Population-scale and single-cell RNA sequencing provides insight into X chromosome inactivation. T. Tukiainen1,2, A. Kirby1,2, T. Lappalainen3,4, A.-C. Villani2,5, R. Satija2, J. Maller1,2, . The GTEx Project Consortium6, A. Regev5, N. Hacohen2,5, D. G. MacArthur1,2 1) Analytic and Translational Genetics Unit, Massachusetts General Hospital, Boston, MA; 2) Program in Medical and Population Genetics, Broad Institute of Harvard and MIT, Cambridge, MA; 3) New York Genome Center, New York, NY; 4) Columbia University, New York, USA, NY; 5) Center for Immunology and Inflammatory Diseases, Massachusetts General Hospital, Charlestown, MA; 6) The Genotype-Tissue Expression (GTEx) Project Consortium.

   X chromosome inactivation (XCI) effectively balances the expression dosage of X chromosome genes between men and women by randomly silencing one of the two X chromosomes in each female cell. However, XCI is incomplete and variable; more than 15% of X chromosome genes have been reported to fully or partially escape from inactivation, but the full extent of XCI and its variation between tissues and individuals remains unclear. We have deployed several complementary approaches based on high-throughput RNA sequencing to comprehensively profile the landscape and variability of escape from XCI. Using detailed gene expression data from the Geuvadis and GTEx consortia we show that approximately half of the reported escape genes demonstrate male/female expression differences detectable at population-level. For these genes sex-biased expression is present and directionally similar across the various tissues studied, suggesting XCI is tightly and uniformly regulated across human tissues. Our analysis also highlights several novel candidate escape genes following similar sex-bias patterns. In line with earlier studies, we observe that even in genes with some degree of XCI escape the expression from the inactive copy of X is rarely as high as from the active X, potentially explaining why not all reported escape genes demonstrate large male/female expression differences. To confirm these observations and assess individual-level variability in escape from XCI we have analyzed high-throughput single-cell RNA-seq data from 192 cells from an exome-sequenced female sample, allowing the direct determination of expression from the inactive and active X chromosomes. In addition, we have assessed the allelic imbalance across the X chromosome in monoclonal tissue samples and in tissue samples showing skewed X inactivation. These analyses highlight well-known escape genes, replicate several of our novel candidates, and also confidently flag several additional candidate XCI escape genes with only modest sex-bias in the population-level analysis, hence extending the number of genes with variable escape and underscoring the large degree of inter-individual variability in X inactivation. Together these analyses provide a comprehensive view of the landscape of escape from XCI, essential for deeper understanding on how the process and escape genes contribute to sexual dimorphism and sex chromosome aneuploidies.