Abstract IA11: Targeting the mechanosensing machinery to limit glioblastoma initiation and invasion

Abstract

Abstract Glioblastoma (GBM) is the most common and deadly primary brain cancer, with a median survival time of less than two years from diagnosis, even with aggressive multimodal care. A pathophysiological hallmark of GBM is the intimate infiltration of tumor cells within the brain, which complicates surgical resection and contributes to the adaptive-evasive response to anti-angiogenic agents. In this sense, GBM may be regarded as a disease of cell migration, raising the possibility that one could target the machinery of cell migration to limit tumor progression. Over the past decade, we have shown that GBM migration is strongly sensitive to biophysical inputs from the environment including tissue stiffness1 and physical confinement.2 We have also shown that microenvironmental stiffness-based cues can act synergistically with canonical mitogenic signaling pathways (e.g. EGFR) to drive GBM progression.3 These efforts, together with seminal studies from other laboratories, have helped stimulate interest in mining the physical microenvironment and associated cellular mechanosensing systems for novel prognostic and therapeutic targets. In this presentation, I will start by discussing our recent efforts to investigate these ideas specifically in the context of patient-derived GBM tumor initiating cells (TICs), a small subset of tumor cells with stem-like properties that are thought to drive tumor recurrence and drug resistance. Surprisingly, we find that TICs are comparatively insensitive to microenvironmental mechanical inputs, escaping the block on motility normally present in highly compliant matrix environments. We show that this stiffness-sensing can be restored through overexpression of contractile agonists, including Rho GTPase, Rho-associated kinase, and myosin light chain kinase. This intervention not only restores suppression of motility on compliant matrices but also significantly retards 3D invasion in vitro and reduces diffuse infiltration in a mouse orthotopic xenograft model. The latter effect is accompanied by a 30% extension in survival. This provides direct evidence that manipulation of the cellular mechanosensing machinery can be used to successfully reduce GBM progression in a preclinical model.4 These and other results have motivated us to develop next-generation culture platforms for tumor mechanobiology that enable discovery and screening in a more scalable and high-throughput platform than traditional approaches. For example, I will discuss a platform we recently introduced that allows orthogonal patterning of extracellular matrix stiffness and adhesive ligand density. We have used this platform to both identify microRNAs whose expression is sensitive to matrix biophysical parameters and explore the role played by macrophage-secreted factors in driving GBM adhesion and mechanosensing.5