The blood-brain barrier (BBB), located at the level of brain capillaries, is a crucial component of the neurogliovascular unit where it is responsible for brain homeostasis maintenance. By limiting the access of molecules to the brain parenchyma, it effectively protects the central nervous system (CNS) from harmful substances, but at the same time represents a major hurdle for potential neuropharmaceuticals to reach their central target. The difficulty to study BBB features and permeabilityin vivo led to the development of different in vitro BBB models. Our model, consisting of a coculture of brain capillary endothelial cells and glial cells, has been extensively characterized: it provides a robust model exhibitingin vivo BBB characteristics and has proved useful in elucidating the cellular interactions at the level of the BBB in physiological and pathological conditions. The initially developed coculture model was completed by the addition of brain pericytes to design three-cell culture models. Indeed, pericytes now appear as important actors of BBB formation and maintenance, but are often absent ofin vitro BBB models. Gathering the three major cell populations forming the BBB - endothelial cells, glial cells and pericytes - seems important to more accurately reproducein vivoconfiguration, with the aim of understanding cellular interactions in physiological and pathological conditions. On the basis of our original coculture model, two different three-cell culture models were designed: while pericytes were cultured distant from endothelial cells in the first model, both were closely associated in the second one. Both models were characterized in terms of endothelial marker expression and morphology, paracellular permeability and expression of efflux pumps, demonstrating that they provide reliable in vitro BBB models. They may be useful in deciphering the contribution of pericytes to the BBB phenotype and in the response of BBB to injury, taking into account the gliovascular component. In most cases, the intended use of in vitro BBB models in drug discovery is to predict whether investigational drugs are likely to achieve relevant CNS exposure to elicit the desired pharmacological effect. However, in vitro BBB models usually do not allow high enough throughput to efficiently evaluate the large number of compounds generated by pharmaceutical companies in early drug discovery stages. Adapted from our original co-culture model, a firstin vitro BBB model has been developed to meet the increasing need for BBB screening solutions, by reducing its format (24-well instead of 6-well format), and limiting time, cost and technical needs. Nevertheless, the associated procedure still required some technical skills to achieve a properin vitro BBB model, related to the critical trypsinization step of endothelial cells. We aimed at making this model even simpler, by providing this model in a frozen format that allows the final user to dedicate minimal time and resource. All tested BBB features attest of the reliability of this newin vitro BBB model to support drug discovery in pharmaceutical but also in smaller biotech companies. The work presented herein aimed at adapting our original coculture model to address different issues. On the one hand, simplifying the associated procedure allows to produce BBB permeability data at a large scale in early drug discovery. On the other hand, integrating pericytes in the model allows to study cellular interactions within the gliovascular component. These models may therefore respectively contribute to the development of neurotherapeutics with improved BBB permeability, and to a better knowledge of cellular intercommuni- cations within the BBB in health and disease.
