MODELING HUMAN BRAIN DEVELOPMENT AND DISEASE USING PRIMARY GLIOBLASTOMA AND HUMAN STEM CELL-DERIVED ORGANOIDS

Human brain development is an intricate process that involves closely timed coordination of cell proliferation, fate specification, neuronal differentiation, morphological transformation, and the integration of diverse cell types. Understanding this process in detail has been constrained by limited access to fetal brain tissue and the inability to study active neurodevelopmental processes in humans at the cellular level. Although model organisms have provided important insights into the mechanisms underlying brain development, these systems fail to fully recapitulate many human-specific features which often relate to disease. To address these challenges, human brain organoids, self-assembled three-dimensional neural aggregates, have been engineered from human pluripotent stem cells to model the architecture and cellular diversity of the developing human brain. Brain organoids have a distinct advantage over monolayer cultures as they recapitulate the structural organization of ventricular proliferative zones and the sequential generation of diverse neural cell types. My early work focused on optimizing brain organoid protocols and applying the model system to better understand aspects of human neural development and disease. I worked to optimize a protocol for hippocampal organoid differentiation to understand human embryonic hippocampal neurogenesis. My later work focused on applying many of same culture principles to engineer and optimize a protocol to grow glioblastoma organoids from surgically resected tumor tissue. By avoiding tissue dissociation and growing the organoids in an optimized medium, our protocol better preserves the cellular heterogeneity and behavior of tumor cells. We used this methodology to generate a biobank of organoids from over 50 patients, and thoroughly characterized many of these organoid samples by immunohistochemistry, mutation analysis, and bulk and single-cell transcriptomics. I applied our new model system to test targeted drugs and immunotherapies, including CAR-T cells and showed that treatment response could be predicted by the mutation and/or expression profile of the tumors. Our model system holds promise for testing targeted therapies in a dish before initiating treatment in patients to determine the best therapeutic options. The first chapter of my thesis describes the current status of the field of human stem cell-derived brain organoids, including an overview of early mammalian brain morphogenesis, recent efforts to generate human brain organoids that model the development of specific brain regions, endeavors to enhance cellular complexity to better mimic the in vivo condition, and neurological diseases modeled using brain organoids. The second chapter deals with my major work in engineering a novel three-dimensional patient-derived glioblastoma organoid model. This work represents a substantial advancement in the field of brain cancer modeling and holds promise for better preserving tumor tissue characteristics for mechanistic studies and therapeutic testing. The third chapter provides detailed methodology for generating and biobanking glioblastoma organoids and co-culturing with CAR-T cells to test immunotherapies. The protocol was written with the goal of improving adaptation and reproducibility in other laboratories. The fourth chapter describes my progress with generating hippocampal organoids from human pluripotent stem cells. The protocol has been optimized to produce medial pallium progenitors and putative hippocampal neurons. I further investigate synaptic connectivity using retrograde tracing and orthotopic xenograft in mouse brains