Identifying ‘Druggable’ Targets and Preclinically Overcoming Non-Genetic/Adaptive Resistance to Menin Inhibitors in AML with MLL1r or mtNPM1

In AML with MLL1 rearrangement, the N-terminus of MLL1 becomes fused to partner proteins, resulting in an aberrant transcriptional program, including dysregulated expression of HOXA9, MEIS1, PBX3, FLT3, MEF2C and CDK6. In mutant (mt) NPM1 AML cells, wild-type MLL1 is the main regulator of HOXA9, MEIS1 and FLT3, promoting self-renewal of myeloid progenitor cells. Menin inhibitors (MI) disrupt the binding of Menin to MLL1, leading to reduced levels/activity of HOXA9 and MEIS1, as well as repression of MLL1 or MLL1 fusion protein (FP) target genes. Clinically, MIs are well-tolerated, and induce remissions in patients with AML harboring MLL1r. However, most patients either fail to respond to MI treatment or eventually relapse. MI-resistance in AML with MLL1r or mtNPM1 is either due to mutations in the MI-binding domain of Menin or is non-genetic/adaptive, due to a dysregulated epigenome/transcriptome/proteome. To probe the mechanisms of resistance to MIs, in the present studies, through repeated shocks with LD90-dose of MI, followed by drug washout and recovery, we generated MI-resistant, MLL1r (MV4-11-MITR) and mtNPM1 (OCI-AML3-MITR) AML cell lines that are tolerant/resistant (TR) to MIs. As compared to the parental (LD50 values of 51.4 nM and 55 nM), MV4-11-MITR and OCI-AML3-MITR cells exhibited LD50 to MI SNDX-50469 of >1 µM. These MITR cells exhibited cross-resistance to other MIs, including ziftomenib and DS1594b. Custom TaqMan assays confirmed that these MITR cells did not display any hotspot mutations in Menin described in patient-derived (PD) AML cells exhibiting MI-resistance. Whole exome sequencing also did not show any new driver mutations that could potentially contribute to the MITR. In contrast, ChIP-Seq analysis of the chromatin mark, H3K27Ac, which highlights active enhancers and promoters, demonstrated that the genes involved in the core transcriptional regulatory circuitry were altered in the MITR versus parental cells. In OCI-AML3-MITR cells, genome-wide ATAC-Seq and RNA-Seq analyses showed concordant reduction in the ATAC-Seq peaks and mRNA expressions, with significant depletion of MEIS1, IGF2BP2, BCL2, and PBX3, but induction of leukemia stem cell-associated CLEC12A and CD244 expressions. In the MITR vs parental cells, mass spectrometry identified MI-resistance induced protein expression-perturbations, including 166 commonly upregulated and 197 commonly downregulated proteins. A domain-specific CRISPR screen in MV4-11-MITR cells identified several ‘druggable’ co-dependencies with MI-treatment, including the histone acetyltransferases EP300 and MOZ, and co-enrichments with MI-treatment, including SMARCA4, CREBBP and BRD4, suggesting these as the potentially ‘druggable’ targets for improving MI activity. Accordingly, co-treatment with FHD-286, the inhibitor of BRG1/BRM (BAF chromatin-remodeler complex ATPases) or the BET inhibitor (BETi), OTX015, combined with SNDX-50469 exerted synergistic lethality in the parental, as well as in the MV4-11-MITR and OCI-AML3-MITR cells. We also compared these MI-based combinations against patient-derived (PD), 6 MI-resistant (MITR) and 4 MI-sensitive AML cell samples, exhibiting MLL1r or mtNPM1 but lacking any MITR-associated Menin mutations. Co-treatment with SNDX-50469 and the HATs GNE049 or GNE781 was synergistically lethal in the MITR cells. Importantly, in NSG mice engrafted with luciferized OCI-AML3-MITR cells, the in vivo co-treatment for 8-weeks with FHD-286 (1.5 mg/kg, QD x 5 days) overcame resistance to the MI SNDX-5613 (50 mg/kg, BID, x 5 days) in the OCI-AML3-MITR xenograft. Moreover, as compared to treatment with each agent alone, co-treatment with FHD-286 and the BETi OTX015 (30 mg/kg, QD x 5 days) significantly reduced OCI-AML3-MITR AML burden and improved survival of the NSG mice (p < 0.05). Notably, in a luciferized MLL1r MITR AML PDX model, compared to treatment with each agent alone, co-treatment with FHD-286 and SNDX-5613 or OTX015 for 8-weeks yielded significantly superior survival of the NSG mice (p < 0.05). These findings demonstrate that co-treatment with FHD-286 overcomes in vitro and in vivo MI-resistance, while significantly improving in vivo efficacy of BETi in the cell-line xenograft and PDX models of MITR AML cells. They also show that combinations of epigenetically targeted agents may be effective in MI-sensitive or -resistant AML with MLL1r or mtNPM1.