Cyclic AMP Signaling Inhibits Megakaryopoiesis by Targeting An E2A-CDKN1A Transcriptional Axis

Abstract 915

The regulation of megakaryopoiesis by cyclic AMP (cAMP) signaling forms the basis for treatment approaches to thrombocytosis in patients with myeloproliferative neoplasms. Anagrelide, a first line therapeutic, is an inhibitor of the cAMP catabolizing enzyme phosphodiesterase III, and lowers platelet counts by inhibiting early stages of megakaryocytic differentiation. The mechanism by which anagrelide inhibits megakaryocytic differentiation has never been fully elucidated. As shown previously in a human CD34+ cell culture model, elevations in cAMP induced by anagrelide as well as the adenylyl cyclase agonist forskolin impaired megakaryocytic differentiation in a protein kinase A (PKA) dependent manner (Rubinstein et al., 2010 ASH Abstract #289). In those experiments, forskolin induced a 50% decrease in expression of the bHLH transcription factor E2A while other members of the E-protein subfamily were unaffected. Additionally, elevated cAMP had no effect on E2A expression in CD34+ derived erythroid cultures. The current study focuses on the functional significance of megakaryocyte-specific E2A downregulation by cAMP signaling. Firstly, to explore the whether E2A levels affect megakaryopoiesis, we utilized an shRNA knockdown approach in human CD34+ cells. Knocking down E2A expression strongly impaired megakaryocytic differentiation, as reflected by CD41 upregulation and cellular enlargement. Conversely, retrovirally enforcing expression of E2A enhanced megakaryocytic differentiation and overrode the inhibitory effects of forskolin. These results using ex vivo culture of human progenitors suggest that the E2A downregulation caused by PKA signaling contributes to the inhibition of megakaryopoiesis by cAMP. In vivo studies showed E2A+/− mice to have normal platelet counts but abnormalities in marrow megakaryopoiesis consisting of decreases in polyploidization, cell size, and CD42 expression. In addition, E2A+/− mice displayed a striking 5-fold increase in the number of splenic CD41+ FSClo cells, possibly representing a compensatory response to the defective megakaryocytic maturation in the marrow. In addressing how E2A may regulate megakaryocytic differentiation, our studies focused on CDKN1A, encoding p21, as a potentially relevant target. CDKN1A is known to be a directly regulated by E2A in hematopoietic progenitors, is strongly upregulated during megakaryopoiesis, and has been implicated in polyploidization. In human CD34+ cells, forskolin treatment completely blocked the CDKN1A upregulation normally seen in late stage megakaryocytic differentiation (day 5). A role for E2A in megakaryocytic regulation of CDKN1A was supported by studies using shRNA knockdown and retrovirally enforced expression. Knockdown of E2A decreased p21 expression, whereas enforced E2A expression partially rescued p21 upregulation in the presence of forskolin. The kinetics of forskolin-mediated megakaryocytic inhibition, with rapid and reversible signaling events affecting delayed transcriptional events, suggested involvement of an epigenetic switch. In preliminary data supporting this notion, the DNA methylation inhibitor 5-aza-2'-deoxycytidine was able to reverse the cAMP induced impairments in both CD41 expression of p21 transcription. In summary, the cAMP induced inhibition of megakaryocytic differentiation in parts acts through the E2A-p21 regulatory axis. As E2A can regulate the epigenetic landscape within its target genes, we postulate that fluctuations in E2A expression following elevations in cAMP may reconfigure the epigenetic topology within critical loci such as CDKN1A and thereby disable the megakaryocytic program.