Development and Application of Novel Methodologies to Interrogate X-chromsome Inactivation.

Abstract

-chromosome inactivation equalizes X-linked gene expression between the mammalian sexes by epigenetically silencing genes of one of the two X-chromosomes in females. Once established, the transcriptional states of the inactive and active X-chromosomes are faithfully transmitted across many cell division cycles, making X-inactivation a paradigm of epigenetic inheritance. In mice, the transcriptional fates of the two X-chromosomes are thought to be established by opposing long non-protein-coding (lnc) RNAs, Xist and Tsix. Xist is exclusively expressed from the inactive X-chromosome and is postulated to trigger X-inactivation. Tsix RNA is expressed solely from the active-X-chromosome and is believed to prevent X-inactivation by repressing Xist. Prior work by our group has shown that X-linked gene silencing is able to initiate in the absence of Xist RNA, calling into question current models of X-inactivation, including the role of the Tsix RNA. Here, I examined the spatial and temporal requirement of the Tsix RNA in X-inactivation. By developing a protocol to image nascent RNA expression in intact mouse embryos, I discovered that Tsix is dispensable during the initiation phase of X-inactivation. Instead, Tsix is required to prevent Xist expression and X-inactivation once X-inactivation has occurred normally. Through these data, I hypothesized that novel X-linked lncRNAs and proteins contribute to X-inactivation. I therefore developed an allele-specific RNA-Seq pipeline to catalog lncRNAs and proteins that are expressed from the inactive X-chromosome, which my data predict are candidate regulators of X-inactivation. I also leveraged this RNA-Seq pipeline to profile X-inactivation defects in mouse trophoblast stem cells (TSCs) lacking EED, a core subunit of Polycomb Repressive Complex 2 (PRC2) that is thought to silence X-linked genes. Unexpectedly, I found that only a small subset of genes is derepressed from the inactive X-chromosome in Eed null TSCs. In wild-type TSCs, these genes are characterized by low-level transcription and open chromatin. Thus, PRC2 prevents induction of basally transcribed X-linked genes but is dispensable in the repression of stringently silenced X-linked genes, providing a novel mode of epigenetic transcriptional repression by PRC2. Future work will characterize novel regulators of X-inactivation discovered through my RNA-Seq pipeline.