Article pubs.acs.org/molecularpharmaceutics
Mathematical Modeling of the in Vitro Hepatic Disposition of Mycophenolic Acid and Its Glucuronide in Sandwich-Cultured Human Hepatocytes Norikazu Matsunaga,†,‡ Sho Wada,† Takeo Nakanishi,† Miho Ikenaga,† Mikio Ogawa,‡ and Ikumi Tamai*,† †
Department of Membrane Transport and Biopharmaceutics, Faculty of Pharmaceutical Sciences, Institute of Medical, Pharmaceutical and Health Sciences, Kanazawa University, Kakuma-machi, Kanazawa, 920-1192, Japan ‡ Pharmacokinetic Research Laboratories, Ono Pharmaceutical Co., Ltd., 17-2, Wadai, Tsukuba, 300-4247, Japan ABSTRACT: In recent years, it has become increasingly important to test the safety of circulating metabolites of novel drugs as part of drug discovery and development programs. Accordingly, it is essential to develop suitable methods for identifying the major metabolites and their disposition in animal species and in humans. Mycophenolic acid (MPA), a selective inosine-5′-monophosphate dehydrogenase (IMPDH) inhibitor, is metabolized by glucuronidation and enterohepatic circulation of MPA-glucuronides is an important factor in the continuous systemic exposure of MPA. In humans, about 90% of the administered MPA dose is finally excreted as MPA phenyl-glucuronide (MPAG) in urine. Notably, the plasma concentration of MPAG is much higher than that of MPA. These factors suggest that, after its formation in hepatocytes, MPAG is excreted into bile and is also transported across the basolateral membrane to enter the circulation. In the present study, we performed metabolic/ hepatobiliary transport studies of MPA and MPAG using sandwich-cultured human hepatocytes (SCHH) and constructed mathematical models of their hepatic disposition. We also performed vesicular transport studies to identify which human multidrug resistance-associated proteins (MRPs) are involved in the transport of MPAG from hepatocytes. MPAG was a preferred substrate for the biliary excretion transporter MRP2 and the hepatic basolateral transporters MRP3 and MRP4 in conventional and metabolic/hepatobiliary transport studies using SCHH and vesicular transport studies using human MRPexpressing membrane vesicles. The resulting mathematical model suggested that the basolateral transport plays an important role in the hepatic disposition of MPAG formed in hepatocytes. Our findings suggest that mathematical modeling of metabolic/ hepatobiliary transport studies using SCH will provide useful information for determining the fate of metabolites formed in hepatocytes. KEYWORDS: sandwich-cultured hepatocytes (SCH), hepatic disposition, mathematical modeling, multidrug resistance-associated protein 2 (MRP2), multidrug resistance-associated protein 3 (MRP3)
■
INTRODUCTION The liver is a key organ involved in controlling the disposition of endogenous and exogenous compounds. Four factors determine hepatic disposition: basolateral uptake, basolateral efflux, intracellular metabolism, and biliary excretion. Characterization of hepatic disposition will provide useful information to help us understand the in vivo pharmacokinetics of many drugs. To determine these factors, several in vitro pharmacokinetic studies are performed using liver-derived fractions or liverspecific enzyme-expressing materials. Subcellular liver fractions, including microsomes and S9, are usually used to determine intrinsic metabolic clearance because of their high enzyme activities, their ease of preparation and preservation, and the ability to predict metabolic clearance in vivo. Although hepatocytes are ideal models for estimating hepatic disposition, their use has been hindered because of their limited availability and the large variability in enzyme activities among lots. Nevertheless, hepatocytes are increasingly being considered for such research following improvements in isolation techniques © 2013 American Chemical Society
that have increased the availability and quality of hepatocytes. In addition, cell culture techniques, including sandwich culture techniques, have remarkably improved. In sandwich culture, hepatocytes are cultivated between two gelled collagen layers. Sandwich-cultured hepatocytes (SCH) maintain their polarity, morphology, and liver-specific metabolic activities.1−3,54 Moreover, SCH develop functional canalicular networks sealed by tight junctions and express hepatic transport proteins that are localized to the correct membrane domains, allowing researchers to assess directional transport functions.4 Because of these unique features, SCH are useful for predicting intrinsic uptake clearance,5,6 intrinsic metabolic clearance,7,8 intrinsic biliary clearance,9,10 drug−drug interactions,11,12 and drug-induced hepatotoxicity.13,14 Several Received: Revised: Accepted: Published: 568
August 27, 2013 November 28, 2013 December 9, 2013 December 9, 2013 dx.doi.org/10.1021/mp400513k | Mol. Pharmaceutics 2014, 11, 568−579
Molecular Pharmaceutics
Article
metabolic/hepatobiliary transport studies have exploited the intrinsic metabolic clearance of SCH and revealed the hepatic disposition of a parent compound and its metabolites,15,16 as well as metabolite-mediated interactions17 and hepatotoxicity.18 Mycophenolic acid (MPA), administered as an ester prodrug (mofetil) or enteric-coated sodium salt, is an uncompetitive, selective, and reversible inosine-5′-monophosphate dehydrogenase (IMPDH) inhibitor that is widely used for immunosuppressive therapy after renal transplantation and in several autoimmune diseases.19,20 MPA is predominantly metabolized by glucuronidation in the liver.21,22 The major metabolite of MPA is MPA phenyl-glucuronide (MPAG), an inactive metabolite for IMPDH, and the minor metabolite is MPA acyl-glucuronide (AcMPA), which shows similar potency to MPA for IMPDH in humans.23,24 These MPA-glucuronides undergo biliary excretion, enter the enterohepatic circulation, and are reabsorbed as MPA, contributing to the continuous systemic exposure of MPA.25,26 However, the systemic exposure of MPAG is also much greater than that of MPA, and the administered MPA is eventually eliminated from the kidney as MPAG.19,27 These findings indicate that MPAG formed in hepatocytes is excreted into bile and is also transported into the circulation across the basolateral membrane. Several metabolic enzymes and transporters are involved in the hepatic disposition of MPA and MPAG. At least five isoforms of UDP-glucuronosyltransferase (UGT) have been implicated in the glucuronidation of MPA to MPAG, with UGT1A9 being the major isoform in the liver.28,29 MPAG is also a substrate of liver-specific organic anion transporting polypeptide (OATP)1B1 and OATP1B3.30,31 In addition, multidrug resistance-associated protein 2 (MRP2) is responsible for the biliary excretion of MPAG.32,33 Although MPAG formed in hepatocytes enters the circulation, the proteins involved in the basolateral efflux of MPAG have not been studied. Therefore, in the present study, we performed metabolic/hepatobiliary transport studies using sandwichcultured human hepatocytes (SCHH) to construct mathematical models of the hepatic disposition of MPA and MPAG. We also performed vesicular transport studies to identify which MRPs are involved in the transport of MPAG in human hepatocytes.
■
Table 1. Characteristics of Cryopreserved Human Hepatocytes lot no.a
age of donor (years)
other information
IZT GHA RTM
44 1 61
single donor, female, Caucasian single donor, female, Caucasian single donor, female, Caucasian
a
The transporter activity and plateability of each lot were confirmed by the supplier.
Sandwich Culture of Hepatocytes. Cryopreserved human hepatocytes were thawed according to supplier’s standard protocol (Life Technologies Corporation). After thawing at 37 °C, the hepatocytes were decanted into CHRM and centrifuged at 60g for 10 min. The supernatant was discarded. After resuspending the hepatocytes in CHPM (7 × 105 viable cells/mL), 0.5 mL of the hepatocyte suspension was added to each well of collagen I-coated BioCoat 24-well plates (BD Biosciences, Bedford, MA). After 4 h of static incubation under an atmosphere of 5% CO2 in air at 37 °C, the incubation medium was replaced to Geltrex (350 μg/mL) containing WME with hepatocyte maintenance supplement. Cultures were then maintained in Geltrex-free WME with hepatocyte maintenance supplement, which was changed every 24 h. SCHH Uptake of MPAG. SCHH on day 4 (Lots IZT, GHA, and RTM) were preincubated with 0.5 mL of Ca2+containing buffer (HBSS[Ca/Mg(+)] containing 20 mM HEPES) for 5 min in an incubator (37 °C, 5% CO2). Then, the buffer was removed, and SCHH were statically incubated with 0.3 mL of 1−1000 μM MPAG in Ca2+-containing buffer at 37 °C for 2 min, because the time-dependent uptake of MPAG was confirmed over 2 min. After incubation, the buffer was removed under suction, and each well was washed twice with the same ice-cold buffer. Then, the cells were destroyed by adding 300 μL of ethanol, and the ethanol was evaporated on a hot plate at 55 °C. Next, 100 μL of 2% SDS was added to each well to lyse the cells, and the lysate was diluted with 200 μL of water. Finally, 200 and 25 μL of the diluted lysate were used for LC-MS/MS (liquid chromatography with tandem mass spectrometry) analysis and to determine the protein concentration, respectively. Biliary Excretion of MPAG in SCHH. The biliary excretion index (BEI) was calculated using B-CLEAR technology (Qualyst, Inc., Research Triangle Park, NC) by dividing the difference in substrate accumulation between Ca2+-containing buffer (cellular plus canalicular accumulation) and Ca2+-free buffer (cellular accumulation) by the accumulation in Ca2+containing buffer.34 SCHH on day 4 were preincubated with 0.5 mL of Ca2+-containing buffer or Ca2+-free buffer (HBSS[Ca/Mg(−)] containing 20 mM HEPES and 1 mM EGTA) at 37 °C for 5 min under 5% CO2 in air. Then, the buffer was removed, and SCHH were incubated with 0.3 mL of the same buffer containing 1 μM MPAG at 37 °C for 3, 5, or 10 min in an incubator (37 °C, 5% CO2). The buffer was aspirated, and SCHH were washed twice with the same ice-cold buffer; the cells were destroyed by adding 300 μL of ethanol, which was evaporated as described above. Next, 100 μL of 2% SDS was added to each well to lyse the cells, and the lysate was diluted with 200 μL of water. Finally, 200 and 25 μL of the diluted lysate were used for LC-MS/MS analysis and to determine the protein concentration, respectively.
MATERIALS AND METHODS
Chemicals. MPA, MPAG, MPA-d3, and MPAG-d3 were purchased from Toronto Research Chemicals Inc. (North York, Canada). Cryopreserved hepatocyte recovery medium (CHRM), cryopreserved hepatocyte plating medium (CHPM), hepatocyte maintenance supplement pack, Geltrex, phenol red-free Williams’ media E (WME), and Hanks’ balanced salt solution (HBSS) were purchased from Life Technologies Corporation (Carlsbad, CA). Human MRP2, MRP3, MRP4, and MRP8-expressing Sf9 membrane vesicles were purchased from Genomembrane Inc. (Yokohama, Japan). All other chemicals and reagents were of special or HPLC grade. Human Hepatocytes. Cryopreserved human hepatocytes were purchased from Celsis In Vitro Technologies (Baltimore, MD). The supplier provided preliminary confirmation of the metabolic enzyme and transporter activities of the hepatocytes. The characteristics of each of the hepatocyte lot are shown in Table 1. 569
dx.doi.org/10.1021/mp400513k | Mol. Pharmaceutics 2014, 11, 568−579
Molecular Pharmaceutics
Article
Disposition of MPA and MPAG in SCHH. Metabolic/ hepatobiliary transport studies with SCHH were performed as previously described, with minor modifications.16 SCHH on day 4 (Lot IZT) were preincubated with Ca2+-containing or Ca2+-free buffer at 37 °C for 5 min and then statically incubated for 1, 3, 5, 10, 15, 20, 30, or 45 min with 0.3 mL of MPA (initial concentration of 1 μM) in the same buffer under 5% CO2 in air at 37 °C. At each time point, the entire volume of the buffer was collected, and the cells were rinsed twice with the same icecold buffer. The rinsed cells were immediately destroyed by adding 300 μL of ethanol, which was evaporated as described above. Then, 100 μL of 2% SDS was added to each well to lyse the cells, and the lysate was diluted with 200 μL of water. We used 200 and 25 μL of the diluted lysate for LC-MS/MS analysis and to determine the protein concentration, respectively. Additionally, 200 μL of the buffer was used for LC-MS/MS analysis. Membrane Vesicle Uptake. Vesicular transport studies were performed by a rapid filtration technique according to the manufacturer’s protocol (Genomembrane Inc.). First, we determined whether MPAG is a substrate of each human MRP by incubating 10 μM MPAG with human MRP2, MRP3, MRP4, or MRP8-expressing Sf9 membrane vesicles (0.5 mg/ mL) at 37 °C for 5 min in the presence of 2 mM glutathione and cofactor (4 mM ATP or AMP) in transport buffer (70 mM KCl, 7.5 mM MgCl2, and 50 mM MOPS-Tris, pH 7.0). After incubation, 40 μL of the reaction mixture was collected and added to 5 volumes of ice-cold stopping buffer, consisting of 70 mM KCl and 40 mM MOPS-Tris (pH 7.0). The terminated reaction mixture was filtered through a presoaked 0.45 μm membrane filter (Millipore, Bradford, MA), and the filter was washed five times with 0.35 mL of ice-cold stopping buffer. The amount of MPAG retained on the membrane filter was measured by LC-MS/MS. MPAG uptake over time was linear until 2 min for human MRP2 and MRP4 and until 40 s for human MRP3. Accordingly, 10−1000 μM MPAG was incubated with each of the MRP-expressing membrane vesicles (0.5 mg/mL) at 37 °C for 2 min (human MRP2 and MRP4) or 0.5 min (human MRP3) to assess the concentration dependence. Analytical Methods. For hepatocyte studies, internal standard in methanol (IS solution) was added to an aliquot of the cell lysate or the buffer, and the resultant solution was diluted 10-fold with 10% methanol. The diluted solution was applied to an OASIS HLB column (Waters, 3 cc, 60 mg) that had been sequentially preconditioned with 2 mL each of methanol and water. The column was washed twice with 2 mL of water and then twice with 2 mL of 10% methanol. The analytes were eluted twice with 2 mL of methanol, and the eluate was evaporated to dryness. The residue was reconstituted with 10 mM ammonium acetate (pH 3.5)/acetonitrile (4:1, v/ v) and injected into a LC-MS/MS system. For vesicular transport studies, the membrane filter trapping the vesicles was collected in a centrifuge tube and dissolved with acetonitrile. Then, the IS solution and methanol were added to the tube and centrifuged (10 000g, 3 min). The entire volume of the supernatant was collected in a glass tube and diluted 10-fold with water. The diluted solution was treated as described for hepatocyte studies and injected into the LC-MS/MS system. Calibration standards (5−1000 nM for hepatocyte studies, 10− 2000 nM for vesicular transport studies) and quality control (QC) standards (three concentrations; low [QCL], middle [QCM], and high [QCH]) were pretreated as described for the
analytical samples. Calibration curves were constructed by the linear least-squares regression (1/x weighting) method of the peak area ratio (analyte/IS) versus nominal concentration. Results obtained for a double-blank sample (neither analyte or IS added) and a blank sample (only IS) were not included in the calibration curve. The back-calculated concentration standards met the following acceptable criteria:
