Functionally Distinct Cell Populations Characterize Cord Blood Derived Megakaryocytes

A fundamental understanding of platelet biology requires functional assessment of megakaryocyte-platelet expressed genes and transcripts. The need for gene functional studies has recently increased with unbiased genome wide associations with platelet phenotypes. Animal models with an over-expressed or knocked-out gene have been particularly valuable systems, although species differences may preclude optimal evaluation. Assessing protein function in human platelets can be accomplished with pharmacologic inhibitors, although inhibitor availability and specificity can plague this approach. Studies with primary human megakaryocytes (MKs) are greatly challenged by inaccessibility, low yields and contamination with other cells. Cultured human megakaryocyte-like cell lines (e.g., Meg-01, HEL and others) have proven valuable, but these cells poorly recapitulate platelet physiology. Increasingly, investigators are using cultured human MKs to study gene function. CD34+ hematopoietic stem cells from human cord blood are routinely used to generate megakaryocytes in culture. This system has advantages of being human, relatively easy to obtain, cultured from primary cells and able to be genetically manipulated. Although cord blood-derived megakaryocytes (CBDMs) have proven useful for studying aspects of megakaryocyte/platelet biology, there is little information about the heterogeneous cell populations observed in CBDMs. In the course of our studies with CBDMs we observed multiple distinct cell populations by flow cytometry. Similar populations were also observed in the megakaryocytic Meg-01 cell line. The largest cells (called P1) were the most abundant (consisting of nearly 100% of cells) at day 3 in culture. P2 cells, which are smaller and more granular than P1, appeared at day 6 and by day 13 were ~50% of the total. P3 appeared at day 6 and are the smallest, with size and granularity roughly similar to platelets; by day 13 these were ~30% of the total. P1 but not P2 nor P3 became CD61/CD41/CD42 positive and CD34 negative over 13 days in culture. Ninety-seven percent and 93% of P2 and P3 cells, respectively, were phosphatidyl serine (PS) positive, whereas 93% of P1 cells were PS negative. Electron microscopy revealed many typical features of bone marrow MKs in the PS negative (P1) cells, including large size, polyploid nucleus, mitochondria and immature granules. However, the demarcation membrane system was poorly developed. Virtually all of the PS positive (P2) cells were apoptotic, lacked granules and had no discernable nuclei. Purification of P1 and P2 populations followed by re-culturing revealed that P1 gives rise to both the P2 and P3 populations, whereas P2 gave rise to no other population. CD34+ cells cultured with the pan-caspase inhibitor Z-VAD-FMK developed a much smaller proportion of P2 cells with a corresponding increase in P1 cells. Stimulation of these 3 populations with collagen related peptide, thrombin, protease activated receptor 1-activating peptide (PAR1-AP) and PAR4-AP showed strong integrin activation in P1 cells, but not in P2 nor P3 cells. To assess which population was susceptible to genetic manipulation, day 3 cultures were infected with a lentiviral construct containing a shRNA to talin. Talin knockdown prevented agonist-induced integrin activation in P1 cells and as expected, had no effect on the P2 or P3 population. In summary, these data indicate that only a portion of CBDMs are functional and only the P1 population should be used to assess MK gene function. Over time, the majority of CBDMs become apoptotic and the smaller P3, platelet-like particles have minimal response to agonist compared to peripheral blood platelets. These results may partially explain the difficulties in generating functional platelets in vitro and are consistent with the need to restrain apoptosis for successful megakaryocytopoiesis.