Deranged MicroRNA 16-2 Expression Contributes to Erythropoiesis in Polycythemia Vera.

Abstract 3896

Poster Board III-832

We previously reported the finding of significantly raised levels of mature microRNA 16 (miR-16) in CD34⁺ cells from PV patients (pts) (Guglielmelli P et al, ASH 2008, 179A). Mature miR-16 can derive from both miR-16-1 at chr13q14.3 and miR-16-2 at chr3q26.1, which differ in their precursor miRNA (pre-miR) sequence. To distinguish between either genes as a source of increased miR-16 levels we quantified each pre-miR in PV CD34⁺; we observed that the premiR-16-1/16-2 ratio decreased from 1.47± 2.38 in controls (ctr) to 0.52± 0.34 in PV (P<.001). To confirm the preferential expression of miR-16-2 we knocked-down miR-16-1 or miR-16-2 using specific siRNAs; in PV CD34⁺ treated with miR16-2 siRNA the miR-16 levels were 90±9% lower than in scramble-treated cells while miR16-1 siRNA caused only a 10±3% reduction (P <.01). Levels of miR-16 in PV CD34⁺ were not correlated to the JAK2V671F burden; also, there was no modification of miR-16 in JAK2V617F mutant HEL or UKE-1 cells after treatment with the JAK2 inhibitor AZD1480. These data suggested that the abnormally increased mature miR-16 in PV are largely derived from miR-16-2. To address underlying mechanisms we performed direct sequencing at the miR-16-1 and miR-16-2 regions without detecting sequence abnormalities; also we excluded gene copy number changes (SNP 6.0 array). The promoter of miR16-2 is not characterized yet, but recent data suggest that it might be the same of the structural maintenance of chromosomes-4 (SMC4) gene, to which miR16-2 is intronic. We have found a significant correlation between miR-16 and SMC4 mRNA levels in PV CD34⁺ (R=0.77, P<.001); furthermore, after induced erythroid differentiation of normal CD34⁺ a concurrent increase of miR-16 and SMC4 mRNA was observed, indirectly supporting a shared regulatory control. Methylation-specific PCR and sequencing at upstream CpG-rich regions discovered that PV patients had unmethylated CpGs compared to control cells, suggesting that methylation may concur to miR-16-2 regulated expression. To evaluate involvement of miR16 in abnormal erythropoiesis of PV, we first analyzed the kinetics of miR-16 during erythroid differentiation of CD34⁺. Levels of miR-16 steadily increased from day 6, when progenitors were switched from proliferative to differentiative phase, up to day 12-14 and, at any time point considered, those measured in PV cells were significantly greater than in ctrs (P= .01). Also, a time-dependent increase of miR16 paralleling the transcription of beta-globin mRNA was observed in hydroxyurea or Na-butyrate treated K562 cells. Then we over-expressed (Amaxa) mature miR16 in normal CD34⁺ at day 6 of culture, and found that the percentage of CD36⁺/GPA⁺ cells increased from 13.6±4.8% in scramble to 28.8±11.3% (P<.05); notably, even in the absence of EPO, some GPA⁺ cells were generated from miR-16-transduced CD34⁺. Furthermore, normal CD34⁺ cells transfected with miR-16 produced significantly increased number of CFU-e and BFU-E (174±100 and 1283±250/10³ CD34+) compared to scramble cells (105±90 and 733±90; P<.01) while myeloid colonies were unchanged. A genome-wide expression profiling in normal CD34⁺ transfected with miR-16-2 siRNA discovered 618 genes significantly de-regulated, among which genes involved in erythroid differentiation such as NFE2, MYB, FGFR2, HLF, GATA-1 and globin family genes (HBA1, HBA2, HBB). Overall, these data support a role of miR-16 in normal erythropoiesis. In case of PV we employed a knock-down strategy using either pre-miR-16-1 or pre-miR-16-2 siRNAs. The use of miR-16-2 siRNA resulted in significant reduction of Epo-independent erythroid colonies (EEC) from GFP-sorted CD34⁺ cells (60±18% lower than scramble) while no change was observed in cell treated with anti-miR-16-1. These data were confirmed in independent experiments using peripheral blood mononuclear cells (MNC): the number of EEC decreased from 207±20 to 170±23 to 95±25/10⁵ MNC in cells transduced with scramble, anti miR-16-1 or miR-16-2, respectively (P<.01 for miR-16-2 vs others). Also the number of BFU-E generated in cytokine-supplemented cultures of PV MNC was significantly reduced by miR-16-2 siRNA (360±80) compared to cells transduced with scramble (540±120) or miR-16-1 (500±87) (P<.001). In summary, these data support a role for deregulated miR-16-2 in abnormal erythropoiesis of PV, and anticipate its possible relevance as a therapeutic target.