SUMMARY The most common genetic alteration in glioblastoma (GBM) is amplification of the receptor tyrosine kinase EGFR. In GBM, some amplified EGFR further mutates to yield exon-deleted EGFRvIII, which is constitutively active and endocytosis impaired, thereby signaling to effector pathways that favor survival over proliferation. It is not clear why EGFRvIII is specifically selected for in this disease, but GBM cells poorly tolerate unbridled signaling from the EGFR effector RAS. Chronic RAS signaling places an oncogene-induced stress on cell membranes that gives rise to excessive micropinocytosis and vacuolization, concluding with an alternative form of cell death called methuosis. The objective of this work is to reconcile EGFR amplification and EGFR/RAS-induced membrane stress through EGFRvIII and the signaling intermediates they share. Preliminary evidence suggests that EGFRvIII may achieve stress-relieving signaling adaptations by rewiring the network of protein-protein interactions at different subcellular locations within GBM cells. Our hypothesis is that EGFRvIII is a GBM-specific adaptive mechanism for overcoming methuosis. We will build a computational model of EGFR/EGFRvIII signaling that tests this hypothesis by accounting for the network of protein interactions and signaling relevant for the methuosis phenotype. The specific aims are to 1) define the key intermolecular interactions in the EGFR signaling network and mechanistically predict the consequences of network adaptations to EGFRvIII expression; 2) map differential EGFR signaling network activation among glioblastoma cells to the methuosis phenotype through a hybrid mechanistic and data-driven computational model; and 3) test model predictions about signaling control of methuosis in vitro and in vivo using new tools to monitor RAS-ERK and AKT activities concurrently and noninvasively. This project brings together a team of investigators with complementary expertise in mechanistic and data-driven computational models of receptor- mediated signaling, protein-protein interactions, in vivo transplantations of GBM, and treatment of GBM patients using investigational approaches. By quantitatively testing the hypothesis about EGFRvIII as a key regulator of oncogene-induced plasma membrane stress in glioblastoma, our collaborative project holds promise for identifying conceptually new approaches for driving alternative cell-death phenotypes in a highly chemotherapy-resistant cancer for which durable therapies are desperately needed.