DNA sequences associated with a ncRNA (Engreitz et al. 2013). Such studies would be useful in determining any sequence specificities of XIST RNA. In mice, RAP showed that Xist RNA initially localises to distal regions across the X chromosome that are not defined by specific sequences; rather, these distal regions are defined by spatial proximity to the Xist locus (Engreitz et al. 2013). In addition to localising to these chromosomes, we also found that XIST could form Cot-1 holes on all of the autosomes, indicating silencing of the repetitive elements. By measuring the intensity of the Cot-1 RNA FISH signal under the XIST signal, we found that two of the integration sites, 3q and 4q, showed greater silencing of Cot-1 DNA. This suggests that XIST RNA is better able to function in these integration sites, possibly reflecting the sequence content on the chromosomes or around the integration site. However, these two integrations were highly perinuclear, and the Cot-1 RNA FISH signal was generally weaker at the nuclear periphery. 42 After correcting for differences in subnuclear positioning, it appears that XIST forms Cot-1 holes on the different autosomes with roughly equal efficiency. This would suggest that when in the same subnuclear context, there aren?t great differences between the autosomes that would allow some to form better Cot-1 holes than others.  Autosomes harbouring XIST transgenes had previously shown that XIST can recruit various proteins that are normally recruited by XIST in X inactivation, including PRC2 (shown by enrichment of H3K27me3, the mark laid down by PRC2), PRC1 (shown by H2AK119u1 enrichment), mH2A, Hnrnp-u and Ash2l.  Recruitment in the HT1080 integration lines of proteins normally recruited to the inactive X would further support that XIST RNA can not only localise to various autosomes but have downstream effects on the autosomes. Therefore, to test for protein recruitment in the HT1080s, two of the integrations, the chromosome three and X integrations, were tested for their ability to recruit PRC2 by testing for H3K27me3 enrichment by immunofluorescence coincident with the XIST RNA FISH signal. We saw that both integrations could recruit this mark. With the same two integration sites, we also tested the ability of XIST to direct the incorporation of the histone variant mH2A into the nucleosomes, and saw that XIST is able to incorporate mH2A into both the X chromosome and chromosome three. This shows that XIST RNA is capable of directing epigenetic features onto autosomes. As with the XIST localisation, this study did not address the extent of the recruitment of mH2A and H3K27me3 to the chromosomes. Chromatin immunoprecipitation followed by DNA sequencing (ChIP-seq) would reveal the extent to which mH2A and H3K27me3 are recruited to the XIST-expressing autosome.  In this study, we showed that XIST-expressing autosomes are often found in the same subnuclear positions where the Xi is normally found (i.e. perinuclear and perinucleolar positions). Some integration sites were most often in the perinuclear category, whereas others were most often in the perinucleolar category. The frequent perinuclear/perinucleolar location of the XIST-expressing autosomes might reflect XIST RNA localisation to the autosomes. To investigate this, we assessed the subnuclear position of the one of the integrations (the chromosome 3 integration) prior to XIST integration and compared this with its position when it is expressing XIST. We saw that the subnuclear positioning changes upon XIST expression, in that the chromosome has an increased association with the nucleolus upon XIST expression. This is in agreement with the fact that deleting Xist in skin fibroblasts results in loss of perinucleolar localisation (Zhang et al. 2007). Perinucleolar localisation has been suggested to be of significance in XCI, as the Xi localises to this region specifically during S phase (Zhang et al. 2007). This suggests that perinucleolar targeting may be required for replication of the Xi. This idea is supported by the enrichment of Snf2h in perinucleolar region (Zhang et al. 2007). 43 Snf2h is the catalytic subunit of a complex required for the progression of the replication fork through condensed chromatin (Collins et al. 2002). In contrast, deletion of Xist does not result in the loss of nucleolar localisation of the Xi in two-cell stage mouse embryos (Namekawa et al. 2010), which in itself would suggest that a component of the Xic other than Xist is responsible for perinucleolar targeting. In this study we saw perinucleolar localisation of chromosomes with only an XIST transgene integrated, with no other region of the XIC present, which would suggest that it is XIST itself that it responsible for perinucleolar targeting. However, in the eight-cell stage mouse embryos both wild-type and Xist-deleted cells lose their nucleolar association. A possibility is that in the two-cell stage a component other than Xist is responsible for the perinucleolar localisation of the Xi and in differentiated cells XIST is required. It is also possible that both XIST and other sequences within the XIC contribute to nucleolar localisation.  In this study, three XIST deletion constructs were analysed for their ability to localise. Deletion of repeats B and C together did not delocalise XIST RNA, nor did deletion of the 3? end, leaving most of exon1(including all the repeats except for repeat E). However, when these two deletions were combined into a single construct (i.e. a construct consisting of exon 1 with repeats B and C removed) the ability of the RNA to localise is severely compromised, and small specks of XIST RNA can be seen covering a large area of the nucleus. However, XIST RNA does not become evenly dispersed throughout the nucleus. It is possible that as XIST RNA drifts away, it is still bound by proteins and this limits its diffusion. This study does not conclusively prove the XIST RNA is delocalised; it is possible that the RNA attaches to the chromosome but the chromosome does not condense, hence the more dispersed pattern of XIST RNA. This could be addressed using RAP (RNA antisense purification), which enables the identification of DNA sequences bound by an RNA. If XIST RNA is still localised, the DNA sequences should map to chromosome three (where the constructs were integrated).  These deletion constructs could also be used to deduce whether various proteins and epigenetic features of the Xi depend on the sequences deleted. This could delineate the roles of the deleted sequences in XCI. As for repeat B/C region, it has been shown that the transcription factor YY1 binds to this region of Xist RNA and tethers it to the inactive X (Jeon & Lee 2011), discussed in more detail below, which is presumably how deletion of this region is contributing to the delocalised XIST signal in the double deletion construct. However, given that the single deletion does not delocalise XIST, this would suggest that YY1 is also binding elsewhere. 44 How the 3? end of XIST contributes to XIST RNA localisation is less obvious. In the paper showing that YY1 tethers XIST RNA to the X chromosome, an Xist construct consisting of exon 1 was able to localise. However, in an assay designed to assess the ability of transgenic Xist to ?squelch? endogenous Xist away from the X and to the transgene-harbouring chromosome) the exon 1 construct was not able to squelch endogenous Xist quite as well as the full length Xist. While this result does not indicate a precise role for the 3? end, it supports a functional role for the 3? end of Xist. Our result, that two regions singly deleted give no phenotype whereas both deletions together do give a phenotype, is consistent with a series of Xist deletions in mouse cells that demonstrated that sequences act in a redundant fashion to localise Xist RNA (Wutz et al. 2002). This redundancy demonstrates the complexity of studying XIST; to assess the function of a particular part of XIST, it is not necessarily enough to delete it alone, as other sequences may be compensating. It would be interesting to evaluate whether this redundancy is a feature of other lncRNAs. The importance of the deleted sequences in localising XIST RNA would be supported by the ?delta delta? construct failing to localise in the other integration sites, or by an XIST construct consisting only of the sequences deleted in the ?delta delta? construct being able to localise. It would be interesting to assess whether the deleted sequences are required for gene silencing to occur; absence of silencing could reflect binding of proteins involved in silencing to the deleted sequences, either directly or indirectly.   Besides YY1, a number of other proteins also participate in XCI. In this study we knocked down several of these proteins, and show that some of these knockdowns result in the disruption of the XIST RNA focus.  We find that knockdown of hnRNP-U/SAF-A disrupts the XIST RNA focus. hnRNP-U is enriched on the Xi (Helbig & Fackelmayer 2003) and, as it has both DNA and RNA binding domains and forms multimers in the presence of nucleic acids (Fackelmayer et al. 1994), is a good candidate for a protein that spreads XIST RNA across the X. Removal of Xist expression results in a loss of hnRNP-U enrichment (Pullirsch et al. 2010) and conversely, removal of hnRNP-U results in delocalisation of XIST/Xist (Hasegawa et al. 2010). Together this suggests a mutual dependency of hnRNP-U and XIST RNA, in that localisation of each one depends on the presence of the other. In support of this, removal of Xist expression results in loss of Xist foci and Hnrnp-u foci at very similar rates (Pullirsch et al. 2010), pointing to a tight association between Xist and Hnrnp-u. The loss of Xist localisation upon Hnrnp-u knockdown was done in a mouse neuroblastoma cell line (Neuro2a cells). As mentioned previously, there are important 45 differences in the process of XCI between mice and humans, demonstrating the importance of studies in human cells. Here we show that hnRNP-U is also required for the formation of an intact XIST focus in human HT1080 cells. YY1, briefly discussed above, is a transcription factor that acts as both an activator and repressor of gene transcription. Yy1 is required for Xist RNA to be tethered to the XIC at the onset of XCI (Jeon & Lee 2011), and knockdown of Yy1 results in delocalisation of Xist RNA. Again, this was done in mouse cells. Here we show that knockdown of YY1 also results in delocalisation of XIST in a human cell line. Yy1 binds to repeat C of mouse Xist mouse, and maybe also repeat B. In the present study we show that an XIST RNA lacking both repeats B and C is able to localise, suggesting that YY1 is also able to tether XIST to the chromosome by binding elsewhere in the RNA molecule. Ash2l is enriched on the Xi (Pullirsch et al. 2010), which is unexpected as it is a TrxG protein involved in laying down a mark associated with active chromatin, H3K4me3 (Steward et al. 2006).  Other TrxG proteins are not enriched on the Xi (Pullirsch et al. 2010), demonstrating that Ash2l is not present on the Xi as part of a TrxG complex. Given this, and also given that H3K4me3 is depleted on the Xi, it has been suggested that Ash2l might serve a structural role in the chromatin to allow access of proteins that repress gene expression (Pullirsch et al. 2010) This hypothesis is based on a study implicating the Drosophila homologue of ASH2L to be functioning in such a way (Angulo et al. 2004). Another possibility is that ASH2L is involved in the localisation of XIST RNA to the chromosome. Ash2l is recruited with very similar kinetics to Hnrnp-u (Pullirsch et al. 2010), which could reflect an involvement of Ash2l in Xist localisation. The disrupted XIST RNA signal upon knockdown of ASH2L shown in this study could reflect a role in XIST localisation. However, another possibility is that ASH2L is involved in condensation of the Xi, similar to SMCHD1 (Nozawa et al. 2013). XIST RNA may be localised to the chromosome, but the XIST RNA FISH signal could appear disrupted due to lack of condensation of the chromosome. The RAP technique could be used to determine if XIST RNA is still associated with the chromosome with knockdown of ASH2L and could thus be used to distinguish between these two possibilities. CUL3 and SPOP, which together form a ubiquitin ligase complex, ubiquitinate two proteins involved in XCI - the BMI1 subunit of PRC1 and also mH2A (Hern?ndez-Mu?oz et al. 2005). Knockdown of CUL3 or SPOP prevents mH2A recruitment to the Xi, which may suggest that mH2A ubiquitination is required for its incorporation into the Xi. 46 ATRX is a chromatin remodeling protein enriched on the Xi (Baumann & De La Fuente 2009). Interestingly, one study has suggested that ATRX is a negative regulator of mH2A incorporation (Ratnakumar et al. 2012), which is surprising given that both proteins are enriched on the Xi. The suggestion was based on the fact that knockdown of ATRX results in an increase in mH2A deposition at telomeres. ATRX, together with Daxx, deposits another histone variant, H3.3. H3.3 and mH2A occur in mutually exclusive nucleosomes (Ratnakumar et al. 2012). Therefore, another explanation for their results is that knockdown of ATRX prevents H3.3 incorporation into nucleosomes, and in the absence of H3.3 mH2A can be incorporated. Seeing as ATRX binds mH2A (Ratnakumar et al. 2012), it remains possible that ATRX is a positive regulator of mH2A incorporation ? perhaps, in the presence of Daxx , ATRX incorporates H3.3, but in the absence of Daxx, ATRX incorporates mH2A. In support of this possibility, CUL3/SPOP, in addition to ubiquitinating mH2A and PRC1, also ubiquitinates Daxx (Kwon et al. 2006), but unlike mH2A and PRC1, ubiquitination of Daxx by CUL3/SPOP results in its degradation. Therefore, one possible scenario for mH2A recruitment to the Xi is that CUL3/SPOP ubiquitinates Daxx, leading to its degradation, and in the absence of Daxx, ATRX is able to incorporate mH2A. The lack of mH2A recruitment to the Xi upon knockdown of CUL3 or SPOP has been suggested to indicate that ubiquitination of mH2A might be involved in its recruitment to the Xi (Hern?ndez-Mu?oz et al. 2005). This may be; however, the possible scenario presented here suggests that another reason that SPOP and CUL3 knockdown results in loss of mH2A enrichment is that without the CUL3/SPOP complex, Daxx is not ubiquitinated and degraded and therefore ATRX, in the presence of Daxx, is unable to recruit mH2A. Hern?ndez-Mu?oz et al. (2005) reported that knockdown of SPOP or CUL3 did not affect XIST coating of the Xi. However, this difference may be due to differences in how the effect on XIST was assessed; in the present study the XIST signal was observed in Photoshop and it was seen that the XIST signal was disrupted; however, the XIST signal does indeed still form a reasonable focus. Also, Hern?ndez-Mu?oz et al. (2005) performed their knockdowns in HEK293 cells. It may be that HT1080 cells are more sensitive to knockdown of CUL3 or SPOP; perhaps not all of the epigenetic features or proteins that would retain XIST on the chromosome are present in the HT1080s.  As the CUL3/SPOP E3 complex ubiquitinates PRC1, which is involved in chromatin compaction (Endoh et al. 2012), one explanation for the disruption of the XIST signal observed here is that knocking down SPOP or CUL3 results in PRC1 that is no longer ubiquitinated and therefore unable to compact chromatin. This could result in the XIST signal appearing to be more dispersed due to the less compacted nature of the chromosome. The ubiquitination activity of PRC1 is not required for chromatin compaction (Endoh et al. 2012; Eskeland 2010); therefore 47 this putative stimulation of PRC1?s compaction activity may well be independent of its ubiquitination activity. mH2A is composed of an N terminal region similar to that of canonical H2A and a large C terminal region. Like PRC1, mH2A is also involved in chromatin compaction. The linker region of mH2A (which connects mH2A?s N terminal histone region and its large C terminal macro domain) results in nucleosome oligomerisation in an in vitro system (Muthurajan et al. 2011). Interestingly, however, the macro domain actually inhibits this oligomerisation. Muthurajan et al. (2011) suggested various mechanisms for the activation of the linker region?s compaction activity, including the possibility of a protein that binds to the macro domain or of a post-translational modification of the macro domain or the linker region, or proteolytic cleavage of the macro domain. Given that the CUL3/SPOP complex ubiquitinates mH2A, it may be that ubiquitination of mH2A is the trigger that prevents the macro domain from inhibiting the compaction activity of the linker region. Interestingly, the amino acid composition of the linker region of mH2A is very similar to that of the C-terminal domain of the linker histone H1 (Muthurajan et al. 2011). Furthermore, mH2A-containing chromatin fractions do not contain histone H1 (Abbott et al. 2004), suggesting that H1 does not associate with mH2A-containing nucleosomes. Muthurajan et al. (2011) suggested that the linker region of mH2A might act as a surrogate histone H1 in nucleosomes lacking histone H1. With this is mind, and taking into account that the CUL3/SPOP complex ubiquitinates both mH2A and PRC1 (both of which have chromatin compaction activity), the CUL3/SPOP complex may contribute to the compaction of both types of nucleosome (i.e. H1-containing and mH2A-containing) in two separate pathways, as follows, and shown diagrammatically in  Figure 4.1 Model for compaction of nucleosomes containing either histone H1 or macroH2A: Compaction of H1-containing nucleosomes: CUL3/SPOP ubiquitinates the BMI subunit of PRC1 (Hern?ndez-Mu?oz et al. 2005). This stimulates the chromatin compaction activity of PRC1 (based on the fact that knocking it down in the present study disrupted the XIST signal). PRC1 ubiquitinates K119 on H2A (which may or may not be a result if CUL3/SPOP action). uH2A increases the association of the C-terminal region of H2A with H1  and decreases H1 phosphorylation, also contributing to compaction (Zhou et al. 2008). Compaction of mH2A-containing nucleosomes: CUL3/SPOP ubiquitinates mH2A (Hern?ndez-Mu?oz et al. 2005). Ubiquitination prevents the C-terminal macro domain from inhibiting the chromatin compaction activity of the linker region of mH2A, resulting in compaction. 48  Figure 4.1 Model for compaction of nucleosomes containing either histone H1 or macroH2A Two putative pathways for the compaction of mH2A-associated or H1-associated nucleosomes. Green boxes show steps that have been shown; red boxes contain speculative parts of the pathways. Small red circles denote ubiquitin.49  In summary, the CUL3/SPOP ubiquitin ligase complex has previously been shown to ubiquitinate PRC1, mH2A and Daxx (Hern?ndez-Mu?oz et al. 2005; Kwon et al. 2006). Both mH2A and PRC1 have chromatin compaction activity. Taking into account that knockdown of either CUL3 or SPOP in the present study disrupted the XIST signal, the roles of each of the three ubiquitination events by the CUL3/SPOP complex are speculated as follows: 1) PRC1: ubiquitination of PRC1 at H1-containing nucleosomes may stimulate PRC1?s chromatin compaction activity. 2) Daxx: ubiquitination of Daxx may result in a depletion of Daxx on the Xi, thereby allowing ATRX to recruit mH2A. 3) mH2A: ubiquitination of mH2A may release the inhibition of the macro domain on the linker region?s chromatin compaction activity. To verify that knockdown of SPOP or CUL3 prevents condensation of the chromosome, as opposed to preventing XIST RNA localisation, it should be confirmed that XIST is still attached to the chromosome. This could be done using the RAP technique (Engreitz 2013).   The initial coating of an X chromosome is an essential step for the silencing of the chromosome. How XIST RNA localises the X chromosome is, however, largely unknown. In this study we have used inducible XIST transgenes in human HT1080 cells to investigate various aspects of XIST RNA localisation. We have shown that XIST is able to localise to a various autosomes, resulting in various features of an inactive X on the autosomes. Using one of the autosomal integrations, we have shown that repeats B and/or C, together with a sequence/sequences downstream of exon 1 of XIST, act in a redundant fashion to localise XIST RNA. Additionally, by knocking down various proteins, we have confirmed that YY1 and hnRNP-U are required for XIST RNA localisation in humans and also show that the CUL3/SPOP ubiquitin ligase complex and the TrxG protein ASH2L are also required for the formation of an intact XIST focus. As the number of identified lncRNAs expands, the mechanisms deduced of how XIST RNA functions may prove useful in elucidating the mechanisms of action of other lncRNAs    50 5 References Abbott, D.W. et al., Structural characterization of macroH2A containing chromatin. Biochemistry,  2004. 43(5), pp.1352?1359. Agrelo, R. et al., SATB1 Defines the Developmental Context for Gene Silencing by Xist in Lymphoma and Embryonic Cells. Developmental Cell, 2009. 16(4), pp.507?516. Angulo, M., Corominas, M. & Serras, F., Activation and repression activities of ash2 in Drosophila wing imaginal discs. Development, 2004. 131(20), pp.4943?4953. Barr, M.L. & Bertram, A Morphological Distinction Between Neurones Of The Male And Female, And The Behaviour Of The Nucleolar Satellite During Accelerated Nucleoprotein Synthesis. Nature, 1949. 163(4148), pp.676?677. Baumann, C. & De La Fuente, R., ATRX marks the inactive X chromosome (Xi) in somatic cells and during imprinted X chromosome inactivation in trophoblast stem cells. Chromosoma, 2009. 118(2), pp.209?222. Beard, C., LI, E. & Jaenisch, R., Loss Of Methylation Activates Xist In Somatic But Not In Embryonic-Cells. Genes & Development, 1995. 9(19), pp.2325?2334. Beletskii, A. et al., PNA interference mapping demonstrates functional domains in the noncoding RNA Xist. Proceedings of the National Academy of Sciences of the United States of America, 2001. 98(16), pp.9215?20. 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. Human Genetics, 1996. 97(5), pp.573?578. Bernstein, E. et al., Mouse Polycomb Proteins Bind Differentially to Methylated Histone H3 and RNA and Are Enriched in Facultative Heterochromatin. Molecular and Cellular Biology, 2006. 26(7), pp.2560?2569. Blewitt, M E et al., An N-ethyl-N-nitrosourea screen for genes involved in variegation in the mouse. Proceedings Of The National Academy Of Sciences Of The United States Of America, 2005. 102(21), pp.7629?7634. Blewitt, Marnie E et al., SmcHD1, containing a structural-maintenance-of-chromosomes hinge domain, has a critical role in X inactivation. Nature Genetics, 2008. 40(5), pp.663?9.  Boggs, B.A. et al., Differentially methylated forms of histone H3 show unique association patterns with inactive human X chromosomes. Nature Genetics, 2002. 30(1), pp.73?76. Borden, J. & Manuelidis, L., Movement Of The X-Chromosome In Epilepsy. Science, 1988. 242(4886), pp.1687?1691. 51 Bourgeois, C.A. et al., New Data On The Insitu Position Of The Inactive X-Chromosome In The Interphase Nucleus Of Human-Fibroblasts. Human Genetics, 1985. 69(2), pp.122?129. Brockdorff, N. et al., The Product Of The Mouse Xist Gene Is A 15 Kb Inactive X-Specific Transcript Containing No Conserved Orf And Located In The Nucleus. Cell, 1992. 71(3), pp.515?526. 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, 1992. 71(3), pp.527?42. Brown, C.J. et al., Localization Of The X-Inactivation Center On The Human X-Chromosome In Xq13. Nature, 1991. 349(6304), pp.82?84. Brown, C.J. & Willard, H.F., The Human X-Inactivation Center Is Not Required For Maintenance Of X-Chromosome Inactivation. Nature, 1994. 368(6467), pp.154?156. Caparros, M.-L. et al., Functional Analysis Of The Highly Conserved Exon IV Of XIST RNA. Cytogenetic And Genome Research, 2002. 99(1-4), pp.99?105.  Chadwick, B.P. & Willard, H F, Chromatin of the Barr body: histone and non-histone proteins associated with or excluded from the inactive X chromosome. Human Molecular Genetics, 2003. 12(17), pp.2167?2178. Chadwick, B.P. & Willard, H F, Multiple spatially distinct types of facultative heterochromatin on the human inactive X chromosome. Proceedings Of The National Academy Of Sciences Of The United States Of America, 2004. 101(50), pp.17450?17455. Chakravarthy, S., Patel, A. & Bowman, G.D., The basic linker of macroH2A stabilizes DNA at the entry/exit site of the nucleosome. Nucleic Acids Research, 2012. 40(17), pp.8285?8295. Chaumeil, J. et al., A novel role for Xist RNA in the formation of a repressive nuclear compartment into which genes are recruited when silenced. Genes & Development, 2006. 20(16), pp.2223?37. Available at:  Chow, J.C. et al., Inducible XIST-dependent X-chromosome inactivation in human somatic cells is reversible. Proceedings Of The National Academy Of Sciences Of The United States Of America, 2007. 104(24), pp.10104?10109. Collins, N. et al., An ACF1-ISWI chromatin-remodeling complex is required for DNA replication through heterochromatin. Nature Genetics, 2002. 32(4), pp.627?632. Costanzi, C. & Pehrson, J.R., Histone macroH2A1 is concentrated in the inactive X chromosome of female mammals. Nature, 1998. 393(6685), pp.599?601. 52 Cowieson, N.P. et al., Dimerisation of a chrome shadow domain and distinctions from the chromodomain as revealed by structural analysis. Current Biology, 2000. 10(9), pp.517?525. Csankovszki, G. et al., Conditional deletion of Xist disrupts histone macroH2A localization but not maintenance of X inactivation. Nature Genetics, 1999. 22(4), pp.323?324. Csankovszki, G., Nagy, A. & Jaenisch, R., Synergism of Xist RNA, DNA methylation, and histone hypoacetylation in maintaining X chromosome inactivation. JOURNAL OF CELL BIOLOGY, 2001. 153(4), pp.773?783. Czermin, B. et al., Drosophila enhancer of Zeste/ESC complexes have a histone H3 methyltransferase activity that marks chromosomal polycomb sites. Cell, 2002. 111(2), pp.185?196. Dong, K.B. et al., DNA methylation in ES cells requires the lysine methyltransferase G9a but not its catalytic activity. EMBO Journal, 2008. 27(20), pp.2691?2701. Duthie, S.M. et al., Xist RNA exhibits a banded localization on the inactive X chromosome and is excluded from autosomal material in cis. Human Molecular Genetics, 1999. 8(2), pp.195?204. Eils, R. et al., Three-dimensional reconstruction of painted human interphase chromosomes: Active and inactive X chromosome territories have similar volumes but differ in shape and surface structure. Journal Of Cell Biology, 1996. 135(6, 1), pp.1427?1440. Endoh, Mitsuhiro et al., Histone H2A mono-ubiquitination is a crucial step to mediate PRC1-dependent repression of developmental genes to maintain ES cell identity. PLoS Genetics, 2012. 8(7), p.e1002774. Eskeland, R. et al., Ring1B Compacts Chromatin Structure and Represses Gene Expression Independent of Histone Ubiquitination. Molecular Cell, 2010. 38(3), pp.452?464. Fackelmayer, F. et al., Nucleic-acid-binding properties of hnRNP-U/SAF-A. European Journal of Biochemistry, 1994. 221(2), pp.749?757. Fang, J. et al., Ring1b-mediated H2A ubiquitination associates with inactive X chromosomes and is involved in initiation of X inactivation. Journal Of Biological Chemistry, 2004. 279(51), pp.52812?52815. Gartler, S.M. & Riggs, A.D., Mammalian X-Chromosome Inactivation. Annual Review Of Genetics, 1983. 17, pp.155?190. 53 Goto, Y. et al., Differential patterns of histone methylation and acetylation distinguish active and repressed alleles at X-linked genes. Cytogenetic And Genome Research, 2002. 99(1-4), pp.66?74. Hall, L.L. et al., An ectopic human XIST gene can induce chromosome inactivation in postdifferentiation human HT-1080 cells. Proceedings of the National Academy of Sciences of the United States of America, 2002. 99(13), pp.8677?82.  Hall, L.L. et al., AURKB-mediated effects on chromatin regulate binding versus release of XIST RNA to the inactive X chromosome. The Journal of Cell Biology, 2009, 186(4), pp.491-507. Hansen, R.S. et al., Escape from gene silencing in ICF syndrome: evidence for advanced replication time as a major determinant. Human Molecular Genetics, 2000. 9(18), pp.2575?2587. Hansen, R.S. et al., The DNMT3B DNA methyltransferase gene is mutated in the ICF immunodeficiency syndrome. Proceedings Of The National Academy Of Sciences Of The United States Of America, 1999. 96(25), pp.14412?14417. Hansen, R.S. & Gartler, S.M., 5-Azacytidine-Induced Reactivation Of The Human X-Chromosome-Linked Pgk1 Gene Is Associated With A Large Region Of Cytosine Demethylation In The 5? CpG Island. Proceedings Of The National Academy Of Sciences Of The United States Of America, 1990. 87(11), pp.4174?4178. Hasegawa, Y. et al., The matrix protein hnRNP U is required for chromosomal localization of Xist RNA. Developmental Cell, 2010. 19(3), pp.469?76. He, Y. & Casaccia-Bonnefil, P., The Yin and Yang of YY1 in the nervous system. Journal Of Neurochemistry, 2008. 106(4), pp.1493?502.  Heard, E. et al., Methylation of Histone H3 at Lys-9 Is an Early Mark on the X Chromosome during X Inactivation. Cell, 2001. 107(6), pp.727?738. Helbig, R. & Fackelmayer, F.O., Scaffold attachment factor A (SAF-A) is concentrated in inactive X chromosome territories through its RGG domain. Chromosoma, 2003. 112(4), pp.173?82.  Hermann, A., Goyal, R. & Jeltsch, A., The Dnmt1 DNA-(cytosine-C5)-methyltransferase methylates DNA processively with high preference for hemimethylated target sites. Journal Of Biological Chemistry, 2004. 279(46), pp.48350?48359. Hern?ndez-Mu?oz, I. et al., Stable X chromosome inactivation involves the PRC1 Polycomb complex and requires histone MACROH2A1 and the CULLIN3/SPOP ubiquitin E3 ligase. Proceedings of the National Academy of Sciences of the United States of America, 2005. 102(21), pp.7635?40. 54 Hoki, Y. et al., Incomplete X-inactivation initiated by a hypomorphic Xist allele in the mouse. Development, 2011. 138(13), pp.2649?59.  Jacobs, S.A. et al., Specificity of the HP1 chromo domain for the methylated N-terminus of histone H3. EMBO Journal, 2001. 20(18), pp.5232?5241. Jeon, Y. & Lee, Jeannie T, YY1 Tethers Xist RNA to the Inactive X Nucleation Center. Cell, 2011. 146(1), pp.119?33.  Jeppesen, P. & Turner, B.M., The Inactive X-Chromosome In Female Mammals Is Distinguished By A Lack Of Histone-H4 Acetylation, A Cytogenetic Marker For Gene-Expression. Cell, 1993. 74(2), pp.281?289. Kay, G.F. et al., Expression Of Xist During Mouse Development Suggests A Role In The Initiation Of X-Chromosome Inactivation. Cell, 1993. 72(2), pp.171?182. Keer, J.T. et al., Genetic-Mapping In The Region Of The Mouse X-Inactivation Center. GENOMICS, 1990. 7(4), pp.566?572. Keohane, A.M. et al., H4 acetylation, XIST RNA and replication timing are coincident and define X; autosome boundaries in two abnormal X chromosomes. Human Molecular Genetics, 1999. 8(2), pp.377?383. Kohlmaier, A. et al., A chromosomal memory triggered by Xist regulates histone methylation in X inactivation. PLOS Biology, 2004. 2(7), pp.991?1003. Kwon, J.E. et al., BTB domain-containing speckle-type POZ protein (SPOP) serves as an adaptor of Daxx for ubiquitination by Cul3-based ubiquitin ligase. Journal Of Biological Chemistry, 2006. 281(18), pp.12664?12672. Lachner, M. et al., Methylation of histone H3 lysine 9 creates a binding site for HP1 proteins. Nature, 2001. 410(6824), pp.116?120. Larsen, F. et al.,. CpG Islands As Gene Markers In The Human Genome. Genomics, 1992. 13(4), pp.1095?1107. Lee, J. T. et al., A 450 kb transgene displays properties of the mammalian X-inactivation center. Cell, 1996. 86(1), pp.83?94. Lee, J. T. & Jaenisch, R., Long-range cis effects of ectopic X-inactivation centres on a mouse autosome. Nature, 1997. 386(6622), pp.275?279. Lennartsson, A. & Ekwall, K., Histone modification patterns and epigenetic codes. Biochimica Et Biophysica Acta, 2009. 1790(9), pp.863?868. Limadefa, A. & Jaworska, H., Late DNA Synthesis In Heterochromatin. Nature, 1968. 217(5124), p.138?142. 55 Luger, K et al., Crystal structure of the nucleosome core particle at 2.8 angstrom resolution. Nature, 1997. 389(6648), pp.251?260. Lyon, M.F., Gene action in the X-chromosome of the mouse (Mus musculus L.). Nature, 1961. 190, pp.373?373. Lyon, M.F. et al., Lack Of Inactivation Of A Mouse X-Linked Gene Physically Separated From The Inactivation Center. Journal Of Embryology And Experimental Morphology, 1986. 97, pp.75?85. De Mello, J.C. et al., Random X Inactivation and Extensive Mosaicism in Human Placenta Revealed by Analysis of Allele-Specific Gene Expression along the X Chromosome. PLOS One, 2010. 5(6). Mermoud, J.E. et al., Histone H3 lysine 9 methylation occurs rapidly at the onset of random X chromosome inactivation. Current Biology, 2002. 12(3), pp.247?251. Mohammad, F. et al., Kcnq1ot1/Lit1 noncoding RNA mediates transcriptional silencing by targeting to the perinucleolar region. Molecular And Cellular Biology, 2008. 28(11), pp.3713?3728. Muthurajan, U.M. et al., The Linker Region of MacroH2A Promotes Self-association of Nucleosomal Arrays. Journal Of Biological Chemistry, 2011. 286(27), pp.23852?23864. Namekawa, S.H. et al., Two-Step Imprinted X Inactivation: Repeat versus Genic Silencing in the Mouse. Molecular And Cellular Biology, 2010. 30(13), pp.3187?3205. De Napoles, M. et al., Polycomb group proteins Ring1A/B link ubiquitylation of histone H2A to heritable gene silencing and X inactivation. Developmental Cell, 2004. 7(5), pp.663?676. Nesterova, T B et al., Characterization of the genomic Xist locus in rodents reveals conservation of overall gene structure and tandem repeats but rapid evolution of unique sequence. Genome Research, 2001. 11(5), pp.833?49.  Nozawa, R.S. et al., Human inactive X chromosome is compacted through a PRC2-independent SMCHD1-HBiX1 pathway. Nature Structural & Molecular Biology, 2013. 20(5), pp.566?73.  Panning, B. & Jaenisch, R., DNA hypomethylation can activate Xist expression and silence X-linked genes. Genes & Development, 1996. 10(16), pp.1991?2002. Penny, G.D. et al., Requirement for Xist in X chromosome inactivation. Nature, 1996. 379(6561), pp.131?137. Peters, A. et al., Histone H3 lysine 9 methylation is an epigenetic imprint of facultative heterochromatin. Nature Genetics, 2002. 30(1), pp.77?80. 56 Peters, A. et al., Partitioning and plasticity of repressive histone methylation states in mammalian chromatin. Molecular Cell, 2003. 12(6), pp.1577?1589. Plath, K. et al., Developmentally regulated alterations in polycomb repressive complex 1 proteins on the inactive X chromosome. Journal Of Cell Biology, 2004. 167(6), pp.1025?1035. Pullirsch, D. et al., The Trithorax group protein Ash2l and Saf-A are recruited to the inactive X chromosome at the onset of stable X inactivation. Development, 2010. 137(6), pp.935?43.  Rastan, S., Non-Random X-Chromosome Inactivation In Mouse X-Autosome Translocation Embryos - Location Of The Inactivation Center. Journal Of Embryology And Experimental Morphology, 1983. 78(DEC), pp.1?22. Rastan, S. & Robertson, E.J., X-Chromosome Deletions In Embryo-Derived (EK) Cell-Lines Associated With Lack Of X-Chromosome Inactivation. Journal Of Embryology And Experimental Morphology, 1985. 90, pp.379?388. Ratnakumar, K. et al., ATRX-mediated chromatin association of histone variant macroH2A1 regulates alpha-globin expression. Genes & Development, 2012. 26(5), pp.433?438. Russell, L.B., Mammalian X-Chromosome Action - Inactivation Limited In Spread And In Region Of Origin. Science, 1963. 140(357), p.976?&. Sarma, Kavitha et al., Locked nucleic acids (LNAs) reveal sequence requirements and kinetics of Xist RNA localization to the X chromosome. Proceedings of the National Academy of Sciences of the United States of America, 2010. 107(51) , pp.22196-22201. Schoeftner, S. et al., Recruitment of PRC1 function at the initiation of X inactivation independent of PRC2 and silencing. EMBO Journal, 2006. 25(13), pp.3110?3122. Schotta, G. et al., A silencing pathway to induce H3-K9 and H4-K20 trimethylation at constitutive heterochromatin. Genes & Development, 2004. 18(11), pp.1251?1262. Senner, C.E. et al., Disruption of a conserved region of Xist exon 1 impairs Xist RNA localisation and X-linked gene silencing during random and imprinted X chromosome inactivation. Development, 2011. 138(8), pp.1541?50.  Sharp, A., Robinson, D.O. & Jacobs, P., Absence of correlation between late-replication and spreading of X inactivation in an X;autosome translocation. Human Genetics, 2001. 109(3), pp.295?302. Silva, J. et al., Establishment of histone H3 methylation on the inactive X chromosome requires transient recruitment of Eed-Enx1 Polycomb group complexes. Developmental Cell, 2003. 4(4), pp.481?495. 57 Stein, R., Razin, A. & Cedar, H., In Vitro Methylation Of The Hamster Adenine Phosphoribosyltransferase Gene Inhibits Its Expression In Mouse L-Cells. Proceedings Of The National Academy Of Sciences Of The United States Of America-Biological Sciences, 1982. 79(11), pp.3418?3422. Steward, M.M. et al., Molecular regulation of H3K4 trimethylation by ASH2L, a shared subunit of MLL complexes. Nature Structural & Molecular Biology, 2006. 13(9), pp.852?854. Thorogood, N.P. & Brown, C.J., Active chromatin marks are retained on X chromosomes lacking gene or repeat silencing despite XIST/Xist expression in somatic cell hybrids. PloS One, 2010. 5(5), p.e10787.  Venolia, L. et al., Transformation With DNA From 5-Azacytidine-Reactivated X-Chromosomes. Proceedings Of The National Academy Of Sciences Of The United States Of America-Biological Sciences, 1982. 79(7), pp.2352?2354. Wang, H.B. et al., Role of histone H2A ubiquitination in polycomb silencing. Nature, 2004. 431(7010), pp.873?878. Wang, J.B. et al., Imprinted X inactivation maintained by a mouse Polycomb group gene. Nature Genetics, 2001. 28(4), pp.371?375. Wutz, A., Rasmussen, T.P. & Jaenisch, R., Chromosomal silencing and localization are mediated by different domains of Xist RNA. Nature Genetics, 2002. 30(2), pp.167?74.  Yan, C. & Boyd, D.D., Histone H3 acetylation and H3 K4 methylation define distinct chromatin regions permissive for transgene expression. Molecular And Cellular Biology, 2006. 26(17), pp.6357?71.  Zhang, L.F., Huynh, K.D. & Lee, J.T., Perinucleolar targeting of the inactive X during S phase: Evidence for a role in the maintenance of silencing. Cell, 2007. 129(4), pp.693?706. Zhao, J. et al., Polycomb Proteins Targeted by a Short Repeat RNA to the Mouse X Chromosome. Science, 2008. 322(5902), pp.750?756. Zhou, W. et al., Histione H2A monoubiquitination represses transcription by inhibiting RNA polymerase II transcriptional elongation. Molecular Cell, 2008. 29(1), pp.69?80.      


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items