The hematopoietic system in the embryo develops in anatomically distinct sites, facilitating rapid generation of erythroid cells and formation of a pool of pluripotent HSCs. The origin of definitive HSCs is not fully resolved, and little is known about how the different fetal hematopoietic microenvironments direct the genesis, maturation, expansion and differentiation of HSCs. In avians, de novo hematopoiesis occurs not only in the yolk sac and the AGM but also in another mesodermal appendage, the allantois. In mammals, the allantois forms the umbilical cord and fetal placenta upon fusion with the chorion. The placenta has not been recognized as a hematopoietic organ, although Melchers reported fetal B-cell potential in murine placenta 25 years ago (Nature 1979, 277:219). Recently, Alvarez-Silva et al. showed that the placenta is a rich source for multipotential hematopoietic progenitors prior to the fetal liver (


We have performed spatial and temporal analysis of HSCs during mouse development and for the first time assessed HSC activity in the placenta. Hematopoietic organs from E10.5-18.5 embryos (CD45.1/CD45.2) were treated with collagenase and transplanted in limiting dilutions (3–1/1000 embryo equivalents, ee) into irradiated CD45.2+ adult hosts with CD45.1+ support BM cells. Reconstitution was analyzed by FACS and HSCs were quantified as repopulating units (RUs/ee = ([reconstituted recipients] /[total recipients]) /[transplanted dose]).

Our data show that the placenta functions as a hematopoietic organ that during midgestation harbors a large pool of pluripotent HSCs. The onset of HSC activity in the placenta parallels that of the AGM starting at E10.5–11.0. However, the placenta HSC pool expands until E12.5–13.5 (>50 RUs) contrasting lack of HSC expansion in the AGM. The expansion of CD34+c-kit+ HSCs in the placenta occurs prior to and during the initial expansion of HSCs in the fetal liver and is not accompanied with myeloerythroid differentiation. A far greater expansion of placenta HSCs compared to that of clonogenic progenitors (17-fold vs. 2-fold at E11.5–12.5) suggests that the placenta provides a favorable niche for HSCs. Indeed, placenta HSCs possess functional properties of authentic adult-type HSCs by providing high level multilineage reconstitution for >5 months and exhibiting self-renewal capacity upon serial transplantation. Importantly, placenta HSCs are distinct from circulating HSCs that appear in low numbers after E11.5. HSC activity in the placenta declines towards the end of gestation while HSCs in the fetal liver and blood continue to increase, possibly reflecting mobilization of placenta HSCs to the fetal liver and other developing hematopoietic organs.

The early onset of HSC activity in the placenta suggests that the allantois and its derivatives may participate in de novo genesis and maturation of HSCs together with the AGM and possibly the yolk sac. As the main blood volume from the dorsal aorta reaches the fetal liver via umbilical vessels and the placenta, placenta may also provide a niche where nascent HSCs, or pre-HSCs, from the AGM colonize for maturation and expansion prior to seeding fetal liver. While further studies are needed to define the precise origin of placenta HSCs and the function of placenta microenvironment as an HSC supportive niche, the unique kinetics and magnitude of HSC activity suggest an important, previously unappreciated role for the placenta in establishing the definitive hematopoietic system.

Author notes

Corresponding author