Dose-Dependent Iron Chelating Effects of Eltrombopag on in Vitro Human Megakaryopoiesis

Eltrombopag (ELT), a small molecular thrombopoietin (TPO) mimetic approved for children and adults, could offer a therapeutic alternative to selected neonates and young children with chronic thrombocytopenias. ELT has also been proposed as a potential anti-cancer drug due to its anti-proliferative effects in tumor cells, which are mediated by its strong iron chelating properties. This raises the potential concern that rapidly proliferating normal cells, like bone marrow cells in neonates or young infants, could also be susceptible to the anti-proliferative effects of ELT. In this study, we first assessed the responses of neonatal (cord blood, CB) and adult (peripheral blood, PB) megakaryocyte (MK) progenitors to increasing concentrations of ELT, TPO and Romiplostim (ROM) in vitro. Consistent with the previously described pattern of neonatal megakaryopoiesis, CB progenitors generated 10-times more MKs, which were of lower ploidy but had higher CD42 surface expression levels than PB-MKs. Unlike TPO or ROM, ELT exhibited dose-dependent opposing effects on in vitro megakaryopoiesis: Low concentrations (≤6µM) stimulated megakaryopoiesis, but high concentrations (30µM) suppressed MK differentiation and proliferation, as evidenced by a severe reduction in the percentage and absolute number of CD41+ cells after 7 days of culture. The toxic effects of high ELT concentrations were not abrogated by the addition of TPO at concentrations achieved in hyporegenerative thrombocytopenias (3 ng/mL), and were reproduced by the addition of the iron chelators deferoxamine (DFO) or deferiprone (DFP) at concentrations of 100µM to MK cultures, suggesting iron deficiency as a potential mechanism. To further assess the iron chelating effects of ELT in human MKs, we used the calcein assay. In K562 cells, as well as in CB MKs, ELT concentrations >10µM reduced the labile intracellular iron pool (LIP) to lower levels than the iron chelator DFP at a concentration of 200 µM, indicating the strength of ELT's iron chelating properties. These concentrations of ELT also resulted in a severe reduction of mitochondrial transmembrane potential, detected by the JC-1 assay, decreased proliferation (EdU click assay) and apoptosis (Annexin-V binding). During MK differentiation, committed MK progenitors (CD34+/CD41+ cells) were the cell population most sensitive to the antiproliferative and apoptotic effects of high-dose ELT. Next, we evaluated whether the iron status of differentiated MK progenitors would influence their responses to ELT. To test this, we generated iron-depleted and iron-repleted CB-MK progenitors by culturing CD34+ cells for 7 days in the presence of apo-transferrin (iron-free) or holo-transferrin (iron-saturated), followed by a 72 hour culture with different concentrations of ELT. These studies revealed a strong iron status/ELT dose interaction (interaction p<0.0001); In iron-depleted cultures, 30 µM ELT and 100 µM DFO similarly impaired MK expansion and induced MK apoptosis, but these effects were attenuated in iron-repleted cultures. Interestingly, iron-depleted adult PB progenitors did not show reduced MK expansion or increased apoptosis in response to 30 µM ELT, suggesting a lower susceptibility of adult megakaryopoiesis to the iron chelating effects of ELT. Taken together, these findings suggest that iron status is an important variable affecting the response to ELT, particularly in neonates and children.