Login / Signup

A protein assembly mediates Xist localization and gene silencing.

Amy Pandya-JonesYolanda MarkakiJacques SerizayTsotne ChitiashviliWalter R Mancia LeonAndrey DamianovConstantinos ChronisBernadett PappChun-Kan ChenRobin McKeeXiao-Jun WangAnthony ChauShan SabriHeinrich LeonhardtSika ZhengMitchell GuttmanDouglas L BlackKathrin Plath
Published in: Nature (2020)
Nuclear compartments have diverse roles in regulating gene expression, yet the molecular forces and components that drive compartment formation remain largely unclear1. The long non-coding RNA Xist establishes an intra-chromosomal compartment by localizing at a high concentration in a territory spatially close to its transcription locus2 and binding diverse proteins3-5 to achieve X-chromosome inactivation (XCI)6,7. The XCI process therefore serves as a paradigm for understanding how RNA-mediated recruitment of various proteins induces a functional compartment. The properties of the inactive X (Xi)-compartment are known to change over time, because after initial Xist spreading and transcriptional shutoff a state is reached in which gene silencing remains stable even if Xist is turned off8. Here we show that the Xist RNA-binding proteins PTBP19, MATR310, TDP-4311 and CELF112 assemble on the multivalent E-repeat element of Xist7 and, via self-aggregation and heterotypic protein-protein interactions, form a condensate1 in the Xi. This condensate is required for gene silencing and for the anchoring of Xist to the Xi territory, and can be sustained in the absence of Xist. Notably, these E-repeat-binding proteins become essential coincident with transition to the Xist-independent XCI phase8, indicating that the condensate seeded by the E-repeat underlies the developmental switch from Xist-dependence to Xist-independence. Taken together, our data show that Xist forms the Xi compartment by seeding a heteromeric condensate that consists of ubiquitous RNA-binding proteins, revealing an unanticipated mechanism for heritable gene silencing.
Keyphrases
  • gene expression
  • long non coding rna
  • poor prognosis
  • machine learning
  • small molecule
  • electronic health record
  • transcription factor
  • oxidative stress
  • genome wide
  • deep learning
  • heat shock