Evolution and targeting of the functional states of glioblastoma

Project Summary The application is focused on glioblastoma multiforme (GBM), one of the most lethal forms of human cancer for which the massive knowledge generated by genomic data has provided little therapeutic improvement. One key element in the success of clinical studies for cancer patients is the selection of homogeneous groups of patients harboring tumors that share identifiable functional vulnerabilities rather than general biomarkers. In GBM, the lack of a functional classifier has hindered the targeting of fundamental cancer-driving mechanisms in well- defined patient subgroups, leading to discouraging results. The proposal is founded on a novel classification of GBM that we have recently proposed. Different from previously established marker-based classification, the new classifier is centered on functional activities of cancer cells that we identified by single cell transcriptomic analysis. The classifier was validated in several cohorts of bulk primary GBM and includes four subtypes, two linked to neurodevelopmental programs, neuronal and proliferative-progenitor, and two characterized by divergent metabolic activities, mitochondrial and glycolytic-plurimetabolic. Notably, the mitochondrial subtype is endowed with a distinct sensitivity to oxidative phosphorylation inhibition and is associated with a better survival, while the glycolytic-plurimetabolic subtype is characterized by redundant metabolic activities. Our preliminary analysis revealed that each functional GBM subtype is association with biologically coherent proteomic and phosphoproteomic features. In this application we will combine innovative computational tools and state-of-the- art experimental models in vitro and in vivo to study the impact of functional cell states of GBM in therapy resistance. We built the research plan with the following aims: i) examine and target the plasticity of the neurodevelopmental glioma states under therapy pressure and the cross-talk with signals from the microenvironment; ii) determine how mitochondrial cells adjust to therapy pressure when treated with mitochondrial inhibitors and the mechanism of induced resistance; iii) retrieve therapeutic intervention points from proteomic data focusing on DNA-PK and PKCd, two protein kinases active selectively in the proliferative/progenitor and glycolytic/plurimetabolic subtypes of GBM, respectively. Experimental validations will be applied to these nodal factors and will be performed by our laboratories, which in the course of many years have generated and perfected the array of experimental tools including sequence-annotated patient-derived models to pursue each question. By integrating novel computational and experimental platforms to study the evolution of distinct GBM subtypes, the proposal is conceptually and technically innovative. The successful outcome of this proposal will be the delivery of key information to decipher evolving tumor dependencies under treatment and accurate therapeutic strategies specifically tailored to distinct subgroups of GBM patients.