(604e) Engineered Brain Microvascular Microenvironments Reveal a Role for Pericytes and Astrocytes in Shaping Vascular Architecture and Glioblastoma Behavior

Problem: Glioblastoma (GBM) is the most common primary malignant brain tumor. With a median survival time of one year and a five-year survival rate less than 10%, the disease is currently incurable. Barriers to treatment include diffuse spreading of tumor cells into the surrounding brain tissue, which limits the effectiveness of surgical resection. Individual tumor cells also display heterogeneous phenotypes, which leads to treatment resistance and recurrence post-therapy. Tumor cells reside in various niche microenvironments throughout the tumor tissue, and these niches differentially influence tumor cell behavior and contribute to phenotypic heterogeneity. However, current in vitro models utilized in the drug discovery pipeline do not incorporate niche components, and this oversight contributes to drug failures in animal models and human clinical trials.

GBM tumor cells have been shown to reside in perivascular niches, which are hypothesized to enhance tumor cell invasion and harbor a subpopulation of aggressive cancer stem cells (CSCs). The mechanisms by which the cellular components of the perivascular niche influence tumor cell behavior are unknown. Most in vitro studies have focused on crosstalk between tumor and endothelial cells; however, pericytes and astrocytes also exist in the perivascular niche and likely influence vascular architecture and resultant tumor cell behavior. Here, we develop a three-dimensional culture model of brain microvasculature to investigate the effects of pericytes and astrocytes on vascular morphogenesis. Then, we utilize the model to identify a role for pericytes and astrocytes in defining the overall influence of the perivascular niche on GBM invasion, proliferation, and chemotherapeutic response.

Methods: Brain microvascular endothelial cells (ECs) were cultured either alone, with pericytes (PCs), or with pericytes and astrocytes (ACs) in methacrylamide-functionalized gelatin (GelMA) hydrogels. Resultant microvascular networks were identified using CD31 staining, imaged, and quantified for metrics of vascular complexity (e.g. length, number of branches). Immunofluorescent staining and PCR were used to assess markers of vascular maturation, such as tight junction expression and basement membrane deposition. Primary GBM cells were co-cultured with ECs, ECs and PCs, or ECs and PCs and ACs as spheroids or single cells. GBM spheroid outgrowth was measured as a metric for invasion, and conditioned media was analyzed using a cytokine array to identify potential secreted mediators of invasion. GBM cells encapsulated as single cells were stained for KI67 expression to identify actively cycling versus quiescent cells, and an EdU pulse was used to characterize the rate of tumor cell proliferation. GBM-vascular co-cultures grown for seven days were then treated with 600 µM temozolomide (TMZ) or DMSO control for 48 hours. Tumor cell response to TMZ was assessed using KI67, EdU, and cPARP staining to assess proliferation status, proliferation rate, and apoptotic status respectively.

Results: Microvascular networks self-assemble upon co-culture of ECs, PCs, and ACs, and total network length increases in the presence of PCs and ACs. PCs and ACs also increase the expression of vascular markers such as laminin, ZO-1, and Claudin 5. GBM spheroids encapsulated in the presence of ECs, PCs, and ACs display increased outgrowth, potentially supported by heightened secretion of MMP9 and IL8. The combined presence of ECs, PCs, and ACs increases the fraction of tumor cells that are quiescent, but also increases the cycling rate of the remaining proliferative cells. Finally, tumor cells do not exhibit a significant increase in the apoptosis-to-proliferation ratio (cPARP/KI67) in the presence of vascular cells, suggesting a potential chemoprotective effect for the perivascular niche microenvironment.

Implications: We demonstrate the establishment of brain-mimetic microvascular networks within a hydrogel culture platform. Such a platform can be used to further investigate brain microvascular development and perturbation during injury or disease. In the presence of vascular cells, GBM cells exhibit heightened invasion and adopt heterogeneous proliferative behaviors, which are both attributes of the native disease. The platform will therefore be useful for identifying strategies to mitigate tumor cell invasion and treatment failure arising from tumor cell heterogeneity.