Genomics, the diversification of mammals, and the evolution of placentation.

When and why did variations in placental structure and function evolve? Such questions cannot be addressed without a reliable version of mammalian phylogeny. Twenty-five years ago, the mammalian tree was reshaped by molecular phylogenetics. Soon it was shown, in contrast to prevailing theories, that the common ancestor of placental mammals had invasive placentation. Subsequently, evolution of many other features of extraembryonic membranes was addressed. This endeavour stimulated research to fill gaps in our knowledge of placental morphology. Last year the mammalian tree was again revised based on a large set of genomic data. With that in mind, this review provides an update on placentation in the nineteen orders of placental mammals, incorporating much recent data. The principal features such as shape, interdigitation, the interhaemal barrier and the yolk sac are summarized in synoptic tables. The evolution of placental traits and its timing is then explored by reference to the revised mammalian tree. Examples are the early appearance of epitheliochorial placentation in the common ancestor of artiodactyls, perissodactyls, pangolins and carnivores (with reversion to invasive forms in the latter) and later refinements such as the binucleate trophoblast cells and placentomes of ruminants. In primates, the intervillous space gradually evolved from the more basic labyrinth whereas trophoblast invasion of the decidua was a late development in humans and great apes. Only seldom can we glimpse the "why" of placental evolution. The best examples concern placental hormones, including some striking examples of convergent evolution such as the chorionic gonadotropins of primates and equids. In concluding, I review current ideas about what drives placental evolution and identify significant gaps in our knowledge of placentation, including several relevant to the evolution of placentation in primates.