The human blood brain barrier has yet to be successfully replicated as an in vitro model. One of the more promising approaches has been to develop an in vitro model derived from human pluripotent stem cells. However, as promising as this model may be, a successful replication of the differentiation method on different kinds of pluripotent stem cell lines have yet to be accomplished. We try to approach the promising method as described by Stebbins et al. (2015) to differentiate human pluripotent stem cells into brain like endothelial cells (BECs).
Five different human pluripotent stem cell lines were screened for the possibility to differentiate into BECs. Tüb1159, Tüb16423, Bioni010-C, WTSli024-A and WTSli002-A stem cell lines were initially seeded on Matrigel cultured with mTESR1 media to confluence, then seeded on Matrigel as a single cell suspension. After two-three days of culture we applied undifferentiated media; DMEM/F12 containing NEAA, glutaMAX, KOSR and β-mercaptoethanol for six days to initiate differentiation. The differentiated stem cells were seeded on permeable supports coated with collagen IV and fibronectin (250.000 cells pr. filter insert) in different culture configurations (mono culture, non-contact co-culture and contact co-culture) with primary rat astrocytes to induce barrier-like properties. Endothelial cell media supplemented with retinoic acid were then applied to the cells to ensure selective expansion of BECs. The different culture configurations were observed for several days after seeding by measuring the trans-endothelial electrical resistance (TEER).
Initial pilot studies have shown a significant difference between stem cells grown as a mono culture and stem cells grown in co-culture with rat astrocytes. The stem cell line WTSli024-A had a significant higher TEER grown as a co-culture compared to mono culture (393± 155 Ω*cm2 and 96± 30 Ω*cm2, respectively). The highest TEER were obtained on day three after cell seeding on permeable supports. Confocal microscopy images of the different culture configurations showed the presence of tight junction proteins Claudin-5 and Occludin as well as efflux transporter P-gp, but did, however, not show a confluent monolayer.
Further studies will include testing other seeding cell densities to see whether this could have an impact on the confluence of the monolayer as well as the investigation of the mRNA- and protein expression levels by qPCR and Western Blot.