Abstract
The transcription factor NF-E2 is overexpressed in the vast majority of patients with Myeloproliferative Neoplasms (MPN). In Essential Thrombocythemia (ET) and Primary Myelofibrosis (PMF) NF-E2 levels are elevated independent of the presence or absence of the JAK2V617Fmutation. We have recently shown that NF-E2 overexpression in a murine model leads to an MPN phenotype followed by spontaneous transformation to acute leukemia in a subset of mice.
We have demonstrated that increased NF-E2 transcription is mediated by the transcription factor and proto-oncogene AML-1, which is overexpressed in MPN patients, again irrespective of the JAK2V617Fstatus. AML-1 binds to a conserved enhancer region located 3500 bp upstream of the NF-E2 transcriptional start site. Since AML-1 responsive genes are often also inversely regulated by the tumor suppressor AML-2 (RUNX3), we investigated whether NF-E2 expression is affected by AML-2 as well.
Using chomatin immunoprecipitation assays (ChIP) we show that AML-2 binds the NF-E2 enhancer in vivo at a site distinct from but in proximity to the three AML-1 sites in the -3500bp region. AML-2 binding strongly represses transcription off the NF-E2 enhancer in reporter gene assays. This repression is abrogated by site directed mutagenesis of the AML-2 recognition sequence. Likewise, lentivirally induced AML-2 expression drastically reduces the amount of NF-E2 protein in erythroid and myeloid cells. These data clearly demonstrate that NF-E2 expression is directly regulated by AML-2. AML-2 thus serves as a repressor on the hematopoietic NF-E2 gene, a function previously noted mainly in solid tumors.
In primary cells from patients with polycythemia vera (PV) AML-2 mRNA expression is significantly reduced. Moreover, ChIP assays demonstrate that in primary PV cells, significantly less AML-2 is bound to the NF-E2 enhancer than in healthy controls. Decreased repression by AML-2 thus cooperates with increased AML-1 induced transcription to elevate NF-E2 levels in PV patients.
It has been demonstrated that AML-2 expression can be regulated by two distinct epigenetic mechanisms. For one, DNA methylation of the AML-2 promoter has been reported to silence AML-2 expression in gastric, colorectal and bladder cancers. On the other hand, aberrant histone methylation in the promoter region can also silence AML-2 expression. Here we show that DNA methylation of the AML-2 promoter is unaltered in PV patients. Rather, PV patients display aberrant histone methylation in the AML-2 promoter. Compared to healthy controls, H3K27me3 is significantly increased and H3K4me3 is significantly decreased in primary PV cells. This results in an inactive chromatin conformation on the AML-2 promoter in PV patients. Moreover, we show here that the histone-lysine-methyl-transferase “enhancer of zeste homologue 2” (EZH2), which confers the K3K27me3 mark, binds to the AML-2 promoter and decreases AML-2 expression. PV patients demonstrate significantly increased levels of EZH2 binding to the AML-2 promoter
Treatment of primary PV cells and MPN cell lines with 2'-deoxy-5-azacytidine (DAC, Decitabine) ex vivo reverses the altered histone methylation, restoring the pattern found in healthy controls and decreases EZH2 binding to the AML-2 promoter. At the same time, Decitabine treatment induces AML-2 protein expression, decreases EZH2 expression and reduces the elevated NF-E2 levels. Moreover, a post-PV AML patient receiving Decitabine, displayed normalization of AML-2 promoter histone methylation and EZH2 binding on day 8 and day 15 of treatment.
Taken together our data demonstrate that epigenetic silencing of AML-2 contributes to the elevated NF-E2 expression observed in MPN patients. Treatment with Decitabine restores physiological histone modifications to the AML-2 locus and decreases NF-E2 levels by reactivating AML-2 expression. These data provide a molecular rationale for extending the clinical investigation of epigenetic modifiers such as Decitabine or Azacitidine, currently used in MDS and AML, to patients with MPNs. A phase I study using Azacitidine in high risk PMF patients is currently being planned by the MPD-RC.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.