Organization of Endothelial Cells, Pericytes, and Astrocytes into a 3D Microfluidic in Vitro Model of the Blood–Brain Barrier

Abstract

The endothelial cells lining the capillaries supplying the brain with oxygen and nutrients form a formidable barrier known as the blood-brain barrier (BBB), which exhibits selective permeability to small drug molecules and virtually impermeable to macromolecular therapeutics. Current in vitro BBB models fail to replicate this restrictive behavior due to poor integration of the endothelial cells with supporting cells (pericytes and astrocytes) following the correct anatomical organization observed in vivo. We report the coculture of mouse brain microvascular endothelial cells (b.End3), pericytes, with/without C8-D1A astrocytes in layered microfluidic channels forming three-dimensional (3D) bi- and triculture models of the BBB. The live/dead assay indicated high viability of all cultured cells up to 21 days. Trans-endothelial electrical resistance (TEER) values confirmed the formation of intact monolayers after 3 days in culture and showed statistically higher values for the triculture model compared to the single and biculture models. Screening the permeability of [(14)C]-mannitol and [(14)C]-urea showed the ability of bi- and triculture models to discriminate between different markers based on their size. Further, permeability of [(14)C]-mannitol across the triculture model after 18 days in culture matched its reported permeability across the BBB in vivo. Mathematical calculations also showed that the radius of the tight junctions pores (R) in the triculture model is similar to the reported diameter of the BBB in vivo. Finally, both the bi- and triculture models exhibited functional expression of the P-glycoprotein efflux pump, which increased with the increase in the number of days in culture. These results collectively indicate that the triculture model is a robust in vitro model of the BBB.

3D brain angiogenesis model to reconstitute functional human blood–brain barrier in vitro

“…With the exception of a recent model developed by Campisi et al (), none of the models have achieved vascular permeability comparable to that of the in vivo BBB. Although our model showed a higher permeability than that found in vivo, the vascular permeability of our 3D tri‐cultured vasculature was lower than those of the majority of previous 3D (Adriani et al, ; Bang et al, ; Herland et al, ; Wang et al, ) or 2D models (Booth & Kim, ; Nakagawa et al, ). Importantly, the significant difference in permeability between monoculture and tri‐culture conditions in our system indicates that endothelial–perivascular interactions may have contributed to the barrier phenotype of the BBB vasculature (Figure ).…”

“…Booth and Kim (2012) presented the first microfluidic BBB model, which had a monolayer of mouse EC cocultured with astrocytes and was able to measure TEER in the presence of fluidic shear stress on the endothelium. A number of studies inspired by this model have employed microfluidic BBB models to study BBB functionality and drug delivery (Achyuta et al, 2013;Brown et al, 2015;Griep et al, 2013;Sellgren, Hawkins, & Grego, 2015;Wang, Khafagy el, Khanafer, Takayama, & ElSayed, 2016). These models demonstrated poor physiological resemblance to the human BBB due to the absence of human primary cells, and failed to accurately represent the 3D structure of the luminal vasculature.…”

Transepithelial/endothelial Electrical Resistance (TEER) theory and applications for microfluidic body-on-a-chip devices

“…Therefore, BBB models are critical in developing drugs that can permeate the BBB and affect the brain. Multiple in vitro models of the BBB have been developed 5,[14][15][16][17] , primarily consisting of endothelial cells and astrocytes seeded on opposite sides of a permeable membrane. In the blood-brain barrier system shown in Figure 4, monitoring of TEER enabled optimization of culture conditions and allowed for drug transport testing in the device 15 .…”

3D brain angiogenesis model to reconstitute maturation of functional human blood-brain barrier in vitro

“…The model introduced fluidic shear stress, which was impossible to attain in previous static models. A number of studies inspired by this model have employed microfluidic BBB models to study BBB functionality and drug delivery [11–15]. These models demonstrated poor physiological resemblance to the human BBB due to absence of human primary cells, and failed to accurately represent the 3D structure of the luminal vasculature within the surrounding 3D extracellular matrix (ECM) microenvironment.…”