PM-12 * USING A ZEBRAFISH PEDIATRIC BRAIN TUMOR MODEL FOR PRE-CLINICAL DRUG SCREENING

Modeling cancer in zebrafish is a rapidly growing area of research due to unique imaging and drug screening attributes of the system, as well as cost. While over 20 different zebrafish cancer models have been established, no genetically engineered models of pediatric brain tumors exist. We are using genomic editing techniques with traditional transgenic approaches to model pediatric brain tumors in zebrafish for gene discovery and drug screening. As Primitive Neuroectodermal Tumors (PNETs) represent the largest group of malignant brain tumors in children, we used sequencing and genomics data to predict oncogenic drivers of different PNET subtypes and generated the first zebrafish PNET model. Specifically, we show that activation of NRAS signaling in embryonic oligodendrocyte precursor cells with p53-deficiency generate oligoneural PNETs along the entire CNS axis, with conserved histological and molecular features to human medulloblastoma and CNS-PNETs, including activated SHH signaling and loss of RB signaling. We have also developed new embryonic brain tumor transplantation assays for high-throughput drug screening on hundreds of animals/day and show MEK activity is essential for the oligoneural CNS-PNET tumor growth in vivo. Thus, MEK inhibitors may represent the first targeted therapy option for children with oligoneural CNS-PNETs. We will present these results and plans for expanding the zebrafish system to generate other classes of brain tumors and for growing human PDX's in zebrafish.