Search

Abstract Background Traffic-generated air pollution-exposure is associated with adverse effects in the central nervous system (CNS) in both human exposures and animal models, including neuroinflammation and neurodegeneration. While alterations in the blood brain barrier (BBB) have been implicated as a potential mechanism of air pollution-induced CNS pathologies, pathways involved have not been elucidated. Objectives To determine whether inhalation exposure to mixed vehicle exhaust (MVE) mediates alterations in BBB permeability, activation of matrix metalloproteinases (MMP) -2 and −9, and altered tight junction (TJ) protein expression. Methods Apolipoprotein (Apo) E −/− and C57Bl6 mice were exposed to either MVE (100 μg/m3 PM) or filtered air (FA) for 6 hr/day for 30 days and resulting BBB permeability, expression of ROS, TJ proteins, markers of neuroinflammation, and MMP activity were assessed. Serum from study mice was applied to an in vitro BBB co-culture model and resulting alterations in transport and permeability were quantified. Results MVE-exposed Apo E −/− mice showed increased BBB permeability, elevated ROS and increased MMP-2 and −9 activity, compared to FA controls. Additionally, cerebral vessels from MVE-exposed mice expressed decreased levels of TJ proteins, occludin and claudin-5, and increased levels of inducible nitric oxide synthase (iNOS) and interleukin (IL)-1β in the parenchyma. Serum from MVE-exposed animals also resulted in increased in vitro BBB permeability and altered P-glycoprotein transport activity. Conclusions These data indicate that inhalation exposure to traffic-generated air pollutants promotes increased MMP activity and degradation of TJ proteins in the cerebral vasculature, resulting in altered BBB permeability and expression of neuroinflammatory markers.

Brain penetration is characterized by its extent and rate and is influenced by drug physicochemical properties, plasma exposure, plasma and brain protein binding and BBB permeability. This raises questions related to physiology, interspecies differences and in vitro/in vivo extrapolation. We herein discuss the use of in vitro human and animal BBB model as a tool to improve CNS compound selection. These cell-based BBB models are characterized by low paracellular permeation, well-developed tight junctions and functional efflux transporters. A study of twenty drugs shows similar compound ranking between rat and human models although with a 2-fold higher permeability in rat. cLogP < 5, PSA < 120 Å, MW < 450 were confirmed as essential for CNS drugs. An in vitro/in vivo correlation in rat (R² = 0.67; P = 2 × 10⁻⁴) was highlighted when in vitro permeability and efflux were considered together with plasma exposure and free fraction. The cell-based BBB model is suitable to optimize CNS-drug selection, to study interspecies differences and then to support human brain exposure prediction.

We have evaluated a novel in vitro cell-based human blood-brain barrier (BBB) model that could predict in vivo human brain penetration for compounds with different BBB permeabilities using the clinical positron emission tomography (PET) data. Comparison studies were also performed to demonstrate that the in vitro cell-based human BBB model resulted in better predictivity over the traditional permeability model in discovery organizations, Caco-2 cells. We evaluated the in vivo BBB permeability of [(18)F] and [(11)C]-compounds in humans by PET imaging. The in vivo plasma-brain exchange parameters used for comparison were determined in humans by PET using a kinetic analysis of the radiotracer binding. For each radiotracer, the parameters were determined by fitting the brain kinetics of the radiotracer using a two-tissue compartment model of the ligand-receptor interaction. Bidirectional transport studies with the same compounds as in in vivo studies were carried out using the in vitro cell-based human BBB model as well as Caco-2 cells. The in vitro cell-based human BBB model has important features of the BBB in vivo and is suitable for discriminating between CNS and non-CNS marketed drugs. A very good correlation (r(2) = 0.90; P < 0.001) was demonstrated between in vitro BBB permeability and in vivo permeability coefficient. In contrast, a poor correlation (r(2) = 0.17) was obtained between Caco-2 data and in vivo human brain penetration. This study highlights the potential of this in vitro cell-based human BBB model in drug discovery and shows that it can be an extremely effective screening tool for CNS programs.