LAUSR

the top half being extraembryonic ectoderm-trophoblast (ExE), the bottom epiblast. The ExE secretes proteases (Furin), which diffuse into the epiblast rim to activate Nodal, resulting in gastrulation. In most other mammals, the pregastrulation embryo resembles a hockey puck, with a disc-shaped epiblast overlain by a single layer of trophoblast cells termed Rauber’s layer (RL). RL is likely homologous to the ExE, as both derive from polar trophectoderm. Yet, before gastrulation, RL disappears. We propose that RL is involved in the induction of gastrulation and for this very reason has to disappear in mammals having a disc-shaped epiblast: The unusual shape of the mouse embryo results in the tip of the epiblast being sufficiently far from the ExE to avoid receiving gastrulation signals. A similar signalfree central area can only be obtained in disc-shaped epiblasts if most of RL disappears, thereby confining Furin-positive trophoblast to the edges of the disc. This idea would imply that if RL did not disappear, the epiblast would be exposed to excessive signalling. We tested this using cattle embryos. FURIN was indeed expressed in RL. We determined that RL disappears by apoptosis and not a reduction in cell proliferation and thus made transgenic cattle embryos overexpressing the antiapoptotic protein BCL2. BCL2-transgenes maintained RL for significantly longer than control transgenic embryos. When examining gastrulation-stage BCL2-overexpressing embryos for expression of the gastrulation marker BRACHYURY, we found evidence for an extended domain of gastrulation induction as well as double axis formation. This is reminiscent of mice, where ectopic Nodal activation leads to axis duplication. So, just as players competing in Rugby and Hockey have to adjust to the different trajectile shapes, so did evolution adapt to speciesspecific pregastrulation embryo shapes, while continuing to use the same molecular players.