Abstract:In mammals, X-linked dosage compensation involves two processes: X-chromosome inactivation (XCI) to balance X chromosome dosage between males and females, and hyperactivation of the remaining X chromosome (Xa-hyperactivation) to achieve X-autosome balance in both sexes. Studies of both processes have largely focused on coding genes and have not accounted for transposable elements (TEs), which comprise 50% of the X-chromosome, despite TEs being suspected to have numerous epigenetic functions. This oversight is due in part to the technical challenge of capturing repeat RNAs, bioinformatically aligning them, and determining allelic origin. To overcome these challenges, here we develop a new bioinformatic pipeline tailored to repetitive elements with capability for allelic discrimination. We then apply the pipeline to our recent So-Smart-Seq analysis of single embryos to comprehensively interrogate whether X-linked TEs are subject to either XCI or Xa-hyperactivation. With regards to XCI, we observe significant differences in TE silencing in parentally driven "imprinted" XCI versus zygotically driven "random" XCI. Chromosomal positioning and genetic background impact TE silencing. We also find that SINEs may influence 3D organization during XCI. In contrast, TEs do not undergo Xa-hyperactivation. Thus, while coding genes are subject to both forms of dosage compensation, TEs participate only in Xi silencing. Evolutionary and functional implications are discussed.

Dosage compensation of transposable elements in mammals