Abstract
Liver and bile secretion can be an important first-pass and clearance route for drug compounds and also the site of several drug-drug interactions. In the clinical program for ximelagatran development, an unexpected effect of erythromycin on the pharmacokinetics of the direct thrombin inhibitor ximelagatran and its metabolites was detected. This interaction was believed to be mediated by inhibition of drug transporters, which normally extrude the drug into the bile. Previous Caco-2 cell experiments indicated the involvement of an active efflux mechanism for ximelagatran, hydroxy-melagatran, and melagatran possibly mediated by P-glycoprotein (P-gp). However, the inhibitors used may not have been specific enough and the possibility that transporters other than P-gp were important in the Caco-2 cell assay cannot be excluded. In this study we used RNA interference, a post-transcriptional gene silencing mechanism in which mRNA is degraded in a sequence-specific manner, to specifically knock down P-gp or multidrug resistance-associated protein 2 (MRP2) transporters in Caco-2 cells. The data obtained from bidirectional transport studies in these cells indicate a clear involvement of P-gp but not of MRP2 in the transport of ximelagatran, hydroxy-melagatran, and melagatran across the apical cell membrane. The present study shows that short hairpin RNA Caco-2 cells are a valuable tool to investigate the contribution of specific transporters in the transcellular transport of drug molecules and to predict potential sites of pharmacokinetic interactions. The results also suggest that inhibition of hepatic P-gp is involved in the erythromycin-ximelagatran interaction seen in clinical studies.
The direct thrombin inhibitor ximelagatran was envisaged to have a low potential for pharmacokinetic drug-drug interactions (DDIs) based on the preclinical information showing that ximelagatran, its intermediate metabolites, and melagatran are neither substrates nor inhibitors of the major cytochrome P450 isoenzymes (Bredberg et al., 2003). The preclinical results were supported by the results from three specific healthy volunteer studies showing no clinically significant DDIs between ximelagatran and diclofenac, diazepam, or nifedipine when coadministered simultaneously (Bredberg et al., 2003). The phase I DDI study showing a clear interaction with erythromycin, therefore, came as a surprise. Melagatran area under the curve increased by 82%, whereas ximelagatran plasma levels seemed to be essentially unchanged. Ximelagatran is bioconverted to melagatran via two intermediates, hydroxy-melagatran and ethyl-melagatran (Fig. 1), and an elevation in the plasma concentrations was seen for both intermediates, after coadministration with erythromycin (Eriksson et al., 2006).
Several in vivo and in vitro studies have been performed to investigate the nature of the erythromycin-ximelagatran interaction (Eriksson et al., 2006; Sjödin et al., 2008). The pharmacokinetic profile in humans indicated that the elevation of ximelagatran and its metabolites in plasma caused by erythromycin occurs during the first pass through the liver (Eriksson et al., 2006). Loc-I-Gut studies both in pigs and humans (Matsson E, Eriksson UG, Huffman KJ, Logan U, Petri N, Fridblum P, and Lennernäs H, manuscript submitted for publication) indicate that erythromycin reduced the biliary secretion of ximelagatran and its metabolites, suggesting that the interaction is likely to be due to inhibition of hepatic transporters (Sjödin et al., 2008) (Fig. 2). The Loc-I-Gut study in pigs also indicated that the intestinal absorption of ximelagatran was not affected. Furthermore, Sjödin et al. (2008) suggested that the erythromycin interaction was not likely to have been caused by direct inhibition of ximelagatran metabolism. The hypothesis that the effect of erythromycin on ximelagatran pharmacokinetic is due to inhibition of biliary transporters is further supported by the Caco-2 cell results showing the presence of an active efflux for ximelagatran, hydroxy-melagatran, and melagatran, possibly mediated by P-glycoprotein (P-gp) (Eriksson et al., 2006). However, Caco-2 cells express a number of transporter proteins and the contribution of different biliary transporters needs to be further investigated.
The aim of the present study was to use the short hairpin RNA (shRNA) Caco-2 knockdowns as a tool to identify biliary efflux transporters and predict potential sites of transporter-mediated pharmacokinetic interactions. Stable Caco-2 MDR1 (ABCB1)/P-gp and MRP2 (ABCC2) knockdowns were established using a lentiviral vector expressing shRNA against MDR1 and MRP2, respectively. The effect of P-gp knockdown on the transport of ximelagatran and its metabolites was studied to investigate whether the interaction seen in vivo between erythromycin and ximelagatran was mediated by P-gp. Furthermore, MRP2 knockdown Caco-2 cells were used to study the contribution of MRP2 in the bidirectional transport of ximelagatran and its metabolites.
Materials and Methods
Chemicals.
Lucifer yellow, Hanks' balanced salt solution (HBSS), Dulbecco's modified Eagle's medium, Dulbecco's phosphate-buffered saline, and HEPES were obtained from Invitrogen (Carlsbad, CA). Fetal bovine serum was obtained from Omega (Tarzana, CA). Penicillin-streptomycin, nonessential amino acids, l-glutamine, and trypsin were obtained from cellgro (Herndon, VA). [3H]Bromosulfophthalein ([3H]BSP) was obtained from International Isotopes Clearing House (Leawood, KS). Digoxin was purchased from Sigma-Aldrich (St. Louis, MO). Ximelagatran, hydroxy-melagatran, ethyl-melagatran, melagatran, and their respective 2H- and 13C-labeled internal standards were supplied by AstraZeneca (Mölndal, Sweden). Standard solutions of the analytes and their internal standards were prepared in 0.01 mol/l HCl. HPLC-grade acetonitrile was purchased from Rathburn (Walkerburn, UK). Ammonium acetate and HCl (Titrisol) were purchased from Merck (Darmstadt, Germany). High-purity water was obtained from an ELGA purification system (ELGA, High Wycombe, UK).
Transduction of Caco-2 Cells with shRNAs.
Parental Caco-2 cells were obtained from American Type Culture Collection (Manassas, VA). The cells were maintained in Dulbecco's modified Eagle's medium with 4.5 g/l glucose, supplemented with 10% fetal bovine serum, 1% nonessential amino acids, 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin at 37°C in a humidified atmosphere of 5% CO2 and 95% air. Lentivirus plasmid vectors containing shRNA inserts targeting human P-gp (GenBank accession number NM_000927) and MRP2 (GenBank accession number NM_000392) genes were obtained from Sigma-Aldrich. The transduction procedure was described previously (Zhang et al., 2009). In brief, the cells were plated in 96-well tissue culture plates at a density of 5 × 106 cell/well and transduced with 1 μg of P-gp- or MRP2-targeting shRNA viral vector. Transduced cells were selected in culture medium containing 10 μg/ml puromycin. As a transduction control, parental Caco-2 cells were also transduced with a lentivirus plasmid vector containing shRNA that does not match any known human genes (Sigma-Aldrich), and the transduction procedures were identical to those used with P-gp and MRP2 shRNA vectors.
RNA Isolation and cDNA Synthesis.
Total RNA was isolated from cells using TRIzol reagent (Invitrogen) according to the manufacturer's protocol. cDNA was prepared from 0.25 μg of total RNA by using a SuperScript III First-Strand Synthesis System for reverse transcription-polymerase chain reaction (PCR) with random hexamer primers according to the manufacturer's protocol (Invitrogen).
Real-Time PCR.
Forty-eight different genes were analyzed with quantitative real-time PCR using a TaqMan Custom Array. The 384-well cards were purchased preloaded with TaqMan Gene Expression Assays with specific primers and probes, and 100 ng of cDNA was added to each lane. The cDNA quantification was performed using an ABI PRISM 7900HT system, and the thermal cycle conditions comprised 2 min at 50°C, 10 min of polymerase activation at 95°C, followed by 40 PCR cycles alternating 95°C for 15 s and 60°C for 1 min. Amplification curves were analyzed using the 7900HT sequence detection software SDS 2.3, and the expression for all genes was normalized against the expression of glyceraldehyde-3-phosphate dehydrogenase in each sample (Applied Biosystems, Foster City, CA). Quantification of relative gene expression was performed using the ΔΔCT method. -Fold change was calculated using vector control Caco-2 cells as a calibrator.
Western Blot Analysis for P-gp and MRP2 Protein Detection.
Cell monolayers were washed with cold PBS and lysed in ice-cold radioimmunoprecipitation assay lysis buffer (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Proteins were isolated following the manufacturer's protocol. Protein concentrations were determined using a Quant-it Protein Assay kit with a Qubit fluorometer (Invitrogen). Equal amounts of protein were loaded and separated on 4 to 20% SDS-polyacrylamide gels and then transferred to polyvinylidene difluoride membranes (Millipore Corporation, Billerica, MA). Membranes blocked with 0.2% I-block solution (Applied Biosystems) were probed sequentially with anti-MRP2 (Abcam, Cambridge, MA), anti-P-gp (clone C219; Abcam), and anti-β-actin antibodies (Sigma-Aldrich). Secondary antibodies were goat anti-mouse IgG linked to horseradish peroxidase for P-gp and β-actin (Zymed, Carlsbad, CA), goat anti-rabbit IgG linked to horseradish peroxidase for MRP2 (Abcam). Protein-antibody complexes were visualized using the SuperSignal West Femto Chemiluminescent Substrate (Pierce Chemical, Rockford, IL).
Transport Assays.
Cells were seeded at 60,000 cells/cm2 onto collagen-coated, microporous, polycarbonate membranes in 12-well Costar Transwell plates (1.13 cm2 insert area, 0.4-μm pore size; BD Biosciences, Bedford, MA). The culture medium was changed 24 h after seeding to remove cell debris and dead cells; afterward the medium was changed every other day. The cells were maintained at 37°C in a humidified atmosphere of 5% CO2 in air for 3 weeks to form confluent monolayers. Batches of cell monolayers were certified by measuring the transepithelial electrical resistance (TEER) values of the monolayers and the apparent permeability coefficient (Papp) of control compounds: lucifer yellow (fluorescent low permeability marker), atenolol (low permeability marker), propranolol (high permeability marker), and digoxin (P-gp probe substrate). The acceptance criteria for acceptable batches of cell monolayers were TEER >450 Ω · cm2, lucifer yellow Papp < 0.4 × 10−6 cm/s, atenolol Papp < 0.5 × 10−6 cm/s, and propranolol Papp between 15 × 10−6 and 25 × 10−6 cm/s. Before the transport experiment, TEER values were measured for all cell monolayers, and only monolayers with TEER values that met the acceptance criteria were selected for the transport study. The transport assay buffer was Hanks' balanced salt solution containing 10 mM HEPES and 15 mM glucose at pH 7.4 (HBSSg buffer). Digoxin as a P-gp probe substrate was assayed at 10 μM, [3H]BSP as an MRP2 substrate was assayed at 0.5 μCi/ml (27 nM), and ximelagatran was assayed at 50 μM. Bidirectional transport experiments were conducted at 37°C up to 3 h. For apical (A)-to-basolateral (B) transport, test compound was dosed to the apical chamber (donor), and blank HBSSg buffer was applied to the basolateral chamber (receiver). For B-to-A transport, test compound was dosed to the basolateral chamber (donor), and blank HBSSg buffer was added to the apical chambers (receiver). In ximelagatran assays, 1% bovine serum albumin was included in the receiver HBSSg buffer. Samples were withdrawn from donor chambers at 1 and 2 h and from the receiver chambers at 1 and 3 h; the chambers were replenished with fresh HBSSg buffer with volumes equal to the sample volumes. Immediately upon collection, ximelagatran samples were combined with an equal volume of freshly prepared acetonitrile containing 1% HCl. At the end of ximelagatran transport experiments, the assay buffer on the cell monolayers was removed by aspiration. The monolayers were then washed three times with ice-cold HBSSg buffer, and the insert filters to which cells attached were removed using a scalpel. They were then submerged in 300 μl of acetonitrile with 1% HCl in Eppendorf vials. The cells were lysed by means of vortexing, sonication, and subsequent storage of the Eppendorf vials at −20°C for 1 h to assure complete lysis. Afterward, vials were centrifuged at 10,000g for 20 min. The supernatants were collected for measurement of cellular accumulation of ximelagatran and metabolites.
Sample Analyses.
Lucifer yellow concentrations in the samples were measured using a FLUOstar fluorescence plate reader (BMG Labtech, Durham, NC) with excitation and emission wavelengths of 485 and 538 nm, respectively. Radioactivity of [3H]BSP samples was measured by liquid scintillation counting (LSC6500; Beckman-Coulter, Fullerton, CA). Concentrations of atenolol, propranolol, and digoxin were measured using reverse-phase HPLC with tandem triple quadruple mass spectrometry (LC-tandem mass spectrometry) methods. Samples were diluted with 85:15 HBSSg-acetonitrile (v/v) as follows: A-to-B receiver samples by 4-fold, B-to-A receiver samples by 10-fold, and all donor samples by 24-fold. The HPLC equipment consisted of a series 200 autosampler and a series 200 micro pump (PerkinElmer Life and Analytical Sciences, Waltham, MA). Chromatography was performed at ambient temperature in the reverse-phase mode using a 30 mm × 2.1 mm i.d. 3-μm Thermo BDS Hypersil C18 analytical column with a guard column (Thermo Fisher Scientific, Waltham, MA). The mobile phase buffer was 40 mM ammonium formate, pH 3.5; the aqueous phase consisted of 90% deionized water and 10% buffer (v/v); the organic phase consisted of 90% acetonitrile and 10% buffer (v/v). Typical gradients started at 100% aqueous phase, linearly changed to 100% organic phase over 1.5 min, held at 100% organic for 0.5 min, then returned back to 100% aqueous phase during 0.4 min, and reequilibrated at 100% aqueous phase for 1.6 min. The flow rate was 300 μl/min, and the injection volume was 5 μl. Mass spectrometry analyses were performed on a PE Sciex API3000 triple quadrupole mass spectrometer equipped with a TurboIonSpray interface (Applied Biosystems/MDS Sciex, Foster City, CA). Instrument settings were optimized for each compound. The multiple reaction monitoring transitions for atenolol, propranolol, and digoxin were 267.2/145.1, 260.2/116.1, and 798.8/651.5, respectively. Equipment operation, data acquisition, and data integration were performed using Analyst version 1.4.2 software (Applied Biosystems).
Ximelagatran and its metabolites were analyzed using LC-mass spectrometry. The LC system consisted of a HTS PAL injector (CTC Analytics, Zwingen, Switzerland) combined with an HP 1100 LC binary pump and column oven (Agilent Technologies Deutschland, Waldbronn, Germany). LC separations were undertaken on a HyPURITY C18 analytical column (100 × 2.1 mm, 5 μm; Thermo Fisher Scientific, Manchester, UK) with a HyPURITY C18 precolumn at 40°C and with a flow rate of 750 μl/min. The mobile phases consist of 10 mM ammonium acetate and 5 mM acetic acid with 10% acetonitrile for mobile phase A and 50% acetonitrile for mobile phase B. The analytes coeluted with their respective isotope-labeled internal standards. Detection was performed with a API4000 triple quadrupole mass spectrometer equipped with an electrospray interface (Applied Biosystems/MDS Sciex, Concord, ON, Canada). Instrument control, data acquisition, and data evaluation were performed using Analyst 1.4 software (Applied Biosystems/MDS Sciex).
Calculations.
The apparent permeability coefficient, Papp, was calculated for Lucifer yellow, atenolol, propranolol, [3H]BSP, and digoxin using the following equation: and the rate of appearance, dQ/dt, was calculated for [3H]BSP, digoxin, ximelagatran, and its metabolites. dCr/dt is the slope of the cumulative concentration in the receiver (Cr) compartment over time in micromoles per second, Vr is the volume of the receiver chamber in cubic centimeters, dQ is Cr × Vr, A is the diffusion area in square centimeters, and C0 is the measured dosing concentration (micromolar). Monolayers with Lucifer yellow Papp > 0.8 × 10−6 cm/s were considered leaky during transport experiments and were excluded from further calculations.
The ratio of appearance rate of [3H]BSP, digoxin, ximelagatran, and its metabolites was defined as follows:
Statistical Analysis.
The significance of the effects of P-gp knockdown on digoxin transport and MRP2 knockdown on [3H]BSP transport compared with vector control cells was calculated with a one-tailed Student's t test. One-way analysis of variance followed by a one-sided Dunnett's post hoc test was used to calculate the statistical significance of the increased A-to-B transport, the decreased B-to-A transport, and the decreased ratio of ximelagatran and its metabolites in P-gp and MRP2 knockdown cells compared with vector control cells.
Results
Suppression of MDR1/P-gp and MRP2 mRNA and Protein Expression Using shRNAs.
Stable Caco-2 cell lines with MDR1/P-gp and MRP2 knockdown were established using shRNA constructs targeting, respectively, MDR1- and MRP2-encoding mRNA. The mRNA levels of MDR1 and MRP2 in knockdown cell lines were compared with the levels expressed in noncoding shRNA vector control cells. MDR1 and MRP2 expressions were reduced by 55 and 65%, respectively, in the corresponding knockdown cell lines cultured on filters for 3 weeks (Fig. 3). Furthermore, the inhibition of P-gp and MRP2 protein expression by transduction with shRNA was confirmed by Western blot analysis (Fig. 4).
Gene Expression after MDR1 and MRP2 Knockdown.
The influence of P-gp and MRP2 knockdown on mRNA expression of other transporters, metabolic enzymes, and transcription factors was determined by using quantitative real-time PCR. Except for the targeted transporters, only 2 of 28 detected genes had more than a 2-fold reduction or 2-fold increase compared with the vector control cells (Fig. 5). mRNA expression of UDP-glucuronosyltransferase 2B7 increased 3.4-fold in P-gp knockdown cells, whereas transthyretin increased 2.6-fold in MRP2 knockdown cells. The mRNA expressions of all other genes studied were not markedly affected by the knockdown of P-gp or MRP2.
Effect of MDR1 and MRP2 Knockdown on Drug Transport.
Vector control, P-gp knockdown, and MRP2 knockdown cells were cultured on Transwell filters enabling bidirectional transport studies of probe substrates and the study compound, ximelagatran. Digoxin and [3H]BSP were used as selective substrates for P-gp- and MRP2-dependent transport, respectively (Fig. 6, A and B). The efflux ratio of the P-gp probe substrate digoxin decreased by 83% in P-gp knockdown cells, whereas the efflux ratio of the MRP2 substrate [3H]BSP was reduced by 44% in MRP2 knockdown cells (Table 1), which is consistent with the reduction seen in mRNA and protein expression presented above. Furthermore, the Papp value in the B-to-A direction for digoxin was decreased from 38 to 9.2 × 10−6 cm/s when knocking down P-gp, whereas MRP2 knockdown reduced the [3H]BSP Papp value from 18 to 9.1 × 10−6 cm/s.
In the cell monolayers, ximelagatran was metabolized to its intermediate metabolites, hydroxy-melagatran and ethyl-melagatran, and its end product melagatran. The rates of appearance of ximelagatran and its metabolites were greater in the B-to-A direction than in the A-to-B direction in vector control cells (Table 1), indicating active net efflux across the apical membrane. Knockdown of P-gp resulted in a significant decreased (p < 0.01) appearance rate on the apical side for ximelagatran, hydroxy-melagatran, and melagatran (B-to-A direction) (Fig. 7A). These results indicate that P-gp is involved in the active efflux of the parent and its metabolites. In addition, both the cellular accumulation of melagatran and hydroxy-melagatran (Fig. 8, A and B) and the appearance rate on the basolateral side of ximelagatran, hydroxy-melagatran, and melagatran (A-to-B direction) (Fig. 7B) were increased because of knockdown of P-gp, further strengthening the conclusion that P-gp is involved in the transport of these substances. There was no detectable appearance rate of ethyl-melagatran on the basolateral side and the intracellular concentration was low (Fig. 8C). Ximelagatran is rapidly metabolized in the cell and only low levels of ximelagatran was detected intracellularly (data not shown). Knockdown of MRP2 did not significantly affect the transport of ximelagatran, hydroxy-melagatran, and ethyl-melagatran in either direction. However, a small significant decrease in the appearance rate of melagatran on the apical side (B-to-A direction) was observed (Fig. 7A).
Discussion
RNA silencing leading to functional inactivation of the target gene is an attractive method for down-regulation of the expression of specific genes. Short interfering RNA (siRNA) can mediate strong and specific suppression of gene expression by sequence-specific cleavage of mRNA, thus blocking the translation into target protein (Yue et al., 2009). The gene silencing technique has great potential in drug efflux and transport-mediated drug-drug interactions studies for elucidating the function of specific transporters in drug disposition.
Caco-2 cells are polarized cells expressing biliary transporters such as P-gp and MRP2 in the apical membrane. These apical expressed transporters are mainly found in excretory organs (gastrointestinal tract, liver, and kidney), counteracting systemic exposure and thus playing an important role for mediating the bioavailability, distribution, and excretion of many xenobiotics and clinically important drugs. Transporter-mediated interactions can be caused by coadministration of a compound of interest with other substrates or inhibitors that bind to and/or interact with the same transporter (Choudhuri and Klaassen, 2006). The possible mechanism(s) of involvement of biliary transporters P-gp and MRP2 in the interaction produced by erythromycin coadministration on ximelagatran pharmacokinetics was investigated in this study using shRNA Caco-2 knockdowns. The results indicated a clear involvement of P-gp but not of MRP2 in the efflux of ximelagatran, hydroxy-melagatran, and melagatran. P-gp-mediated efflux can thus be concluded to be important for the biliary secretion of ximelagatran and its metabolites. However, this study does not rule out the possibility that another hepatic transporter, e.g., breast cancer resistance protein, bile salt export pump, and multidrug and toxic extrusion 1 may be of importance in the disposition of ximelagatran and its metabolites.
Erythromycin and quinidine were previously found to inhibit the in vitro transport of ximelagatran, hydroxy-melagatran, and melagatran across Caco-2 cell monolayers, suggesting the involvement of P-gp (Eriksson et al., 2006). However, several factors may have confounded the results from those Caco-2 cell transporter studies using inhibitors. The inhibitors may not be selective and the in vitro concentrations used in such experiments are often high. Thus, it cannot be excluded that these inhibitors blocked the function of several other transporter proteins expressed in Caco-2 cells, e.g., breast cancer resistance protein, MRP3, organic anion-transporting polypeptide-B, and peptide transporter 1 (Watanabe et al., 2005; Hilgendorf et al., 2007; Wang et al., 2008).
The use of the post-transcriptional gene silencing by RNA interference to specifically knock down drug transporters is a rapidly growing field of research (Celius et al., 2004; Hua et al., 2005; Tian et al., 2005; Watanabe et al., 2005; Li et al., 2006; Yue et al., 2009; Zhang et al., 2009). Even though siRNA and shRNA have emerged as powerful tools to study gene function in a sequence-specific manner, off-target effects have been observed in several studies (Jackson et al., 2006; Li et al., 2006; Alemán et al., 2007; Tschuch et al., 2008). Nonspecific effects can depend on both knockdown of nontarget mRNA (Jackson et al., 2006) or compensatory effects causing up-regulation of other genes (Chen et al., 2005). As mentioned previously, the influence of P-gp and MRP2 knockdown on mRNA expression of other transporters, metabolic enzymes, and transcription factors was analyzed in this study. Most of the 28 genes investigated were not markedly affected by the knockdown of P-gp or MRP2. However, UDP-glucuronosyltransferase 2B7 and transthyretin mRNA expression increased more than 2-fold in P-gp and MRP2 knockdown cells, respectively. There are several examples showing that loss of function or down-regulation of transporters or metabolic enzymes can be compensated for by up-regulation of other transporters or metabolic enzymes (Higuchi et al., 2004; Chen et al., 2005; Jeong et al., 2005). Therefore, it is important to investigate changes in the expression of nontarget genes and elucidate whether changes affect the disposition of the study compound. The nontarget effects investigated in this study on Caco-2 cells are not likely to be important for the transport of ximelagatran and its metabolites.
Because the in silico design of siRNA sequences does not entirely eliminate unintended siRNA-mRNA interactions (Jackson et al., 2006; Tschuch et al., 2008), additional approaches are advocated. The degree of nonspecific effects correlates with concentrations of siRNA (Tschuch et al., 2008) and one suggestion is to concentrate the on-target effects while diluting the off-target effects by using endonuclease-prepared siRNA, a mixture of siRNA targeting the same mRNA (Kittler et al., 2005; Li et al., 2006; Tan et al., 2007). Furthermore, Echeverri et al. (2006) suggested that at least two distinct siRNA or shRNA sequences and/or rescue experiments should be performed to avoid false-positive results. Even though straightforward approaches to achieve the right phenotype and simultaneously avoid off-target effects are lacking, the above-mentioned procedures together with properly designed control experiments should be a good strategy to generate siRNA knockdowns with low rates of false-positive and false-negative results.
In conclusion, the results from the Caco-2 cell transport experiments indicate that ximelagatran, hydroxy-melagatran, and melagatran are transported by P-gp but not by MRP2 because the efflux across P-gp but not MRP2 knockdown Caco-2 cell monolayers was clearly reduced. This work, together with previous in vivo and in vitro studies, indicates that inhibition of hepatic P-gp contributed to the increased plasma levels of melagatran seen in healthy subjects coadministered erythromycin. Furthermore, this study shows that knock down of drug transporters using shRNA is a valuable tool to predict potential sites of transporter-mediated pharmacokinetic interactions in drug discovery and development.
Acknowledgments.
We thank Samanth M. Allen, Erica A. Weiskicher, Rebecca A. George, and Yuehua Huang (Absorption Systems) for their expert technical assistance.
Footnotes
This study was supported in part by a research grant from the Food and Drug Administration [Grant 1R43-FD003482-01].
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
doi:10.1124/dmd.109.029967.
-
ABBREVIATIONS:
- DDI
- drug-drug interaction
- P-gp
- P-glycoprotein
- shRNA
- short hairpin RNA
- MDR1
- multidrug resistance 1
- MRP2
- multidrug resistance-associated protein 2
- HBSS
- Hanks' balanced salt solution
- BSP
- bromosulfophthalein
- HPLC
- high-performance liquid chromatography
- PCR
- polymerase chain reaction
- CT
- cycle threshold
- TEER
- transepithelial electrical resistance
- A
- apical
- B
- basolateral
- LC
- liquid chromatography
- siRNA
- small interfering RNA.
- Received October 9, 2009.
- Accepted December 18, 2009.
- Copyright © 2010 by The American Society for Pharmacology and Experimental Therapeutics