Carolyn Janet Brown

X-chromosome inactivation (XCI) is the process by which one of the X chromosomes in XX females is silenced to express similar levels of X-linked genes with XY males. This silencing is incomplete as some genes escape from XCI and other genes vary their XCI status across populations, tissues or samples. Here I derive consensus XCI status calls in humans, extend XCI status calls across species, and determine the relationship between XCI status and various epigenetic marks.I aggregated XCI status calls from multiple studies, deriving XCI status calls for 639 human genes. I found 12% of genes escaping from XCI, 8% variably escaping XCI, and 7% discordant across studies. To make XCI status calls across species I obtained DNA methylation data for 12 species, allowing us to generate an average of 387 XCI calls per species. Overall, 12% of genes escaped XCI, with mouse an outlier with only 5%. Of the genes with predictions across at least four species, 74.8% of them were entirely consistent and only 6% had more than one inconsistent species. Many genes were seen to have primate-specific escape from XCI, while only one gene had an artiodactyla-specific XCI status. The consensus XCI status calls were compared to DNA methylation and commonly analyzed histone marks. I found the expected trend where repressive marks were enriched at genes subject to XCI and activating marks were enriched at genes escaping XCI; however, the histone marks had a large overlap between levels seen at genes subject to XCI and genes escaping from XCI. Only DNAme could accurately predict an individual gene’s XCI status. I combined the marks and found that we could make XCI status calls with 75% accuracy for genes escaping from XCI and 90% accuracy for genes subject to XCI. The marks with the greatest contribution to this predictor were DNAme, H3K27me3 and H3K4me3. The results of these projects further our understanding of which genes escape from XCI, which may be important for analysis of sexual dimorphism and further provide us a means to examine how silencing may be regulated in humans and across mammals.

View record

X-chromosome inactivation establishes dosage compensation between the sexes of eutherian mammals through the long non-coding RNA gene, XIST. During development one of the female X chromosomes up-regulates XIST, which coats that chromosome and causes the large-scale silencing of genes. The X chromosome being inactivated by XIST is repositioned within the nucleus, condensed and enriched with heterochromatin associated factors. Understanding the mechanisms of how XIST functions has provided insights into how non-coding RNAs regulate cellular biology, the process of X-chromosome inactivation and novel tools for regulating the epigenome.The function of XIST was investigated using an inducible XIST cDNA construct integrated into the autosome of a male fibrosarcoma cell line, HT1080. The chromatin domain surrounding the XIST loci had a significant effect on its activity, and of all the autosomal loci the 8p integration site functioned most effectively and was thus used for further studies. A series of isogenic inducible partial XIST constructs were created by modifying the Full length XIST construct in 8p to study the importance of individual regions of XIST. The functions dependent on each region of XIST were identified and the relationships between these identified processes were then examined through the use of chemical inhibitors. XIST silencing of genes was demonstrated to depend upon two distinct regions at the extreme ends of the transcript, but the internal sequences spanning these regions were dispensable. Silencing occurred without obvious dependence on chromatin modifications such as those established by the two polycomb group complexes, that in turn relied on ¬distinct regions of XIST suggesting entirely independent mechanisms. Both polycomb complexes were crucial, along with additional elements, for the recruitment of additional heterochromatin factors. This study in human differentiated cells yielded important insights beyond those seen in mouse differentiating cells. The results of this thesis revealed the regions of XIST that were both crucial and dispensable for its activity, and offer novel insights into the mechanisms that lead to chromosome inactivation.

View record

A long-standing question concerning X-chromosome inactivation has been how some genes avoid the otherwise stable chromosome-wide heterochromatinization of the inactive X. As 20% or more of human X-linked genes escape from inactivation, such genes are important contributors to sex differences in gene expression, and identifying the mechanism by which these exceptions occur will inform our understanding of X-inactivation and broader questions of epigenetic regulation. While bioinformatic studies have generated a list of candidate features, the nature of the elements or definitive evidence that any one particular element is necessary or sufficient for a gene to escape, is still elusive and requires experimental validation. Mouse models offer a well-characterized and readily manipulated system in which to study X-inactivation and escape, but have far fewer genes and gene clusters that escape than humans. Given these differences, it was unclear whether the mechanism of escape gene regulation is conserved between species, and thus, this thesis addresses conservation of the escape process and the potential to model human escape gene regulation using mouse systems.Bacterial artificial chromosomes carrying genes known to escape from X-inactivation in humans were targeted to the Hprt locus and studied on the inactive X in mice. They were examined for escape by expression and inactivation-associated DNA methylation of promoter CpG islands. Expression from the inactive X and corresponding low promoter DNA methylation of human gene RPS4X demonstrated that the mouse system is capable of recognizing human elements. Furthermore, the escape status of the transgene remained stable between developmental time points, tissues, and individual females. A second human escape gene, KDM5C, was targeted to the Hprt locus and was surprisingly subject to inactivation, suggesting that its mechanism of escape was not conserved or that the critical elements for escape were not contained in the transgene. To further interrogate the escape elements involved in both human genes analyzed, as well as additional constructs of interest, a docking site at Hprt was generated in a female mouse embryonic stem cell line. Overall, this thesis contributed to the development of approaches to examine human escape from inactivation, and characterization of two human escape regions.

View record

The process of X-chromosome inactivation achieves dosage compensation betweenmammalian males and females. In females one X chromosome is transcriptionally silencedthrough a variety of epigenetic modifications including DNA methylation. Most X-linked genesare subject to X-chromosome inactivation and only expressed from the active X chromosome.On the inactive X chromosome, the CpG island promoters of genes subject to X-chromosomeinactivation are methylated in their promoter regions, while genes which escape from Xchromosomeinactivation have unmethylated CpG island promoters on both the active andinactive X chromosomes.The first objective of this thesis was to determine if the DNA methylation of CpG islandpromoters could be used to accurately predict X chromosome inactivation status. The secondobjective was to use DNA methylation to predict X-chromosome inactivation status in a varietyof tissues. A comparison of blood, muscle, kidney and neural tissues revealed tissue-specificX-chromosome inactivation, in which 12% of genes escaped from X-chromosome inactivation insome, but not all, tissues. X-linked DNA methylation analysis of placental tissues predicted fourtimes higher escape from X-chromosome inactivation than in any other tissue. Despite thehypomethylation of repetitive elements on both the X chromosome and the autosomes, nochanges were detected in the frequency or intensity of placental Cot-1 holes.The third objective of this thesis was to use DNA methylation to investigate X-chromosomeinactivation in female samples with chromosomally abnormal karyotypes. The spread of Xchromosomeinactivation into the autosomal portion of six unbalanced X;autosometranslocations revealed similarities between X;autosome translocations involving the sameautosome and therefore suggested a role for DNA sequence in influencing X-chromosomeinactivation status of genes. Autosomal genes that escaped from inactivation were found tohave significantly lower L1 and LTR but higher Alu content than genes which were subject toinactivation. Lastly, DNA methylation was used to predict the number of inactive Xchromosomes in triploid placental samples. Triploid samples provide an excellent system inwhich to study the counting step of X-chromosome inactivation and DNA methylation analysisprovides a means to determine the number of inactive X chromosomes using only a DNAsample.

View record

X-chromosome inactivation ensures equal expression of mammalian male and female X-linked genes by transcriptionally silencing one X chromosome in each female cell. The pivotal molecule responsible for the silencing is a long non-coding RNA XIST; however, an all-encompassing model explaining how XIST induces silencing of the whole X chromosome is yet to emerge. This thesis aims to broaden our understanding of XIST action in humans by leveraging an inducible XIST transgene capable of silencing downstream reporters to identify sequences within XIST and XIST-interacting proteins critical for gene silencing.First, we demonstrate that the repeat A region of XIST is necessary and sufficient to induce gene silencing, at least locally, irrespective of the makeup of the surrounding chromatin, and that XIST induces silencing of a distal gene in one of the HT1080 cell lines. Second, we show that individual repeats of a consensus repeat A sequence contribute additively to silencing. Mutations within a construct consisting of two repeat A units both demonstrate that the two palindromic sequences within the repeat A units spanning ‘ATCG’ and ‘ATAC’ tetranucleotides are critical for repeat A function and add to the evidence that the first palindrome forms a hairpin, rather than engaging in pairing between repeat A units.Third, we explore which proteins are critical for XIST-induced silencing. We show that histone deacetylation, an early mark of an X-chromosome inactivation, is likely a consequence, and not the cause of XIST-induced silencing. We next demonstrate that in the transgenic HT1080 system, gene silencing is not accompanied by recruitment of the H3K27me3 repressive histone mark and XIST induces silencing independently of its previously reported associations with the polycomb repressive complex 2 (PRC2). Finally, we performed siRNA-mediated knock-down of 31 proteins previously implicated to play a role in X-chromosome inactivation. Our results show that proteins involved in XIST RNA localization (YY1), chromatin organization (SATB2, HNRNPU), and cell cycle (ATM), as well as an E3 ubiquitin ligase (SPOP) contribute to XIST-induced gene silencing in the HT1080 system. Thus, we demonstrate that the repeat A alone induces gene silencing and identify candidate pathways critical for its function.

View record

Mammalian dosage compensation of X-linked genes is achieved between XX females and XY males by silencing one of the two X chromosomes. It is the expression of a functional non-coding RNA transcript, XIST that is responsible for the initiation of silencing during X-chromosome inactivation. XIST expression and subsequent in cis localization to the future inactive X chromosome initiates a cascade of epigenetic events that leads to the formation of facultative heterochromatin. The exact role the XIST RNA plays in establishing and maintaining the inactive state is uncertain. It is hypothesized that the XIST RNA sets up a repressive nuclear compartment and transcriptionally silences through the recruitment of factors required for setting up the heterochromatic state of the inactive X. In this thesis I address XIST/Xist’s recruitment of factors with two separate approaches: 1) I ask whether Xist, in the absence of silencing is able to recruit epigenetic marks in a somatic cell hybrid system, 2) I evaluate a system whereby the XIST RNA is tagged and can be isolated to identify novel RNA-protein interactions. To assess epigenetic features that may be directly recruited by the expression of the XIST/Xist RNA, I have analyzed mouse/human somatic cell hybrids where XIST/Xist expression and silencing are disconnected. Loss of active chromatin marks, H3 acetylation and H3 lysine 4 methylation, was not observed with XIST/Xist expression; nor was there a gain of DNA methylation or the silencing of the Cot-1 fraction. Therefore, these marks of heterochromatin are not solely dependent upon XIST/Xist expression in a somatic cell.The isolation of the XIST RNA with its interacting partners would allow us to better understand the mechanism by which XIST acts to silence the inactive X. I have developed a MS2 stem loop-tagged XIST RNA and integrated it into an inducible XIST system. The MS2 tagging system is a valid system for pulling out the XIST RNA and identifying interacting proteins since the tagged RNA was able to interact with whatever factor(s) are required to silence the Cot-1 fraction and form a localized RNA signal.

View record

Dosage compensation in female mammals is achieved by XIST/Xist RNA mediatedtranscriptional silencing of one of the two X chromosomes. Several developmental specific cis-acting regulatory elements for mouse XIST have been demonstrated. One of these regulatory elements is the mouse Tsix locus, which is transcribed antisense to Xist and represses Xist on one ofthe X chromosomes at the onset of X inactivation. A transcript antisense to human XIST has been shown; however, its functional significance has been repeatedly challenged. My thesis aims to uncover cis-acting regulatory elements for human XIST and determine whether the elements arecomparable to those found in mice.Currently, multi-copy integrations of a transgene containing human XIST into male mouse embryonic stem (ES) cells or into male human fibrosarcoma cells are the model systems of choice for studying the initiation of human X inactivation because the transcription antisense to humanXIST can be detected by RT-PCR from the transgenes. Using DNase I hypersensitivity (HS) mapping, I found one HS site on the human transgene in mouse ES cells located approximately 11 kb downstream of XIST3’ end. While this HS site does not correspond to the transcription starts of TSIX previously described, it encompasses a small cluster of CTCF binding sites based on in silico search. Besides the downstream HS site, I discovered two HS sites in differentiated cell lines. One of the HS sites is immediately upstream of the XlST transcription start site. The other HS site (HS 101), located approximately 65 kb upstream of human XIST transcription start, resides within a region that shares above 70% sequence identity with cow and dog but not mouse. I analyzed these upstream HS sites and found that HS 101 exhibits bi-directional promoter and enhancer activity. My thesis revealed three previously unidentified HS sites flanking the human XIST locus; the presence of only one ES cell specific HS site downstream of XIST3’ end is in sharp contrast to theseven sites reported in mouse. The results suggest that mouse Xist and human XIST are regulated differently. To account for the differences in regulatory elements, I propose an alternative model for human XIST regulation.

View record