Targeting mTORC1 translational control in FOXO1 fusion positive rhabdomyosarcoma

PROJECT SUMMARY PAX3-FOXO1 and related fusions create oncogenic transcription factors that remodel chromatin to drive a subset of childhood rhabdomyosarcoma (RMS) with dismal cure rates. Therapies to target PAX3-FOXO1 are lacking. Recently, we established mTORC1 as a PAX3-FOXO1 genetic dependency that is exploitable with third generation bi-steric mTORC1 inhibitors, now in phase 1 testing. Unlike rapamycin, RMC-6272 prevents mTORC1-driven assembly of eIF4F and ensuing cap-dependent translation. This corresponds to striking efficacy in patient-derived xenografts (PDX): RMC-6272 induces prolonged remissions, while rapamycin analogs only stabilize disease, mirroring their modest effects in patients. However, PDX eventually regrow and show resistance to RMC-6272 re-treatment, making it clear that a knowledge gap prevents us from effectively targeting PAX3-FOXO1 via mTORC1. What is the molecular basis for PAX3-FOXO1 “addiction” to cap-dependent translation, and how can we rationally combine this agent to overcome resistance and enable curative therapy? This proposal addresses these questions with the goal of providing a molecularly informed strategy to aid ongoing clinical development of bi-steric mTORC1 inhibitors in childhood cancer. Aim 1 will define the molecular basis for RMC-6272 efficacy in fusion positive RMS. Excitingly, informatic and proteomic data lead us to the anchoring hypothesis that PAX3-FOXO1 requires cap-dependent translation for its own expression. We will use precise biochemical assays of mRNA translation and protein synthesis to test this. Building on the observation that patient-derived xenografts (PDX) driven by less frequent PAX7-FOXO1 fusions show reduced response to RMC-6272, we will define molecular features in the untranslated regions (UTRs) of PAX3 and PAX7 fusions that require mTORC1 and eIF4F for their translation, and test whether UTR sequence determines RMC-6272 efficacy. Aim 2 will define the best means to augment RMC-6272 efficacy in PDX. Single agents are rarely successful in curing even genetically simple cancers like fusion positive RMS, and indeed we observe the emergence of resistance to RMC-6272 monotherapy. We find that BET inhibitors (which decrease PAX3-FOXO1 expression) and RAS inhibitors (which blunt feedback activation of MAPK) each are synergistic with RMC-6272. We hypothesize that such mechanism-based combinations will offer curative treatment and will compare them to combinations with chemotherapy that are the mainstay of relapsed RMS studies. Using pharmacodynamic modeling in PDX, we will identify safe and optimal biologic dosing to combine agents, then find the strategy that best prolongs survival and prevents resistance to translate into clinical trials. Completion of these aims will detail how eIF4F orchestrates oncogene output in a frequently lethal pediatric cancer and define strategies to further enhance these effects to offer curative treatment. Beyond guiding bi-steric mTORC1 inhibitor use in RMS, the knowledge gained will enable study of translational control as a therapeutic target in other transcription factor-driven malignancies based on determinants of response we define here.