ABSTRACT: BET bromodomain inhibitors block binding of bromodomain 1 and 2 (BD1, BD2) of BET family proteins to chromatin, and have demonstrated clinical on-target activity in NUT carcinoma (NC), a BRD-NUT fusion oncoprotein-driven cancer. However, toxicity due to inhibition of BD1 has limited the effectiveness of these compounds. A second generation, highly BD2-selective BET inhibitor, DC-9476, was developed to treat non-neoplastic conditions while minimizing toxicity. We find that, in contrast to non-NC cells, NC cells are uniquely sensitive to inhibition by DC-9476 and other BD2-selective inhibitors, which phenocopy the pro-differentiation, growth-arresting effects of non-selective (pan-BD) BET bromodomain inhibition or BRD4-NUT knockdown. We performed total RNA bulk RNAseq to determine what transcriptional changes occur in response to treatment with DC-9476 over a time course (18h, 36h, 48h) in 10-15, 10326 NC cells, and HCC-95 squamous lung non-NC cells. While there were not differentially expressed gene in HCC-95 cells, both 10326 and 10-15 cells showed transcriptional changes consistent with squamous/epithlelial differentiation, and downregulation of translation-associated genes. To understand the basis for this differential sensitivity, we performed a CRISPR knockout screen, which identified the BRD4-targeting E3 ligase, SPOP, as the top hit that confers resistance of NC to DC-9476, suggesting that proteasomal degradation of BRD4/BRD4-NUT is a key vulnerability of NC to BD2-selective BET inhibition. Indeed, treatment with diverse BD2-selective BET inhibitors, but not BD1-selective inhibitors, leads to proteasomal degradation of wild type BRD4 and BRD4-NUT, both of which are required for NC growth, in vitro and in vivo. Importantly, knockout of SPOP prevents loss of BRD4 and BRD4-NUT, and nearly completely rescues NC cells from differentiation and growth arrest induced by BD2 inhibition. Moreover, deletion of the canonical SPOP degron in exogenously expressed BRD4-NUT prevents its degradation from BD2 inhibition. To determine what effects BD2i (DC-9476) has on the chromatin occupancy of BRD4/BRD3-NUT, BRD4L, and H3K27ac, we performed CUT&RUN on NC cells, 10-15 and 10326, and non-NC cells, HCC-95. We found that BD2 inhibition also displaces BRD4, BRD4-NUT, and BRD3-NUT from chromatin, suggesting that degradation is coupled to chromatin release. We leveraged this unique vulnerability in pre-clinical NC models and found that BD2-selective inhibitors achieved significantly greater inhibition of growth and prolonged survival than older generation pan-BD BET inhibitors
