Dysregulated Expression of MicroRNA-16 Contributes to Abnormal Erythropoiesis in Patients with Polycythemia Vera

Discovery of JAK2V617F mutation in myeloproliferative neoplasms (MPN) raised intriguing question on how a single mutation might underlie different clinical phenotypes. Micro-RNAs are important regulators of gene expression, intervening in normal and neoplastic hematopoiesis. An abnormal expression profile of miRNAs in primary myelofibrosis granulocytes (Guglielmelli P, Exp Hematol 2007) and in polycythemia vera (PV) erythroid cells has been described (Bruchova H, Exp Hematol 2007; Haematologica 2008); therefore, dysregulation of miRNAs might contribute to MPN pathogenesis. In a global analysis of miRNA expression in CD34⁺ cells from PV pts we found abnormally increased levels of miR-16; of note, defects in miR-16 causing reduced expression have been implicated in the pathogenesis of chronic lymphocytic leukemia (CLL). There are two miR-16 genes, miR-16-1 (in a cluster with miR-15a on chr 13, mutated/deleted in CLL) and miR-16-2 (in a cluster with miR-15b on chr 3), which differ each to the other in their pre-miR sequence while the mature sequence is conserved. Using stem-looped primer technology and real-time PCR, we determined that expression levels of miR-16 in CD34⁺ cells from 15 PV pts were 10 to 30-fold higher than in CD34⁺ cells from controls or reactive erythrocytosis (P<.0001); the increase was specific for miR-16 since levels of mature miR-15a and miR-15b were unchanged. miR-16 was also slightly increased in PV granulocytes (about 2-fold vs controls; P=.026). Levels of miR-16 were independent of the JAK2V671F allele burden (r=0.4). Sequencing the entire pre- and mature miR-16-1/ miR-15a and miR-16-2/miR15b regions disclosed no abnormality, in contrast with CLL for miR-16-1. We also determined gene copy number by RT-PCR, but found no evidence of miR-16-1/2 copy number changes. Since the mature sequence is shared by the two miR-16, we designed a RT-PCR approach to determine the levels of their respective pre-microRNAs. We found that levels of pre-miR-16-2 were significantly increased in PV CD34⁺ cells compared to controls, while pre-miR-16-1 did not show appreciable changes; there was also a significant correlation between the levels of pre-miR16-2 (unlike premiR- 16-1) and those of total mature miR-16 in PV CD34⁺ cells (r=0.7, P<.01). The preferential contribution of miR-16-2 to abnormally high total miR-16 levels in PV CD34⁺ cells was supported by knock-down experiments using siRNAs specific for either miR-16-1 or miR-16-2. In fact, while in normal CD34⁺ cells the two siRNAs produced a comparable reduction of total miR16 levels, in PV cells only the miR-16-2 siRNA significantly reduced the abnormally increased mature miR-16. It has been previously shown that miR-16 increases in latest stages of normal erythropoiesis in vitro; we determined the kinetics of miR-16 during cultures of CD34⁺ cells induced to differentiate along erythroid lineage from normal subjects and PV pts. In both control and PV cells there was a prompt down-regulation of miR-16 soon after starting culture; thereafter, miR-16 levels started to increase progressively from day 6 reaching a maximum on day 12–14; at any time point considered, the miR-16 levels measured in PV cells were significantly greater than controls. To address the function of miR-16 in erythroid differentiation, we over-expressed (Amaxa technology) mature miR16 in normal CD34⁺ cells both at the beginning of culture and at day 6, when progenitors were switched from a proliferative (in the presence of IL-3, IL-6, SCF) to a differentiative (plus Epo) culture phase. No significant modification compared to mock controls was observed in cells transfected at day 0, while in those transfected at day 6 the percentage of CD36⁺/GPA⁺ cells in the presence of EPO at least doubled; notably, also in the absence of EPO GPA+ cells, virtually absence in mocktreated cultures, were generated. Finally, preliminary evidences indicated that knockingdown miR-16-2 in PV CD34⁺ cells resulted in significant reduction of erythroid colony formation. Overall, these data indicate dysregulated expression of miR-16 in CD34⁺ and in in-vitro generated erythroid cells from PV patients, and point to miR-16-2 as being specifically involved; we also provided evidence for a role of miR-16 in normal erythroid differentiation and for its possible involvement in the abnormal erythropoiesis of PV.