X-chromosome inactivation (XCI) in female mammals allows dosage compensation for X-linked gene products between the sexes1. The developmental regulation of this process has been extensively investigated in mice, where the X chromosome of paternal origin (Xp) is silenced during early embryogenesis owing to imprinted expression of the regulatory RNA, Xist (X-inactive specific transcript). Paternal XCI is reversed in the inner cell mass of the blastocyst and random XCI subsequently occurs in epiblast cells. Here we show that other eutherian mammals have very different strategies for initiating XCI. In rabbits and humans, the Xist homologue is not subject to imprinting and XCI begins later than in mice. Furthermore,Xist is upregulated on both X chromosomes in a high proportion of rabbit andhuman embryo cells, even in the inner cell mass. In rabbits, this triggers XCI on both X chromosomes in some cells. In humans, chromosome-wide XCI has not initiated even by the blastocyst stage, despite the upregulation of XIST. The choice of which X chromosome will finally become inactive thus occurs downstream of Xist upregulation in both rabbits and humans, unlike in mice. Our study demonstrates the remarkable diversity in XCI regulation and highlights differences between mammals in their requirement for dosage compensation during early embryogenesis

In somatic cells, the tetraspanins CD81 and CD9 associate with each other, with additional tetraspanins and with non-tetraspanin molecules to form proteolipidic complexes. Here we show that CD81 is expressed on the surface of oocytes where it associates with tetraspanin-enriched membrane structures. A major CD9 and CD81 partner, CD9P-1, is also expressed by oocytes. Deletion of CD81 gene in mice results in a 40% reduction of female fertility. In vitro insemination indicated that this infertility is due to a deficiency of oocytes to fuse with sperm. While the fertility of CD9-/- mice is severely but not completely impaired, double knock-out CD9-/- CD81-/- mice were completely infertile indicating that CD9 and CD81 play complementary roles in sperm-egg fusion. Finally, a fraction of CD9 was transferred from CD81-/- oocytes to sperm present in the perivitelline space indicating that the defect of fusion of CD81-/- oocytes does not result from an impaired initial gamete interaction.