Evaluation of Blood-Brain Barrier In Vitro Models and Application for Studying Barrier Disruption Induced by Gram-Positive Bacteria

The blood-brain barrier (BBB), located at the level of brain capillary endothelial cells (BCECs), separates the cerebral compartment from the systemic circulation, playing a fundamental role in maintaining the homeostasis of the central nervous system (CNS). The important role that BBB plays both under physiological and pathological conditions lead scientist to look for in vitro models to study the cellular and molecular mechanisms responsible for the permeability variations of this barrier. In regulatory toxicology and in the context of the current European Union political scenario, the development and validation of in vitro models is of outmost importance. The European Centre for the Validation of Alternative Methods (ECVAM) organised in 2003 a workshop on BBB in vitro methods and their application in toxicology; to discuss the in vitro models available and their application in integrated testing strategies. Taking into consideration the outcomes of this workshop and according to the 3Rs concept (reduction, refinement and replacement), we replaced in a well-established BBB in vitro model the primary glial cells (GCs) necessary for the differentiation of BCECs with the C6 glial cell line, to avoid the use of animals. For the first time, we compared directly the structural and functional differentiation of BCECs induced by C6 cells towards GCs. Trans-endothelial electrical resistance (TEER) measurements showed that in the presence of C6 cells the values were always lower than in the presence of GCs. Permeability of the BCECs to both radioactive sucrose and FITCinulin was 2.5-fold higher when cells were co-cultured with C6 than with GCs. Immunocytochemistry studies showed less developed tight junction pattern in the presence of C6. P-gp expression and activity were lower in BCECs co-cultured with C6 than with GCs. The levels of VEGF in the culture medium were 40-fold higher in the presence of C6, suggesting that VEGF was one of the factors responsible for impairing the endothelial barrier co-cultured with C6 cells. Therefore, C6 cell line failed to replace primary GCs in a reliable BBB in vitro model. Furthermore, we used the BBB model consisting of BBCECs co-cultured with primary GCs to investigate the effects of the Gram-positive bacterial cell wall components lipoteichoic acid (LTA) and muramyl dipeptide (MDP) on the structure and function of BBB in vitro. The activation of GCs with LTA disrupted BBB integrity and LTA effect was potentiated by MDP. Immunocytochemistry analysis for tight junction associated 8 proteins showed a delocalisation of AHNAK, revealing that LTA altered the tight junction pattern. LTA-activated GCs produced nitric oxide (NO) and the proinflammatory cytokines TNF-a and IL-1b, which contributed to LTA-induced BBB disruption, since the direct treatment of the endothelial monolayer with TNF-a, IL-1b or a NO donor increased BBB permeability. In addition, the pre-treatment of LTAactivated GCs with antibodies against these two cytokines blocked LTA effect and the presence of 1400W, inhibitor of inducible NO synthase (iNOS), partially reversed LTAinduced decreased TEER. This study showed for the first time that LTA impaired the BBB in vitro through glia activation and suggested that free-radical scavengers and inhibitors of iNOS and of proinflammatory cytokines could be major targets for the adjunctive therapy of CNS pathologies induced by Gram-positive bacteria.

BOVERI Monica; 

2006-05-22

JRC33179