The identification of G9A/EHMT2 reveals a novel interaction between epigenetics and metabolism in T-cell acute lymphoblastic leukemia.

Introduction: The identification of leukemia epigenetic mechanisms has fueled the search of drugs that target enzymes involved in DNA methylation and post-translational histone modification. Although, the epigenetic-metabolomics interplay that sustains cancer cell proliferation, pluripotency and resistance to therapy are far to be dissected. Here we characterize G9a/EHMT2, a protein lysine methyltransferases, as an epigenetic and metabolic target in T-cell acute lymphoblastic leukemia (TALL). Methods: To identify new targets in T-ALL we intersected an epigenome-centered shRNA and a low throughput small molecules screen of epigenetic modifiers with the computational analysis of EHMT2 expression in publically available cancer and normal tissue databases. G9a/EHMT2 emerged as “druggable” target in T-ALL. To demonstrate that G9a is required for T-ALL growth we performed both chemical and genetic studies and described the phenotypic consequences of G9a loss. Morphological appearance of cells treated with G9a inhibitors guided functional and biochemical experiments to justify the role of G9a in the regulation of glycogen metabolism. Results: We previously demonstrated that G9a/GLP inhibitors significantly inhibited cell viability compared to other epigenetic modifiers including histone acetyltransferase (HAT), p300 modulators, EZH2 and DOT1L inhibitors. G9a/EHMT2 and GLP/EHMT1 are conserved protein lysine methyltransferases that localize in euchromatin regions and regulates gene expression and chromosome structure through de novo monoand dimethylation of histone H3 lysine 9. Furthermore, the intersection of publicly available databases demonstrated that EHMT2 is a preferential marker of lymphoid differentiation and that is highly expressed in TALL compared to other cancer subtypes or healthy bone marrow cells. Moreover, an epigenome-centered shRNA screen and low-throughput sgRNA CRISPR validation studies confirmed that T-ALL proliferation depends on G9a expression. We showed that inhibition of G9a impairs T-ALL viability and triggersthe formation of cytoplasmic vacuoles, lysosomes and apoptotic bodies as demonstrated by transmission electron microscopy (TEM). These autophagic vacuoles resulted positive to the Periodic Acid-Schiff reaction and together with the detection of electrodense bodies observed in TEM strongly supported the evidence of intracellular glycogen granules accumulation. Consequently, we speculated that G9a loss might alter glycogen metabolism. Thus, we measured the activity of glycogen synthase kinase-3 (GSK3) by western blotting. We observed G9a inhibition causes an increase of the phospho-Ser9/21 GSK3 (inhibitory sites) upon drug treatment, hence promoting glycogen synthesis. Conclusions: The intersection of multiple chemical-genetic-computational approaches led to the identification of G9a/EHMT2 as a target in T-ALL and revealed a new epigenetic control of glycogen metabolism in this aggressive subset of leukemia