(261g) Novel Blood-Brain Barrier Models Derived From Stem Cell Sources

Over the past two decades, numerous advances have been made in drug development to treat a wide variety of diseases. However, little progress has been made in treating many serious brain diseases because of the inability of ~98% of small molecule drugs and essentially 100% of all protein therapeutics to cross the blood-brain barrier (BBB) and reach their target. The BBB is composed of endothelial cells (ECs) which line brain capillaries and control trafficking between the brain and bloodstream via specialized barrier properties; this barrier is necessary to maintain brain health and homeostasis but presents a significant bottleneck for brain drug delivery. Thus, a long-standing goal of the brain research community has been the creation of a robust human in vitro BBB model to study the mechanisms of BBB development and regulation, with hopes that a model which accurately represents the in vivo BBB may have use in high throughput screening of neuropharmaceuticals that can readily traverse the BBB.

A typical in vitro model consists of brain endothelial cells co-cultured with astrocytes and neurons, brain cells that have been shown to upregulate the BBB phenotype. However, brain endothelial cells isolated from tissue tend to possess poor barrier properties, are typically not of human origin, and are of extremely low yield. Moreover, human astrocytes and neurons are also not readily available for co-culture. To circumvent these issues, we have taken advantage of human stem cell technology, utilizing pluripotent stem cells (hPSCs) to generate BBB-specific endothelial cells and neural progenitor cells (NPCs) to generate astrocytes and neurons, with the design goal of creating a fully human BBB model.

The onset of BBB properties in vivo is dependent on cues provided by the developing embryonic brain. Thus, to recapitulate BBB development in vitro, we developed a novel two-step method to first convert hPSCs to embryonic brain cells, followed by selective expansion of primitive ECs within this neural microenvironment. We have verified the embryonic brain cells provide relevant BBB-inductive cues via the Wnt/β-catenin pathway and the ECs, upon receiving these cues, acquire a combination of protein markers that is characteristic of the BBB. After purification, these hPSC-derived brain-specific ECs are shown to possess superior barrier characteristics compared to other reported human in vitro BBB models, including high trans-endothelial electrical resistance (1450±140 ohmsxcm²), representative permeability to drugs within a wide range of lipophilicity (100-fold permeability difference between hydrophobic and hydrophilic compounds), and functional expression of influx and efflux transporters such as GLUT-1, p-glycoprotein, breast cancer resistance protein, and members of the multidrug resistance protein family.

In parallel, we established differentiation protocols for converting NPCs to mixtures of astrocytes and neurons optimized for co-culture with the brain-specific ECs. In particular, we performed proof-of-principle experiments to demonstrate that both rodent and human NPCs differentiated to astrocytes and neurons can upregulate BBB properties in primary rodent brain endothelial cells, a phenomenon that was contingent on NPC maturation state and final astrocyte to neuron ratio. We are currently in the process of combining the hPSC and NPC platforms to create a fully humanized BBB model that will accelerate the discovery and characterization of novel brain-penetrating therapeutics. Further, we believe the hPSC platform represents an excellent model system for studying BBB formation and maturation, which could have clinical relevance for the treatment of a variety of neurological diseases.