Pre-Clinical Efficacy of Targeting Baf Complexes through Inhibition of the Dual Atpases BRG1 and BRM By FHD-286 in Cellular Models of AML of Diverse Genetic Background

ATP-dependent chromatin remodeling BAF (BRG1/BRM-associated factor) complexes enable transcription factors and co-factors to gain access to enhancer/promoter DNA and modulate transcription. BAF complexes include the core ATPase BRG1 (SMARCA4) or BRM (SMARCA2), which contain the catalytic ATPase domain and a bromodomain (BD). AML stem/progenitor cells express and are dependent on BRG1/BRM activity. Small molecule, BD-targeted inhibitors of the dual BRG1 and BRM ATPase activity have been developed. They repress BRG1/BRM-dependent gene-expressions. FHD-286 is a potent, selective, small molecule, oral inhibitor of BRG1/BRM in early clinical development as a therapy for AML. However, AML sub-type-specific efficacy of FHD-286 is unclear and FHD-286-induced gene-expression perturbations that correlate with its anti-AML efficacy have yet to be determined. Evaluation of the CRISPR-screen dependency map (DepMap) showed greater dependency of numerous AML cell lines on SMARCA4 expression. In the present studies, we probed the in vitro and in vivo efficacy of FHD-286 in inducing differentiation and loss of viability, as well as their molecular correlates in AML cell lines and patient-derived (PD) AML cells. Exposure to FHD-286 (10 to 100 nM) for 4 to 7 days overcame differentiation block and significantly induced CD11b expression and morphologic features of differentiation in AML cell lines with MLL-r, mtNPM1 and chromosome 3q26 lesions (with EVI1 overexpression). A similar exposure to FHD-286 induced loss of viability in the AML cell lines and PD AML cells. Following treatment with 100 nM FHD-286 for 16 hours, RNA-Seq analysis of MOLM13 cells demonstrated significant reduction in the normalized enrichment scores for expressions of gene-sets of targets of MYC, mTORC1, E2F, Interferon-gamma, IL6-JAK-STAT3, as well as of inflammatory response and oxidative phosphorylation genes. QPCR analyses determined significant reduction in mRNA expression of MYC, SPI1 and BCL2 genes. Western analyses showed that treatment with FHD-286 significantly increased p21, p27, PU.1 and CD11b expressions, while reducing expressions of c-Myc and BCL2. Based on these observations, and clinical efficacy of the combination of venetoclax and decitabine/azacitidine, we determined the in vitro lethal activity of co-treatment with FHD-286 and venetoclax or decitabine against AML cell lines and PD AML cells. Notably, co-treatment with FHD-286 and venetoclax or decitabine exerted synergistic lethality against AML cell lines and PD AML cells, especially those expressing MLL-r, mtNPM1 or EVI1 (Delta synergy scores > 5 by the ZIP method). Based on the known efficacy of the Menin inhibitor SNDX-50469 in AML with MLL-r or mtNPM1, we also found that co-treatment with FHD-286 and SNDX-50469 was synergistically lethal against AML cell lines and PD AML cells with MLL-r or mtNPM1. Since treatment with BET (bromodomain and extraterminal) protein inhibitor also inhibits c-Myc and BCL2 expression and was shown to be lethally active in AML cells with EVI1 overexpression, or with MLL-r or mtNPM1, we also found that co-treatment with FHD-286 and BET protein inhibitor OTX015 exerted synergistic lethality against AML cell lines and PD AML cells with chromosome 3q26 lesions and EVI1 overexpression, or with MLL-r or mtNPM1. Finally, in a luciferase-transduced, patient-derived xenograft (PDX) model of AML cells with MLL-AF9 and FLT3, KMT2C/2D and NOTCH2 mutations, we determined that treatment with FHD-286 administered orally alone for 4 to 6 weeks was significantly effective in reducing AML burden and improving overall survival of the mice. Additionally, co-treatment with FHD-286 and venetoclax or decitabine or OTX015, as compared to each drug alone or vehicle control, significantly reduced the AML burden and improved median and overall survival of the immune-depleted mice, without inducing significant toxicity. Taken together, these findings highlight the promise of FHD-286 treatment alone and in rational combinations in exerting significant anti-AML efficacy against cellular models of AML, especially those with MLL-r, mtNPM1 or chromosome 3q26 lesions and EVI1 overexpression.