Role of MDR3 in drug-induced cholestatic liver injury

Abstract

Bile formation is an important physiological process, coordinated by the concerted activity of transmembrane proteins, localized at the canalicular membrane of hepatocytes, belonging to the superfamily of ABC (ATP binding cassette) transporters. These transporters move the primary bile constituents into the canaliculus, a tight intercellular space between apical membranes of adjacent hepatocytes. BSEP (bile salts export pump) pumps bile salts; while MDR3 (multidrug resistance protein 3) and ABCG5/G8 translocate phosphatidylcholine and cholesterol, respectively, making them available for extraction by bile salts. Dysfunctions in the biliary bile salt, phospholipid and cholesterol secretion lead to cholestasis or result in cholesterol crystallization followed by cholelithiasis (gallstone disease). Several drugs may lead to cholestasis by functionally impairing the activity of canalicular transporters. While drug-induced cholestasis due to the inhibition of BSEP is a well investigated mechanism of acquired intrahepatic cholestasis, limited information on the interaction of drugs with lipid transporters such as MDR3 exists, and the role of MDR3 in the pathogenesis of drug-induced cholestasis is poorly understood. The aim of this project was to study the interaction of drugs in a newly established polarized cell line system that expresses the key players involved in canalicular secretion. In the first part of this work, the model LLC-PK1 cell line stably transfected with human NTCP (Na+-taurocholate cotransporting polypeptide), BSEP, MDR3 and ABCG5/G8 and grown in the Transwell® system was functionally characterized. The established cell line was shown to display vectorial bile salt transport and specific phosphatidylcholine secretion into the apical compartment. Hence this model system simulates the in vivo situation at the canaliculus. Subsequently, using the PC derivative C6-NBD-PC an assay for testing the effect of compounds on MDR3 activity was established and validated. In the second part of this work, several hepatotoxic as well as potential hepatoprotective (e.g. beneficial on biliary secretion or on bile ducts) drugs were tested in our model system. Of the several hepatotoxic drugs tested, the antifungal azoles, posaconazole, itraconazole and ketoconazole significantly inhibited MDR3-mediated phospholipid secretion as well as BSEP-mediated bile salt secretion. Surprisingly, in parallel to MDR3 inhibition, posaconazole and itraconazole induced MDR3 protein levels by an unknown mechanism. In contrast, other hepatotoxic drugs such as co-amoxicillin, troglitazone and nefadozone did not affect MDR3 function suggesting that the hepatotoxicity of these compounds is not due to the inhibition of this transporter. Similarly, also the potential hepatoprotective compounds tested did not affect MDR3 activity in our model. In conclusion, the generated in vitro model to study bile salt and phospholipid secretion is functional and can be considered an inexpensive and eproducible system that facilitate parallel screening for BSEP and MDR3 inhibitors, to study and predict drug-induced liver injury caused by inhibition of hepatobiliary transporters.