Abcb1 in Pigs: Molecular cloning, tissues distribution, functional analysis, and its effect on pharmacokinetics of enrofloxacin

P-glycoprotein (P-gp) is one of the best-known ATP-dependent efflux transporters, contributing to differences in pharmacokinetics and drug-drug interactions. Until now, studies on pig P-gp have been scarce. In our studies, the full-length porcine P-gp cDNA was cloned and expressed in a Madin-Darby Canine Kidney (MDCK) cell line. P-gp expression was then determined in tissues and its role in the pharmacokinetics of oral enrofloxacin in pigs was studied. The coding region of pig Abcb1 gene was 3,861 bp, encoding 1,286 amino acid residues (Mw = 141,966). Phylogenetic analysis indicated a close evolutionary relationship between porcine P-gp and those of cow and sheep. Pig P-gp was successfully stably overexpressed in MDCK cells and had efflux activity for rhodamine 123, a substrate of P-gp. Tissue distribution analysis indicated that P-gp was highly expressed in brain capillaries, small intestine, and liver. In MDCK-pAbcb1 cells, enrofloxacin was transported by P-gp with net efflux ratio of 2.48 and the efflux function was blocked by P-gp inhibitor verapamil. High expression of P-gp in the small intestine could modify the pharmacokinetics of orally administrated enrofloxacin by increasing the Cmax, AUC and Ka, which was demonstrated using verapamil, an inhibitor of P-gp.

Scientific RepoRts | 6:32244 | DOI: 10.1038/srep32244 mouse (NM011075.2), respectively. The deduced amino acid sequence of porcine P-gp was 1,286 amino acid residues in length and the estimated molecular weight was 141.966 kDa and the theoretical isoelectric point was  Table 1. Primers used in this study. 8.99. The protein sequence of porcine P-gp shared 89, 88 and 87% identity to that of human (NP000918.2), cow (XP590317.6) and sheep (CAM33439.1), respectively ( Fig. 2A). Unique to the sequences of P-gp in ruminants (sheep and cow) is a 4-amino acids deletion at position 22-25 of the amino acid sequence.
A phylogenetic tree was constructed using the neighbor-joining (NJ) method with a 1,000 bootstrap test based on the multiple alignments, which indicated that the selected 14 protein sequences of P-gp were clustered into two groups: avian and mammal. The porcine P-gp was located in vertebrates group and was closely related to cow and sheep, suggesting a close evolutionary relationship between them (Fig. 2B).

Establishment and characterization of an MDCK cell line stably-transfected with porcine
Abcb1. The pig Abcb1 gene was amplified using primers shown in Table 1 with XhoI and XbaI sites, respectively, and a recombinant pcDNA3.1-pAbcb1 plasmid was constructed. MDCK cells were transfected with pcD-NA3.1-pAbcb1. A single cell colony resistant to G418 was selected as a stable transfectant (MDCK-pAbcb1). The presence of Abcb1 was confirmed with qRT-PCR and western blot (Fig. 5A,B). The accumulation of Rho123 (substrate of human P-gp) was measured to investigate whether MDCK-pAbcb1 cell line has enhanced porcine P-gp protein function. Figure 5C,D show that MDCK-pAbcb1 cells accumulated less Rho123 than MDCK cells (control), indicating a significantly enhanced extruding capacity of Rho123 in MDCK-pAbcb1 cells (p < 0.01). Furthermore, verapamil (100 μ M), an inhibitor of human P-gp, significantly (p < 0.01) increased Rho123 accumulation in MDCK-pAbcb1 cells, demonstrating that such an accumulation could be reversed by verapamil.
Expression of P-gp in pigs. Porcine P-gp mRNA and protein were analyzed by qRT-PCR and Western blot, respectively. As shown in Fig. 6A, Abcb1 mRNA was detected in all tested tissues. In the CNS, the brain capillaries had the most Abcb1 mRNA and the cerebellum had the least. Abcb1 mRNA levels in the cerebral cortex, cerebellum, midbrain, hypothalamus and hippocampus was very low. In the peripheral tissues, Abcb1 were highly expressed in the jejunum, ileum, colon and liver. Abcb1 mRNA transcription in the ileum was greater than in the cecum and kidney (p = 0.005, p = 0.001) and significantly higher than that in the duodenum and rectum (p = 0.017, p = 0.014). Abcb1 mRNA in the jejunum, colon and liver was significantly higher than that in the kidney (p = 0.046, p = 0.018, p = 0.030, respectively). P-gp protein (shown in Fig. 6C) was ~170 kDa as detected in different pig tissues, indicating that antibody Mdr-1 directed against the human isoform could also recognize porcine P-gp. Protein expression in tissues   Figure S2). Samples were derived from the same experiment and blots were processed simultaneously. (C) Accumulation of 5 μ M Rho123 in MDCK and MDCK-pAbcb1 cells with/without verapamil (100 μ M) during 2 h. Mean ± SEM; n = 3. (D) Rho123 accumulation at 2 h with flow cytometry. Histogram shows fluorescence (x-axis) representing Rho123 accumulation plotted as a function of the number of cells (y-axis) and is representative of three independent experiments. appears in Fig. 6B. The most P-gp protein was in the brain capillaries and the least was observed in the hypothalamus. The P-gp protein level was correlated to its mRNA level (r = 0.863, p < 0.05).

The effect of P-gp on in vitro transport and in vivo pharmacokinetics of enrofloxacin.
Enrofloxacin was measured using HPLC, and the lowest limit of detection (LOD) of enrofloxacin was 0.02 μ g/mL, based on a signal-to-noise ratio > 3. The lowest limit of quantification (LOQ) was 0.05 μ g/mL with a signal-to-noise ratio > 10. The mean percent recovery of enrofloxacin exceeded 81%, and intra-assay and inter-assay reproducibility had a relative standard deviation (RSD) < 11%. Assay linearity was good over 0.05-10 μ g/mL with r 2 = 0.9999.
MDCK and MDCK-pAbcb1 cells were grown to confluent monolayers on porous membrane filters, and vectorial transport of the enrofloxacin (12 μ M) across the cells was determined. In the MDCK parental cell line, apically and basolaterally directed translocation of enrofloxacin were similar. However, in the pAbcb1 expressing MDCK cells, P app (BL-AP) was significantly higher than P app (AP-BL) (P < 0.01). The P-gp-mediated transport was completely inhibited by its inhibitor verapamil, resulting in a similar efflux ratio, compared to that of the MDCK parental cell line (Table 2). These results showed highly efficient transport of enrofloxacin by porcine P-gp.
To further confirm whether P-gp affects the pharmacokinetics of enrofloxacin in pigs, verapamil (an inhibitor of P-gp) was administrated prior to enrofloxacin, and the plasma concentration-time curve and pharmacokinetics of enrofloxacin are summarized in Fig. 7 and Table 3, respectively. Compared with the control group, C max , AUC and Ka of enrofloxacin were increased, whereas T peak and T 1/2ka of enrofloxacin were shortened in animals when co-administrated with verapamil. In summary, enrofloxacin pharmacokinetics was altered by P-gp inhibition, indicating that porcine P-gp can influence enrofloxacin behavior in vivo.  Figure S3). Samples were derived from the same experiment and blots were processed simultaneously.  Table 2. Bidirectional transport assay in MDCK and MDCK-pAbcb1 cells (mean ± SEM, n = 3). Statistical significance was analyzed by Student's t-test. ## P < 0.01 significant difference of enrofloxacin transport in MDCK and MDCK-pAbcb1 cells; **P < 0.01 significant difference of enrofloxacin transport in presence/ absence of verapamil.

Discussion
Oral drug administration is convenient intake route for humans, as well as for animals. The effects of P-gp on drug pharmacokinetics are important for the evaluation of the capacity of oral drug intake in pigs. The localization of P-gp in different tissues in pigs was studied in our previous work 15 ; however, how P-gp functions in pigs is unclear.
To address this deficit, we cloned full-length cDNA of porcine P-gp, and characterized the transport activity of P-gp by exogenously expressing the Abcb1 gene in MDCK cells and measuring the porcine P-gp at the protein and transcriptional levels. Then, we assessed its role in the pharmacokinetics of oral enrofloxacin in pigs. This represents the first cloning of full-length porcine cDNA of P-gp. The porcine P-gp shared high similarities with P-gps from cow, sheep, human and mouse, suggesting a similar physiological role in these animals. Drug-binding pockets are predicted to reside in transmembrane segments, which formed a translocation path for drug substrates as they exit the membranes 16 . The nucleotide binding domain may bind ATP or its analogs and can hydrolyze ATP [17][18][19][20] . There is "cross-talk" between nucleotide binding and transmembrane domains as evidenced by cross-linking studies 21 , and transmembrane domains of porcine P-gp were generated here with Protter protein display software. The prediction showed that porcine P-gp contains twelve transmembrane domains which was similar to human P-gp, but glycosylation site prediction revealed seven potential putative N-linked glycosylation sites on porcine P-gp, fewer than that in human P-gp. Some studies have shown that N-glycosylation can participate in many biological processes such as regulation of intracellular targeting, protein folding and maintenance of protein stability 22,23 . However, other studies have reported that N-linked glycosylation is not essential for protein expression, plasma membrane localization or overall function 24,25 . Further studies are necessary to clarify whether glycosylation affects the function of porcine P-gp.
To better understand how porcine P-gp transports various hydrophobic compounds, we predicted residues that participate in drug binding. Our results indicated that amino acids participating in substrate binding were located at TMD1 sites 5 and 6 and TMD2 sites 7, 11 and 12 of porcine P-gp, whereas previous studies have shown that human P-gp substrate binding sites were located in TMD1 sites 5 and 6 and TMD2 sites 11 and 12 26 . This may   influence substrate types. Site-directed mutagenesis studies of Abcb1 can be performed to prove whether these residues in porcine P-gp participate in ligand binding. Knowing substrate binding sites may enable development of drugs that bypass recognition of P-gp. Human colorectal adenocarcinoma cell line (Caco-2) has been approved by the US Food and Drug Administration (FDA) for in vitro transport studies of P-gp 27,28 . However, pharmacokinetics might be different because of the specific functions of P-gp across different species 29 . An alternative approach was to establish a porcine P-gp expression system. The MDCK cell line is an approved model for overexpressing Abcb1 originating from other species due to its negligible expression of endogenous transporters 30,31 . MDCK cell line overexpression porcine P-gp could be used to screen substrate drugs and potential inhibitors or inducers.
The efflux efficiency of P-gp substrate is determined by the P-gp number or density on the cell membrane 32 . Differences expression pattern of P-gp in tissues may influence therapeutic outcomes by affecting the accumulation of drugs in specific tissues. Transcription of Abcb1 had been measured using RT-PCR in the liver and brain capillary endothelial cells (pBCECs) 33 and in the kidneys 34 . We measured Abcb1 mRNA level in the CNS, intestine, liver and kidneys. The data demonstrated that its expression in intestine was not uniform. In the peripheral tissues, the greatest expression was in the ileum. The porcine liver also expressed relatively high Abcb1 mRNA, which was consistent with previous data from human and canines 35,36 . Renal Abcb1 mRNA is lower in the pig, compared to that in humans 36,37 . P-gp in intestinal epithelial cells may prevent pharmacotherapeutic agents from entering the systemic circulation and therefore change their pharmacokinetics 38 . Hepatic metabolism is significant, and high expression level of P-gp is vital for biliary drug, toxicant, and metabolite excretion 39 . More Abcb1 mRNA in the intestine and liver suggests that porcine P-gp may limit absorption and facilitate secretion of substrates into the intestines. In addition, we measured P-gp protein expression level in 60-day old piglets and found that this was coincident with mRNA levels, except for the ileum and jejunum, suggesting a tissue-specific post-transcriptional mode of regulation. Hence, RNA analysis of pig tissue samples should be done in conjunction with protein analysis, because RNA and protein may differ among certain organs.
Of note, the P-gp expression pattern in tissues affects pharmacokinetics of enrofloxacin, a fluoroquinolone (FQ) widely used in veterinary medicine either orally or parenterally. Sorgel's group 40 and others [41][42][43] reported that the concentrations of ciprofloxacin and danofloxacin-mesylate in the gut lumen are higher than those in plasma. Other studies proved that ciprofloxacin and danofloxacin-mesylate were substrates of multiple human ABC transporters [44][45][46] , indicating that intestinal efflux is the underlying mechanism for danofloxacin-mesylate and ciprofloxacin secretion into the lumen. In this study, the net efflux ratio of enrofloxacin across MDCK-pAbcb1 cells was more than 2 and basolateral-to-apical transport of enrofloxacin could be counteracted by the P-gp inhibitor verapamil. Although the concentration of enrofloxacin was not measured in the luminal compartment, its concentration in plasma was increased when animals received verapamil. These results suggest that enrofloxacin might be a substrate of porcine P-gp, similar to the data obtained in chickens 47 . Initially, we treated piglets with the P-gp inhibitor PSC-833, but some animals began vomiting. Thus, PSC-833 was replaced with verapamil, which could inhibit P-gp function and be monitored in an MDCK-pAbcb1 cell line.
Our data also suggested that co-administration of P-gp inhibitors may be an alternative strategy to improve oral bioavailability and therapeutic efficacy of enrofloxacin when used to treat systemic bacterial infection in the swine industry. Three generations of P-gp inhibitors have been identified but none of them show improved therapeutic efficacy due to their broad activity and toxicity 48 . Berberine can significantly inhibit P-gp (data not shown) so it may be a potential P-gp inhibitor as it is not absorbed after oral administration and only inhibits P-gp in the gastrointestinal tract.
In conclusion, the complete cDNA of porcine P-gp was cloned for the first time. The protein structure was modeled and analyzed. The phylogenetic analysis on P-gp proteins from different species was performed, and its protein expression and mRNA levels in different tissues were studied. Also, the effect of P-gp on enrofloxacin transmembrane transport and in vivo pharmacokinetics was evaluated. More studies are required to fully elucidate the complex epigenetic regulation of porcine P-gp during development, tissue-specific expression patterns, and the contribution of epigenetics to species variability in drug disposition and therapeutic response.

Cells. MDCK (Madin-Darby canine kidney) and IPEC-J2 cells were obtained from Shanghai Institute of Cell
Biology, the Chinese Academy of Sciences (Shanghai, China) and grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS, HyClone Laboratories Inc.). Cells were maintained at 37 °C in a humidified atmosphere of 5% CO 2 and sub-cultured once 80% confluent.
Animals. Sixty-day-old healthy crossbred pigs (large white x Landrace x Duroc, 20 ± 2 kg) were bought from Jiangsu Agricultural Academy and reared under standard conditions of light and temperature. Feed and water were provided ad libitum during the study. All pigs were fed according to the breeding standards of the Chinese Local Pigs and National Research Council (NRC). Animal use and handling protocols were approved by the regional Animal Ethics Committee and Nanjing Agricultural University. The protocol of the study was conducted in accordance with guidelines of the regional Animal Ethics Committee and Nanjing Agricultural University.
Total RNA extraction and cDNA synthesis. All pigs were anesthetized and killed by decapitation.
Samples of liver, kidney, intestines and brain tissues were taken immediately (within 5-10 min after death). Brain capillaries were isolated from ~10 g of fresh cerebral cortices according to published methods with minor modification 49 . In brief, meninges and choroid plexus from the brains were removed and cerebral matter was homogenized in 20 volumes of cold Ringer's solution containing 10 mM HEPES (Santa Cruz, Heidelberg, Germany) at pH 7.4 in a glass homogenizer. Ten upwards and downwards strokes were applied during the homogenization. The homogenates were collected and filtered initially through a 150 μ m nylon mesh, the filtrate was then re-filtered Scientific RepoRts | 6:32244 | DOI: 10.1038/srep32244 through a 60 μ m nylon mesh. The brain capillaries are trapped on the 60 μ m nylon meshes and were collected in the tube. All samples were frozen in liquid nitrogen and then stored at − 80 °C until RNA and protein extraction.
Total RNA was extracted from tissues using TRIZOL (TaKaRa, Japan) following the manufacturer's instructions. Quality and concentration were measured using a photometer (Eppendorf Biophotometer, Germany). RNA integrity was assessed with RNA electrophoresis. cDNA templates were synthesized from 1 μ g of total RNA using a HISCRIPT 1 st strand cDNA synthesis kit (Vazyme, Nanjing, China).
Cloning and sequencing of full-length cDNA of porcine Abcb1. PCR primers (Table 1) toward three overlapping cDNA fragments of porcine Abcb1 (F1, F2, F3, shown in Fig. 1A) were designed based on sequence alignments of conserved regions from various species (Human, NM_000927.4; sheep, NM_001009790.1; dog, NM_001003215.1; mouse, NM_011075.2). Three overlapping fragments of porcine Abcb1 were amplified by touchdown PCR using PrimeSTAR GXL DNA Polymerase (TaKaRa, Japan) under these conditions: 95 °C for 5 min, 36 cycles of 95 °C for 30 s, 65-55 °C for 30 s (initial annealing temperature of 65 °C was reduced by 2 °C after every six cycles to 55 °C for the final six cycles), 72 °C for 1 min, and an extra extension of 15 min at 72 °C. PCR products were verified using published methods 50 . After the three products were confirmed to be part of the Abcb1 gene by sequencing, full-length cDNA was obtained from the three fragments using fusion PCR as follows: 95 °C for 5 min, 30 cycles of 95 °C for 30 s, 62.5 °C for 30 s, 72 °C for 1 min, and 72 °C for 15 min. Final PCR products were confirmed by sequencing and the sequence was submitted to GenBank to obtain an accession number.
Bioinformatics analysis of porcine Abcb1 and its protein structural model. cDNA sequence analysis was conducted using NCBI BLAST (http://www.ncbi.nlm.nih.gov/blast). Calculated molecular weights and predicted isoelectric points were obtained with the ExPASy online server (http://www.expasy.org/tools/). Multiple alignments of P-gp protein sequences were generated using the Bio Edit program. A phylogenetic analysis was used to determine relatedness of porcine P-gp to other mammalian P-gp sequences, and alignment data were imported into MEGA version 4.1. The phylogenetic tree was constructed using the neighbor-joining method with Poisson correction 51 . Bootstrap analysis was performed using 1,000 replicates.

Establishment of an MDCK cell line stably-transfected with porcine Abcb1. Primers used to
amplify pig Abcb1 complementary DNAs (cDNAs) were designed based on our previously submitted sequence (GenBank ID: KP233220) containing an XhoI site within the sense primer and an Xba I site within the antisense primer (See Table 1 for primers). PCR amplifications were performed as previously described in the Methods. Pig Abcb1 full cDNA was cloned into the plasmid pcDNA3.1 and transformed into Escherichia coli DH5α cells (Vazyme, China). Plasmid was purified using an Omega Endo-Free Plasmid Mini Kit and was sequenced using vector primers to confirm inserted genes (Invitrogen, Shanghai, China).
To generate cell lines stably expressing porcine P-gp, an MDCK cell line was transfected with pcD-NA3.1-pAbcb1 plasmid using LipofectAMINE 2000 reagent (Invitrogen) according to the manufacturer's instructions. Stable transfectants were selected with G418 (600 μ g/mL, Gibco). When single-cell stable colonies resistant to G418 were observed, cells were sub-cultured again and distinct colonies were isolated using cloning cylinders. Cells were grown in the presence of G418 to maintain expression of transfected genes. MDCK -pAbcb1 cells were propagated and sub-cultured in the same medium with the addition of G418 (300 μ g/mL). MDCK, IPEC-J2 and three colonies of MDCK -pAbcb1 were used to measure exogenous expression of porcine P-gp. qRT-PCR and western blot were used according to published methods 54 , with different primers (Table 1)  Tissues distribution of P-gp (Abcb1) in pigs. Pigs (N = 6) were used to measure constitutive P-gp (Abcb1) expression in different tissues with real time quantitative PCR and western blot. Abcb1 mRNA was measured with qRT-PCR and gene-specific primers for PCR were designed according to the cloned sequence of porcine Abcb1 (KP233220 in GenBank). GAPDH was as internal control for each sample. Primer pairs for genes appear in Table 1. qRT-PCR was performed on a CFX96 Real-Time PCR Detection System (Bio-Rad, USA), and SYBR Green was the dye (Toyobo, Japan). mRNA was measured using 12.5 μ L SYBR Green real-time PCR Master Mix in a 25 μ L volume with 2 μ L cDNA. Each run consisted of an initial 1 min activation cycle at 95 °C, followed by 40 cycles of denaturation for 20 s at 95 °C, annealing for 30 s at 60 °C, and elongation for 31 s at 72 °C. Uniform amplification of products was analyzed according to melting curves of the amplified products. Abcb1 mRNA relative expression in different tissues was measured using the 2 −ΔΔCt method 55 . P-gp relative expression was measured with standard western blot with minor modifications 50 . Tissue protein was extracted using a membrane and cytosol protein extraction kit (Beyotime, Haimen, China). Briefly, equal amounts of membrane proteins (5 μ g/lane) were fractionated with SDS-PAGE and then transferred onto PVDF membranes (BIO-RAD, USA). Membranes were blocked with 5% bovine serum albumin and incubated with primary antibody at the appropriate dilution (Mdr-1, Santa Cruz, Germany, 1:200; β -actin, TransGen, China, Scientific RepoRts | 6:32244 | DOI: 10.1038/srep32244 1:5,000). Membranes were incubated overnight at 4 °C, followed washing three times with PBST for 15 min. Thereafter, membranes were incubated with horseradish peroxidase-labeled goat anti-mouse IgG secondary antibody (1:5,000, Boster, Wuhan, China) for 1 h at 37 °C, then washed three times with PBST for 15 min and two times with PBS for 10 min. Interactive bands were visualized with ECL (Vazyme, China) and scanned with a Tanon 5200 chemiluminescent imaging system (Tanon, China). Relative protein was quantified by calculating densitometric values of target bands using Image J software.
Experimental design of enrofloxacin bidirectional transport assay and pharmacokinetic studies in pigs. Transport assays were carried out as described 45 with minor modifications. MDCK and MDCK-pAbcb1 with a density of 6.6 × 10 4 cells/insert were seeded on a polyethylene terephthalate membrane insert (Millicell cell culture inserts, 1.0 μ m pore size, 6.5 mm diameter) in 24-well culture plates. Cell monolayers with TEER values > 150 Ω.cm 2 were used in the study 56 . Two hours before the experiment, medium in both apical and basolateral sides of the monolayer was replaced by HBSS transport buffer, either with or without 100 μ M verapamil. The experiment was started (t = 0) by replacing the medium in either the apical (AP) or basolateral (BL) compartment with fresh HBSS buffer containing 12 μ M enrofloxacin in the presence or absence of 100 μ M verapamil. Transwells were shaken gently at 50 rpm. At 120 min, 0.20 ml samples were withdrawn from the receiver chambers. Samples were stored at − 20 °C. All experiments were conducted in triplicate.
To evaluate the effect of P-gp on the pharmacokinetics of enrofloxacin, animals (N = 8 total) were randomized to 2 groups. Group 1 animals received enrofloxacin (10 mg/kg, po), and each animal in group 2 was pre-treated with verapamil (10 mg/kg, po) and then enrofloxacin (10 mg/kg, po). Blood samples were collected from the jugular vein 20 min before and at 0.083, 0.25, 0.5, 0.75, 1, 1.5, 2, 3, 4, 5, 6, 8, 12 and 24 h after enrofloxacin administration. Samples were centrifuged at 1,500 × g for 10 min and plasma was harvested and aliquoted for storage at − 80 °C before HPLC analysis.
HPLC analysis of enrofloxacin. HPLC analysis of enrofloxacin was performed as published with minor modification 47,57 . Concentrations of enrofloxacin were measured using an Agilent 1200 HPLC system. In brief, plasma was thawed at room temperature and centrifuged at 2,000 × g for 5 min. The supernatant (0.5 mL) was mixed with acetonitrile and separated into organic and water phases by centrifugation. The organic phase was evaporated to dryness under a nitrogen stream and residue was resuspended with mobile phase solution. The samples from the transport assays were directly evaporated to dryness under a nitrogen stream and 40 μ L of the mobile phase solution was used to concentrate the samples for 5 times. Twenty microliters of the mixture were injected into the HPLC column (Kromasil C18 columns, 5 mm particle size, 250 × 4.6 mm). The mobile phase composition was 0.1 M phosphoric acid (adjusted pH to 3.5 with triethylamine)/acetonitrile (84:16, v/v) and the flow rate of mobile phase was set to 0.85 mL/min. UV absorbance was measured at 278 nm. Assay validation including recovery rate, inter-and intra-assay precision, accuracy and assay linearity of enrofloxacin were assessed using published methods 58 . P app Calculation and Pharmacokinetic analysis. Apparent permeability coefficients (P app ) were calculated using the following equation: P app = (dQ/dt)/(A × C 0 ), where A is the area of filter membrane, C 0 is the initial concentration of the test drug, dQ is the amount of transported drug, and dt is time elapsed. The efflux ratio (ER) was calculated from (P app B → A)/( P app A → B) 59 . Where P app B → A and P app A → B are BL to AP and AP to BL apparent permeability coefficients, respectively. The net efflux ratio was calculated using the following equation: Net efflux ratio = efflux ratio in MDCK-pAbcb1/efflux ratio in MDCK.
Pharmacokinetics were calculated for individual data using 3p97 practical pharmacokinetic software (Version97, Chinese Pharmacologic Association, Beijing, China). The best fit compartment model was assessed according to Akaike's information criterion.
Statistical analysis. All data were analyzed for statistical significance using SPSS Statistics 17.0 (version 17.0, SPSS Inc., Chicago, IL, USA). The differences between means were considered significant at p < 0.05 and very significant at p < 0.01. Data were shown as means ± SEM.