In this regard, the placentas associated with EPO-/- and EPO-R-/- knockout mice were reported to be of normal size at gestational Days 13-15 when the embryos died in utero apparently of severe anemia. However, it was not reported whether the morphology and distribution of the various populations of trophoblast cells were normal. An autocrine role for EPO, whereby the hormone contributes to the survival, proliferation, and differentiation of tro-phoblast cells, may be analogous to its role in early colony-forming  and erythroleukemic cells, as well as human hepatocellular carcinoma cells.
Placental trophoblast and other cells expressed both EPO and EPO-R with the only apparent exception being the fetoplacental vascular endothelium, which expressed the receptor, but not the ligand, as assessed by immunocytochem-istry; present study). In addition to putative autocrine roles, another potential action of trophoblast-derived EPO is that it might interact with EPO-R on endothelium to stimulate fetoplacental and uterine vascular development, since EPO has a capacity to promote angiogenesis. Although the expression of EPO has been recently described in the endodermal cells of the mouse vascular yolk sac, possibly trophoblast-derived EPO contributes to primitive erythropoiesis and vasculogenesis in the human yolk sac or promotes the differentiation of hemangioblasts identified in the villous core. Further, results of elegant studies by Dancis and colleagues have suggested that perhaps the trophoblast cell has the capability to respond to erythroid differentiation signals, since reconstitution of irradiated isogeneic mice with placental cells leads to the formation of hematopoietic colonies in the spleen. Clearly, this new and nonclassical site of EPO and EPO-R expression on cells composing the human placenta opens up the possibility of physiological roles for this hormone in addition to erythropoiesis.