Comparison of three in vitro models expressing the membrane drug transporter OATP1B1

Abstract

Background: Drug transporters have been increasingly recognized as important determinants of variable drug disposition and response. For the pharmaceutical industry it is of importance to develop in vitro models, which could predict possible transporter-associated genetic variability and drug-drug interactions of a new compound. The purpose of this study was to compare uptake kinetics of estradiol-17â-D-glucuronide, atorvastatin and pravastatin in three in vitro models expressing the membrane drug transporter OATP1B1.

Methods: Two in vitro models overexpressing OATP1B1 were applied in this study; HEK293 cells transfected with the gene encoding OATP1B1 and Xenopus laevis oocytes injected with OATP1B1 cDNA. HEK293 cells transfected with the empty vector pT-REX and water injected oocytes were used as control systems. HepaRG cells, the third model applied in this study, express naturally occurring levels of OATP1B1 and other transporters, and studies at 4°C were used as control. Uptake studies were performed with three known substrates for OATP1B1; estradiol-17â-D-glucuronide, atorvastatin and pravastatin. Time- and concentration dependent uptake was studied. Km values were estimated by non-linear regression analysis of the active uptake calculated in concentration dependence studies. The active uptake was calculated as the difference between intrasystemic accumulation of test substance in the model system and in the respective control system.

Results: Estradiol-17â-D-glucuronide exhibited active uptake in all time- and concentration dependence uptake studies performed on the three models. The active transport of estradiol-17â-D-glucuronide was saturated in the concentration dependence studies, with estimated Km values of 5.6 ± 0.3 ìM, 5.9 ± 3.8 ìM and 22.3 ± 7.1 ìM in HEK293 cells, oocytes and HepaRG cells, respectively. Atorvastatin showed active uptake in all time dependence uptake studies. In the concentration dependence uptake studies, atorvastatin exhibited no active uptake in HEK293 cells and oocytes, while in the HepaRG cells, a non-saturated active uptake was demonstrated. In the time dependence uptake studies with pravastatin, an active uptake was shown in HEK293 and HepaRG cells, while a limited active uptake was demonstrated in the oocytes. On the contrary, concentration dependence uptake studies showed no active uptake of pravastatin in HEK293 and HepaRG cells, but a saturated active uptake in oocytes with estimated Km value of 54.8 ± 31.4 ìM.

Discussion and conclusion: Estradiol-17â-D-glucuronide was the only substrate showing consistent time- and concentration dependent results in all tested models. Results from the studies with atorvastatin and pravastatin were inconsistent between the models. Furthermore, concentration dependence and saturation of uptake was difficult to obtain even if a time dependent uptake was recorded. The inconsistent results with atorvastatin and pravastatin could be due to the higher lipophilicity and greater extent of passive diffusion compared to estradiol-17â-D-glucuronide. It could therefore be hypothesized that when a substrate exhibits greater passive diffusion than active uptake, the models would not be sufficiently sensitive to separate the active uptake from the passive uptake. As drugs are generally more lipophilic than estradiol-17â-D-glucuronide, none of the in vitro transport models studied are expected to be appropriate for routine screening of new drugs as possible substrates for OATP1B1.