Investigating Glioblastoma Microenvironment and Cellular Interactions with 3D Bioprinted Tumor Models
Investigating Glioblastoma Microenvironment and Cellular Interactions with 3D Bioprinted Tumor Models
Glioblastoma (GBM) is the most common and fatal adult primary brain cancer. Maximal safe surgical resection followed by concurrent chemotherapy and radiotherapy results in a five-year survival rate less than 7% for patients afflicted with GBM, and therapeutic advances to treat glioblastoma remain stagnant. In vivo, complex cell-matrix and cellular interactions among tumor cells, stromal cells, and the extracellular matrix (ECM) lead to a dynamic and heterogeneous GBM tumor ecosystem, characterized by significant immune infiltration and immunosuppression. Traditional model systems such as 2D cell culture or animal models either lack sufficient complexity to mimic the pathophysiological GBM microenvironment or face challenges with engraftment, lack of normal immune interactions, and low throughput nature of animal experiments. The discrepancies between drug evaluation results from pre-clinical models and actual clinical outcomes have also led to failure of many compounds in clinical trials. In this dissertation, ECM and cellular aspects of GBM are independently investigated using 3D models generated with a light-based 3D bioprinting technique and brain tumor-relevant biomaterials, revealing important criteria for fabricating biomimetic GBM models. The models serve as scalable and physiologic platforms to interrogate the role of different factors, including matrix properties and cellular crosstalk, in various tumor progression events, such as tumor migration, functional dependencies, drug responses, and immunologic interactions within a species-matched condition. In addition, a computational method based on machine learning algorithms is developed to predict drug sensitivity of monocyte or microglia infiltrated GBM tissues to small molecule probes and reveal features that contribute to the outcome of the prediction. The integration of computational tool with more clinically relevant tumor models could be a promising solution to investigate the disease mechanisms and accelerate the drug development for GBM.