PDF, English Restricted access: Repository staff only until 11 August 2026. Login+Download (29MB) | Terms of use

Abstract

During female mammalian development one of the two X chromosomes becomes silenced through X chromosome inactivation (XCI). XCI ensures proper X-linked gene dosage between females (XX) and males (XY), and is essential for female development. Intriguingly, some X-linked genes escape the chromosome-wide gene silencing during XCI (i.e., the “escapees”) and are expressed from both the active (Xa) and inactive X chromosome (Xi) within the same nucleus. The relevance of escape from XCI for female development and sexual dimorphism is becoming increasingly recognized, but the mechanisms remain elusive. Escapees can be classified into constitutive (3 - 5% of X-linked genes in mice), which escape in all tissues and individuals and tissue- and/or cell-specific facultative escapees (> 20%). Constitutive escapees have Y chromosome homologs and thus expression from the Xi is presumably required for higher CTCF binding on the Xi in close proximity to escapees in NPCs proper female development. Expression from the Xi of a subset of facultative escapees on the other hand, may be due to inefficient maintenance of XCI silencing, and this in some cases could underlie sex-biased disorders. Clusters of facultative escapees have been found to reside in 3D chromatin domains that resemble Topologically Associating Domains (TADs) on the Xi. The Xi is largely depleted for TADs. However, a few TAD-like domains can be detected on the Xi that correlate with expressed clusters of escapees, suggesting that there might be an interplay between 3D organization and transcriptional activity of facultative escapees. The main goal of my PhD project was to explore the relationship between gene activity and 3D chromosome topology on the Xi. To this end, I first established a fast and straightforward Capture Hi-C protocol, which allows for the allele-specific interrogation of the local 3D genome topology of megabase-sized regions of interest. The allele-specific analysis of this protocol relies on the mapping of known single nucleotide polymorphisms (SNPs) in F1 hybrid mouse embryonic stem cells (mESCs), generated from a cross between two highly polymorphic mouse strains. Compared to genomewide Hi-C this new easy-to-follow and time-effective Capture Hi-C protocol with its commercial capture strategy enables increased resolution at a reduced cost, thereby making it accessible for a wide variety of applications. I used this high-resolution Capture Hi-C protocol to investigate the high-resolution 3D genome topology of facultative escape clusters in several clonal neural progenitor cell (NPC) lines, where different subsets of facultative escapees are found, showing clone-specific, transcription-correlated local 3D genome topology at these clusters. I further characterized these facultative escape clusters by performing CUT&RUN for different chromatin marks and structural factors to investigate differences in these features at escape clusters compared to the silenced chromatin of the Xi. I observed NPC clone-specific escape from XCI as well as structural features, with active loci in the different NPC clones being exclusively engaged in TAD-like structures and long-range loops. To test the role of structural factors such as the CCCTC-binding factor (CTCF) in these TAD-like structures and in the maintenance of escapee expression, I established clonal NPC cell lines with a degron-based system that allowed acute and reversible depletion of CTCF. Surprisingly, depletion of CTCF for several days did not result in a major effect on 3D chromatin structure nor in the expression of escapees on the Xi, despite higher CTCF binding on the Xi in close proximity to escapees in NPCs. This implies that escapee expression patterns and 3D structure on the Xi are largely independent of CTCF in NPCs. This argues for a transcription-induced local 3D genome topology at the facultative escape clusters of the Xi, rather than a CTCF-loop extrusion-based model. Furthermore, I found that depletion of CTCF did not lead to a spread of escape status into nearby regions of the Xi. Thus, contrary to previous hypotheses, CTCF does not play a role as a boundary factor that prevents spread from facultative or constitutive escapee regions into neighbouring silent genes. I also explored the novel and unexpected role that the XCI master regulatory RNA, Xist, plays in modulating the expression of escapees in NPCs. The prevailing view has been that Xist RNA together with XCI initiation partners including SPEN, only plays a major role in establishing XCI during early embryogenesis, and that X-linked genes then remain repressed throughout life due to epigenetic mechanisms, mainly at the chromatin level. However, recent evidence has revealed that reduced Xist RNA levels can in fact increase escapee expression in somatic cells, and that even a slight change in the dosage of some escapee-encoded proteins, can in turn influence cell states that impact tissue homeostasis and disease susceptibility. I examined NPCs with an inducible Xist gene on the Xi, which allowed me to increase Xist levels in NPCs and assess the impact on facultative and constitutive escapees. Remarkably, increased Xist RNA levels resulted in efficient silencing of escapee genes. Furthermore, I showed that this silencing is SPEN-dependent implying that a similar mechanism to the initiation of XCI is used in NPCs. Although both constitutive and facultative escapees are downregulated, constitutive escapees seem to be more resistant to complete silencing via Xist even after prolonged exposure. On the other hand, facultative escapees can be irreversibly silenced. Using Capture Hi-C, I further show that higher levels of Xist in NPCs lead to loss of local 3D genome topology at facultative escape clusters. This reinforces the hypothesis from the first part of my PhD that TAD-like structures on the Xi are transcription-induced structures. Taken together, I have been able to show during my PhD that facultative escape from XCI is highly variable between different clonal NPC lines and that although this correlates with CTCF binding on the Xi and occurs in the context of local TAD-like structures on the Xi, both escape and associated 3D structures seem to be maintained largely independently of CTCF. I have also shown that increased Xist RNA levels in NPCs can impose silencing of both facultative and constitutive escapees. Xist RNA together with its partner SPEN, thus play a significant role in initiating the silencing of escapees, even in a somatic context. This discovery has implications for physiology and disease, as natural variations in Xist levels during development or in adult tissues will likely influence the extent to which genes escape XCI, thereby profoundly impacting global gene expression, differentiation potential, and tissue homeostasis.