Skip to main content

Exploring mechanisms of gene repression and escape during X-chromosome inactivation

Periodic Reporting for period 3 - XPRESS (Exploring mechanisms of gene repression and escape during X-chromosome inactivation)

Reporting period: 2019-01-01 to 2020-04-30

"Epigenetic mechanisms ensure that patterns of gene expression that are established during development can be appropriately maintained; they are also involved in preserving genome integrity and ensuring the appropriate expression and organisation of the genome at different stages in the life of organisms. The focus of this grant concerns a mammalian paradigm for epigenetic processes: X-chromosome inactivation (XCI), whereby one of the two X chromosomes in females is silenced to ensure that most X-chromosome gene expression is equal between XX females and XY males. X inactivation is a developmentally regulated process that is accomplished within the space of a few cell divisions during early embryogenesis. A remarkable feature of this process is that the two homologous X chromosomes are treated differently, while residing in the same nucleus. This differential treatment becomes clonally heritable, and is maintained throughout life but is reversed in the germ line. X inactivation results in the silencing of over 1000 genes and the formation of a condensed chromatin structure also known as the Barr body. However, past work from our lab and others, has shown that not all genes are silenced at the same time and that some genes can escape from X inactivation, either constitutively, or else in a facultative, lineage-specific fashion. So far, almost nothing is known about the processes of gene silencing and escape during X inactivation. X-chromosome-wide silencing is dependent on the action of a unique non-coding RNA, Xist (X-inactive-specific-transcript). When XCI is induced during differentiation, Xist is up-regulated from one of the two X chromosomes, coating it in cis and resulting in gene silencing. Several decades of work have sought to explore the mechanisms of Xist RNA action and its capacity to silence genes.

In this ERC grant we are exploring the molecular mechanisms ensuring gene silencing on the X during X inactivation, as well as the mechanisms that allow some genes to be expressed (ie escape) from an otherwise heterochromatic X. Different levels of epigenetic control are being explored: local chromatin accessibility modifications, transcription factor binding, chromosome organisation. We are looking at chromosome wide changes at the nucleotide level using genomics approaches, and the factors involved using both biochemical and imaging approaches. We are also using genetic engineering and fluorescently labelled cells as a powerful and easy way to test for the functions of cis and trans regulatory candidates. This research has already allowed us to uncover some of the key factors involved in the initiation of XCI and the changes that occur at the chromatin level during the gene silencing as well as escape from repression. We have provided a detailed molecular view (epigenomic ""roadmap"") of the earliest chromatin events following Xist RNA coating of the chromosome in cis. These early events led us to predict that specific chromatin modifiers (histone deacetylases) might be involved and we went on to show that this is indeed the case. We have also uncovered novel factors that bind to the Xist RNA and participate in gene silencing. Importantly, we have identified factors that may account for the differences between genes on the X in their silencing kinetics. We are now investigating the molecular basis for differential gene repression and escape from XCI, taking into account the regulatory landscape and 3D architecture of different loci, the chromatin status and the possible pre-marking by certain factors, as well as the trans-acting factors that become recruited when Xist RNA coats the chromosome. Our results provide new insights into mechanisms of gene regulation in general and how they relate to non-coding RNAs, chromosome architecture and chromatin states. Our findings are relevant not only to the silencing of the X chromosome in mammals, but more generally for our understanding of how and why genes become silenced and others remain expressed in diseases such as cancer, where epigenetic mechanisms become disrupted. Indeed some of the key factors involved in XCI that we are studying are also implicated in cancer (eg SPEN/SHARP, RNA binding proteins, HDACs, Polycomb group factors).
X-chromosome inactivation (XCI) is a paradigm for developmental gene regulation and epigenetic processes. In this grant we have set out to dissect the molecular mechanisms that underlie the XCI process at the level of genes, using an integrated approach to define the relative roles of the non-coding Xist RNA, that triggers XCI, as well as transcription factors, chromatin and chromosome structure.
We are defining both general principles underlying gene silencing and escape in XCI, as well as locus specific features that could serve as new paradigms.
The specific objectives were to:
(1) Define epigenomic and transcriptomic dynamics during XCI
(2) Identify trans-acting factors involved in gene silencing and escape
(3) Explore how chromosomal organization influence the transcriptional behavior of X-linked loci
All of the projects are progressing well. The main accomplishments and ongoing work are summarised below for each of the aims.

AIM 1: Defining the mechanisms of X-linked gene silencing
Aim 1.1 was to investigate transcriptional silencing events during the onset of XCI. This is almost complete with three manuscripts completed. Using polymorphic embryonic stem cells (TX1072) harboring a doxycycline-inducible Xist gene on one X chromosome we could study the transcriptional and chromatin events linked to XCI at high resolution in time (scale of hours), using mRNAseq, GroSeq, Ttseq and ChIPseq. We uncovered chromatin features that predict rates of gene silencing (submitted); we demonstrated the role of histone deacetylation as an early step in silencing of some genes and also in allowing the spread of polycomb marks into active gene bodies (Zylicz, Bousard et al, under revision); we also examine how the different regions of Xist RNA impact on chromatin changes during XCI (Da Rocha et al, in preparation).
Aim 1.2 was to investigate candidate proteins involved in Xist RNA mediated gene silencing.
We have used both knock out and Auxin-degron approaches to investigate proteins known to be involved in Xist-dependent silencing. These include HDAC3 which we show participates in XCI, in part through activation of prebound HDAC3 on the X chromosome (Zylicz, Bousard et al, under revision); SPEN which is a key protein involved in Xist-mediated gene silencing has been intensively investigated using a degron approach in ESCs, an in vivo KO in mice, as well as functional dissection of different protein domains and biochemical approaches. We have gained important insights into Spen’s mechanism of action (Dossin et al, in preparation). Other factors that are recruited by Xist RNA (Fus, PRC1) are also under investigation.
Aim 1.3 concerns genetic screens to identify new factors implicated in XCI using multiple endogenous X-linked genes that we fluorescently tagged in TX1072 cells, as a read-out for silencing or escape from XCI. An initial genome-wide knock out (GeCKO) approach failed, due to poor infection rates in TX1072 cells (see Section 2.2). Alternative esiRNA screens were successfully set up in collaboration with Frank Buchholz, Biotec Dresden. Validation of novel candidate factors that affect different X-linked genes is ongoing. In parallel, a small-scale screen of 40 candidates that bind the A-repeat of Xist RNA (its silencing region) identified by the Shkumatava lab (Curie Institute), was performed using esiRNAs in G6pd-GFP/Tomato tagged TX1072cells, by FACS analysis identifying known and novel Xist RNA binding factors (Grandgeorge, Pinheiro et al, under revision).

AIM 2: Mechanisms of escape from X-chromosome inactivation
Aim 2.1 was to investigate the timing and nature of escape from XCI. The transcriptomic and chromatin profiling in Aim 1.1 allowed us to identify some characteristics of regions that can escape constitutively from XCI. We have demonstrated that the X-linked DXZ4 locus is involved in sub-dividing the inactive X into two large megadomains, and is also potentially implicated in efficiency of facultative escape in ESCs differentiated into neural progenitor cells (Giorgetti et al, Nature 2016). We also uncovered a striking link between facultative escape and local TAD-like organisation. We have now shown that in vivo, the DXZ4 locus has a significant impact on facultative escape and we have identified unusual chromatin features of the locus that may explain the mechanism through which this occurs (Attia et al, in preparation). Aim 2.2 involving genetic screens to identify factors involved in facultative (eg Mecp2) and constitute (Jarid1c) escape from XCI (as in Aim 1.2) is underway. Aim 2.3 aims to define the cis-regulatory elements that underlie escape from XCI and is currently ongoing. We have also established the Capture-HiC protocol to examine 3D architecture in different contexts across loci of interest (genes that escape constitutively or facultatively).

AIM 3: Functional assays exploring cis and trans regulation of X-linked genes during XCI
We have created tools for live cell imaging of the X chromosome in ESCs and embryos to visualize and track the movement of genomic loci of interest during XCI. Repetitive operator arrays (TetO, CumateO) have been introduced close to loci of interest for their in vivo visualization, via expression of the operator binding proteins fused to fluorescent proteins. Multiple loci that either show silencing upon XCI, or else constitutive or facultative escape have been successfully tagged and visualised. We can now explore their mobility and 3D architecture, in particular when loci escape or are silenced. This can be compared to dynamics on the active X. We are currently tracking the dynamics of the mentioned loci using fast time lapse live cell microscopy. We have also tagged Xist using the Bgl-EGFP system to track Xist accumulation in ESCs and developing embryos. This allows the tracking of Xist induction in living cells and is being combined with fluorescently labelled antibody fragments (Mintbodies, collaboration with H. Kimura) that detect H3K27me3, H4K20me1, H3K9ac and RNA polymerase II. Extensive live-imaging has been performed to track dynamic changes occurring upon Xist induction. Extensive effort is now targeted towards implementing automated analysis tools. The operator tagged loci can also be targeted by transcriptional regulators and chromatin factors (including those identified in the screens in Aims 1 and 2) using fusions with operator binding proteins. This will be one of the approaches to further explore mechanisms of silencing and escape from XCI. Finally, we have established ex-vivo culture of peri-implantation embryos in collaboration with M. Zernicka Goetz as proposed and the imaging as well as single/low-cell molecular profiling will now be carried out. Our encouraging initial results led us to cross our mouse lines to homozygosity in order to simultaneously track Xist (XistBGL-GFP) and H4K20me1 accumulation (mintbody-mCherry) We are now ready to implement this line for extensive live embryo imaging and compare the findings with in vitro ESC imaging datasets.

In summary Aim 1 is almost complete, Aim 2.1 and 2.2 are in the process of completion and Aims 2.3 and Aim 3 are ongoing.
The conclusions from this first part of our ERC grant have major implications not just in the context of the X chromosome but more generally. Insights from X inactivation help our understanding of gene regulation during embryogenesis and should reveal some of the basic principles underlying gene expression such as the roles of chromatin structure, chromosome organization and non-coding RNAs in enabling or preventing transcription. The results of this grant are providing us with unprecedented insights into the mechanisms underlying the dynamic changes in gene expression occur in the space of just minutes or hours, during the initiation of X inactivation. Our combination of genome wide data production, genetic manipulation and live cell imaging should provide a paradigm for integrative data analysis and modeling of heterochromatin formation. By integrating all of our genome-wide information on the chromatin and transcriptional status of the X chromosome upon Xist RNA coating, we have been able to establish some of the key events that underlie the onset of XCI. The genome-wide data sets will provide the first-ever molecular roadmap of the epigenetic process of X inactivation and should reveal key principles guiding gene-silencing events as well as mechanisms through which genes can be expressed even in an otherwise heterochromatic state. This roadmap will also provide a powerful resource forming the basis for investigation of the regulators identified in our screens and others, for factors involved in gene activation and repression, as well as for non-coding RNA function and chromosome biology. The screens performed within this grant, as well as previous work from several laboratories, searching for proteins implicated in X inactivation, have led to discoveries about the very strategies for gene silencing, involving many unexpected levels of epigenetic regulation including RNA modifications, RNA metabolism, chromatin modifications, chromosome organisation and nuclear organisation. Furthermore, some of the proteins involved in X inactivation (such as SPEN) appear to be directly implicated in cancers such as leukemia. Indeed we hope that our work will lead to new hypotheses that can be tested using animal models and human cell lines, exploring the roles that SPEN and other Xist RNA associated factors have in cancer. Such work should open up potential avenues of research with prognostic and therapeutic value.