QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: dmd.106.010801v1 35/4/539    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Cao, J. Articles by Zhou, S.-F. Search for Related Content PubMed PubMed Citation Articles by Cao, J. Articles by Zhou, S.-F.

Role of P-glycoprotein in the Intestinal Absorption of Glabridin, an Active Flavonoid from the Root of Glycyrrhiza glabra

Department of General Surgery, the First Municipal Hospital of Guangzhou, Guangzhou, China (J.C.); Departments of Pharmacy (X.C.) and Nephrology (X.-Q.Y.), the First Affiliated Hospital, Sun Yat-sen University, Guangzhou, China; Department of Pharmacology and Toxicology, Australian Institute of Chinese Medicine, Sydney, New South Wales, Australia (J.L., S.-F.Z.); Department of Biochemistry, School of Life Sciences (A.-L.X.), and Institute of Clinical Pharmacology, School of Pharmaceutical Sciences (M.H.), Sun Yat-sen University, Guangzhou, China; Departments of Pharmacy (E.C.) and Biological Sciences (F.-S.S.), Faculty of Science, National University of Singapore, Singapore; School of Pharmacy, Faculty of Medical and Health Sciences, the University of Auckland, Auckland, New Zealand (J.-Y.W.); School of Medicine, Deakin University, Waurn Ponds, Victoria, Australia (W.D.); Department of Molecular & Clinical Pharmacology, Guangdong Provincial Cardiovascular Institute, Guangzhou, China (X.-Y.Y.); and Department of Maternal Medicine, Obstetrics and Gynecology Hospital, Fudan University, Shanghai, China (X.-T.L.)

(Received April 24, 2006; accepted January 3, 2007)

	Abstract

Top Abstract Materials and Methods Results Discussion References

 
Glabridin is a major constituent of the root of Glycyrrhiza glabra, which is commonly used in the treatment of cardiovascular and central nervous system diseases. This study aimed to investigate the role of P-glycoprotein (PgP/MDR1) in the intestinal absorption of glabridin. The systemic bioavailability of glabridin was approximately 7.5% in rats, but increased when combined with verapamil. In single-pass perfused rat ileum with mesenteric vein cannulation, the permeability coefficient of glabridin based on drug disappearance in luminal perfusates (P_lumen) was approximately 7-fold higher than that based on drug appearance in the blood (P_blood). Glabridin was mainly metabolized by glucuronidation, and the metabolic capacity of intestine microsomes was 1/15 to 1/20 of that in liver microsomes. Polarized transport of glabridin was found in Caco-2 and MDCKII monolayers. Addition of verapamil in both apical (AP) and basolateral (BL) sides abolished the polarized transport of glabridin across Caco-2 cells. Incubation of verapamil significantly altered the intracellular accumulation and efflux of glabridin in Caco-2 cells. The transport of glabridin in the BL-AP direction was significantly higher in MDCKII cells overexpressing PgP/MDR1 than in the control cells. Glabridin inhibited PgP-mediated transport of digoxin with an IC₅₀ value of 2.56 µM, but stimulated PgP/MDR1 ATPase activity with a K_m of 25.1 µM. The plasma AUC_0–24h of glabridin in mdr1a(–/–) mice was 3.8-fold higher than that in wild-type mice. These findings indicate that glabridin is a substrate for PgP and that both PgP/MDR1-mediated efflux and first-pass metabolism contribute to the low oral bioavailability of glabridin.

View larger version (9K): [in this window] [in a new window]   FIG. 1. Chemical structures of glabridin and mefenamic acid.

P-glycoprotein (PgP/MDR1) was initially found to be expressed at high levels in many tumor cell lines and conferred multidrug resistance (MDR) to a variety of anticancer drugs, including vinca alkaloids, epipodophyllotoxins, taxanes, and anthracyclines. It is a 170- to 180-kDa plasma membrane protein encoded by the human MDR1 and MDR3 genes, the murine mdr1a, mdr1b, and mdr2 genes, and rat pgp1, pgp2, and pgp3 genes (Borst and Elferink, 2002). However, only the expression of human MDR1, and rodent mdr1a and mdr1b appears to selectively confer MDR. PgP/MDR1 is made up of two symmetrical halves, each of which contains six membrane-spanning domains and an ATP-binding site. Besides its role in conferring MDR, PgP/MDR1 has also been found in many normal tissues, including the intestinal epithelium, brain capillary endothelial cells, hepatocytes, and renal tubular cells, suggesting its important role in drug absorption, elimination, and distribution (Borst and Elferink, 2002). PgP/MDR1 and other apical ATP-binding cassette (ABC) transporters such as multidrug resistance-associated protein 1 (MRP1) and breast cancer resistance protein act as a major intestinal barrier, limiting the absorption of a number of lipophilic drugs (Fromm, 2004; Kunta and Sinko, 2004). Oral administration remains the most popular route for most prescribed drugs and herbal medicines due to its convenience and noninvasiveness. However, poor oral absorption is a concern for many drugs and herbal constituents.

Despite the wide use of licorice, the pharmacokinetic properties of glabridin and its interaction with PgP/MDR1 have not been investigated in animals and humans. Data on its absorption, metabolism, distribution, and elimination are lacking. This prompted us to investigate the role of PgP/MDR1 in its intestinal absorption using several in vitro and in vivo models: 1) single-pass rat intestinal perfusion with mesenteric vein cannulated; 2) Caco-2 cell monolayers; 3) monolayers of MDCKII cells overexpressing human PgP/MDR1; 4) healthy rats in the presence and absence of combined verapamil; and 5) mdr1a knockout mice. The single-pass rat intestinal perfusion model allows determination of intestinal absorption of compounds without concern for confounding effects from hepatic first-pass effect. We examined the effects of glabridin on PgP/MDR1-mediated digoxin transport in Caco-2 cells and on PgP/MDR1 ATPase activity. In addition, the metabolism of glabridin in rat intestinal and hepatic microsomes was investigated.

	Materials and Methods

Top Abstract Materials and Methods Results Discussion References

 
Chemicals and Reagents. Glabridin [(R)-4-(3,4-dihydro-8,8-dimethyl)-2H,8H-benzo[1,2-b:3,4-b']dipyran-3yl)-1,3-benzenediol] extracted and purified from the root of Salvia miltiorrhiza was purchased from the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). This compound has a purity >99.0%, as determined by high performance liquid chromatography (HPLC). Dulbecco's modified Eagle's medium, fetal bovine serum, 0.05% trypsin-EDTA, penicillin-streptomycin, nonessential amino acids, and sterilized Hanks' balanced salt solution (HBSS; pH 7.0) containing 25 mM HEPES and 25 mM glucose were obtained from Invitrogen (Carlsbad, CA). Krebs-Ringer buffer (pH 6.8) was used as buffer in the rat intestinal perfusion study. Phenol red, trypan blue, propranolol, dimethyl sulfoxide (DMSO), D-saccharic acid 1,4-lactone, verapamil hydrochloride, Brij 58 (polyoxyethylene 20 cetyl ether), nifedipine, probenecid, quinidine sulfate dihydrate, indomethacin, diclofenac, mefenamic acid, mannitol, and antipyrine were all purchased from Sigma-Aldrich (St. Louis, MO). -Nicotinamide adenine dinucleotide phosphate in reduced form (NADPH) and uridine diphosphate glucuronic acid (UDPGA) were purchased from Roche Diagnostics Ltd. (Sydney, Australia). The leukotriene D4 receptor antagonist, 3-[[[3-[2-(7-chloro-2-quinolinyl)-(E)-ethenyl]phenyl][[3-(dimethylamino)-3-oxopropyl] thio]methyl]thio]propionic acid (MK-571 or L-660,711), was a gift from Dr. Ford Hutchinson (Merck Frosst Canada, Inc., Kirkland, QC, Canada). Tissue culture plastics and 0.4-µm pore size 12-mm i.d. Transwell polycarbonate inserts were obtained from Corning Co. (Corning, NY). The control MDCKII cells with empty vector and their human MDR1 recombinantly transfected derivative, MDR1-MDCKII, were obtained as a kind gift from Professor Piet Borst (The Netherlands Cancer Institute, Amsterdam, The Netherlands). The water used was purified by a Milli-Q purification system (Millipore, Billerica, MA). All other chemicals and reagents were of analytical or HPLC grade as appropriate.

Animals. Male healthy Sprague-Dawley rats (200–260 g) were kept in a room under controlled temperature (22 ± 1°C) and automatic day-night rhythm (12-h cycle) and housed in wire-bottom cages with paper underneath. The ethical approval of this study was obtained from the Ethical Committee of the Australian Institute of Chinese Medicine, Sydney, Australia. Animals were treated humanely, and the animal experiments were performed in accordance with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the National Institutes of Health of the United States of America (National Institutes of Health publication 85-23, 1985).

Cell Culture. Caco-2 cells were obtained from The American Type Culture Collection (Manassas, VA). Caco-2, control MDCKII, and MDR1-MDCKII cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 1% nonessential amino acids, and 100 U/ml penicillin and gentamicin in an atmosphere of 5% CO₂ and 90% relative humidity at 37°C. The expression levels of MDR1 in Caco-2 cells, control MDCKII cells, and the MDR1-transfected MDCKII cells were monitored every two to four passages by Western blotting analysis.

Cytotoxicity Assay. The cytotoxic effect of glabridin on various cells examined was determined using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazonium bromide (MTT) assay. Cells were exposed to the drug for 48 h, and the absorbance of formazan, a metabolite of MTT, was measured at a wavelength of 595 nm using a microplate reader (Tecan Instrument Inc., Research Triangle Park, NC). The MTT assays were the means of at least six independent experiments, each performed in replicates of eight for each drug concentration.

Systemic Bioavailability of Glabridin in Rats. The animals were fasted overnight with free access to water before drug administration. Glabridin was freshly prepared by dissolving in DMSO and then diluted with distilled water, resulting in a final DMSO concentration of 0.2% (v/v). Rats were randomized to three groups (n = 6) to receive 5 or 20 mg/kg glabridin by gavage, or 5 mg/kg by i.v. bolus injection through the tail vein. Blood samples were collected through jugular vein cannulation into heparinized tubes at predetermined times over 24 h after drug administration. Plasma was obtained by centrifugation at 5000g for 6 min at 4°C and the plasma was then transferred to clean 1.5-ml tubes. All samples were stored at –20°C until analysis.

Preparation of Rat Intestinal and Hepatic Microsomes. Hepatic and intestinal mucosal tissues were collected from healthy male Sprague-Dawley rats (190–250 g) and stored at –80°C. Hepatic and intestinal microsomes were prepared by differential centrifugation as described previously (Zhou et al., 2000). The rat liver homogenates were centrifuged at 9000g for 20 min at 4°C. The supernatant was then centrifuged at 105,000g for 1 h at 4°C using an ultracentrifuge with a Type 70 Ti rotor (Beckman Coulter, Inc., Fullerton, CA). Microsomal protein concentration was determined by the bicinchoninic acid method (Smith et al., 1985). Tissue microsomes were stored at –80°C until use.

In Vitro Metabolism and Metabolic Inhibition Study in Rat Hepatic and Intestinal Microsomes. Initial incubations containing either 1.0 mg/ml rat hepatic microsomes or 5.0 mg/ml rat intestinal microsomes in the presence of NADPH or UDPGA were performed to investigate whether phase I and/or glucuronidation reaction were involved in the metabolism of glabridin. The depletion of substrate was determined to monitor glabridin metabolism. A single ion monitoring determined the possible formation of new metabolites of glabridin in intestinal and hepatic microsomes using liquid chromatographymass spectrometry (LC-MS). Once any metabolite from phase I or phase II reaction was detected, the incubation conditions in hepatic and intestinal microsomes were examined with respect to microsomal protein concentration and incubation time. Typical incubations (200 µl) for glabridin glucuronidation contained hepatic (0.1 mg/ml) or intestinal (1.0 mg/ml) microsomal protein, 10 mM UDPGA, 5 mM MgCl₂, 0.1 mg/ml D-saccharic acid 1,4-lactone, Brij 58 (0.1–0.2:1, ratio of Brij 58 over microsome, w/w), and glabridin (0.1–50 µM) in 0.1 M phosphate buffer (pH 6.8). D-Saccharic acid 1,4-lactone was used to inhibit the activity of ß-glucuronidase in microsomes. Typical microsomal incubations (200 µl) for oxidation contained hepatic (1.0 mg/ml) or intestinal (5.0 mg/ml) microsomal protein, 0.5 mM NADPH, 5 mM MgCl₂, and glabridin (0.1–50 µM) in 0.1 M phosphate buffer (pH 7.4). All incubations were performed in triplicate, initiated by the addition of UDPGA or NADPH, and conducted at 37°C in a shaking water bath for 30 min. Incubations were stopped by cooling on ice and adding 400 µl of an ice-cold acetonitrile/methanol mixture (3:1 v/v) containing 2 µM internal standard (IS), and vortexing vigorously. Mixtures were centrifuged at 5000g for 10 min to remove the precipitated microsomal protein. The supernatant was removed and evaporated under nitrogen and the residue reconstituted with 100 µl of mobile phase, and 20 to 50 µl was injected into the LC-MS apparatus for the determination of glabridin and identification of metabolites. The inhibition of hepatic glucuronidation of glabridin in vitro by various compounds including verapamil, indomethacin, and diclofenac (all at 100 µM) was investigated at various concentrations.

Single-Pass Intestinal Perfusion Experiments. The surgical procedures were conducted carefully to prepare the single-pass intestinal perfusion with mesenteric vein cannulation as described previously (Zhang et al., 2006). In brief, rats were anesthetized with 0.5 ml of a cocktail containing ketamine at 75 mg/kg and xylazine (Sigma-Aldrich) at 5 mg/kg by i.p. injection, and three essential procedures were performed for animals undergoing in situ intestinal perfusion: jugular vein cannulation for infusion of blood collected from the donor rats, isolation of an ileal segment for glabridin perfusion, and cannulation of the mesenteric vein for continuous collection of blood samples. Glabridin was infused at 0.1, 0.5, or 2.0 µM with or without the presence of verapamil at 100 µM (a PgP inhibitor), probenecid (200 µM, a MRP1 inhibitor), MK-571 (100 µM, a MRP1/2 inhibitor), or celecoxib (100 µM, a MRP4 inhibitor). The in situ intestinal perfusion was initiated by infusing glabridin solution from the syringe pump at 1.0 ml/min for 4 min followed by perfusion at 0.25 ml/min for the remainder of the experiment using a syringe pump ("22" pump; Harvard Apparatus, Holliston, MA). The blood from the mesenteric vein was collected into a heparinized 1.5-ml tube at 5-min intervals over 60 min. In the meantime, perfusate samples were also collected from the outflow of the segment outlet every 5 min into 1.5-ml test tubes over 60 min. The collected blood samples were immediately centrifuged at 5000g for 6 min and the resultant plasma was transferred to a clean 1.5-ml tube and stored at –20°C until analysis.

The perfusion solution containing 100 µM(34.5 µg/ml) phenol red (i.e., phenolsulfonphthalein) was used as a nonabsorbable marker for measuring water flux and to correct for changes in the water flux across the incised ileal segment (Zhang et al., 2006). Additional control experiments were conducted to examine the disappearance (in ileal lumen) and appearance (in mesenteric vein blood) rates of the passive transcellular (antipyrine, 100 µM) and paracellular (mannitol, 1.0 mM) markers to validate our single-pass rat ileum perfusion system (n = 6 per group). The gut does not metabolize mannitol and antipyrine and absorbs these two compounds in an unchanged form. The effects of verapamil at 100 µM, probenecid at 200 µM, MK-571 at 100 µM, and celecoxib at 100 µM on the intestinal transport of both probe compounds, antipyrine and mannitol, were also investigated.

Uptake and Efflux Study of Glabridin by Cells. The uptake and efflux of glabridin by Caco-2, MDCKII, and MDR1-MDCKII cells were examined in confluent cell cultures grown on 60-mm plastic culture dishes (Corning) as described previously (El Hafny et al., 1997; Zhou et al., 2005). For the uptake assay, exponentially growing cells were exposed to 0 to 50 µM glabridin over 120 min at 37°C. For the concentration-dependence study, the incubation time was 30 min. The medium was aspirated off at indicated times, and the dishes were rapidly rinsed five times with 50 ml of ice-cold phosphate-buffered saline (PBS). HPLC analysis ensured that the final wash contained no residual glabridin. The cells were harvested and each cell pellet was suspended in 200 µl of extraction solution (acetonitrile/methanol 1:1, v/v, with 0.01 N HCl) with the addition of 10 µl of 1.0 mg/ml mefenamic acid, which was used as IS. For the efflux assay, glabridin (0.1–50 µM) was added to confluent cell cultures grown on 60-mm plastic culture dishes (Corning) before three washes with 20 ml of warm PBS and incubated for 120 min. After five washes of the cells with 4°C PBS to eliminate the extracellular drug, cells were incubated in culture medium for 20 min at 37°C. After centrifugation of the cells in culture medium at 5000g for 10 min, the supernatant was dried using a rotary concentrator and the residues were reconstituted with the mobile phase; 10 to 20 µl was then injected into the LC-MS system for glabridin concentration determination. The cellular uptake and efflux of glabridin was expressed as ng/min/mg cellular proteins and corrected by subtraction of the mean extracellular adsorption value of [¹⁴C]sucrose by cells tested.

The effects of various ATP inhibitors (sodium azide at 10 mM and 2,4-dinitrophenol at 5 mM), and PgP and MRP inhibitors, including verapamil, nifedipine (both PgP inhibitors, 100 µM), MK-571 (a MRP1/2 inhibitor, 100 µM), probenecid (a MRP1 inhibitor, 200 µM), and celecoxib (a MRP4 inhibitor, 100 µM), on glabridin cellular uptake and efflux were investigated in Caco-2 cells. All inhibitors were freshly prepared by dissolving in DMSO and then diluted by PBS. The final concentration of DMSO was 0.2% (v/v). These inhibitors at indicated concentrations showed little cytotoxicity (<8.0%) to the cells tested using the MTT assay when incubated for 2 h. For uptake inhibition assay, all inhibitors were preincubated with cells for 2 h and coincubated for 30 min after addition of glabridin. For efflux inhibition assay, glabridin at 0.1 or 1.0 µM was added to the cells and incubated for 120 min to achieve maximum drug uptake. After five washes with ice-cold PBS, the cells were incubated in the presence of an inhibitor for 30 min at 37°C. Thereafter, cells were washed with cold PBS buffer five times. The cells were then harvested, lysed by sonication, and extracted using ice-cold acetonitrile/methanol mixture (1:1 v/v, with 0.01 N HCl) as described above. All uptake and efflux assays in the absence and presence of inhibitor were studied in at least three independent experiments. DMSO at a final concentration of 0.2% (v/v) in the culture buffer, used to dissolve all the inhibitors, did not change the accumulation of glabridin in Caco-2 and MDCKII cells.

The uptake of a known PgP substrate, daunomycin, was performed with or without 100 µM verapamil as described above. Preliminary experiments showed that daunomycin uptake was at equilibrium after 60 to 90 min of incubation in Caco-2 cells. Cells were thus incubated for 30 min with 1.0 µM daunomycin (0.3 µCi/well [³H]daunomycin; PerkinElmer Life and Analytical Sciences, Boston, MA) and unlabeled daunomycin with or without 100 µM verapamil. Cells were then washed with HBSS and further processed as described above. In addition, control uptake assays were performed using the extracellular marker [¹⁴C]sucrose (565 mCi/mmol) and [³H]propranolol (both from GE Healthcare, Buckinghamshire, UK). For the efflux assay, [³H]vinblastine (GE Healthcare) was used as a model substrate, and the radioactivity was determined by an LC-6000 liquid scintillation counter (Beckman Coulter, Inc.).

Transport Study of Glabridin in Caco-2, Control MDCKII, and MDR1-MDCKII Monolayers. For the transport studies of glabridin, Caco-2 cells, control MDCKI cells, or MDR1-MDCKII cells were seeded at a density of 5 to 10 x 10⁵ cells/well onto polycarbonate membrane Transwell inserts (Corning) in 12-well plates. The effective transepithelial electric resistance (TEER) of the monolayers (equal to total TEER value – value in empty filter membranes) was examined routinely before and after the experiment using a Millicell ERS apparatus (Millipore). Caco-2 cells were used for transport experiments 21 days after cell seeding, when the effective TEER values typically exceeded 260 to 350 · cm². The transport experiments in Caco-2 cells were conducted on cells between passages 30 and 35. For MDCKII and MDR1-MDCKII cells, only cells at passages 5 to 9 were used for transport studies after receipt from The Netherlands Cancer Institute. Cells were used in transport experiments at days 5 to 7 after cell seeding, when the effective TEER values for MDCKII monolayers were typically 40 to 60 · cm² and 120 to 150 · cm² for MDR1-MDCKII monolayers. [¹⁴C]Mannitol (GE Healthcare) was used as a probe for paracellular transport, and a value of 0.5% per hour indicated acceptable integrality for the monolayers examined. The transport of glabridin by Caco-2, control MDCKII, and MDR1-MDCKII monolayers was investigated on an orbital shaker as described previously (Zhang et al., 2006). In brief, the monolayers were washed twice with warm HBSS containing 25 mM HEPES (pH 7.0) before the transport experiments. A pH of 7.0 was chosen because it was close to the ileum pH value, and this pH resulted in maximum apical (AP) to basolateral (BL) and BL to AP transport of glabridin. Nine independent incubations were performed in triplicate for all experiments.

The effects of pH, Na⁺, temperature, and ATP on glabridin transport across Caco-2 monolayers were investigated. The effect of apical or basolateral pH (5.5–7.4) on the AP to BL and BL to AP flux of glabridin at 0.1 and 1.0 µM was examined at pH 7.4 for the receiving side. The pH was altered by substituting appropriate amounts of HEPES in the incubation medium by equimolar (25 mM) 2-[N-morpholino]ethanesulfonic acid. In experiments to investigate the effect of Na⁺ on the flux of glabridin at 0.1 and 1.0 µM across the Caco-2 monolayers, sodium chloride in the HBSS was replaced by equimolar amounts (140 mM) of potassium chloride. In addition, the permeability of glabridin from AP to BL and BL to AP was measured after incubation for 30 min at 4°C or 37°C. To determine whether there was an energy dependence of glabridin flux in Caco-2 cells, the transport medium depleted in glucose was used in both sides of the monolayers. Like all other ABC transporters, PgP acts as an efflux pump by exporting its substrate from the membrane or cell cytosol to the exterior of the cells, with ATP hydrolysis as the driving force. Sodium azide (10 mM) or 2,4-dinitrophenol (5.0 mM) (both ATPase inhibitors) was added to both the AP and BL sides and the monolayers were incubated for 30 min at 37°C. In experiments to investigate the effects of verapamil (100 µM), probenecid (200 µM), MK-571 (100 µM), and celecoxib (100 µM) on the transport of glabridin (0.1 and 1.0 µM) from AP to BL and BL to AP directions in the Caco-2 monolayers, the inhibitor was added to the incubation medium on both the AP and BL sides of the monolayers throughout the experiment and also preincubated with the cells for 2 h before addition of glabridin. The cells were further incubated for 30 min after addition of glabridin at 0.1 or 1.0 µM. All inhibitors were freshly prepared using DMSO before the experiment, with a final DMSO concentration of 0.2% (v/v). Vehicle (0.2% DMSO) was used for the control inserts. All collected transport buffer samples containing glabridin were stored at –20°C until analysis using a validated LC-MS method. All incubations were performed in triplicate to nine times. To avoid interday cell-cell variations, the transport experiments for the determination of transport kinetics or inhibition by various compounds were conducted on the same day using the same batch of cellular monolayers.

Inhibition of Digoxin Transport by Glabridin in Vitro. Caco-2 cells were grown and cultured on 0.4-µm polycarbonate membrane Transwell inserts (Corning). Transport of [³H]digoxin (GE Healthcare) at 5.0 µM (15 Ci/mmol) was determined by its addition to the BL side of the Caco-2 monolayer and by measuring the transport of radioactivity into the receiving compartment over 1 h, in the absence or presence of glabridin (0.1–100 µM), or the positive control, verapamil (0.25–100 µM), added in both sides.

P-glycoprotein ATPase Activity Assay. The effect of glabridin on PgP/MDR1 ATPase activity was determined using the ABC Transporter ATPase assay reagent kit consisting of human PgP/MDR1 membrane and PgP/MDR1-negative control membrane fractions, buffers, solutions, and relevant reagents (Nacalai Tesque Inc., Kyoto, Japan). In brief, human PgP/MDR1 membrane or PgP/MDR1-negative control membrane (20 µg) was preincubated at 37°C for 5 min in 50 µl of reaction buffer and glabridin (0.1–100 µM) in the presence or absence of 50 µM sodium orthovanadate in a 96-well microplate (Invitrogen). The ATPase reaction was initiated by the addition of 25 µl of 10 mM MgATP solution. The reaction was stopped after 30 min by addition of 30 µl of 10% (w/v) sodium dodecyl sulfate solution, and the amount of phosphate was determined immediately by adding 200 µl of detection solution (1:4, 35 mM ammonium molybdate in 15 mM zinc acetate, pH 5.0, and 10% ascorbic acid, pH 5.0) and incubated at 37°C for 20 min in the dark. The amount of inorganic phosphate complex was determined by measuring the absorbance at 750 nm wavelength by comparing the absorbance to a blank phosphate buffer (pH 7.4) as a standard using a microplate spectrometer (Tecan Instrument Inc.). The vanadate-sensitive ATP hydrolysis was determined by subtracting the value obtained with the vanadate-coincubated membrane fractions from vanadate-free membrane fractions. Verapamil (0.25–100 µM) was used as the positive control in these experiments. To estimate the reaction kinetics by glabridin and verapamil, the ATP hydrolysis rate was fitted to several nonlinear kinetic models using the Prism 3.0 program (GraphPad Software Inc., San Diego, CA).

Effects of Coadministered Verapamil on the Plasma Pharmacokinetics of Glabridin in Rats. In separate kinetic experiments, we examined the effects of coadministered verapamil at 25 or 100 mg/kg on the plasma pharmacokinetics of glabridin in healthy male Sprague-Dawley rats. The dose of verapamil selected was approximately the maximum tolerated dose, as assessed in pilot experiments in male Sprague-Dawley rats. Rats were randomized to receive the following different treatments (n = 6 per group): glabridin at 5 mg/kg by gavage plus water (0.3 ml, control vehicle), and glabridin at 5 mg/kg by gavage in combination with verapamil at 50 or 100 mg/kg by oral gavage dissolved in water. The inhibitor was administered 2 h before glabridin dosing. Blood was collected as described above. The concentrations of glabridin in plasma were determined by LC-MS.

Pharmacokinetic Study of Glabridin in mdr1a(–/–) Gene-Deficient and Wild-Type Mice. FVB/NJ (20–35 g) and mdr1a gene-deficient mice (25–30 g) were purchased from The Jackson Laboratory (Bar Harbor, ME) and Taconic Farms, Inc. (Germantown, NY), respectively. The mice were treated with oral glabridin at 5 mg/kg by gavage (n = 4 per time point). At predetermined time points, the mice were sacrificed by neck dislocation. Blood was immediately collected and plasma was obtained by centrifugation at 5000g for 10 min at 4°C, and all tissues were processed as described above. The concentration of glabridin in plasma and tissues was determined by LC-MS analysis.

Liquid Chromatography-Mass Spectrometry and High Performance Liquid Chromatography Analysis of Glabridin. The concentrations of glabridin in rat and mouse plasma, perfusates, and transport medium in cellular monolayers and cellular lysates were determined by an LC-MS system equipped with an Agilent 1100 LC (Agilent Technologies, Palo Alto, CA) connected to an Applied Biosystems Q-Trap 4000 mass spectrometer (Applied Biosystems, Foster City, CA) through an electrospray ionization source. Chromatographic separation was achieved using a C18 HyperClone ODS column (200 mm x 4.6 mm i.d.) (Phenomenex, Torrance, CA) preceded by a Phenomenex C18 guard cartridge at room temperature (22°C). Mobile phase composed of methanol and 0.1% (v/v) formic acid (85:15, v/v) had a flow rate of 0.2 ml/min. Injection volume was 20 µl of a sample kept in an autosampler set at 10°C. Air set at 600°C and a pressure of 70 psi was used for heating, and the nebulizing gas was set at 40 psi. The capillary temperature was 450°C and the spray voltage was 4000 V. The product ions were recorded using a negative ion detection mode. The monitored ions and collision energy were m/z 323.1201.3 and 30 eV for glabridin, and m/z 240.1196.1 and 25 eV for the IS. The lower limit of quantitation of glabridin was 0.025 to 0.05 ng/ml in rat and mouse plasma, perfusates, and other matrices tested in this study. Glabridin was recovered by >95%, and was stable when kept at 10°C for 36 h, at –20°C for 3 months, and after five to eight freeze-thaw cycles.

The antipyrine and mannitol in perfusates were determined by a validated HPLC method as described previously (Miki et al., 1996; Hung et al., 2001). The lower limits of quantitation for antipyrine and mannitol were 30.0 and 195.0 ng/ml, respectively.

Pharmacokinetic Calculation. The plasma concentration-time curves of glabridin in rats were obtained by plotting the mean plasma concentrations of glabridin versus time on a semilogarithmic scale. Pharmacokinetics parameters were calculated by standard model-independent pharmacokinetic formulae using the WinNonlin program (Pharsight Inc., Mountain View, CA). The elimination half-life (t_ß) value was calculated as 0.693/ß, where ß is the elimination rate constant calculated from the terminal linear portion of the log plasma concentration-time curve. The total areas under the plasma concentration-time curve from time 0 to the last quantifiable time point (AUC_0-t) and from time 0 to infinity (AUC_0-) were calculated using the log trapezoidal rule. The AUC₀ was calculated by the following equation:

(1)

where C_t is the last measurable plasma concentration. The maximum plasma concentration (C_max) for glabridin was obtained by visual inspection of the plasma concentration-time curve, whereas the initial drug concentration (the extrapolated concentration at zero time) of the drug after i.v injection was calculated by back-extrapolation of the plasma concentration-time curve to the y-axis. The plasma clearance (CL) was estimated by dividing the total administered dose by the AUC_0-. The volume of distribution (V_d) was calculated by dividing CL by ß.

The systemic bioavailability (F, %) of glabridin after oral administration was determined as follows:

(2)

where AUC^p.o. and AUC^i.v. are the areas under the plasma concentration curves calculated after oral and i.v. administration, respectively.

Data Analysis. Data are presented as mean ± S.D. The initial statistical analysis to evaluate the differences in the mean kinetic parameters among the different groups was carried out by a one-way analysis of variance followed by a post hoc test (Dunnett's multiple comparison test). Student's t test was conducted for the between-group comparisons with a significance level of P < 0.05.

The permeability values of glabridin across rat ileum were calculated based on the disappearance of the drug from the lumen (P_lumen) as well as the appearance of drug in the blood (P_blood) using the following equations (Zhang et al., 2006):

(3)

(4)

where r is the radius of the intestinal lumen (0.18 cm), l is the length of the segment (cm), Q is the flow rate of drug through the intestinal segment, C₀ is the concentration of glabridin at the start of perfusion (in the syringe), C_lumen is the steady-state concentration of glabridin exiting the lumen, dX/dt is the rate of glabridin appearance in mesenteric venous blood (µg/s), and A is the surface area of the ileal segment (= 2 rl,cm²).

The apparent permeability coefficient (P_app) in cellular monolayers is expressed in centimeters per second and calculated as the following equation:

(5)

where Q/t is the permeability rate (µg/s), A is the surface area of the membrane (cm²), and C₀ is the initial drug concentration in the donor chamber (µg/ml). Samples from the 30-min point were used for P_app calculations since, at this time, steady state has been achieved. The net BL to AP efflux of glabridin (R_net) was determined by calculating the ratio of P_app in the BL to AP direction versus P_app in the AP to BL direction (P_app(BL-AP)/P_app(AP-BL)) as eq. 6 (Zhou et al., 2005).

(6)

The passive diffusion flux rate (excluding the influence of efflux transporter) of glabridin in Caco-2 monolayers was estimated by conducting the transport experiment in the presence of verapamil (100 µM), assuming that other transporters play a minimal role in glabridin transport. The active transport flux rates were then estimated by subtracting the passive diffusion rates from total flux rates. Several models to describe the kinetics of the calculated active transport and ATPase stimulation activity of glabridin (single- and two-binding sites with and without a nonsaturable component, substrate inhibition, and the sigmoid models) were fitted with the following models and using the Prism 3.0 program (GraphPad Software Inc., San Diego, CA).

(7)

(8)

(9)

(10)

(11)

(12)

where v is the rate of glabridin transport or ATP hydrolysis; V_max is the maximum velocity; K_m, Michaelis-Menten constant; [S], the substrate concentration; K_d, the nonsaturable component; K_is, the substrate inhibition constant; h', the Hill coefficient for cooperative substrate binding; and subscripts 1 and 2 represent the first and the second type of enzyme binding sites. The choice of model was confirmed by F test and comparison of Akaike's information criterion values (Yamaoka et al., 1978).

Apparent inhibition constant (K_i) was estimated using eqs. 13 to 15 as previously described (Zhou et al., 2005).

(13)

(14)

(15)

where P_I and P₀ are the P_app values of glabridin in the direction of BL to AP in the presence and absence of the inhibitor, respectively; and P_I/P₀ is a reflection of the inhibitory effect of the test compound on the active BL to AP transport of glabridin across the Caco-2 monolayers. [I] is the concentration of inhibitor in the donor and the receiver side. P_app1 is the total transport in the absence of any inhibitory compound, P_app2 is the total transport in the presence of a potential inhibitor, and P_app3 is the passive diffusion component.

	Results

Top Abstract Materials and Methods Results Discussion References

 
Oral Bioavailability of Glabridin in Rats. The representative plasma concentration-time profiles of glabridin after oral (5 and 20 mg/kg) and i.v. (5 mg/kg) administration in rats are shown in Fig. 2 and the pharmacokinetic parameters of glabridin are listed in Table 1. After oral administration of glabridin at 5 or 20 mg, the C_max, T_max, and AUC_0–24h of glabridin were 15.10 ± 4.72 ng/ml and 60.41 ± 18.87 ng/ml; 4.33 ± 1.86 h and 4.50 ± 2.17 h; and 96.49 ± 32.34 ng/ml · h and 387.98 ± 137.28 ng/ml · h, respectively. These results showed that the oral pharmacokinetics of glabridin is linear (dose-independent), as indicated by the proportional increase of C_max and AUC of glabridin when its oral dose was increased from 5 mg/kg to 20 mg/kg. The t_ß of glabridin was 2.38 to 2.41 h after administration at 5 to 20 mg/kg.

View larger version (26K): [in this window] [in a new window]   FIG. 2. The representative plasma concentration-time profiles of glabridin after oral administration at 5 and 20 mg/kg and intravenous bolus injection at 5 mg/kg (n = 6 per group).

In addition, after i.v. bolus injection of glabridin at 5 mg/kg, the C_max, AUC_0–24h, t_ß, CL, and V_d were 1.92 h, 1301.48 ± 375.79 ng/ml · h, 1.92 h, 59.01 ml/min/kg and 2.72 l/kg, respectively. Thus, the F values of glabridin in rats were 7.45% and 7.44%, respectively, after oral dosing of glabridin at 5 and 20 mg/kg. These findings indicate that the oral absorption and oral bioavailability of glabridin are low and dose-independent.

In Vitro Metabolism of Glabridin in Rat Intestinal and Hepatic Microsomes. In rat intestinal and hepatic microsomes, no oxidative metabolites of glabridin were observed when glabridin was incubated in the presence of NADPH using LC-MS and HPLC methods. Marked formation of glabridin glucuronide was found when glabridin was incubated with rat hepatic microsomes in the presence of UDPGA. The formation of glabridin glucuronide, when monitored by relative metabolite ion density using LC-MS, was linear up to 60 min incubation time and 50 µM substrate concentration. A minimal peak of glabridin glucuronide using HPLC and a weak signal of glabridin glucuronide ion using LC-MS were observed when the substrate was incubated in rat intestinal microsomes. The formation rate of glabridin glucuronide in rat intestinal microsomes was approximately 1/15 to 1/20 of that when incubated with rat hepatic microsomes. In addition, diclofenac and indomethacin (both at 100 µM) inhibited the formation of glabridin glucuronide in rat hepatic and intestinal microsomes by 68.5 ± 15.2% and 72.3 ± 14.3%, respectively. However, verapamil at 100 µM did not inhibit glabridin glucuronidation in both intestinal and hepatic microsomes.

Transport of Glabridin in Single-Pass Perfusion Study of Rat Ileum. There was insignificant loss of glabridin (<6.0%) when the drug was perfused through the perfusion apparatus used in this study, indicating that there was no significant adsorption of glabridin to the tubing wall of the system. The compound was stable in the perfusion buffer as well as intestinal perfusate at 37°C for at least 12 h.

There were no oxidative metabolites formed in the perfusates or mesenteric vein blood after glabridin was loaded, as determined by both HPLC and LC-MS analysis. No detectable peak of glabridin glucuronide in the perfusates or mesenteric vein blood was found using HPLC analysis, but a weak signal of glabridin glucuronide ion using LC-MS was observed. This indicated that the glucuronidation of glabridin by rat perfused ileum segment was minimal or just detectable, and gut metabolism had a minor impact on the determination of permeability coefficients of glabridin in this model.

For intestinal perfusions with glabridin, samples were collected from the outlet of the ileal segment and mesenteric vein at 5-min intervals over 60 min. The permeability values of glabridin at 0.1, 0.5, and 2.0 µM are shown in Table 2. The permeability of glabridin based on the luminal disappearance of the compound was estimated at steady state (i.e., samples obtained from 30 to 60 min). The P_lumen values of glabridin were 6.51 ± 0.72 x 10^–4, 8.22 ± 0.91 x 10^–4, and 11.54 ± 1.31 x 10^–4 cm/s, respectively, when the concentrations of glabridin in perfusates were 0.1, 0.5 and 2.0 µM. With the increase of glabridin concentration in perfusates, the P_lumen values significantly increased (P < 0.05). The appearance of glabridin in mesenteric vein blood increased when the drug concentration was 0.1, 0.5, and 2.0 µM in perfusates. The concentration-dependent increase in P_lumen of glabridin may reflect the relatively low to moderate intrinsic permeability of glabridin and possible involvement of a saturable active mechanism for its intestinal transport. As for the permeability based on appearance of glabridin in the mesenteric blood (P_blood), concentration-dependent increases in permeability were evident, whereas the P_blood values at 0.1, 0.5, and 2.0 µM were 6.5- to 7.0-fold lower than P_lumen (P < 0.05).

Verapamil, probenecid, MK-571, and celecoxib all did not significantly alter the P_lumen values of glabridin at 0.1, 0.5, or 2.0 µM(P > 0.05). However, the P_blood values of glabridin at 0.1, 0.5, and 2.0 µM increased significantly in the presence of 100 µM verapamil, with the ratio of P_lumen/P_blood decreased from 7.5 to 8.0 to 3.0 to 3.3 (Table 2). However, probenecid, MK-571, and celecoxib insignificantly affected the P_blood values at all substrate concentrations tested. These data suggest that PgP-mediated efflux effectively limited the absorption of glabridin across the intestinal wall, which was at least partially reversed by verapamil, a known PgP inhibitor, but not by MRP inhibitors including probenecid (MRP1), MK-571 (MRP1/2), and celecoxib (MRP4).

View larger version (19K): [in this window] [in a new window]   FIG. 3. Effects of incubation time (A and C) and substrate concentration (B and D) on the uptake and efflux of glabridin in Caco-2 cells cultured in HBSS. The curves in plots B and D represent the best fit of the one-binding-site model. E, the kinetic profiles of glabridin efflux from Caco-2 cells when preloaded at 1.0 µM and the intracellular concentrations of glabridin were monitored by LC-MS over 120 min of incubation. The data are the mean ± S.D. of at least six to nine determinations.

Cytotoxicity and Metabolism of Glabridin in Caco-2, Control MDCKII, and MDR1-MDCKII Cells. Glabridin at 0.1 to 100 µM did not show significant cytotoxicity (<10%) to Caco-2, control MDCKII, and MDR1-MDCKII cells when incubated for up to 48 h as determined by the MTT assay.

No detectable oxidative metabolites were observed when glabridin at 0.1 to 100 µM was incubated with Caco-2, control MDCKII, and MDR1-MDCKII cells for 2 to 48 h as determined by HPLC and LC-MS analysis. A minor signal was detected for glabridin glucuronide ion when the substrate was incubated with Caco-2 for 2 to 48 h, as determined by LC-MS analysis. The minimal formation of glabridin glucuronides in Caco-2 cells indicated that Caco-2 cells had a weak metabolic ability for glabridin, possibly due to low levels or activity of uridine diphosphate glucuronosyltransferases that metabolize this compound, and metabolism had a minor impact on the assessment of transport of glabridin in this model.

View larger version (17K): [in this window] [in a new window]   FIG. 4. Effects of sodium azide, 2,4-dinitrophenol, verapamil, nifedipine, probenecid, MK-571, and celecoxib on the uptake in Caco-2 cells (A) and efflux (B) of glabridin at 1.0 µM from Caco-2 cells. The inhibitor was preincubated with the cells for 2 h and incubated further for 30 min after addition of glabridin for uptake inhibition assay. For efflux inhibition assay, the cells were washed five times with ice-cold PBS after incubation with glabridin at 1.0 µM for 120 min and incubated further in the presence of an inhibitor for 30 min at 37°C. The data are the mean ± S.D. of at least six to nine determinations. *, P < 0.05; **, P < 0.01.

View larger version (21K): [in this window] [in a new window]   FIG. 5. Transport of glabridin across Caco-2 monolayers. A, effect of concentration of glabridin (0.1–100 µM) on the permeability (Papp) of glabridin from the apical (AP) to basolateral (BL) and BL to AP side; and B, effect of incubation time (0–60 min) on the permeability (Papp) of glabridin at 1.0 µM in the AP to BL and BL-AP directions. Glabridin was loaded on either the AP or BL side and incubated at 37°C. Samples from the receiving side were collected at indicated time points and glabridin was determined by LC-MS. Data are the mean ± S.D. from at least six to nine determinations. **, P < 0.01.

View larger version (20K): [in this window] [in a new window]   FIG. 6. Model fitting for the passive and active transport of glabridin from apical (AP) to basolateral (BL) and BL to AP sides in Caco-2 monolayers. The passive and active BL-AP and AP-BL transport was calculated only based on verapamil inhibition of PgP in Caco-2 cells assuming that the role of other transporters is minimal and ignored. Glabridin (0.1–100 µM) was loaded on either the AP or BL side and incubated at 37°C over 30 min. Samples from the receiving side were collected and glabridin was determined by LC-MS. Data are the mean ± S.D. from six to nine independent determinations. The curve represents the best fit of one saturable transport model.

Uptake of the probe markers sucrose and propranolol into Caco-2 cells was determined upon incubation up to 120 min. Propranolol penetrated into Caco-2 cells to a low extent (0.25 ± 0.03 to 0.44 ± 0.52 ng/min/mg cellular protein), and diffusion of sucrose into the cells was minimal (0.04 ± 0.00 to 0.07 ± 0.01 ng/min/mg cellular protein). The uptake of sucrose and propranolol did not increase with incubation time and substrate concentration in Caco-2 cells. In addition, when daunomycin was used as a model PgP/MDR1 substrate, its uptake by Caco-2 cells was significantly increased by 75.2 ± 9.4% when coincubated with 100 µM verapamil. As well, the efflux of the model substrate vinblastine was also dependent on the substrate concentration and incubation time, but its cellular efflux was characterized by a biexponential kinetics with half-lives of 9.8 min and 2.7 h, respectively.

Transport of Glabridin in Caco-2 Monolayers. After incubation of glabridin at 0.1 to 100 µM loaded at either the AP or BL side, the sample was collected from the receiving side for LC-MS analysis. No detectable metabolites were observed when glabridin was loaded on the AP or BL side at all concentrations over 60 min. The time course and concentration effect of glabridin flux from AP to BL or BL to AP have been examined and the results are shown in Fig. 5. After AP or BL drug loading, glabridin appeared on the receiving side by 5 min. The flux rate (ng/min/cm²) of glabridin from the AP to BL or BL to AP side was largely proportional to glabridin concentrations over 0.1 to 100 µM and was linear up to 60 min of incubation time. The transport rate of glabridin across Caco-2 monolayers from the BL to AP side was significantly (P < 0.05) higher than that from the AP to BL side. The P_app of glabridin from the BL to AP side (8.12–25.63 x 10^–5 cm/s) was approximately 3.3- to 8.4-fold higher than those from the AP to BL side (0.87–5.76 x 10^–5 cm/s) with a marked decrease in P_app values for both directional transport at increasing glabridin concentrations (Fig. 5). The R_net values ranged from 4.3 to 9.4. These results demonstrated a polarization in the Caco-2 permeability toward glabridin and a predominantly secretory rather than absorptive transport. The BL to AP efflux rate of glabridin increased with increasing glabridin concentrations over 0.1 to 100 µM but appeared saturable when glabridin concentration was 10 µM, as indicated by a nonproportional increase in the efflux (data not shown). Consistently, there was a significant decrease in P_app values for BL to AP flux at glabridin concentrations 10 µM(P < 0.001).

View larger version (19K): [in this window] [in a new window]   FIG. 7. Effect of various inhibitors, including sodium azide (10 mM), 2,4-dinitrophenol (5 mM), verapamil (100 µM), probenecid (200 µM), MK-571 (100 µM), and celecoxib (100 µM) on the apical (AP) to basolateral (BL) and BL to AP transport of glabridin at 0.1 µM (A) or 1.0 µM (B) in Caco-2 monolayers upon incubation for 30 min. Glabridin (0.1 or 1.0 µM) was loaded on either the apical or basolateral side and incubated for 30 min at 37°C. Transport studies were undertaken in the absence or presence of the potential inhibitor added to both the apical and basal compartments. All inhibitors were preincubated for 2 h with Caco-2 cells. Samples from the receiving side were collected and glabridin concentrations were determined by LC-MS methods. Data are the mean ± S.D. from at least six to nine determinations. *, P < 0.05; **, P < 0.01.

View larger version (26K): [in this window] [in a new window]   FIG. 8. Effects of incubation time (A and C) and substrate concentration (B and D) on the uptake and efflux of glabridin in control MDCKII and MDR1-MDCKII cells cultured in HBSS. The curves in plots B and D represent the best fit of a one-binding-site model. The data are the mean ± S.D. of at least six to nine determinations. *, P < 0.05; **, P < 0.01.

Reducing the apical pH to 5.5 to 6.5 caused a significant (P < 0.05) increase in glabridin flux by 24.5 to 56.8% at 0.1 and 1.0 µM from the AP to BL or BL to AP side compared with the values at pH 7.4. A maximum P_app was observed at pH 7.0 for both AP-BL and BL-AP transport at 0.1 and 1.0 µM substrate concentration. Lower pH may reduce the ionization of glabridin and thus increase its intestinal transport. The substitution of sodium salts in the transport medium with potassium salts had no significant effect on the flux of glabridin for either the AP to BL or BL to AP direction, suggesting that the active transporter system for glabridin was sodium-independent. Reducing the incubation temperature from 37°C to 4°C significantly (P < 0.05) decreased the flux of glabridin at 0.1 and 1.0 µM from the AP to BL or BL to AP side, with a 42.2 to 75.5% reduction of the P_app values (data not shown). Moreover, the absence of glucose in the transport medium did not significantly affect the AP to BL flux of glabridin at either 0.1 or 1.0 µM. In contrast, depletion of glucose significantly (P < 0.05) decreased the BL to AP flux of glabridin at 0.1 and 1.0 µM by 50.5 to 65.2%. These results indicated that the transport of glabridin across Caco-2 monolayers was pH-, energy-, and temperature-dependent, but not sodium-dependent.

The effects of ATP inhibitors and various ABC transporter inhibitors on the transport of glabridin (0.1 and 1.0 µM) in Caco-2 monolayers were also investigated. Addition of the transport buffer at both sides with sodium azide (10 mM), 2,4-dinitrophenol (1 mM), or verapamil (100 µM) significantly (P < 0.05) increased the AB to BL flux of glabridin at 0.1 µM by 52.8%, 38.6%, and 82.4%, respectively (P < 0.05) (Fig. 7). In contrast, these compounds caused a significant (P < 0.05) decrease in the BL to AP flux of glabridin at 0.1 µMby 48.4%, 43.3%, and 53.1%, respectively. Similar results were observed when the concentration of glabridin was increased to 1.0 µMinthe presence of the above inhibitors. The estimated K_i values based on eqs. 13 to 15 for sodium azide, 2,4-dinitrophenol, and verapamil were 8.8 mM, 1.5 mM, and 6.8 µM, respectively. However, probenecid, MK-571, and celecoxib slightly altered the AP-BL and BL-AP P_app values (P > 0.05), suggesting that MRPs play a minor or negligible role in the intestinal transport of glabridin.

Uptake and Efflux of Glabridin in Control MDCKII and MDR1-MDCKII Cells. As shown in Fig. 8, the intracellular accumulation and efflux amounts of glabridin in both control MDCKII and MDR1-MDCKII cells were linear up to 120 min of incubation time. The accumulation and efflux rate of glabridin in both control MDCKII and MDR1-MDCKII cells over 0.1 to 50 µM increased in a concentration-dependent manner, following the Michaelis-Menten kinetics with the one-binding-site model being the best fit (Fig. 8). The estimated K_m and V_max in both control MDCKII and MDR1-MDCKII cells for glabridin uptake were 12.75 ± 1.43 and 10.91 ± 2.23 µM, and 16.22 ± 0.70 and 4.47 ± 0.34 ng/min/mg cellular proteins, respectively (Fig. 8). The uptake of glabridin by control MDCKII cells was significantly (approximately 3-fold) higher than that in MDR1-MDCKII cells (P < 0.05). The estimated K_m and V_max in both control MDCKII and MDR1-MDCKII cells were 5.70 ± 1.75 and 5.60 ± 0.73 µM, and 8.44 ± 0.81 and 60.03 ± 2.44 pg/min/mg cellular proteins, respectively. The efflux rate of glabridin from MDR1-MDCKII cells was significantly (approximately 5- to 7-fold) higher than that in control MDCKII cells (P < 0.05 or < 0.01).

View larger version (19K): [in this window] [in a new window]   FIG. 9. Transport of glabridin (0.1–50 µM) from the apical (AP) to basolateral (BL) and BL to AP side in control MDCKII and MDR1-MDCKII monolayers. Glabridin was loaded on either the AP or BL side and incubated at 37°C for 15 min. Samples from the receiving side were collected at indicated time points and glabridin was determined by LC-MS. Data are the mean ± S.D. from at least six to nine determinations. **, P < 0.01; ***, P < 0.001.

Uptake of the probe markers sucrose and propranolol into both control MDCKII and MDR1-MDCKII cells was also determined upon incubation up to 120 min. Propranolol penetrated into both cell lines to a low degree with a value of 0.18 ± 0.02 to 0.37 ± 0.40 ng/min/mg cellular protein, and diffusion of sucrose into the cells was minimal (0.03 ± 0.00 to 0.06 ± 0.01 ng/min/mg cellular protein). The uptake of sucrose and propranolol did not increase with incubation time and substrate concentration in either cell line. These findings indicate that PgP/MDR1 did not significantly affect the uptake of both sucrose and propranolol. In addition, the efflux of the probe drug, vinblastine, from MDR1-MDCKII cells was significantly (approximately 4- to 8-fold) higher than that in control MDCKII cells (data not shown).

Transport of Glabridin in Control MDCKII and MDR1-MDCKII Monolayers. To further investigate the nature of the polarized transport of glabridin, transport studies were conducted in control MDCKII and MDR1-MDCKII cells which stably and functionally overexpress the human PgP/MDR1. Our Western blotting analysis demonstrated that control MDCKII cells expressed constitutive canine P-glycoprotein at a much lower level than that in the recombinant MDR1-MDCKII cells (data not shown).

View larger version (18K): [in this window] [in a new window]   FIG. 10. Effects of glabridin on the transport of digoxin in Caco-2 monolayers. The transport of [3H]digoxin at 5.0 µM (15 Ci/mmol) was determined by its addition to the basolateral (BL) side of the Caco-2 monolayer and by measuring the transport of radioactivity in the other compartment (apical side) over 1 h, in the absence or presence of glabridin (0.1–100 µM), or verapamil (0.25–100 µM), added in both sides. **, P < 0.01.

At all substrate concentrations, the permeability of glabridin in the BL-AP direction in the MDR1-MDCKII cells was significantly (P < 0.01 or 0.001) greater than that in the BL-AP direction in control MDCKII cells. As for the Caco-2 data, increased glabridin concentration also resulted in lower P_app values in both AP-BL and BL-AP directions for both MDCKII and MDR1-MDCKII monolayers.

Inhibition of P-glycoprotein-Mediated Digoxin Transport by Glabridin. We examined the effects of glabridin on PgP-mediated transport of the probe digoxin in Caco-2 monolayers. The results are shown in Fig. 10. Glabridin inhibited digoxin transport in a concentration-dependent manner with an IC₅₀ value of 2.56 ± 0.04 µM. In addition, verapamil exhibited potent inhibitory effects on PgP-mediated transport of digoxin, with an IC₅₀ value of 2.34 ± 0.03 µM. These results indicated that glabridin was a potent PgP inhibitor in vitro.

Stimulation of P-glycoprotein ATPase Activity by Glabridin. The affinity of glabridin to PgP/MDR1 was assessed by the ATPase activity assay. The plot of ATP hydrolysis as a function of glabridin concentrations over 0.1 to 100 µM demonstrated a concentration-dependent stimulation of vanadate-sensitive PgP/MDR1 ATPase activity (Fig. 11). The one-binding-site model was the best fit for this reaction. The estimated K_m and V_max values of PgP/MDR1-mediated ATP hydrolysis by glabridin were 25.05 ± 2.86 µM and 80.50 ± 3.47 nmol/min/mg protein, respectively. In addition, a significant stimulatory effect was exhibited by verapamil over 0.25 to 100 µM with a K_m and V_max of 5.98 ± 0.7 µM and 89.34 ± 2.84 nmol/min/mg protein, respectively. The estimated K_m value for verapamil to PgP/MDR1-mediated ATP hydrolysis was in agreement with previously reported values (4.06–6.10 µM) (Adachi et al., 2001; Ohashi et al., 2006).

View larger version (16K): [in this window] [in a new window]   FIG. 11. Concentration-dependent stimulation of PgP ATPase activity in isolated membrane fractions from human MDR1 expressing cells by glabridin (A) and verapamil (B). The ATPase activity in PgP membrane was determined by measuring vanadate-sensitive inorganic phosphate release, using 4 mM MgATP, as described in Materials and Methods. Data are the mean ± S.D. from at least three independent experiments. The vanadate-sensitive ATP hydrolysis was determined by subtracting the values obtained with the vanadate-coincubated PgP membrane fractions from vanadate-free membrane fractions. One-binding site model is the best fit for the ATP hydrolysis reaction stimulated by both glabridin and verapamil. *, P < 0.05; **, P < 0.01.

Effect of Coadministered Verapamil on the Plasma Pharmacokinetics of Glabridin in Rats. The plasma concentration-time profiles when glabridin was coadministered alone or in combination with verapamil at 25 or 100 mg/kg are shown in Fig. 12 and the pharmacokinetic parameters are shown in Table 3. Combined verapamil significantly (P < 0.05) increased the plasma C_max and AUC_0–24h of glabridin in a dose-dependent manner. Coadministered verapamil at 25 and 100 mg/kg caused a significant increase (P < 0.05) in plasma C_max [from 15.02 ± 4.59 (control) to 19.01 ± 5.38 and 25.38 ± 7.62 ng/ml, respectively] and AUC_0–24h [from 97.41 ± 34.58 (control) to 116.21 ± 36.36 and 170.52 ± 52.28 ng/ml · h, respectively], compared with the control rats receiving glabridin alone. The oral bioavailability of glabridin was correspondingly increased from 7.55% to 9.02% and 13.19%, respectively. In addition, coadministered verapamil increased the t_ß values of glabridin in a dose-dependent manner. Furthermore, the T_max of glabridin was 4.15 ± 0.97 h, whereas it was significantly (P < 0.05) decreased to 2.25 ± 0.52 and 2.15 ± 0.46 h when glabridin was coadministered with 25 and 100 mg/kg verapamil, respectively.

View larger version (25K): [in this window] [in a new window]   FIG. 12. The plasma concentration-time profiles in rats treated with glabridin alone or glabridin in combination with 5 or 25 mg/kg verapamil dosed orally. Rats were treated with glabridin at 5 mg/kg by gavage and the plasma concentrations of glabridin over 24 h were determined by LC-MS analysis. Data are the mean ± S.D. of six rats.

A Comparison of Plasma Pharmacokinetics of Glabridin in mdr1a(–/–) and Wild-Type Mice. To further investigate the impact of PgP/MDR1 on the plasma pharmacokinetics, we compared the pharmacokinetics of glabridin in mdr1a(–/–) and wild-type mice. The results are shown in Fig. 13 and Table 4. The plasma pharmacokinetics of glabridin in mdr1a(–/–) mice were significantly different from those in wild-type mice. The plasma AUC_0–24h and C_max of glabridin in mdr1a(–/–) mice were 3.77- and 2.83-fold higher, respectively, than those in wild-type mice, with significantly longer elimination half-life observed in mdr1a(–/–) mice compared with wild-type mice (3.54 ± 1.14 versus 2.97 ± 0.89 h). These findings provided further evidence that PgP/MDR1 had an important impact on the oral bioavailability and elimination in vivo.

View larger version (19K): [in this window] [in a new window]   FIG. 13. The plasma concentration-time profiles of glabridin in mdr1a(–/–) and wild-type mice. Mice were treated with glabridin at 5 mg/kg by gavage and the plasma concentrations of glabridin over 24 h were determined by