Abstract by Burak Ozgür

Introduction and aims
The barrier function of the brain capillaries (the blood-brain barrier, BBB) helps maintain a stable composition of the brain interstitial fluid by serving as a gatekeeper between the blood circulation and the brain parenchyma. The physical tightness of the barrier and the expression of efflux transporters hinder the brain entry of a large range of solutes, while nutrients, micronutrients, and hormones may enter via specialized transport proteins. The BBB presents a major obstacle when developing new drugs to treat central nervous system (CNS) disorders. In vitro models of the BBB that recapitulate the in vivo counterpart can provide valuable insights into the mechanisms from which drugs can enter the brain. The overall aim of the project was to generate in vitro models of the BBB, which showed similarity to the native BBB in terms of transporter expression, phenotype, and drug permeation properties. The work presented in this thesis dealt with the characterization and development of two BBB in vitro models: primary cultures of bovine brain endothelial cells and endothelial-like cells obtained from human-induced pluripotent stem cells (hIPSCs). The first sub-aim was to investigate whether the downregulation of transporters and BBB-markers reported for primary cultures of brain endothelial cells, could be rescued by co-culture with astrocytes or by culturing the endothelial cells in a hypoxic environment (1 % of O₂). The second aim was to investigate if endothelial-like cells, differentiated from hIPSCs, could be used to generate a functional BBB model. The resulting models were characterized in terms of barrier tightness, expression of BBB marker proteins, and functional expression of important transporters and receptors present at the brain endothelium.

Results
The primary bovine brain endothelial cells showed a decreased expression of transport proteins and endothelial phenotype markers when cultured following conventional protocols. The transcript levels of solute carriers and ABCB1 were downregulated in cultured brain capillary endothelial cells compared to those determined in fresh and intact brain capillaries. Co-culture with astrocytes resulted in inducing effects on the barrier tightness and the transcript levels of some genes, albeit the overall impact was minor. Interestingly, an enhanced BBB phenotype was obtained by growing and differentiating brain capillary endothelial cells under hypoxic conditions without compromising the barrier integrity. This is the first study that employs hypoxia during the growth and differentiation of primary endothelial cells to improve BBB phenotype. The hypoxia-differentiation led to upregulation of GLUT-1, LAT-1, P-gp, and TfR at the transcript level. The upregulation of GLUT-1, P-gp, and TfR could be confirmed at the protein level as well. The observed effects of hypoxia were found to be HIF-1α mediated since the chemical activation of HIF-1α using desferrioxamine (DFO) resulted in similar effects, and DFO actually increased the barrier tightness at the lower concentrations (up to 20 µM). Similarly, it was observed that exposure of mature monolayers of brain capillary endothelial cells to hypoxic conditions resulted in enhanced barrier properties coinciding with increased transcript levels of CLDN-5.

The hIPSC derived endothelial-like monolayers using two different stem cell lines (SBAD0201 and Bioni010-C) exhibited relatively high barrier tightness compared to other existing BBB in vitro models. The hIPSCs derived BBB models expressed various endothelial cell markers and tight junction proteins. However, relatively low expression levels of CLDN-5 were observed in both models, and CDH1 (E-cadherin) was detected at the mRNA level in SBAD0201 derived monolayers. The SLC transporters, GLUT-1 and LAT-1, were found in high abundance and functionally expressed, as demonstrated by transcellular transport studies using substrates and specific inhibitors of the two transporters. Interestingly, the efflux transporters, P-gp and BCRP, were not functionally expressed in hIPSC derived cell monolayers. Transcellular transport study using the [¹²⁵I]-Angiopep-2, sensitive to the addition of non-labeled Angiopep-2, indicated functional expression of the receptor system, LRP-1. 

Conclusions
The results obtained in this study confirm that primary brain capillary endothelial cells undergo a phenotypic alteration due to dedifferentiation during in vitro culture. Co-culture with rat astrocytes in the culture could not rescue the BBB phenotype in the endothelial cells. It was demonstrated for the first time that prolonged hypoxia during culture does not cause any changes to morphology or the tight junction integrity in cultured primary brain capillary endothelial cells. Cells grown and differentiated under hypoxia did actually exhibit enhanced BBB phenotype relative to cells cultured under normoxic conditions, as demonstrated by the upregulation of the BBB transporters, GLUT-1, P-gp, LAT-1, and TFR. These events were found to be HIF-1α mediated as the effects of hypoxia on the endothelial cells were mimicked by the addition of DFO at lower concentrations. The findings partly support why brain capillaries show functional barrier properties with a functional expression of various transporters during the early brain vascularization during embryogenesis. Furthermore, the study provides a novel methodological tool for enhancing the BBB phenotype in primary brain endothelial culture. Lastly, the observed functional tightness and functional expression of SLC-transporters and LRP-1 confirmed that the SBAD0201 and Bioni010-C derived cell monolayers may serve as a useful screening tool for investigations on SLC mediated transport as well as LRP1 mediated transcytosis. However, the models based on hIPSCs in their current state lack functional P-gp expression while also expressing epithelial cell markers. Thus, caution must be taken when studying different aspects of the BBB.