Random X-chromosome inactivation (XCI) is a dosage compensation process by which females transcriptionally silence one of their two X chromosomes to achieve the equal X-linked gene expression level with XY males. The X-inactivation center (Xic) is the X-linked sequence necessary and sufficient for counting, choice and cis -spreading of silencing. Xist, Tsix and Xite are three well-studied non-coding RNA genes within the Xic that function in cis. On the future active X (Xa), Xite activates Tsix, which in turn represses Xist. On the future inactive X (Xi), Xite and Tsix are turned off, which allows Xist accumulation. Following Xist -mediated gene silencing, the Xi is maintained in a stable heterochromatic state that is inherited in all subsequent cell divisions. Only one X chromosome remains active while the other becomes inactive. Therefore, the two Xs must be coordinated in trans to ensure mutually exclusive silencing. I hypothesized that interchromosomal pairing mediates this communication. To examine the potential spatial association of the two Xs, fluorescence in situ hybridization (FISH) and chromosome conformation capture (3C) were used in mouse embryonic stem (ES) cells at different phases of XCI. The two Xs indeed transiently interact in trans at the onset of XCI, but not in the maintenance phase. The pairing is specific to the Xic. Deleting Tsix or Xite perturbs pairing and counting/choice, while their autosomal insertion induces de novo X-autosome pairing. Ectopic X-autosome interactions inhibit endogenous X-X pairing and block the onset of XCI. Thus, I propose that Tsix and Xite regulate counting and mutually exclusive choice through X-X pairing.

In principle, X-X pairing can be regulated by both cis -acting DNA elements and trans -acting factors. Using transgenic analyses, the counting/choice/pairing elements have been mapped to a bipartite domain in Tsix and Xite. Discrete genetic elements flanking Tsix 's minimal promoter may individually be competent in inducing de novo X-autosome pairing and repressing X-X pairing/counting/choice. Because multiple CTCF and YY1 binding sites exist in pairing elements, the role of transcription factors was investigated by transient depletion through RNA interference (RNAi). CTCF but not YY1 is required specifically for initiation of pairing/initiation of XCI. Transcriptional inhibition experiment further implicates that X-X pairing may depend on active transcription. Taken together, I propose that CTCF and transcription-coupled factors play critical roles in the pairing process at Xic, which facilitates mutually exclusive choice based on the cooperative action of several functionally redundant genetic elements within Tsix and Xite.