High PRMT1 Expression Promotes the Progression of Acute Megakaryocytic Leukemia Via Controlling Metabolic Pathways

Metabolic reprogramming is needed not only to accommodate but also to drive leukemia progression. Yet very little is known on genetic factors other than IDH1 mutations, which can drive leukemogenesis via metabolic reprogramming. Here, we will present data to suggest that protein arginine methyltransferases 1 (PRMT1) is a driver for acute megakaryocytic leukemia via reprogramming metabolism. PRMT1 is highly expressed in megakaryocyte-erythrocyte progenitor cells and downregulated during the terminal differentiation of megakaryocytes. Constitutively high expression of PRMT1 in acute megakaryoblastic leukemia (AMKL) blocks megakaryocyte differentiation. PRMT1 upregulates RBM15 protein level via methylation-dependent ubiquitylation pathway (Zheng et al. Elife, 2015). In this presentation, we discovered that metabolic stress such as hypoxia downregulates PRMT1 protein level. Thus, metabolic stress is the upstream signal for the PRMT1-RBM15 pathway. We have identified that RBM15 specifically binds to 3'UTR of mRNAs of genes involved in metabolic pathways. Using RNA-immunoprecipitation with anti-RBM15 antibody and real-time PCR assays, we validated that RBM15 binds to mRNAs of genes involved in fatty acid oxidation and glycolysis. We transduced PRMT1 into an RBM15-MKL1 expressing cell line 6133. Overexpression of PRMT1 renders 6133 cells to grow in a cytokine-independent manner with increased mitochondria biogenesis, which in turn produces higher concentration of ATP in our metabolomic analysis. Based on the analysis of metabolomics data and RBM15-target genes, we conclude that PRMT1 promotes the usage of glucose as bioenergy via oxidative phosphorylation and inhibits fatty acid oxidation. Given that acetyl-coA is higher in PRMT1 expressing 6133 cells, we asked whether histone acetylation is upregulated in PRMT1 overexpressed 6133 cells. Indeed, we found higher histone acetylation level in PRMT1 highly expressed cells. We also found that propionylated histone is reduced, which is consistent with reduced fatty acid oxidation. Propionyl-CoA molecules are produced from fatty acids with odd carbon numbers. Thus PRMT1-mediated metabolic reprogramming changes epigenetic programming during leukemia progression. Intriguing, we also found PRMT1 overexpression enhances histone H3S10 phosphorylation via methylation-dependent ubiquitylation of DUSP4. DUSP4 promotes polyploidy during megakaryocyte differentiation. Thus PRMT1 caused profound epigenetic changes to promote leukemogenesis. In this vein, we established mouse AMKL models by bone marrow transplantation of 6133 cells as well as human AMKL patient samples respectively. Using this mouse model, we tested PRMT1 inhibitors, acetyltransferase inhibitors as well as other metabolic inhibitors. Treating cells with PRMT1 inhibitors as well as metabolic inhibitors promote MK differentiation of AMKL leukemia cells. Metabolomics analysis of cells recovered from mouse models will be discussed in the presentation. In summary, our data demonstrated that PRMT1 is a major sensor for metabolic stress and that PRMT1 in turn reprograms metabolic pathways to bring epigenetic changes in leukemogenesis. Therefore, targeting PRMT1 and downstream PRMT1-regulated metabolic pathways will offer new avenues in treating acute megakaryocytic leukemia and other hematological malignancies with defective megakaryocyte differentiation.