EPEN-31. Developmental and Oncogenic Transcription Factor Circuits as Dependencies in Ependymoma

Brain tumors are the most common cause of cancer death in children. ZFTA-RELA gene fusion is one the most potent drivers of cancer and is sufficient to induce tumors when expressed during brain development. ZFTA-RELA (denoted ZRFUS) fusion is the most frequent events that occurs in an aggressive childhood brain tumor called ependymoma (> 70% of cases). ZFTA recruits RELA to novel DNA binding sites and is necessary to activate ependymoma oncogene transcription. There are currently no targetable treatments for ependymoma, thus studying the mechanisms that regulate ZRFUS oncogenic programs may yield opportunities to develop effective therapies. To study proteins that regulate gene expression programs in brain cancer, the Mack lab and others have comprehensively characterized the active chromatin landscapes of several adult and pediatric brain cancers. This genome-wide analysis has identified highly active TFs, termed core regulatory circuit (CRC) TFs that govern gene expression programs such as MYC, GLI2, SOX2, and OLIG1/2, previously described in brain tumors such as glioblastoma and medulloblastoma. Critically, a glial cell fate specification TF, SOX9, showed the highest levels of activity in ependymoma. A functional RNA interference screen of CRC TFs prioritized SOX9 as the top cancer dependency gene required for ZRFUS ependymoma cell proliferation. To study ZRFUS ependymoma, we developed one of the first genetic mouse models of the disease, and show in preliminary data, that SOX9 knockout abolishes tumor initiation. Surprisingly, SOX9 KO has no impact on tumor initiation in an aggressive glioma model, suggesting tumor-specific contributions of SOX9. This concept is supported by our data that shows SOX9 co-recruitment to a vast majority of ZRFUS binding sites in the genome. Our data supports that SOX9 regulates ZFTA-RELA target cistrome; presenting a potential pathway that may be explored for therapeutic benefit.