Despite decades of biochemical, proteomic and genetic characterizations of the Xist longnon-coding RNA, the master regulator of X-chromosome inactivation (XCI), the
spatiotemporal kinetics of Xist expression and how it coordinates the dynamic recruitment
of protein effectors to direct robust gene silencing across the X-chromosome remain
unexplored. Furthermore, most studies of Xist are conducted in the mouse system, so its
role in mediating X-chromosome dampening (XCD), a unique mode of X-chromosome
dosage compensation in humans, is still an open question. Therefore, in this thesis, I
sought to plug this lack of understanding by (i) elucidating the molecular mechanism by
which the chromosome-wide silencing compartment on the inactive X-chromosome (Xi)
is formed and maintained across XCI, (ii) determining if XIST is responsible for XCD in
human embryonic development, and (iii) characterizing the link between the genomewide localization patterns of Xist and its effect on gene silencing.
iii
In Chapter 2, we employed sophisticated super-resolution microscopy and single
particle tracking techniques in living mouse embryonic stem cells (ESCs) undergoing XCI
to precisely interrogate the spatiotemporal kinetics of Xist upregulation and its association
with key protein partners that direct various aspects of XCI, and found that Xist first forms
distinct foci that are locally restricted to about 50 sites across the Xi. Each site contains
two Xist molecules, which seed the spatial concentration of integral protein interactors
into what we term as a supra-molecular complex (SMC), made up of CIZ1 and CELF1 for
localization of Xist across the Xi, PCGF5 and other components of Polycomb Repressive
Complex 1 (PRC1) for inducing heterochromatinization and Xi compaction, and the
transcriptional repressor SPEN, which is aided by its intrinsically disordered regions
(IDRs) to promote phase separation-mediated aggregation of more SPEN molecules for
the faithful and robust maintenance of X-linked gene silencing.
In Chapter 3, we demonstrated the requirement of XIST in regulating XCD, as
CRISPR-Cas9-mediated ablation of XIST in various human pluripotent stem cell (hPSC)
lines that model XCD in vitro resulted in the loss of transcriptional downregulation on the
dampened X-chromosome (Xd). By applying a high throughput genomics method that
maps the chromatin binding of XIST, we discovered increased XIST enrichment on the Xi
relative to the Xd, suggesting that XIST may be unable to exert complete silencing due to
its reduced accumulation on the Xd. Moreover, we characterized a novel function of XIST,
which can surprisingly propagate to certain autosomal regions beyond the X-chromosome
territory, as well as repress some of the autosomal genes that it targets, resulting in a sex
iv
imbalance of the XIST-enriched autosomal gene expression levels that is also observed
in human pre-implantation embryos in vivo.
In Chapter 4, we built on our findings from Chapter 3 and showed that like in hPSCs
in which the Xd exhibits a lower enrichment of XIST binding, early differentiating mouse
ESCs also display diminished levels of Xist accumulation on the Xi, accompanied by
reduced X-linked gene silencing, compared to the higher levels of Xist enrichment in late
differentiated cells with complete XCI. Additionally, we characterized the distribution of
Xist to autosomes in the mouse system, and illustrated how Xist spreading to the A- and
B-compartments of the genome can be used to predict its differential ability to trigger gene
silencing. Furthermore, we highlighted the remarkable ability of Xist to enact partial Xlinked gene repression in a mouse somatic cell line, which has surpassed the narrow
developmental window surrounding pluripotency that was previously established in the
XCI field for Xist-mediated silencing.
Collectively, our findings gathered from these research projects have greatly
expanded on the current knowledge about the molecular landscape established by Xist
in setting up the Xi compartment, revealed insights on how XIST controls both XCI and
XCD, and examined the genome-wide localization of Xist/XIST and its consequent gene
silencing dynamics. In addition to addressing several open questions in the XCI field, our
knowledge gained from this body of work on the Xist lncRNA may be extended to
characterize other lncRNAs that contribute to the establishment and/or maintenance of
gene regulatory nuclear compartments.