<p>This thesis examines the transport of exogenous molecules across the blood-brain barrier (BBB), focusing on active efflux, using positron emission tomography (PET), computer simulation and modelling. P-glycoprotein (P-gp) inhibition was studied using [¹¹C]verapamil and [¹¹C]hydroxyurea was investigated as a new marker for active efflux transport. Simulations were carried out to explore the importance of the efflux transporter location in the BBB. Brain concentrations of [¹¹C]verapamil, [¹¹C]GR205171 and [¹⁸F]altanserin were compared in various laboratory animal species and in humans.</p><p>A central aspect of the studies has been the novel combination of dynamic PET imaging of the brain pharmacokinetics of a labelled drug, administered through an exponential infusion scheme allowing time-resolved consequence analysis of P-gp inhibition, and mathematical modelling of the obtained data. The methods are applicable to drugs under development and can be used not only in rodents but also in higher species, potentially even in humans, to investigate the effects of P-gp or other transporters on drug uptake in the brain.</p><p>The inhibition of P-gp by cyclosporin A (CsA) and the subsequent change in brain concentrations of [¹¹C]verapamil occurred rapidly in the sense that [¹¹C]verapamil uptake increased rapidly after CsA administration but also in the sense that the increased uptake was rapidly reversible. The P-gp inhibition was best described by an inhibitory indirect effect model in which CsA decreased the transport of [¹¹C]verapamil out of the brain. The model indicated that approximately 90% of the transport of [¹¹C]verapamil was P-gp-mediated. The low brain concentrations of [¹¹C]hydroxyurea appeared to be a result of slow transport across the BBB rather than active efflux. This exemplifies why the extent and the rate of brain uptake should be approached as two separate phenomena. The brain-to-plasma concentration ratios for the three studied radiotracers differed about 10-fold be-tween species, with lower concentrations in rodents than in humans, monkeys and pigs. The increase in brain concentrations after P-gp inhibition was somewhat greater in rats than in the other species. </p><p>The findings demonstrate a need to include the dynamics of efflux inhibition in the experimental design and stress the importance of the choice of species in preclinical studies of new drug candidates. </p>

This thesis examines the transport of exogenous molecules across the blood-brain barrier (BBB), focusing on active efflux, using positron emission tomography (PET), computer simulation and modelling. P-glycoprotein (P-gp) inhibition was studied using [11C]verapamil and [11C]hydroxyurea was investigated as a new marker for active efflux transport. Simulations were carried out to explore the importance of the efflux transporter location in the BBB. Brain concentrations of [11C]verapamil, [11C]GR205171 and [18F]altanserin were compared in various laboratory animal species and in humans. A central aspect of the studies has been the novel combination of dynamic PET imaging of the brain pharmacokinetics of a labelled drug, administered through an exponential infusion scheme allowing time-resolved consequence analysis of P-gp inhibition, and mathematical modelling of the obtained data. The methods are applicable to drugs under development and can be used not only in rodents but also in higher species, potentially even in humans, to investigate the effects of P-gp or other transporters on drug uptake in the brain. The inhibition of P-gp by cyclosporin A (CsA) and the subsequent change in brain concentrations of [11C]verapamil occurred rapidly in the sense that [11C]verapamil uptake increased rapidly after CsA administration but also in the sense that the increased uptake was rapidly reversible. The P-gp inhibition was best described by an inhibitory indirect effect model in which CsA decreased the transport of [11C]verapamil out of the brain. The model indicated that approximately 90% of the transport of [11C]verapamil was P-gp-mediated. The low brain concentrations of [11C]hydroxyurea appeared to be a result of slow transport across the BBB rather than active efflux. This exemplifies why the extent and the rate of brain uptake should be approached as two separate phenomena. The brain-to-plasma concentration ratios for the three studied radiotracers differed about 10-fold be-tween species, with lower concentrations in rodents than in humans, monkeys and pigs. The increase in brain concentrations after P-gp inhibition was somewhat greater in rats than in the other species. The findings demonstrate a need to include the dynamics of efflux inhibition in the experimental design and stress the importance of the choice of species in preclinical studies of new drug candidates.

Pseudomonas aeruginosa est un pathogène opportuniste majeur de l'Homme capable de mettre en jeu tout un ensemble de mécanismes pour résister aux antibiotiques. Parmi eux, le système d'efflux actif MexXY(OprM) s'oppose, lorsqu'il est surproduit, à l'accumulation intracellulaire de différents composés dont certains, comme les aminosides et les fluoroquinolones, sont largement utilisés dans le traitement des infections à P. aeruginosa. Chez les souches cliniques, la surproduction de la pompe MexXY résulte de mutations soit dans le gène répresseur de l'opéron mexXY, mexZ (mutants agrZ) soit dans des loci génétiques encore inconnus (mutants agrUI). Au cours de ce travail, nous avons carcatérisés deux types de mutants agr W dérivés de la souche de référence PAO 1. Les premiers, agr WI, présentent une augmentation de la résistance aux antibiotiques substrats de la pompe Mex.XY comparable à celle observée chez les mutants agrZ tandis que les seconds, agrW2, sont en plus résistants aux carbapénèmes et aux peptides cationiques ( colistine ). Par une approche de séquençage à haut débit des génomes, nous avons identifié chez les mutants agrWI une mutation dans deux des quatre allèles codant pour la sous-unité ribosomale 23S et chez les mutants agrW2 une mutation dans le régulateur de réponse d'un système à deux composants dénommé ParR. A l'aide d'expériences de RT-qPCR, d'inactivation et de complémentation génique deux voies distinctes d'activation de MexXY/OprM ont été identifiées. Parallèlement, la comparaison des transcriptomes globaux des mutants agrWI, agrW2 avec celui de la souche PAOl nous a permis de mieux comprendre dans quel processus cellulaire s'intègre la pompe Mex.XY /OprM. / Pseudomonas aeruginosa is a nosocomial pathogen naturally resistant to many antibiotics thanks to numerous resistant mechanisms. Among them, overproduction of the MexXY/OprM efflux system leads to decrease significantly the susceptibility of P. aeruginosa to aminoglycosides and fluoroquinolones. In clinical strains, upregulation of this pump often results from mutations occurrinJ in mexZ ( agrZ mutants), the local repressor gene of the mexXY operon. Analysis of MexXY­overproducing mutants selected in vitro from the reference strain PAO 1 led to identification of two new classes of mutants (agrWmutants) harboring an intact mexZ gene. The first, named agrWI mutants, shows an increase resistance to Mex.XY substrates similar to that observed in agrZ mutants while the second, dubbed agrW2, are more resistant to carbapenems and cationic peptides (colistin) in addition to aminoglycosides and fluoroquinolones. Whole-genome sequencing experiments revealed in agrWI mutants a mutation in two of the four alleles encoding the 23S ribosomal subunit and in agr W2 mutants, a mutation in the response regulator of a two-component system called ParR. By using RT-qPCR, inactivation and complementation experiments, two distinct activation pathway of the MexXY /OprM efflux system have been identified. Meanwhile, transcriptomic profiles of agrWJ and agrW2 mutants compared with the PAOl reference strain has allowed us to better understand the physiologie function of the MexXY/OprM efflux pump