Up to 1 in 6 people worldwide, 1 billion people, are estimated to have a neurological disorder, costing over $500B in the US alone. For most neurological diseases there is still no pharmacologic treatment available that can slow or stop neuronal damage. New models are critically needed that more faithfully recapitulate human disease, providing a tool for enhanced discovery of biomarkers and targets, effective therapeutic development, and personalized drug screening. Current studies aimed at identifying genetic variants and profiling cell type-specific gene activity are shining new light on potential mechanisms. However, more advanced systems capable of functionally assessing the consequences of these variants are needed to dissect disease etiology and evaluate the efficacy of potential interventions. For example, in neurological disease, patient heterogeneity makes the interpretation of genetic variants formidable, often implicated genes are expressed in non-coding regions and with cell type specificity, and many putative disease mechanisms involve regulatory mechanisms not well-conserved in animal models. Therefore, to establish an advanced preclinical model of the brain that could be used to functionally assess genetic variants associated with neurological disease, I have developed a multicellular integrated brain model, miBrain, that incorporates brain-resident immune, neuronal, glial, and vascular cell types of human patient-specific origin and with 3D tissue structural organization. To construct this model, I differentiated each of the six major brain cell types from iPSCs, encapsulated them in a novel biomaterial scaffold I developed with cell-instructive and brain-mimetic cues, Neuromatrix Hydrogel, and optimized a co-culture method to form a 3D engineered brain tissue mimic. miBrains consist of 3D immune-glial-neurovascular units with enhanced cell- and tissue-scale phenotypes inclusive of microglial immune cells with upregulated homeostatic signatures in healthy conditions, contiguous lumenized vascular networks and blood-brain barrier (BBB), neurovascular units, and interconnected myelinated neuronal networks. To enable perfusion of the BBB within the miBrain, I developed a novel microfluidic platform, the GelChip, via a 3D Printing strategy, to support 3D cellular network formation within complex co-cultures and engineered hydrogels and leveraged this to establish the miBrain-on-Chip. I have harnessed these systems to model APOE4 risk for Alzheimer’s Disease, recapitulating canonical disease hallmarks of increased reactive astrocytes, amyloid aggregates, and neuronal tau phosphorylation. Further, harnessing the modularity of the system to probe cell type-specific effects, I found that APOE4 astrocytes are sufficient to increase neuronal tau phosphorylation via crosstalk with microglia. Thus, I have established a novel preclinical brain model with broad utility for dissecting disease mechanisms, assessing delivery to the brain, and probing barrier function and other hallmarks in contexts of disease. This system could be harnessed to probe a wide range genetic variants, identify nodes for intervention, and test potential therapeutic strategies, with an approach that could be readily adapted to target other tissues.
