Several recent studies have demonstrated substantial biological differences between cord blood (CB)- and adult peripheral blood (PB)-derived megakaryocytes (MKs). Specifically, neonatal (CB) progenitors proliferate at a much higher rate than adult (PB) progenitors, and generate 10-fold more MKs per progenitor when cultured with thrombopoietin (Tpo). The highly proliferative neonatal MKs undergo cytoplasmic maturation without polyploidization, which ultimately generates large numbers of small, low ploidy, but fully mature MKs. Adult MKs, in contrast, undergo successive rounds of endomitosis to reach much higher ploidy levels, and maturation is coupled with polyploidization, so that MKs with the highest ploidy levels are also the most mature. The molecular mechanisms underlying these developmental differences are just beginning to be elucidated. Here, we investigated the effects of cortistatin A (CA), a highly specific small molecule inhibitor of cyclin-dependent kinase 8 (CDK8) and its paralog CDK19, on megakaryopoiesis. CDK8 and CDK19, together with CCNC (Cyclin C), MED12/MED12L and MED13/MED13L, form "CDK modules" which can associate with Mediator, a 26-subunit complex that acts as a bridge between transcription factors and the transcriptional machinery to coordinate gene expression. The Mediator complex has been implicated in developmental disorders and cancer. CB-derived MK progenitors treated with CA from day 7 to day 14 of culture exhibited a dose-dependent reduction in proliferation (6.2±1.7 vs. 24.9±2.2 fold expansion in treated vs. control cultures; p=0.003), accompanied by an increase in ploidy levels to those comparable to adult PB-derived MKs (34±6% vs. 8.9±0.9% MKs with ploidy ≥8N in treated vs. control cultures; p=0.014). MK maturation, evaluated by CD42b surface expression level, also increased with advancing ploidy in CA-treated MKs, in a manner similar to that observed in adult MKs, and CA-treated mature MKs were capable of pro-platelet formation in vitro. These changes were not observed when undifferentiated CB-derived CD34+ or CD41-negative cells were treated, indicating that the effect was specific to committed MK progenitors. CA treatment induced the expected decrease in STAT-1 phosphorylation at Serine 727, a specific site of CDK8-mediated phosphorylation, confirming effective CDK8 inhibition in treated MKs. Next, we used microarray to evaluate the gene expression profile of CB-derived MKs following CA treatment for 4 and 8 hours vs. untreated cells. These studies revealed significant CA-induced changes in the MK gene expression profile. By Gene Set Enrichment Analysis (GSEA), CA treatment significantly upregulated genes that were downregulated in CB- vs. PB-derived MKs. Furthermore, genes upregulated by CA in CB-derived MKs were also upregulated in the megakaryoblastic cell line SET2 (Pelish, Liau et al., Nature, in press), suggesting that the gene expression program affected in SET2 cells is likely the same one affected in neonatal MKs. At the protein level, we observed time- and dose-dependent increases in RUNX-1 in CA-treated vs. control MKs. In summary, this study demonstrated that treatment with CA induced a phenotypic switch from neonatal to adult-like megakaryopoiesis, accompanied by changes in the MK gene expression profile. These findings indicate a novel role for Mediator kinases in the regulation of megakaryopoiesis, and potentially on the developmental differences between neonatal and adult MKs. These studies open the door to a better understanding of and to potential novel therapies for a number of developmental stage-specific megakaryocyte and platelet disorders, which exclusively or more severely affect neonates and infants, including the thrombocytopenia/absent radius (TAR) syndrome and the transient myeloproliferative disorder associated with trisomy 21 and GATA1s mutations.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.