P10.04.A MYC-DRIVEN PEDIATRIC BRAIN TUMORS SHARE A COMMON PHOTORECEPTOR IDENTITY

Abstract BACKGROUND Medulloblastoma (MB), pineoblastoma (PB) and retinoblastoma (RB), three types of malignant neural pediatric brain tumors are anatomically situated in different parts of brain but share a common feature - elevated expression of a photoreceptor ‘program’. MYC-amplified MBs and PBs are often histologically indistinguishable. Both grow aggressively, often relapse after standard treatment and have a dismal prognosis. Trilateral RBs are rare but often present with aggressive eye tumors first and later on PBs. It is important to understand how these photoreceptor-positive tumors arise and to know if they could be similarly treated. CRX, a transcription factor critical for retinal and pineal cell development is also a master regulator for MB-Group3 tumor maintenance. METHODS AND RESULTS To understand if MYC-driven MB, PB and RB can arise from specific photoreceptor-positive progenitors in the developing brain, we performed Crx-lineage tracing with tamoxifen injections. We could confirm that Crx-positive cells exist in retina and in pineal gland progenitors as expected. However, Crx-traced cells also marked granule neurons in the flocculonodular lobe of the cerebellum, which is a recently suggested site of MB-Group3 origin arising from the developing rhombic lip. To investigate if the Myc oncogene can generate different types of pediatric brain tumors in Crx-positive cells from retina, pineal gland and cerebellum, XMYCT58A-Tomato and XMYC was established by crossing Crx-CreERT2 strain with a LSL-MycT58A strain and Rosa26LSL-tdTomato stain. Here stabilized Myc oncogene was turned on with tamoxifen injection after birth and red tumor cell development could be followed. XMYCT58A-Tomato mice developed brain tumors from retina, pineal gland and cerebellum after 4-6 months. Histologically, tumors were non-glial (GFAP-, Olig2-), showed neuronal activity (Neurod1+) and stained positive for photoreceptor markers (CRX+, OTX2+) similar to both MB-Group3 and PB-MYC. CONCLUSIONS Our data suggest that MYC-driven Group 3 MB, PB and RB all could originate photoreceptor positive cells, which has implications for future research and the development of novel treatments targeting these devastating childhood malignancies.