Specific patterns of H3K79 methylation influence genetic interaction of oncogenes in AML.

Understanding mechanisms of cooperation between oncogenes is critical for the development of novel therapies and rational combinations. Acute myeloid leukemia (AML) cells with KMT2A-fusions and KMT2A partial tandem duplications (KMT2APTD) are known to depend on the histone methyltransferase DOT1L, which methylates histone 3 lysine 79 (H3K79). About 30% of KMT2APTD AMLs carry mutations in IDH1/2 (mIDH1/2). Previous studies showed that 2-hydroxyglutarate produced by mIDH1/2 increases H3K79 methylation, and mIDH1/2 patient samples are sensitive to DOT1L inhibition. Together, these findings suggested that stabilization or increases in H3K79 methylation associated with IDH mutations support the proliferation of leukemias dependent on this mark. However, we found that mIDH1/2 and KMT2A alterations failed to cooperate in an experimental model. Instead, mIDH1/2 and 2-hydroxyglutarate exert toxic effects, specifically on KMT2A-rearranged AML cells (fusions/partial tandem duplications). Mechanistically, we uncover an epigenetic barrier to efficient cooperation; mIDH1/2 expression is associated with high global histone 3 lysine 79 dimethylation (H3K79me2) levels, whereas global H3K79me2 is obligate low in KMT2A-rearranged AML. Increasing H3K79me2 levels, specifically in KMT2A-rearrangement leukemias, resulted in transcriptional downregulation of KMT2A target genes and impaired leukemia cell growth. Our study details a complex genetic and epigenetic interaction of 2 classes of oncogenes, IDH1/2 mutations and KMT2A rearrangements, that is unexpected based on the high percentage of IDH mutations in KMT2APTD AML. KMT2A rearrangements are associated with a trend toward lower response rates to mIDH1/2 inhibitors. The substantial adaptation that has to occur for 2 initially counteracting mutations to be tolerated within the same leukemic cell may provide at least a partial explanation for this observation.

Epigenetic regulation of protein translation in KMT2A-rearranged AML.

Inhibition of the H3K79 histone methyltransferase DOT1L has exhibited encouraging preclinical and early clinical activity in KMT2A (MLL)-rearranged leukemia, supporting the development of combinatorial therapies. Here, we investigated two novel combinations: dual inhibition of the histone methyltransferases DOT1L and EZH2, and the combination with a protein synthesis inhibitor. EZH2 is the catalytic subunit in the polycomb repressive complex 2 (PRC2), and inhibition of EZH2 has been reported to have preclinical activity in KMT2A-r leukemia. When combined with DOT1L inhibition, however, we observed both synergistic and antagonistic effects. Interestingly, antagonistic effects were not due to PRC2-mediated de-repression of HOXA9. HOXA cluster genes are key canonical targets of both KMT2A and the PRC2 complex. The independence of the HOXA cluster from PRC2 repression in KMT2A-r leukemia thus affords important insights into leukemia biology. Further studies revealed that EZH2 inhibition counteracted the effect of DOT1L inhibition on ribosomal gene expression. We thus identified a previously unrecognized role of DOT1L in regulating protein production. Decreased translation was one of the earliest effects measurable after DOT1L inhibition and specific to KMT2A-rearranged cell lines. H3K79me2 chromatin immunoprecipitation sequencing patterns over ribosomal genes were similar to those of the canonical KMT2A-fusion target genes in primary AML patient samples. The effects of DOT1L inhibition on ribosomal gene expression prompted us to evaluate the combination of EPZ5676 with a protein translation inhibitor. EPZ5676 was synergistic with the protein translation inhibitor homoharringtonine (omacetaxine), supporting further preclinical/clinical development of this combination. In summary, we discovered a novel epigenetic regulation of a metabolic process-protein synthesis-that plays a role in leukemogenesis and affords a combinatorial therapeutic opportunity.

Transient potential receptor melastatin-2 (Trpm2) does not influence murine MLL-AF9-driven AML leukemogenesis or in vitro response to chemotherapy.

Transient potential receptor melastatin-2 (TRPM2) is a nonselective cationic, Ca(2+)-permeable transmembrane pore that is preferentially expressed in cells of the myeloid lineage and modulates signaling pathways converging into NF-kB. This is of potential interest for acute myeloid leukemia (AML) therapy, as NF-κB signaling is emerging as a key pathway, mediating drug resistance and leukemia-initiating cell survival in AML. Inhibition of NF-κB signaling has been found to be synergistic with chemotherapy. TRPM2 is overexpressed in AML compared with normal bone marrow, with the highest levels in the FAB M3-6 subtypes. To determine the effect of TRPM2 depletions in a defined genetic model, we established MLL-AF9-driven AML on a Trpm2(-/-) genetic background. Trpm2(-/-) MLL-AF9 leukemias displayed reduced NF-κB phosphorylation as well as nuclear translocation. In vivo, primary and secondary recipients of Trpm2(-/-) MLL-AF9 leukemias exhibit increased latency compared with recipients of wild-type leukemia cells. However, the difference in latency was small and was lost in tertiary transplants. The effect of loss of Trpm2 in a BCR-ABL/NUP98-HOXA9 fusion model was even smaller. Given reports that loss or inhibition of TRPM2 enhanced killing by DNA-damaging agents in neuroblastoma, breast cancer, and prostate cancer cell lines, we exposed Trpm2(-/-) and Trpm2(wt) primary MLL-AF9 leukemias to doxorubicin, cytarabine, and etoposide, but found no difference in IC50 values. The in vitro response to decitabine was also unaffected. In summary, Trpm2 does not seem to play a major role in myeloid leukemogenesis. Additionally, loss of Trpm2 does not augment the cytotoxicity of standard AML chemotherapeutic agents.