Synergistic Efficacy of Dual Menin and PRMT5 Targeting Against NPM1 mutated and KMT2A-rearranged Leukemia

Menin inhibitors (Men-i) were developed as a treatment for KMT2A-rearranged (KMT2A-r) leukemias, where they inhibit the interaction of the adaptor protein Menin (encoded by the MEN1 gene) with an oncogenic KMT2A-fusion protein. In previous work, we discovered that NPM1 mutated (NPM1mut) AML is similarly dependent on the interaction of Menin with the wildtype histone methyltransferase KMT2A, and Menin inhibition causes suppression of a MEIS1 and HOX-driven leukemic gene expression and induces differentiation (Kühn et al. Cancer Discov. 2016). Although various novel Men-i demonstrated dramatic efficacy against NPM1mut or KMT2A-r leukemias in clinical trials, questions remain concerning the exact mechanism of action. Expressly, it is incompletely understood how the Menin-KMT2A chromatin complex drives leukemic gene expression and what other protein domains and complex binding proteins may contribute to leukemogenesis.

To assess the functional relevance of protein domains in MEN1 and KMT2A and to identify new therapeutic opportunities in NPM1mut AML, we employed a tiling CRISPR/Cas9 domain scan using a library of 2,041 sgRNAs targeting the MEN1 and KMT2A coding sequence every eight nucleotides on average. We identified several protein domains that represent strong dependencies in NPM1mut OCI-AML3 cells, including the CXXC but not the SET domain in KMT2A. Some domains were mapped to putative binding regions of predicted interaction partners of Menin or KMT2A. This included PRMT5 an arginine methyltransferase, which was reported to bind Menin in endocrine tissues and was implicated in the pathogenesis of myeloid neoplasms. As CRISPR/Cas9 directed PRMT5 knockout was lethal to the OCI-AML3 cells and we had found combined Menin and PRMT5 inhibition (PRMT5-i) to be synergistic in a previous drug screen, we sought to combine these drugs to enhance anti-leukemic activity. First, we assessed the PRMT5-i JNJ-64619178 (JNJ) and GSK3326595 (GSK) as single drugs. Both were efficacious in various AML cell lines, particularly in NPM1mut and KMT2A-r cells with IC50 values below two and 100nM, respectively. There was significant drug synergy of both PRMT5-i with each of two novel Men-i VTP50469 (VTP) and KO-539, while we found only moderate growth inhibition and no synergy in HL60 cells (wildtype for KMT2A and NPM1) that served as negative control. Next, we assessed combined Men-i and PRMT5-i on non-genetic Men-i resistant NPM1mut AML cells developed from single-cell clones exposed to increasing VTP concentrations that lack any known Men-i resistance mutations. PRMT5-i (JNJ and GSK) partially re-sensitized these resistant cells to Men-i.

To explore the effects of combined Men-i and PRMT5-i on leukemogenic gene expression, we performed RNAseq upon combination, single drug, or vehicle control treatment. These findings were integrated with genome-wide Menin and KMT2A binding determined by ChIPseq. While we observed dramatic transcriptional suppression of Menin/KMT2A binding targets, including MEIS1 and PBX3, upon single Men-i, PRMT5-i did not enhance this effect. However, E2F target genes were strongly downregulated upon PRMT5-i treatment, as previously reported in JAK2 mutated myeloid neoplasms, pointing to suppression of complementary pathways as a potential mechanism of drug synergy. Combination treatment also resulted in increased expression of interferon response genes compared to single drugs or vehicle control and resulted in significantly enhanced apoptosis and differentiation, as evidenced by Annexin V and CD11b staining.

To assess the in vivo efficacy of PRMT5-i, we treated NPM1mut AML patient-derived xenograft mice for 17 days with GSK and observed a significantly improved survival of the treated animals compared to the drug vehicle control. Next, we assessed the Men-i and PRMT5-i combination in an aggressive cell line-derived NPM1mut AML xenograft model. Four weeks of in vivo treatment significantly reduced leukemia burden in peripheral blood and bone marrow and significantly enhanced mice survival compared to single drugs and vehicle control.

The presented data underscore the therapeutic potential of combined Menin and PRMT5 inhibition as a novel synergistic approach to targeting NPM1mut and KMT2A-r leukemias and is already available for clinical investigation.