Preclinically Effective Menin Inhibitor SNDX-50469 and SNDX-5613-Based Combinations Against MLL1-Rearranged (MLL-r) or NPM1-Mutant AML Models

MLL1 (KMT2A) is a transcriptional regulator and a histone-lysine-N-methyltransferase. N-terminus of MLL1 binds to the scaffold protein Menin via the menin binding domain (MBD). Menin-KMT2A complex regulates expression of the leukemogenic Homeobox A9 (HOXA9) and its co-factor MEIS1 in myeloid stem progenitor cells. In MLL1-rearranged (MLL-r) AML (~10% of AML), the N-terminus of MLL1 gene is fused to the C-terminus of any of over 80 fusion partners, including AF4, AF9, ENL and ELL (~70% of MLL-r AML), which are part of (and recruit) the super elongation complex (SEC) (including AFF1/4 and pTEFb) and DOT1L to induce H3K4Me3 and H3K79Me2 marks on active chromatin. MLL1 fusion protein (MLL-FP) dysregulates gene expression of HOXA genes and their co-factors MEIS1, PBX3, as well as of MEF2C and CDK6. In AML with mutant (mt) NPM1 (NPM1c), MLL1 is the main oncogenic driver of HOXA9, MEIS1 and FLT3, promoting self-renewal of myeloid progenitor cells. Conditional knockout of MEN1 prevents MLL-r or NPM1c AML. Treatment with orally bioavailable Menin inhibitor, SNDX-50469 or SNDX-5613, disrupts binding of Menin to its binding pocket in MLL1/2 and MLL1-FP, evicting Menin from the chromatin, reducing MLL1/2 and MLL1-FP binding to targets, as well as inducing differentiation and apoptosis in AML cells expressing MLL-FP or NPM1c. SNDX-50469 or SNDX-5613 treatment also exhibits impressive in vivo anti-AML efficacy in AML models expressing MLL-FP and NPM1c. In the phase I/II AUGMENT-101 clinical trial, SNDX-5613 monotherapy was well tolerated and achieved high rates of objective remissions in patients with previously treated relapsed/refractory AML harboring MLL-r or NPM1c. Focus of the present studies was to further elucidate epigenetic, gene-expression and anti-AML effects, as well as identify novel synergistic combinations of SNDX-50469 or SNDX-5613 with other targeted agents to achieve superior in vivo efficacy in AML expressing MLL-r or NPM1c. Utilizing ATAC and RNA-Seq in AML MOLM13 cells (expressing MLL-AF9 and FLT3-ITD), we first determined that exposure to SNDX-50469 simultaneously reduced ATAC peaks and mRNA expression of HOXA9, MEIS1, PBX3, FLT3, MEF2C and CDK6. Concurrent ChIP-Seq analysis demonstrated reduction in H3K27Ac peaks and reduced activity of super enhancers of MEIS1, PBX3, MEF2C, MYC and MYB. Single-cell RNA-Seq analysis of AML cells with MLL-AF9 and mtFLT3 also confirmed that SNDX-50469 treatment caused reduction in mRNA expression of MLL-FP targets in AML stem/progenitor cells, including HOXA9, LAMP5, PBX3, MEF2C, MEIS1, FLT3, MYB and CDK6. Consistent with this, Western analysis in MOLM13 and OCI-AML3 cells (NPM1c and homozygous NRAS mutation) following SNDX-50469 treatment showed markedly reduced protein levels of MEIS1, FLT3, CDK6, and BCL2, with increased levels of MCL1 and CD11b. Mass cytometry (CyTOF) analysis of patient-derived (PD) MLL-r or mtNPM1 AML samples also confirmed that SNDX-50469 treatment attenuated protein levels of Menin, MEIS1, MEF2C, PBX3, RUNX1, FLT3, CDK6, Bcl-xL and BCL2 in phenotypically characterized AML stem cells (with high expression of CLEC12A, CD123, CD244, and CD99, but low expression of CD11b). Notably, in vitro treatment of cell lines and PD AML cells from MLL-r or NPM1c AML with SNDX-50469 or SNDX-5613 in combination with venetoclax, or OTX015 (pan-BET protein inhibitor), or abemaciclib (CDK6 inhibitor) for 72 to 96 hours induced synergistic loss of viability, as determined by SynergyFinder. No loss of viability was observed in normal CD34+ progenitor cells or AML cells lacking MLL-FP or NPM1c. We next determined in vivo anti-leukemia efficacy of SNDX-5613 alone (50 to 75 mg/kg, PO, BID) or its co-treatment with venetoclax (30 mg/kg, PO, daily) or OTX015 (30 mg/kg, PO, daily) for 3 to 4 weeks in NSG mice engrafted with either MOLM13-GFP/Luciferase xenograft or with PD NPM1c and mtFLT3-GFP/Luciferase AML xenograft. Co-treatment of engrafted mice with SNDX-5613 and venetoclax or OTX015 caused significantly greater reduction in AML burden and increased overall survival without weight loss or other toxicities than treatment with each agent alone (p< 0.005). These preclinical findings highlight the molecular correlates of anti-AML efficacy of SNDX-50469 or SNDX-5613. They also demonstrate greater in vitro and in vivo efficacy of SNDX-5613-based combinations with inhibitors of BCL2, BET proteins and CDK6 against AML harboring MLL1-FP or NPM1c.