Dynamicdosage compensation of transposable elementsin mammals

In mammals, X-linked dosage compensation involves two processes: X-inactivation to balance X chromosome dosage between males and females, and X-hyperactivation to achieve X-autosome balance. Transposable elements (TEs) are repetitive mobile DNA sequences and major sources of X chromosome dosage. Current studies in X-linked dosage compensation mainly focus on genes, with TEs still remaining elusive. Here we use “So-Smart-Seq” to comprehensively determine the allelic dynamics of TEs in the context of imprinted and random X-chromosome inactivation (XCI). We find that chromosomal loci and genetic backgrounds significantly impact TE silencing, and reveal putative roles of SINEs in shaping 3D chromatin architectures. We show remarkable difference in TE silencing between two forms of XCI and discover that X-hyperactivation is absent in TEs. Our data provide deep insight for TE studies and greatly contribute to complete understanding of X-chromosome wide dosage compensation. Overall design: For the third-party data reanalysis, 188 GSM samples (GSM7980593-7980780) from GSE168455 were reanalyzed. Allelic expression of transposable elements during imprinted XCI was extracted from data GSE168455 that reveal full allelic transcriptome of single preimplantation embryos at various stages from WT and Xist mutant mouse. The allelic ratio between two X chromosome was calculated for each TE family. To acquire allelic expression dynamics of transposable elements in random XCI, allelic bulk RNAseq was performed in differentiating TST ES cells, a mus/cas hybrid ES cell line where X chromosome from mus allele is progressively silenced during random XCI. Expression of transposable elements were determined and the allelic ratio between two X chromosome was calculated for each TE family. *************************************************************** The table below lists GEO accessions reused/reanalyzed for this study. ***************************************************************