Complete Myeloid Potential Arises First in the Mammalian Yolk Sac

In the mouse embryo, hematopoietic function is required by E10.5 (embryonic day 10.5) before adult-repopulating hematopoietic stem cells (HSC) exist. The earliest erythroid function is provided by a wave of primitive erythroid progenitors that arise at E7.5, in association with megakaryocyte and macrophage progenitors. Intriguingly, a second wave of hematopoietic potential arises between the first primitive hematopoietic wave and functional HSC formation. This second progenitor wave also forms in the yolk sac but is distinguished from the primitive wave by its slightly later onset (E8.25), generation of definitive erythroid cells, and its additional association with granulocyte and mast cell progenitors. The proposed function of these “wave 2” progenitors is to colonize the newly formed fetal liver (beginning at E10) and differentiate into the first mature definitive erythroid cells observed in circulation at E12. However, it is unclear how much definitive hematopoiesis arising in the yolk sac recapitulates the paradigm of later HSC-derived myeloid potential, progenitor hierarchy and immunophenotype. To investigate this question, we examined markers of adult myeloid progenitor maturation in the yolk sac and early fetal liver. As previously described by others, all definitive hematopoietic progenitors in the yolk sac, unlike those in the bone marrow, express CD41, which we found associated with Fc gamma receptor expression (FcγR, CD16/32) beginning at E8.5. By E9.5, definitive hematopoietic progenitors can be identified by their surface co-expression of ckit, CD41, FcγR, as well as endoglin. When cultured in vitro, these cells can differentiate into all myeloid lineages, including neutrophils, eosinophils, basophils and mast cells as identified by morphology, immunophenotype and gene expression. Preliminary clonal analysis confirms that a common erythroid/granulocyte progenitor exists in this population. Consistent with adult myelopoiesis, we found a qualitative association of higher FcγR expression with granulocyte fate and higher endoglin expression associated with erythroid fate. However, the unusual co-expression of these four markers and the prevalence of erythroid fate, even in FcγR^hi cells, suggest the definitive hematopoietic progenitors in the yolk sac may be quite plastic and highly predisposed to an erythroid fate. Consistent with the concept that these “wave 2” progenitors colonize the fetal liver, we also found similar ckit⁺CD41⁺FcγR⁺endoglin⁺ cells in the early liver (E11.5) with the potential to produce a variety of myeloid cells when cultured in vitro. The emergence of enucleated definitive erythrocytes by E12, within 24 hours HSC fetal liver colonization, implies that these first erythrocytes are derived from the yolk sac definitive progenitors found in the liver by E10.5. We therefore asked whether the multiple myeloid potentials associated with “wave 2” progenitors are similarily realized the early fetal liver. Beginning at E11.5 we found a population of Gr1⁺Mac1⁺ cells in the liver with morphological and histological characteristics of neutrophils, and increase 100-fold in number between E12.5 and E14.5. In contrast, we did not observe eosinophils, basophils or mast cells by morphology, immunophenotype or by RT-PCR for lineage-specific messages. We conclude that complete definitive myeloid potential first arises in the yolk sac from a unique population of ckit⁺FcγR⁺CD41⁺endoglin⁺ progenitors. Our data suggest that these progenitors then enter the fetal liver and differentiate into a subset of their potential fates producing the first mature definitive erythromyeloid cells. This second wave of hematopoietic progenitors emerging from the yolk sac thus serves as a novel model of mulitpotential definitive hematopoiesis.