2025 AIChE Annual Meeting

(634d) Engineering a 3D Human Brain-on-Chip to Model Cerebrovascular Dysfunction in Alzheimer’s Disease

Authors

Alice Stanton, Stanford University
Md. Rezaul Islam, Massachusetts Institute of Technology
Aaron Choi, Massachusetts Institute of Technology
Eulim Kang, Massachusetts Institute of Technology
Robert Langer, Massachusetts Institute of Technology
Blood-brain barrier (BBB) dysfunction is implicated early in the progression of neurodegenerative diseases, including Alzheimer’s. However, our understanding of BBB and brain function in both healthy and diseased states remains limited by the models and tools used to study them. The BBB consists of a network of high-curvature microvessels that exhibit exceptionally low paracellular permeability and transcellular transport in humans compared to other species, creating a significant challenge in treating neurological disorders. The current reliance on 2D cellular models and animal studies fails to replicate the structural and functional complexity of the human BBB and its integration with neuronal cells, underscoring the urgent need for advanced in vitro platforms that accurately mimic the human brain microenvironment.

To address this challenge, we developed a 3D microfluidics-based brain-tissue model derived from human induced pluripotent stem cells (iPSCs). This approach leverages iPSCs to create a well-controlled system in which specific risk genes can be systematically studied. Key brain cell types—including endothelial cells, astrocytes, pericytes, microglia, neurons, and oligodendrocytes—are differentiated separately from iPSCs, encapsulated in a hydrogel, and self-assemble into 3D cellular architectures. This results in a fully integrated BBB with perfusable microvasculature in close proximity to neuronal cell types, resembling the organization of the human brain. Morphological characterization confirms vessel parameters consistent with human brain physiology, including junctional protein expression and neurovascular coupling. Functional assays demonstrate selective permeability that mirrors the in vivo BBB. As a case study, we apply this model to explore how capillary endothelial cells comprising the BBB respond to Alzheimer's associated insults on the peripheral side and alterations in glial support on the brain side. We identify transcriptomic and cellular alterations in the BBB in the context of Alzheimer’s disease and APOE4 risk-gene carriers.

In summary, this brain-on-chip model is uniquely suited to capture emergent tissue behaviors within the context of healthy vs. diseased states. By bridging molecular insights with mechanistic understanding, this platform offers a transformative tool for investigating neurodegenerative mechanisms and informing the development of brain-targeted therapeutic interventions.