PM-23 * EXPLORATION OF BASIC AND THERAPEUTIC MECHANISMS IN NOVEL MODELS OF RTK/RAS-NETWORK-DRIVEN GLIOBLASTOMA

Abstract

Glioblastoma (GBM) remains fatal despite intensive treatments including standard-of-care surgery, radiation, and chemotherapy. Modern targeted therapies have also failed to produce effective results in clinical trials, possibly due to a lack of relevant preclinical models for drug discovery/assessment. Approximately 90% GBMs harbor aberrations in RTK/RAS-signaling, including EGFR amplification/gain-of-function mutations (57%), PDGFRa amplification (10%), and NF1 deletion/loss-of-function mutations (10%). KRAS mutations also occur, but at low frequency (1%). Frequent alteration of RTK/RAS-signaling emphasizes a great need for several distinct RTK/RAS-network-driven preclinical GBM models that mirror the range of human GBM lesions, since therapeutic responses may differ depending on the major network driver in a given tumor. We have recently generated a genetically-engineered mouse model of GBM via adult-inducible astrocyte-specific perturbation of RB, KRAS and PTEN. The triple mutant mice develop GBMs that faithfully recapitulate the histopathology and molecular characteristics of human primary GBMs. Interestingly, without RAS activation, only low-grade disease is induced. We established astrocytes from the triple mutant mouse model [RB-inactivated/KRASG12D/+/PTEN+/-; RKP astrocytes)] that elicit GBM upon intracranial transplantation, as well as the RB/PTEN double mutant mice [RB-inactivated/PTEN+/-; RP astrocytes], which do not produce GBM. Thus, the RP cells provide an ideal platform for evaluation of multiple RAS-activating events in causation of GBM. We are currently introducing the most frequent events found in human GBMs, including EGFR, ERFRvIII and PDGFRa activation and NF1 deficiency, into RP cells in vitro. The ability of each event to convert the RP cells into GBM-forming astrocytes by orthotopic transplantation is being assessed. This novel in vitro-to-allograft approach provides direct comparisons of etiologies driven by various RAS-activating events in the same genetic background and will potentially lead to basic discovery of distinct mechanisms in GBM tumorigenesis. Furthermore, this approach will generate relevant panels of RTK/RAS-network-driven GBM models for robust preclinical therapeutic discovery and biomarker development.