Inducible Gata1 Suppression As a Novel Strategy to Expand Physiologic Megakaryocyte Production from Embryonic Stem Cells

Embryonic stem (ES) and induced pluripotent stem (iPS) cells represent potential sources of megakaryocytes and platelets for transfusion therapy. However, most current ES/iPS cell differentiation protocols are limited by low yields of hematopoietic progeny, including platelet-releasing megakaryocytes. Mutations in the mouse and human genes encoding transcription factor GATA1 cause accumulation of proliferating, developmentally arrested megakaryocytes. Previously, we reported that in vitro differentiation of Gata1-null murine ES cells generated self-renewing hematopoietic progenitors termed G1ME cells that differentiated into erythroblasts and megakaryocytes upon restoration of Gata1 cDNA by retroviral transfer. However, terminal maturation of Gata1-rescued megakaryocytes was aberrant with immature morphology and no proplatelet formation, presumably due to non-physiological expression of GATA1. We now engineered wild type (WT) murine ES cells that express doxycycline (dox)-regulated Gata1 short hairpin (sh) RNAs to develop a strategy for Gata1-blockade that upon its release, restores physiologic GATA1 expression during megakaryopoiesis. In vitro hematopoietic differentiation of control scramble shRNA-expressing ES cells with dox and thrombopoietin (TPO) produced megakaryocytes that underwent senescence after 7 days. Under similar differentiation conditions, Gata1 shRNA-expressing ES cells produced immature hematopoietic progenitors, termed G1ME2 cells, which replicated continuously for more than 40 days, resulting in ~10¹³-fold expansion (N=4 separate experiments). Upon dox withdrawal with multi-lineage cytokines present (EPO, TPO, SCF, GMCSF and IL3), endogenous GATA1 expression was restored to G1ME2 cells followed by differentiation into erythroblasts and megakaryocytes, but no myeloid cells. In clonal methylcellulose assays, dox-deprived G1ME2 cells produced a mixture of erythroid, megakaryocytic and erythro-megakaryocytic colonies. In liquid culture with TPO alone, dox-deprived G1ME2 cells formed mature megakaryocytes in 5-6 days, as determined by morphology, ultrastructure, acetylcholinesterase staining, upregulated megakaryocytic gene expression (Vwf, Pf4, Gp1ba, Selp, Ppbp), CD42b surface expression, increased DNA ploidy and proplatelet production. Compared to G1ME cells rescued with Gata1 cDNA retrovirus, dox-deprived G1ME2 cells exhibited more robust megakaryocytic maturation, similar to that of megakaryocytes produced from cultured fetal liver. Importantly, G1ME2 cell-derived megakaryocytes generated proplatelets in vitro and functional platelets in vivo (~40 platelets/megakaryocyte with a circulating half life of 5-6 hours). These platelets were actively incorporated into growing arteriolar thrombi at sites of laser injury and subsequently expressed the platelet activation marker p-selectin (N=3-4 separate experiments). Our findings indicate that precise timing and magnitude of a transcription factor is required for proper terminal hematopoiesis. We illustrate this principle using a novel, readily reproducible strategy to expand ES cell-derived megakaryocyte-erythroid progenitors and direct their differentiation into megakaryocytes and then into functional platelets in clinically relevant numbers.