Human TSIX- an evolutionary relic?
- Cathy Holding
© BioMed Central Ltd 2001
Received: 6 December 2001
Published: 20 December 2001
Analysis of the human TSIX transcript shows it to be a truncated and probably non-functional evolutionary relic
Significance and context
X-inactivation in female mammals is the means by which they compensate for having an additional X chromosome relative to males. The mechanisms of silencing extra X chromosomes are not known, but current thinking is that an RNA transcript originates from the XIST (X-inactive specific transcript) gene located in the X-inactivation center of one X chromosome, and this results in the silencing of that chromosome - unless its expression is blocked. In mice, this block has been suggested to be provided by the antisense RNA transcript of Xist - namely Tsix - which originates at a point downstream of the Xist locus and extends all the way back over the gene to cover the promoter.
Transcripts from the human XIST region were studied by inserting concatamerized human XIST genomic regions into mouse embryonic stem cells and observing transcription directly in these cells and also in the chimeric mice derived from them. Migeon et al. found a human transcript that could be the counterpart of the mouse Tsix, but with crucial differences. Analysis of its sequence has thrown up some fascinating insights into the mechanisms of X-inactivation and specifically into the action of the mouse antisense Tsix transcript. Briefly, only one antisense transcript from around the human XIST region was detected, arising from a point downstream of the XIST gene and extending upstream but not as far as its murine counterpart, and not covering the promoter region of the human XIST gene. Furthermore, a GC-rich region is missing from the 5' end of this antisense gene, a region whose counterpart in the mouse is differentially methylated. Finally, a massive rearrangement of the region downstream of the XIST gene was revealed in humans compared to mice.
The major conclusions drawn by Migeon et al. are that a functional human TSIX transcript probably did exist once upon a time, but that some kind of major X-chromosome rearrangement event(s) severely truncated the 5' end of the antisense gene. This resulted directly or indirectly in the loss of the imprintable CpG island, while also breaking up another postulated gene, the human version of the mouse testis-specific factor Tsx, into several pieces and scattering them around the X chromosome. Migeon et al. suggest that the human TSIX gene cannot get imprinted, that it is expressed after the X-inactivation event, and that its transcript does not interact with the XIST promoter, so that it may be a nonfunctioning relic. They argue that, together with other data, this provides evidence for the existence of other as-yet undiscovered factors involved in X-inactivation in humans.
The paper includes an excellent and highly readable discussion, with a full overview of the most recent work to date. As far as the implications for mouse X-inactivation are concerned, my reading of this paper is that mouse Tsix may reinforce non-random X-inactivation in some of its tissues by the differential imprinting of a CpG island in exon 2 of the Tsix gene, and that it suppresses Xist gene expression by interacting directly with the promoter - but it is not sufficient on its own to have a role in random X-inactivation.