Modelling glioblastoma tumour-host cell interactions using adult brain organotypic slice co-culture

Glioblastoma (GBM) is an aggressive incurable brain cancer. The cells that fuel the growth of tumours resemble neural stem cells found in the developing and adult mammalian forebrain. These are referred to as GBM stem cells (GSCs). Similar to neural stem cells, GSCs exhibit a variety of phenotypic states: dormant, quiescent, proliferative and differentiating. How environmental cues within the brain influence these distinct states is not well understood. Laboratory models of GBM tumours can be generated using either genetically engineered mouse models, or via intracranial transplantation of cultured tumour initiating cells (mouse or human). Unfortunately, these approaches are expensive, time-consuming, low-throughput and ill-suited for monitoring of live cell behaviours. Here we explored whole adult brain coronal organotypic slices as a complementary strategy to remove the experimental bottleneck. Mouse adult brain slices remain viable in a neural stem cell serum-free basal media for several weeks. GSCs can therefore be easily microinjected into specific anatomical sites ex vivo. We demonstrated distinct responses of engrafted GSCs to different microenvironments in the brain. Within the subependymal zone - one of the adult neural stem cell niches - a subset of injected tumour cells could effectively engraft and respond to endothelial niche signals. GSCs transplanted slices were treated with the anti-mitotic drug temozolomide as proof-of-principle of the utility in modelling responses to existing treatments. Thus, engraftment of mouse or human GSCs onto whole brain coronal organotypic brain slices provides a convenient experimental model for studies of GSC-host interactions and preclinical testing of candidate therapeutic agents.