Project Summary Abstract. NUT carcinoma (NC, formerly NUT midline carcinoma) is an aggressive squamous carcinoma, affecting all ages, that is highly (>90%) lethal. The over-arching goal of this proposal is to improve survival of these patients through mechanism-driven identification of key oncogenic targets. NC is defined by NUTM1-fusion oncogenes, most commonly (~78%) encoding the BRD4-NUT protein. BRD4-NUT drives NC growth through the blockade of differentiation and maintenance of proliferation by forming very large (100kb-2MB), acetyl-histone-rich super-enhancers called megadomains (MD) that we described. MDs arise from the recruitment of p300, a histone acetyl-transferase (HAT), by NUT to acetylated histones bound by the dual bromodomains of BRD4, a BET family protein. We have shown that BRD4-NUT drives transcription of key oncogenic targets through the additional recruitment of numerous transcriptional activators by BRD4. Our demonstration that treatment with BET bromodomain inhibitors (BETi) that competitively inhibit binding of BET bromodomains to chromatin, can inhibit growth of NC in humans led to a new field investigating the role of BRD4 in cancer. Unfortunately, the efficacy of BETi monotherapy is limited by toxicity. In this next phase, we are investigating novel oncogenic mechanisms whose targeting can synergize with BETi therapeutically. Our preliminary data, together with what is established, has led us to formulate a mechanistic two-domain model of how NC growth is epigenetically driven. The model will be tested in the aims below with the overall goal of identifying a therapeutic combination of compounds optimized for maximally selective, synergistic tumor inhibition with minimized toxicity. The model proposes that the first, transcriptionally repressive domain formed by PRC2 enables NC growth by repressing the expression of pro-differentiation, tumor suppressive genes (tested in aim 1). The second, physically separate permissive MD is comprised of BRD4-NUT-p300 bound to acetylated histones that drive transcription of oncogenic genes. In this model, MD formation requires cooperative binding of both BRD4-NUT bromodomains (BD1 and BD2) to acetylated histones (tested in aim 2); and the acetyl- histone-permissive state that enables MD formation requires mono-methylation of H3K79 by DOT1L (aim 3). Inhibition of both of these domains would 1. repress transcription of oncogenic genes, such as MYC, and 2. de- repress transcription of genes critical for differentiation, senescence, and exit from cell cycle. Indeed, we have found that co-inhibition of EZH2 and BRD4-NUT is synergistic in vitro and in vivo. Specific Aim 1. Determine the oncogenic role of EZH2 in NUT carcinoma. Specific Aim 2. Determine the oncogenic roles of BRD4 bromodomains 1 and 2 in NUT carcinoma. Specific Aim 3. What is the role of DOT1L in BRD4-NUT megadomain formation and oncogenic function?