Glioblastoma multiforme (GBM), a grade IV astrocytoma, is the most common and deadly type of primary malignant brain tumor, with a patient’s median survival rate ranging from 15 to 17 months. The current treatment for GBM involves tumor resection surgery based on MRI image analysis, followed by radiotherapy and treatment with temozolomide. However, the gradual development of tumor resistance to temozolomide is frequent in GBM patients leading to subsequent tumor regrowth/relapse. For this reason, the development of more effective therapeutic approaches for GBM is of critical importance. Low tumor oxygenation, also known as hypoxia, constitutes a major concern for GBM patients, since it promotes cancer cell spreading (invasion) into the healthy brain tissue in order to evade this adverse microenvironment. Tumor invasion not only constitutes a major obstacle to surgery, radiotherapy, and chemotherapy, but it is also the main cause of death in GBM patients. Understanding how hypoxia triggers the GBM cells to become invasive is paramount to developing novel and more effective therapies against this devastating disease. In this review, we will present a comprehensive examination of the available literature focused on investigating how GBM hypoxia triggers an invasive cancer cell phenotype and the role of these invasive proteins in GBM progression.

The role of astrocytes in the glioblastoma (GBM) microenvironment is poorly understood; particularly with regard to cell invasion and drug resistance. To assess this role of astrocytes in GBMs we established an all human 2D co-culture model and a 3D hyaluronic acid-gelatin based hydrogel model (HyStem™-HP) with different ratios of GBM cells to astrocytes. A contact co-culture of fluorescently labelled GBM cells and astrocytes showed that the latter promotes tumour growth and migration of GBM cells. Notably, the presence of non-neoplastic astrocytes in direct contact, even in low amounts in co-culture, elicited drug resistance in GBM. Recent studies showed that non-neoplastic cells can transfer mitochondria along tunneling nanotubes (TNT) and rescue damaged target cancer cells. In these studies, we explored TNT formation and mitochondrial transfer using 2D and 3D in vitro co-culture models of GBM and astrocytes. TNT formation occurs in glial fibrillary acidic protein (GFAP) positive “reactive” astrocytes after 48 h co-culture and the increase of TNT formations was greater in 3D hyaluronic acid-gelatin based hydrogel models. This study shows that human astrocytes in the tumour microenvironment, both in 2D and 3D in vitro co-culture models, could form TNT connections with GBM cells. We postulate that the association on TNT delivery non-neoplastic mitochondria via a TNT connection may be related to GBM drug response as well as proliferation and migration.

Despite the importance of the tumor microenvironment in regulating tumor progression, few in vitro models have been developed to understand the effects of non‐neoplastic cells and extracellular matrix (ECM) on drug resistance in glioblastoma (GBM) cells. Using CellTrace‐labeled human GBM and microglial (MG) cells, we established a 2D co‐culture including various ratios of the two cell types. Viability, proliferation, migration, and drug response assays were carried out to assess the role of MG. A 3D model was then established using a hyaluronic acid‐gelatin hydrogel to culture a mixture of GBM and MG and evaluate drug resistance. A contact co‐culture of fluorescently labeled GBM and MG demonstrated that MG cells modestly promoted tumor cell proliferation (17%‐30% increase) and greater migration of GBM cells (>1.5‐fold increase). Notably, the presence of MG elicited drug resistance even when in a low ratio (10%‐20%) relative to co‐cultured tumor cells. The protective effect of MG on GBM was greater in the 3D model (>100% survival of GBM when challenged with cytotoxics). This new 3D human model demonstrated the influence of non‐neoplastic cells and matrix on chemoresistance of GBM cells to three agents with different mechanisms of action suggesting that such sophisticated in vitro approaches may facilitate improved preclinical testing.