Methylation of Progesterone Receptor Promoter-Associated CpG Island in Polycythemia Vera and Related Myeloproliferative Disorders.

Polycythemia vera (PV), essential thrombocythemia (ET) and idiopathic myelofibrosis (MF) are clonal myeloproliferative disorders (MPD). A recently discovered activating mutation of JAK2 tyrosine kinase has been found in most patients with polycythemia vera (PV), in about half of those with essential thrombocythemia (ET) and myelofibrosis (MF), and in 10–20% patients with chronic myelomonocytic leukemia, Philadelphia-negative CML, atypical or unclassified MPD and megakaryocytic leukemia. It is not known what other factors determine the disease phenotype of PV, MF, and other MPD, and what factors other than JAK2 lead to disease progression. Very little is known about epigenetic changes in PV. DNA methylation of promoter-associated CpG islands is a well-recognized mechanism of epigenetic silencing used by tumors for evasion from regulatory mechanisms, and it is an alternative to genetic lesions in cancer causation. Using a genome-wide screen for differentially methylated CpG islands, we found methylation of progesterone receptor promoter region (PGR) in PV granulocytes. We then developed pyrosequencing assays for quantitative detection of PGR methylation in bisulfite-treated PCR-amplified DNA. The PGR methylation above normal control levels was observed in ET (2/12 patients, 17%), PV (10/22 patients, 45%), MF (8/12 patients, 67%), and patients with acute myeloid leukemia and antecedent PV (6/7 patients, 86%). We compared the levels of PGR methylation in MPD with the mutation status of JAK2. The 1849G>T JAK2 mutation was present in 16/27 (59%) MPD patients with unmethylated PGR and 21/26 (80%) patients with methylated PGR; the difference not statistically significant; p=0.135. The role of progesterone receptor signaling in hematopoiesis is not known. Using real time quantitative RT-PCR assay for progesterone receptor expression we found detectable levels in granulocytes from 4/5 normal individuals while the expression in granulocytes from 5/5 PV patients was not detectable. To assess the functional significance of progesterone receptor silencing, we explored the effect of mifepristone, a progesterone receptor antagonist, on the response of BFU-E progenitors to erythropoietin. Mifepristone increased the sensitivity of BFU-E progenitors from normal blood to low concentrations of erythropoietin (60–250 mU/ml) suggesting that disabling of progesterone receptor may increase the response of hematopoietic cells to proliferative stimuli. In conclusion, our data show that PGR methylation is present in half of PV patients and it is even more frequent in MF and PV transformed to AML. Silencing of progesterone receptor by methylation may be an epigenetic change contributing to MPD phenotype and transformation to leukemia.