Directin vivomapping of functional suppressors in glioblastoma genome

Abstract

AbstractGlioblastoma (GBM) is one of the deadliest cancers, with limited effective treatments and single-digit five-year survival1-7. A causative understanding of genetic factors that regulate GBM formation is of central importance8-19. However, a global, quantitative and functional understanding of gliomagenesis in the native brain environment has been lacking due to multiple challenges. Here, we developed an adeno-associated virus (AAV) mediated autochthonous CRISPR screen and directly mapped functional suppressors in the GBM genome. Stereotaxic delivery of an AAV library targeting significantly mutated genes into fully immunocompetent conditional Cas9 mice robustly led to gliomagenesis, resulting in tumors that recapitulate features of human GBM. Targeted capture sequencing revealed deep mutational profiles with diverse patterns across mice, uncoveringin vivoroles of previously uncharacterized factors in GBM such as immune regulatorB2m,zinc finger proteinZc3h13,transcription repressorCic,epigenetic regulatorsMll2/3andArid1b,alongside canonical tumor suppressorsNf1andPten. Comparative cancer genomics showed that the mutation frequencies across all genes tested in mice significantly correlate with those in human from two independent patient cohorts. Co-mutation analysis identified frequently co-occurring driver combinations, which were validated using AAV minipools, such asMll2, B2m-Nf1,Mll3-Nf1andZc3h13-Rb1. Distinct fromNf1-oncotype tumors,Rb1-oncotype tumors exhibit undifferentiated histopathology phenotype and aberrant activation of developmental reprogramming signatures such asHomeoboxgene clusters. The secondary addition ofZc3h13orPtenmutations drastically altered the gene expression profiles ofRb1mutants and rendered them more resistant to the GBM chemotherapeutic temozolomide. Our study provides a systematic functional landscape of GBM suppressors directlyin vivo, opening new paths for high-throughput molecular mapping and cancer phenotyping.