Credentialing murine models for glioblastoma preclinical drug development

摘要

ABSTRACT Precision medicine promises to revolutionize oncology by targeting drugs to specific mutations. However, targeted drugs have failed to produce durable clinical responses when used as single agents in glioblastoma (GBM), the most common and deadly primary brain tumor. Genetically engineered mouse (GEM) models are essential for functional validation of such mutations, but technical limitations have prevented their widespread use in preclinical cancer drug development. The Miller Lab has developed non- germline GEM (nGEM) models. The Berens Lab has performed comprehensive genomic and chemovulnerability profiling in a genetically diverse and faithful panel of patient-derived human xenograft (PDX) models. The Johnson Lab has developed a novel chemical proteomics method, multiplex inhibitor beads coupled with mass spectrometry (MIB-MS), to assess the activation state of the cellular kinome en masse and has shown that dynamic kinome reprogramming contributes to targeted drug resistance. In this Multi-PI project, we will combine our expertise to address the following Aims: (1) To credential PDX models against human GBM by kinome proteomics; (2) To develop genetically-matched nGEM models from distinct cells of origin; and (3) To credential PDX and nGEM models by dynamic kinome profiling. We will develop a genetically diverse panel of nGEM models with defined driver mutations and cellular origins that will be useful adjunct to PDX for preclinical drug development. We will then credential both PDX and nGEM models against resected GBM specimens using cross-species genome, transcriptome, and kinome, and drug response profiling. Models will be genomically matched to their human counterparts and used to develop rational combination therapies that combat single agent resistance mechanisms in genomically- defined tumor subtypes. This work will therefore help realize the promise of precision medicine in neuro- oncology.