Xist imprinting is promoted by the hemizygous (unpaired) state in the male germ line

Abstract

This contribution is part of the special series of Inaugural Articles by members of the National Academy of Sciences elected in 2015. Contributed by Jeannie T. Lee, October 1, 2015 (sent for review July 29, 2015; reviewed by Nora Engel and Kathrin Plath)The long noncoding X-inactivation-specific transcript (Xist gene) is responsible for mammalian X-chromosome dosage compensation between the sexes, the process by which one of the two X chromosomes is inactivated in the female soma. Xist is essential for both the random and imprinted forms of X-chromosome inactivation. In the imprinted form, Xist is paternally marked to be expressed in female embryos. To investigate the mechanism of Xist imprinting, we introduce Xist transgenes (Tg) into the male germ line. Although ectopic high-level Xist expression on autosomes can be compatible with viability, transgenic animals demonstrate reduced fitness, subfertility, defective meiotic pairing, and other germ-cell abnormalities. In the progeny, paternal-specific expression is recapitulated by the 200-kb Xist Tg. However, Xist imprinting occurs efficiently only when it is in an unpaired or unpartnered state during male meiosis. When transmitted from a hemizygous father (+/Tg), the Xist Tg demonstrates paternalspecific expression in the early embryo. When transmitted by a homozygous father (Tg/Tg), the Tg fails to show imprinted expression. Thus, Xist imprinting is directed by sequences within a 200-kb X-linked region, and the hemizygous (unpaired) state of the Xist region promotes its imprinting in the male germ line.Xist | imprinting | X inactivation | transgenerational inheritance | meiotic silencing by unpaired DNA

Developmental regulation of X-chromosome inactivation

“…A similar study using single--copy transgenes has previously concluded that this is evidence that the imposed imprint is exclusively a maternal, repressive one and that Xist is expressed by default when inherited paternally without the need of a paternal imprint [36]. In the more recent study [44], however, in which multicopy transgenes were used, a different conclusion was drawn, which implied both a maternal and a paternal imprint (Fig. 1).…”

“…However, it seems that H3K9me3 is not the primary imprinting mark, as there is no difference in H3K9me3 at the Xist promoter before and after oocyte growth, when the imprint is established [43]. On the other hand, the Xist locus appears more condensed after imprint establishment during oocyte growth, which could influence Xist repression on the Xm, although the opposite has been postulated for Xist expression from the Xp (see below [44] Another critical regulator with a function in imprinted XCI is Tsix, the noncoding antisense regulator of Xist. Tsix is expressed from the Xm during imprinted XCI and mouse embryos with a Tsix mutation on the Xm express Xist from both the paternal and maternal X--chromosomes in extraembryonic tissues and die during early postimplantation development [45,46].…”

“…1C) come from a recent study using transgenic experiments [44]. Transgenes, containing the Xic region including Xist, Jpx, Tsix and Xite, have been inserted on autosomes and showed correct imprinted Xist expression -silent Xist when inherited through the maternal and active Xist when through the paternal germline.…”

“…However also the imprints on Xist --the repressive imprint on the Xm [39] and the facilitating one on the Xp [44] need to be erased, in order to allow for random choice in the next round of XCI. Not all XCR events necessarily need to be coupled or take place at the same time.…”

“…Despite the gain in knowledge we have about epigenetic marks detected in the preimplantation embryo, we still don't know, which type of marks are established in the germline and transmitted through the gametes. Interestingly, both during establishment of the paternal permissive, as well as the maternal repressive imprint on Xist, the compaction state of the Xic might play a role [43,44]. How these states correlate with potential key epigenetic imprints remains to be determined and finding the imprints will also allow establishing, how long into development these imprints persist and at which stage they are erased.…”