TMOD-28. AUTHENTIC HUMAN GLIOMA MODELING USING GENETICALLY ENGINEERED INDUCED PLURIPOTENT STEM CELLS

Abstract Many mouse and human glioma models have been utilized to study the genetic alterations involved in the genesis of these tumors, but they have not been fully evaluated for how closely they recapitulate pathobiology, including tumor heterogeneity, which is an inherent feature making patient treatment difficult. Here we present new glioma models using genetically engineered human pluripotent stem cells, in which authentic pathobiology is recapitulated through precision gene editing. Specifically, we show that neural progenitor cells (NPCs) derived from human induced pluripotent stem cells (iPSCs), with different combinations of genetic drivers introduced by CRISPR/Cas9-mediated editing give rise to distinct intracranial tumors recapitulating authentic pathobiology of the disease when engrafted in immunocompromised mice. NPCs deficient in PTEN and NF1, a genotype associated with the mesenchymal molecular subtype, and NPCs deficient in TP53 and expressing a PDGFRA activating mutation (PDGFRAΔ8–9), a genotype associated with the proneural glioblastoma molecular subtype, give rise to distinct tumors resembling glioblastoma. Both models presented inter and intra-tumor heterogeneity based on single-cell RNA sequencing, with the former model showing proneural signature and the latter a mesenchymal signature, and both having different degrees of cycling, and stem cell-enriched populations. The TP53/PDGFRA model had more clonal variability related with striking karyotype abnormalities including extrachromosomal DNA. Additionally, we expanded this approach to pediatric gliomas. Brainstem tumors derived from NPCs introduced with TP53 R248Q and H3F3A K27M mutations presented features of glial tumors with global expression of histone H3 K27M accompanied by suppression of histone H3K27 trimethylation, compatible with H3 K27M-mutant pediatric diffuse midline glioma. Using these isogenic human brain tumor models, we aim to advance our understanding of the pathobiology associated with different driver mutations and further, to provide a platform for development of targeted therapy.