PROJECT SUMMARY Anaplastic thyroid cancer (ATC) is a lethal entity characterized by frequent mutations in the BRAF oncogene. Until recently there were no treatment options to substantially prolong survival in ATC patients. Unlike BRAFV600E- driven melanomas, thyroid cancers with BRAFV600E mutations are only modestly responsive to targeted kinase inhibitor monotherapy, which emphasizes the significant context dependent differences in the molecular mechanisms that compensate for oncogenic signaling blockade. The recent approval of dual BRAF + MEK inhibition is a critical addition to the arsenal against advanced ATC, with results showing improved tumor control. Despite the initial success of targeted therapies, acquisition of resistance and relapse occur frequently and severely limit complete responses and cures. Such therapy-induced adaptive mechanisms include the rewiring of metabolic pathways that protect from BRAF pathway inhibition, however, it is not known if metabolic alterations occur in ATCs during the evolution of acquired resistance and if they have any potential clinical utility. In preliminary experiments we found that MEK inhibitors induce lipolysis and lipid droplet (LD) loss in ATC cells. LD are dynamic organelles participating in crucial cellular functions including metabolism, protein trafficking, membrane synthesis and signaling. We generated in vitro models of acquired resistance to MEK inhibitors with re-established LD abundance and analyzed their gene expression patterns to identify candidate vulnerabilities that can be pharmacologically targeted with clinically approved therapies. We found that expression of the DNA repair enzyme MGMT was downregulated in LDhigh but not LDlow MEK inhibitor resistant (MEKiR) clones. Here we propose to rigorously evaluate the hypothesis that LD abundance and MGMT expression can serve as biomarkers indicating alkylating agent sensitivity in preclinical models of MEKi resistant thyroid cancers. In aim 1 we will identify metabolic determinants of MGMT expression in MEKiR by uncovering nutritional requirements and epigenetic marks that control MGMT expression and sensitivity to alkylating agents. We will engineer isogenic cell lines with perturbed expression of the rate limiting triglyceride lipase ATGL/PNPLA2 to ask if LD accumulation is necessary or sufficient for MGMT silencing, and will assess if protein- protein interactions regulate ATGL activity in MEKiR. In aim 2 we will quantify the efficacy of alkylating agents in xenotransplanted MEKiR model tumors. Further, we will generate novel in vivo preclinical models of MEKi resistance by intermittent trametinib treatment of tumors in mice, and will validate LD abundance as a biomarker for alkylating agent efficacy by determining tumoral LD turnover, MGMT expression and response to alkylating agent therapy in these tumors.