Identification Of Bipotential Lin-CD34+CD38- Integrin α2- Erythrocyte-Megakaryocyte Progenitors In Human Bone Marrow

Abstract Human hematopoietic progenitor cells with megakaryocyte and erythroid commitment have been defined as the Lin-CD34+CD38+CD123-CD45RA- and in addition, as CD110+ (Manz et al. 2002; Edvardsson et al. 2006). However, previous colony assays have shown that bipotential megakaryocyte-erythroid progenitors (MEPs) also reside in the stem cell-enriched CD34+CD38- fraction in bone marrow (BM) (Debili et al., 1996). So far, the phenotype of this MEP population has remained obscure. We have here studied the expression of integrin α2 chain in normal human bone marrow CD34+CD38- stem and progenitors. Integrin α2 chain was expressed in most BM CD34+CD38- cells (96.3±3.8; mean±SD), in contrast to a more restricted expression in cord blood CD34+CD38- cells (Wong, et al. 2013). In order to investigate the potential functional differences associated with the expression of integrin α2 on BM CD34+CD38- cells, we characterized the integrin α2+ and α2- cell populations by in vitro LTC-IC, single cell colony assays and gene expression analysis. The LTC-IC progenitors were exclusively integrin α2+. In contrast, the integrin α2- fraction was highly enriched in erythroid BFU-E progenitors. In single cell colony assay in methylcellulose, 21.96% of single Lin-CD34+CD38- integrin α2- cells gave rise to BFU-E, as compared to 0.65% of integrin α2+ cells. Likewise, the frequency of megakaryocyte colony forming cells, analyzed by a serum free collagen based culture system, was significantly higher in the integrin α2- fraction, suggesting an existence of the bipotential MEP in the integrin α2- fraction. To confirm the existence of bipotential MEPs in the Lin-CD34+CD38-CD45RA- integrin α2- cell population, we performed single cell clonogenic assay using serum free medium with cytokines (SCF, TPO, IL-3 and EPO) in Terasaki plates. At day 12-14 of culture 30-40% of the cells showed clonal growth. The erythroid and megakaryocytic differentiation of the single cell derived clones was assessed by their immunophenotype and morphology. Interestingly, FACS analysis of the 32 clones indicated that all of them contained erythroid cells (CD235+CD41-). Most importantly, 22 of the 32 clones contained both erythroid and megakaryocytic cells (CD235-CD41+), showing a high degree of bipotential MEP activity in the integrin α2- population. Minimal myeloid differentiation (CD15/CD33/CD66b+ cells <0.5%) was seen in only 3 of the 32 clones. During the differentiation the integrin α2 receptor was upregulated in CD41+ megakaryocytic cells but not in the CD235+ erythroid cells. Furthermore, 83 clones derived from single cells, including clones with very limited number of cells, were individually transferred to cytospin slides for evaluation of their morphology. The small clones invariably consisted of megakaryocytic cells. 48.19% of all clones analyzed were identified as bipotent MEPs. This finding was further supported by upregulation of HBD and downregulation of GATA3, HLF in the integrin α2- cells, revealed by gene expression analysis. Taken together, our data demonstrate a high frequency of bipotential MEP progenitors in adult BM Lin-CD34+CD38-CD45RA- integrin α2- cell fraction and suggest that BM Lin-CD34+CD38- integrin α2- cells are transcriptionally primed towards development into erythroid and megakaryocytic cells. The identification of this novel bipotential MEP population provides a means for further analysis on lineage fate decisions of primitive erythroid and megakaryocytic cells and may facilitate studies aiming to expand and differentiate these lineages for clinical transplantation and transfusion trials. Disclosures: Ekblom: Novartis and BMS: Honoraria.