Use of Mouse Embryonic Stem Cell Differentiation To Model Primitive Erythropoiesis In Vitro.

Primitive erythroid cells (EryP) are among the earliest differentiated cells to form in the mammalian embryo but their maturation and terminal differentiation remain poorly understood. Our studies of embryonic peripheral blood identified a stepwise morphological progression in EryP development from E9.5 to E12.5. During this time, EryP undergo synchronous changes in morphology accompanied by changes in surface antigen expression. Terminal maturation is marked by enucleation. By E12.5, adult definitive erythrocytes (EryD) begin to appear in the blood stream. EryD are small, enucleated cells. At this time, enucleated EryP are not easily distinguished from EryD. Therefore, we developed a transgenic (Tg) mouse model (epsilon-globin::GFP) in which green fluorescent protein (GFP) is expressed exclusively within EryP (Fraser et al. (2007) Blood 109: 343–352). This transgenic line made it possible to study primitive erythropoiesis at an unprecedented resolution. We next developed murine embryonic stem (ES) cell lines in which a GFP reporter is expressed exclusively within the nuclei of EryP by using a histone H2B fusion protein (H2B-EGFP) that labels DNA during all phases of the cell cycle. Six (6) e-globin::H2B-EGFP Tg ES cell lines were characterized for expression of the GFP reporter using fluorescence microscopy and FACS. Tg ES cell-derived embryoid bodies (EBs) were harvested at different times (days 2–10) to cover the period when we would expect EryP progenitors to form and differentiate. GFP+ cells were purified from these staged EBs by FACS sorting and cytospin preparations were stained with Giemsa to identify developmental landmarks. Three lines showed comparable peak expression (7–10% of all EB cells were GFP+); one of these was chosen for more detailed analysis. The kinetics of appearance of green fluorescence was roughly concurrent with development of red pigment (indicating hemoglobinization). Culture conditions were optimized for activation of the reporter during differentiation, as were conditions for FACS isolation of viable GFP+ EB-derived cells. We found that the ES cell-derived GFP+/EryP were strikingly similar morphologically to FACS-sorted EryP from E8.5 mouse embryos. Also as observed for EryP in the mouse embryo, several cell surface antigens were upregulated on GFP+/EryP during ES cell differentiation. On the basis of the morphology of Giemsa-stained FACS-purified GFP+/EryP, their apparently synchronous maturation, and upregulation of surface antigens that are also upregulated during EryP development in vivo, we conclude that the maturation of ES-derived EryP is very similar to that of EryP in the mouse embryo. The ability to scale up will be a major advantage of the ES cell system for future microarray analysis. For example, one 6 cm plate of day 8 EBs (from 100,000 ES plated) yielded 60,000 GFP+/EryP; in contrast, E8.5 embryos yielded approximately 3,000 GFP+/EryP per embryo (FACS-sorted cells from pooled embryos). Thus, one would need 120 E8.5 embryos to obtain the same number of GFP+/EryP as from one 6 cm plate of ES-derived EBs. A long-term goal of these studies will be to identify molecular signatures of progenitors versus more committed EryP and thereby gain insights into pathways and regulatory molecules involved in their development.