Our work suggested recurrent DNA double-stranded (DSBs) in developing neural stem and progenitor cells (NSPCs) may predispose to genomic variations associated with medulloblastomas (MB), the major pediatric brain cancer. To search for such DSBs, we developed and applied a sensitive genome-wide DSB-detection method to discover recurrent DSB clusters (“RDCs”) in mouse NSPC genomes. All RDCs were in genes, of which most have roles neural cell communication and/or are implicated in neuropsychiatric diseases and cancer.  Based on robust preliminary data, we propose three inter-related specific aims that will elucidate mechanisms of RDC generation, extend studies to human neural progenitors, and evaluate contributions of RDCs to recurrent genomic rearrangements in MB. In Aim 1 we will evaluate roles of transcription and replication stress in NSPC RDC formation, elucidate relationships between RDC chromosome domain structure with RDC DSB generation/resolution and gene expression, and develop a new approach to assay NSPC RDC formation in vivo to better elucidate developmental and disease implications.  Aim 2 studies will elucidate human RDC genes, which is necessary to directly assess their relevance to genomic alterations found in neuropsychiatric diseases and cancer. We will use our recently developed chromosome-specific targeting approach to identify and characterize human NSPC RDCs. Aim 3 will exploit a mouse MB model developed in our lab in which tumors harbor recurrent chromosomal rearrangements similar to those found in an aggressive class of human MBs to test the hypothesis that RDC DSBs may contribute to the generation of recurrent genomic variations found in MBs. To achieve this goal, we are collaborating with Peter Lichter (DKFZ, Heidelberg, Germany) to perform whole genome sequencing of mouse MBs from our model. If correlations between recurrent MB genomic breakpoints and RDC locations are established, we will test implicated RDC function with an approach adapted from our MB model.