La survenue de lésions hépatiques représente une cause majeure de retrait des médicaments au cours de leur développement et après leur mise sur le marché. La manifestation la plus fréquente des effets secondaires liés aux médicaments est lacholestase qui résulte d’un blocage de la sécrétion biliaire. Ils peuvent être prévisibles, généralement dépendants de la dose, ou dans certains cas n’être observés que chez un nombre restreint de patients traités, par exemple avec la chlorpromazine (CPZ), un neuroleptique. Il s’agit alors d’une hépatotoxicité idiosyncratique. Notre travail a eu pour but d’induire une cholestase avec ce médicament et d’étudier les mécanismes impliqués, en présence ou non d’un stress inflammatoire en utilisant comme modèle expérimental les cellules hépatiques différenciées HepaRG dérivées d’un cholangio-hépatocarcinome humain. Nous avons tout d’abord validé ce modèle en montrant que les principaux transporteurs d’influx et d’efflux canaliculaires et basolatéraux sont bien localizes dans les domaines membranaires appropriés, et que les canalicules biliaires sont fonctionnels et fermés comme dans les hépatocytes humains cultivés en sandwich, le modèle de référence. Le traitement par CPZ à une concentration élevée (50μM) entraine après 15min la génération d’un stress oxydant associé à une altération du potentiel membranaire mitochondrial et de la distribution péricanaliculaire des microfilaments de F-actine et à une inhibition de l’efflux canaliculaire de l’acide taurocholique. Après 24h, on observe notamment une inhibition de l’expression des deux principaux transporteurs canaliculaires, BSEP et MDR3, du transporter d’influx NTCP et une surexpression du transporteur basolatéral MRP4. Ces effets suggèrent une réponse compensatrice des cellules face à l’accumulation intracellulaire des acides biliaires. L’inflammation est considérée comme un facteur de susceptibilité dans l’hépatotoxicité idiosyncratique. Nous avons recherché si dans un context inflammatoire induit par l’IL-6 et l’IL-1β, les effets cytotoxiques et cholestatiques de CPZ sont aggravés. Après un prétraitement de 24h par les deux cytokines proinflammatoires, les cellules HepaRG, ont été co-exposées à 20μM CPZ pendant 1 à 5 jours. Bien que les cytokines aient induit un stress inflammatoire et inhibé le métabolisme de la CPZ et les transcrits de CYP3A4 et CYP1A2, deux principaux CYPs impliqués dans le métabolisme de ce médicament, la modulation des effets cytotoxiques et cholestatiques de la CPZ observés est restée limitée, y compris après 5 jours. Une cytotoxicité accrue de 20% et une amplification de l’inhibition des transcrits et de l’activité de NTCP ainsi que la dérégulation de l’expression d’autres gènes liés à la cholestase, ont été constatées suite au co-traitement à CPZ et aux cytokines. Au total, nos résultats montrent qu’il est possible d’induire une cholestase in vitro à partir des cellules HepaRG et que la cholestase induite par CPZ a pour origine l’induction d’un stress oxydant. Ils montrent en outre que l’étude de certains facteurs de susceptibilité peut être envisagée. / Drug-induced liver injury is the major cause of drug withdrawal during development and marketing process. The most common manifestation of adverse drug reactions is cholestasis, which results from alteration of bile flow. Adverse drug reactions are usually classified either as dose-dependent and reproducible (intrinsic) or unpredictable (idiosyncratic) occurring only in certain susceptible patients as observed with chlorpromazine (CPZ), a neuroleptic drug. Our work aimed to induce cholestasis with this drug and to study the mechanisms involved in the presence or absence of an inflammatory stress using differentiated HepaRG liver cells derived from a human cholangio-hepatocarcinoma as an experimental model. We firstly validated this cell model by demonstrating that the major canalicular and basolateral influx and efflux transporters are localized to the appropriate membrane domains, and that the bile canaliculi are functional and closed as in sandwichcultured human hepatocytes, the reference model. Treatment with CPZ at a high concentration (50μM) induces, as early as 15min, generation of oxidative stress which is associated with altered mitochondrial membrane potential, disruption of the pericanalicular F-actin cytoskeleton distribution and inhibition of canalicular efflux of taurocholic acid. After 24-hour treatment with CPZ, mRNA expression of the two main canalicular bile transporters, BSEP and MDR3, and of the main influx transporter, NTCP, was decreased. By contrast, expression of MRP4 mRNA, a basolateral transporter, was increased. These latter events likely represent hepatoprotective responses which aim to reduce intrahepatic accumulation of toxic BA. Inflammation is considered as a factor of susceptibility to idiosyncratic hepatotoxicity. We investigated whether in an inflammatory stress induced by IL-6 and IL-1β, cytotoxic and cholestatic effects of CPZ are exacerbated. After a 24 hour pre-treatment by either pro-inflammatory cytokines, HepaRG cells were co-exposed to 20μM CPZ for 1 to 5 days. Although cytokines have induced inflammatory stress and inhibited the metabolism of CPZ and transcripts of CYP3A4 and CYP1A2, two main CYPs involved in the metabolism of this drug, the modulation of cytotoxic and cholestatic effects of CPZ was limited, even after 5 daily treatments. Increased cytotoxicity by 20 %, amplification of NTCP mRNA and activity inhibition and deregulation of the expression of other genes associated with cholestasis, were observed in CPZ- and cytokine-co-treated cells. Altogether, our results show that it is possible to induce in vitro cholestasis using HepaRG cells and that CPZ-induced cholestasis depends on the generation of oxidative stress. They also show that the certain susceptibility factors may be investigated.

Biliary excretion is an important elimination route for many drugs and metabolites. For such compounds, it is important to know the extent of excretion and drug exposure in the bile, e.g., for the risk assessment of drug interactions, liver toxicity and the effects of genetic variants. In this thesis, duodenal aspiration of bile was performed in healthy volunteers and complemented with experiments in an in vivo model in pigs to increase the understanding of the intestinal and hepatobiliary disposition of two direct thrombin inhibitors. The compounds investigated, ximelagatran and AZD0837, are both prodrugs that require bioactivation to exert their pharmacological effect. Upon co-administration with erythromycin and ketoconazole, respectively, altered plasma exposure to ximelagatran and AZD0837 and their respective metabolites has been observed. The main objective of this thesis was to characterize the biliary excretion of the compounds, and investigate whether this elimination route explains the observed drug-drug interactions. High plasma-to-bile AUC ratios were observed, in particular for ximelagatran, its active metabolite melagatran, and AR-H067637, the active metabolite of AZD0837. These high ratios indicate the involvement of active transporters in the biliary excretion of the compounds, which is important since transporters constitute possible sites for drug interactions. The effects of erythromycin and ketoconazole on the plasma exposure of the prodrugs and metabolites were confirmed in both the pig and the clinical studies. The changes seen in plasma for ximelagatran and its metabolites were partly explained by reduced biliary clearance. Inhibited CYP3A4 metabolism likely caused the elevated plasma levels of AZD0837, whereas reduced biliary clearance was seen for AR-H067637 suggesting an effect on its excretion into bile. In summary, the studies led to mechanistic insights in the hepatobiliary disposition of ximelagatran and AZD0837, and demonstrate the value of combined clinical and animal studies for the investigation of the biliary drug excretion.

In vivo studies of hepatobiliary disposition are challenging. The hepatobiliary system is complex, as its physiological localization, complex cellular structure with numerous transporters and enzymes, and the interindividual variability in protein expression and biliary flow will all affect the in vivo disposition of a drug under investigation. The research included in this thesis has focused on the involvement of hepatic transport proteins in the hepatobiliary disposition of rosuvastatin. The impact that several transport inhibitors had on the pharmacokinetics of rosuvastatin was investigated in healthy volunteers and in pigs. The effects were considerable, following inhibition of sinusoidal transport proteins by cyclosporine and rifampicin. These inhibitors significantly reduced the hepatic extraction of rosuvastatin by 50 and 35%, respectively, and the plasma exposure increased by factors of 9.1 and 6.3, respectively. Drug-drug interactions (DDI) resulting in markedly higher plasma exposures are important from a drug safety perspective as increased extrahepatic exposure of statins is associated with an increased risk of severe side-effects, such as myopathy which in rare cases could develop into rhabdomyolysis. The DDI caused by cyclosporine and rifampicin can probably be attributed to inhibition of hepatic uptake transporters. In contrast, inhibition of canalicular transporters by imatinib did not significantly affect the pharmacokinetics of rosuvastatin, which suggests that the intracellular concentration of the inhibitor in the hepatocyte was insufficient to affect the transport of rosuvastatin, or that imatinib is not a sufficiently potent inhibitor in vivo. Furthermore, gemfibrozil administered as a single dose into the jejunum in healthy volunteers and pigs did not affect the plasma or biliary pharmacokinetics of rosuvastatin. The previously reported DDI in humans upon repeated dosing with gemfibrozil might be explained by the accumulation of metabolites able to affect the disposition of rosuvastatin. The investigations presented in this thesis conclude that transport proteins are of considerable importance for the hepatobiliary disposition of rosuvastatin in vivo. The Loc-I-Gut catheter can be applied for the investigation of biliary accumulation and to determine bile specific metabolites, however it has limitations when conducting quantitative measurements. In the porcine model, hepatic bile can be collected for up to six hours and enables the determination of the hepatic extraction in vivo.

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) techniques offer the opportunity to understand the physiological processes involved in the distribution of the contrast agent in vivo. This work utilises a liver specific contrast agent (gadoxetate) and demonstrates the potential use of these techniques to study transporter-mediated process in vivo. In vitro experiments investigated gadoxetate’s interaction with uptake and efflux transporters at the cellular level, ideally a prerequisite to understand the contribution of transporter proteins in in vivo pharmacokinetics. MRI was used to measure the accumulation of gadoxetate in fresh rat hepatocytes. Furthermore, LC-MS/MS methodology was optimised in conjunction with two in vitro systems to determine the role of transporters in gadoxetate uptake and efflux. HEK-OATP1B1 transfected cells were used to optimise the LC-MS/MS technique and Caco-2 cell monolayers were used to examine whether gadoxetate is a substrate of the efflux transporters, Pgp and BCRP. Subsequent studies demonstrated the use of DCE-MRI techniques to study transporter-mediated processes. Two pharmacokinetic models were proposed to quantify the uptake and efflux of gadoxetate in vivo. The suitability of the models in describing the liver concentration profiles of gadoxetate was assessed in pre-clinical and clinical reproducibility studies. Further pre-clinical experiments demonstrated the ability of the proposed DCE-MRI techniques to monitor changes in the uptake and efflux rate estimates of gadoxetate into hepatocytes, through co-administration of the transporter inhibitor, rifampicin, at two doses. The work presented demonstrates the potential use of DCE-MRI techniques as a diagnostic probe to assess transporter-mediated processes and drug-drug interactions (DDIs) in vivo.

No description available.