2026 Theses Doctoral

iPSC-derived models of the human neurovascular unit in health and familial Alzheimer’s disease

Cui, Aomeng

Brain microvascular endothelial cells (BMECs) comprise the blood-brain barrier (BBB), a highly specialized structure that protects the brain from harmful pathogens and closely regulates the passage of molecules. In neurological and neurodegenerative diseases, BBB dysfunction can exacerbate disease pathogenesis. Alzheimer’s disease (AD), the leading cause of dementia in the world, is frequently associated with early vascular and BBB deficits. However, studies of the human BBB in health and disease have been limited by the lack of robust protocols to generate induced pluripotent stem cell (iPSC) derived BMECs. Existing methods generate cells that do not fully recapitulate key BMEC functions or the BBB transcriptome. The development of a novel, thoroughly characterized iPSC-derived model of the BBB enables investigation of the mechanisms driving early BBB dysfunction in AD.

In Chapter 1, I introduce the specialized features of the blood-brain barrier and its development and maintenance by expression of brain-specific endothelial cell markers and interactions with astrocytes and pericytes in the neurovascular unit (NVU). Next, I describe current in vitro models of the BBB and existing 3D models. Lastly, I introduce the features of AD and describe the existing literature on BBB dysfunction in AD.
In Chapter 2, I report that the combination of directed differentiation of iPSCs into BBB-primed ECs and transcription factor reprogramming with two brain EC markers, FOXF2 and ZIC3 generates reprogrammed BMECs (rBMECs). I also describe the molecular profile of rBMECs by characterizing expression of tight junctions and response to angiogenic stimuli. Transcriptomic analysis reveals that rBMECs are brain-specific ECs that express a subset of key BBB markers.
Next, I asked in Chapter 3 whether rBMECs exhibit key BBB features, such as a low level of paracellular and transcellular permeability, low transcytosis, and elevated activity of drug efflux transporter P-glycoprotein. Finally, I asked whether rBMECs could interact with iPSC-derived pericytes and astrocytes when cocultured in a 3D microfluidic system with flow. Notably, rBMECs formed tubules with open lumens and made close contacts with pericytes and astrocytes. Barrier integrity for rBMECs cocultured with pericytes and astrocytes was improved, and culture in a 3D system with shear stress increased expression of BBB-relevant genes.
In Chapter 4, I asked whether familial AD-causing mutation APPLon has effects on cells of the neurovascular unit, beyond its known effects on neurons. Transcriptional analysis revealed APPLon induces cell-specific changes to junctional markers in rBMECs, reactivity in astrocytes, and phenotype switching in VSMCs. APPLon also induces global pathway changes in all NVU cells relating to angiogenesis and extracellular matrix. Functionally, APPLon rBMECs have decreased barrier integrity. APPLon VSMCs have decreased migration following a wound scratch, while APPLon pericytes have decreased proliferation, suggesting that APPLon disrupts key functions in multiple NVU cells.
In Chapter 5, I summarize the major findings from my research and future directions for how a clearer understanding of early BBB dysfunction using more faithful human models can potentially guide future therapeutic development in AD.