Mechanistic Basis and Efficacy of Targeting β-Catenin-TCF7L2-JMJD6-MYC Axis to Overcome Resistance to BET Inhibitors

Hematopoietic progenitor cells of myeloproliferative neoplasms with myelofibrosis (MPN-MF) exhibit constitutive activation of JAK-STAT5/3 and NFkB signaling. Transformation of MPN-MF to AML (post-MPN sAML) occurs in up to 15% of patients with MPN-MF. Standard induction anti-AML chemotherapy and the JAK1 & 2 inhibitor (JAKi) ruxolitinib are ineffective in post-MPN sAML. BET protein BRD4 is a non-oncogene addiction target in AML, and treatment with acetyl-lysine mimetic BET protein inhibitor (BETi) disrupts binding of BRD4 to acetylated chromatin and transcription factors (TFs). This attenuates transcription of super-enhancer regulated oncogenes, including MYC, Bcl-xL, PIM1 and CDK4/6, inhibiting growth and survival of post-MPN sAML blasts. BETi treatment also inhibits binding of BRD4 to acetylated RELA (NFkB-p65), inhibiting its transcriptional activity and attenuating levels of its target cytokines. However, BETi treatment induces BRD4, potentially reducing BETi activity in repressing oncogenes. Preclinical and clinical studies have demonstrated that innate or adaptive BETi-resistance is common in sAML cells. To model BETi-resistance, we repeatedly exposed (10 times) secondary (s) AML SET2 and HEL92.1.7 (HEL) cells to 1.0 µM of the BETi OTX015 for 48 hours followed by full recovery, thus generating BETi persister-resistant (BETi-P/R) SET2-P/R and HEL-P/R cells. These cells showed > 10-fold resistance to OTX015 and cross-resistance to other BETis. Compared to the parental controls, BETi-P/R cells lacked additional genetic alterations or altered levels of TRIM33, SPOP, DUB3 or phosphorylated BRD4 (previously described mechanisms of BETi-resistance). However, ATAC-Seq and ChIP-Seq (H3K27Ac mark) analyses demonstrated that, as compared to their parental controls, BETi-P/R cells showed gain of peaks and active enhancers with enrichment of STAT5, MYC, PU.1 and GATA2 binding sites. Newly gained peaks were in the enhancers of JAK1/2, RUNX1, PU.1, MYC and BCL2L1. RNA-Seq determined mRNA level alterations, included induction of gene-sets involving MYC/MAX, STAT5, NFkB and TCF7L2 targets. QPCR and Western analyses confirmed increase in the mRNA and protein levels of TCF7L2, JMJD6, c-Myc, Survivin and PIM1 in HEL-P/R over HEL92.1.7 cells. Expression of the arginine demethylase JMJD6, recruited by BRD4 to regulate enhancer-mediated transcriptional pause-release, was also increased. This was associated with increased expression of the nuclear β-catenin-TCF7L2 targets, including Cyclin D1, TERT, survivin, c-Myc and PU.1. Patient-derived human AML blasts that exhibited innate resistance ex vivo to BETi, also demonstrated increased expression of TCF7L2, JMJD6 and c-Myc. We next probed the mechanistic role of the β-catenin-JMJD6-TCF7L2-MYC axis in conferring BETi-resistance. CRISPR-Cas9-mediated knockout of TCF7L2 or JMJD6 significantly reversed BETi-resistance in BETi-P/R sAML cells (p < 0.001). Conversely, ectopic overexpression of TCF7L2 or JMJD6 significantly conferred BETi-persister-resistance in HEL and SET2 cells (p < 0.001). Notably, confocal microscopy demonstrated increased binding of β-catenin with TBL1 and TCF7L2 in the nucleus of BETi-P/R sAML cells. BC2059, which disrupts binding of nuclear β-catenin with TBL1 and TCF7L2, depleted β-catenin levels and exerted similar lethality in BETi-P/R sAML and control sAML cells. shRNA-mediated knockdown of BRD4 and treatment with BRD4-PROTAC (proteolysis-targeting chimera) ARV-771 (Arvinas, Inc.) that degrades BRD4/3/2, also induced similar levels of apoptosis in BETi-P/R and control sAML cells. Co-treatment with ARV-771 and BC2059 synergistically induced lethality in BETi-P/R sAML cells as well as in patient-derived, CD34+ sAML BPCs (combination indices < 1.0). This was associated with marked attenuation of c-Myc, TCF4, Survivin, CDK6, PIM1 and Bcl-xL levels. Also, compared to each agent alone, in vivo treatment with ARV-771 (30 mg/kg SQ daily x 5, per week) and BC2059 (30 mg/kg IP BIW per week) for 3 weeks, significantly reduced sAML burden and improved survival of NSG mice engrafted with HEL-P/R cells (p < 0.01). Collectively, these findings underscore that increased levels and activity of β-catenin-TCF7L2-JMJD6-MYC axis is mechanistically responsible for BETi-P/R, and co-targeting with BRD4 degrader and β-catenin-TCF7L2 inhibitor is a promising therapeutic strategy against BETi-P/R sAML BPCs.