Kinetics of Xist-induced gene silencing depends on combinations of epigenetic and genomic features

To initiate X-chromosome inactivation (XCI), the long non-coding RNA Xist mediates chromosome-wide gene silencing of one X chromosome in female mammals to equalize gene dosage between the sexes. The efficiency of gene silencing, however is highly variable across genes, with some genes even escaping XCI in somatic cells. A gene’s susceptibility to Xist-mediated silencing appears to be determined by a complex interplay of epigenetic and genomic features; however, the underlying rules remain poorly understood. We have quantified chromosome-wide gene silencing kinetics at the level of the nascent transcriptome using allele-specific Precision nuclear Run-On sequencing (PRO-seq). We have developed a Random Forest machine learning model that can predict the measured silencing dynamics based on a large set of epigenetic and genomic features and tested its predictive power experimentally. While the genomic distance to the Xist locus is the prime determinant of the speed of gene silencing, we find that also pre-marking of gene promoters with polycomb complexes is associated with fast silencing. Moreover, a series of features associated with active transcription and the O-GlcNAc transferase Ogt are enriched at rapidly silenced genes. Our machine learning approach can thus uncover the complex combinatorial rules underlying gene silencing during X inactivation. In this study we set out to identify the genetic and chromatin features of X-linked loci, or more precisely the enrichment or depletion of certain features, that predispose genes on the chromosome X to be efficiently silenced or avoid silencing. For this, we measured chromosome-wide silencing dynamics of X-linked genes following induction of Xist expression, using allele-specific Precision nuclear Run-On sequencing (PRO-seq) or RNAseq. 8 Timepoints after Dox induced Xist expression in undifferentiated ES cells were included in the PROseq analysis, of which the uninduced and 24 hrs induced samples were performed in replicate. For mRNAseq analysis, two dox-induced Xist time courses were performed: One in undifferentiated cells, over 6 timepoints, all in replicate, to measure the silencing rates of X-linked genes without interference of differentiation, and the second in differentiating cells, over 6 timepoints, all in replicate, to measure the effect of differentiation on the Xist induced silencing of X-linked genes.