RESUMEN
The growing array of in vitro models of the blood-brain barrier (BBB) which have been used makes it difficult to draw firm conclusions concerning the BBB penetration of HIV-1 protease inhibitors. What is needed is a combined in vivo and in vitro study on biological models that mimic as closely as possible the normal human BBB, to establish whether and how indinavir crosses the BBB. We developed a new human BBB model using primary endothelial cells and astrocytes. The biological relevance of this model was checked with respect on the one hand, to the close relationship between the log of drug permeability coefficient normalized to molecular weight and the log of the 1-octanol/water partition coefficient, and on the other hand to the functional P-glycoprotein (P-gp) expression. We employed this model to perform transport studies with indinavir and showed that the rate of in vitro indinavir transport from the basal to apical compartment was higher than the rate of apical to basal transport. Pretreatment of the BBB model with the P-gp inhibitor, quinidine, significantly increased apical to basal transport. Intracellular indinavir accumulation was increased in BBB as a result of inhibition of active transport. These data were correlated with the indinavir-mediated P-gp ATPase modulation showing that indinavir specifically interacted with a binding site on P-gp. Moreover, the activation of P-gp ATPase by indinavir was inhibited by quinidine. In addition, the in vivo brain to plasma concentration ratio of indinavir into mice showed that indinavir concentration was up to five times higher in the brain of mdr1a(-/-) mice than in the brain of mdr1a(+/+) mice. All these results confirm the role of P-gp in preventing the passage of indinavir across BBB and thus its entry into the central nervous system (CNS). Our human BBB model represents a useful tool for the evaluation of drug penetration into the CNS.