Abstract
In vivo screening highlighted ependymoma oncogenes and tumor suppressor genes within recurrent CNAs.
Major finding: In vivo screening highlighted ependymoma oncogenes and tumor suppressor genes within recurrent CNAs.
Concept: Candidate genes were validated in mouse models and converge on common cellular functions.
Impact: Disruption of vesicle trafficking and cholesterol biosynthesis promotes ependymoma tumorigenesis.
Mapping of recurrent DNA copy-number alterations (CNA), including gain or loss of specific chromosome regions in tumor cells, has been instrumental in highlighting oncogenes and tumor-suppressor genes (TSG). However, in some cases, such as in ependymoma brain tumors, CNAs may span large regions, making it difficult to identify cancer driver genes. Mohankumar, Currle, and colleagues used a series of murine models of ependymoma to screen 84 candidate oncogenes and 39 candidate TSGs, which were previously shown to reside within recurrent CNAs in ependymomas and a variety of other cancers. Fluorescently labeled cerebral, hindbrain, or spinal neural stem cells were engineered to express candidate ependymoma oncogenes and implanted as pools into the cerebrum of mice. Six different pools of cells induced tumor formation, and expression analysis revealed tumor-specific DNA enrichment of 14 candidate oncogenes. Further in vivo validation of individual candidates confirmed the ability of eight oncogenes, including ZNF668, BCL7C, and RAB3A, to promote ependymoma formation in the cerebrum and, in some cases, the hindbrain and spine. A similar in vivo strategy was used to identify and validate ten ependymoma TSGs, including several established TSGs and previously uncharacterized TSGs such as ALDH3A1, ACTR1A, SNX6, ULK2, and PCMT1. Remarkably, many ependymoma TSGs functionally converged with the identified oncogenes on common cellular functions, in particular DNA modification and repair, vesicle trafficking, and cholesterol biosynthesis. Indeed, Rab3a gain or loss of Ulk2 or Snx6 enhanced exocytosis or impaired endocytosis, respectively, resulting in prolonged oncogenic growth factor receptor signaling and increased sensitivity to a pan–fibroblast growth factor receptor inhibitor. In addition, downregulation of cholesterol biosynthesis machinery sensitized ependymoma cells to multiple statins in vitro. Together, these data highlight a cross-species in vivo approach to identify ependymoma driver genes in recurrent CNAs and suggest potential therapeutic targets for the treatment of this brain tumor.