Selective inhibition of MDR1 (ABCB1) by HM30181 increases oral bioavailability and therapeutic efficacy of paclitaxel

Abstract

Multi-drug resistance 1 (MDR1, ABCB1), also known as P-glycoprotein (P-gp), restricts intestinal uptake of many drugs, and contributes to cellular resistance to cancer chemotherapy. In this study, we examined the pharmacologic characteristics of HM30181, a newly developed MDR1 inhibitor, and tested its capacity to increase the oral bioavailability and efficacy of paclitaxel, an anti-cancer drug usually given by intravenous injection. In the ATPase assay using MDR1-enriched vesicles, HM30181 showed the highest potency (IC50 = 0.63 nM) among several MDR1 inhibitors, including cycloporin A, XR9576, and GF120918, and effectively blocked transepithelial transport of paclitaxel in MDCK monolayers (IC50 = 35.4 nM). The ATPase inhibitory activity of HM30181 was highly selective to MDR1. HM30181 did not inhibit MRP1 (ABCC1), MRP2 (ABCC2), and MRP3 (ABCC3), and partially inhibited BCRP (ABCG2) only at very high concentrations. Importantly, co-administration of HM30181 (10 mg/kg) greatly increased oral bioavailability of paclitaxel from 3.4% to 41.3% in rats. Moreover, oral co-administration of paclitaxel and HM30181 showed a tumor- inhibitory strength equal or superior to that of intravenous paclitaxel in the xenograft model in nude mice. These results identify HM30181 as a highly selective and potent inhibitor of MDR1, which in combination with paclitaxel, may provide an orally effective anti-tumor regimen. 

1. Introduction

The multi-drug resistance 1 (MDR1, ABCB1) protein, also known as P-glycoprotein (P-gp), belongs to the ATP-binding cassette (ABC) transporter superfamily and functions as an organic transporter (Higgins, 1992). The ABC proteins regulate essential processes in the alimentary tract, liver, kidney and immune system, and protect organisms from a wide range of toxic compounds (Dietrich et al., 2003; El-Sheikh et al., 2008; Hung et al., 1998). Among them, MDR1 and some of the multi-drug resistance proteins (MRPs) function as xenobiotic export pumps. MDR1 was originally identified as a gene that confers multi-drug resistance on cancer cells, and the MRPs, as a second type of drug pump, inducing resistance independently of MDR1. However, these transporters not only mediate drug resistance in cancer cells, but also perform the native physiological role of protecting an organism from toxic substances (Deeley et al., 2006; Dietrich et al., 2003).

The MDR1 gene product, a protein of 170 kDa, exports hydrophobic or amphiphilic xenobiotic substances, including the vinca alkaloids, colchicine, daunorubicin, and taxanes, by ATP-dependent mechanism (Shirasaka et al., 2006). The overexpression of MDR1 on the surface of tumor cells is one of the most common causes of multi-drug resistance, which occurs through efflux of drugs from the cancer cells (Ambudkar et al., 1999; Gottesman et al., 2002). MDR1 is also highly expressed in the gut, liver and kidneys, organs that control absorption, distribution and excretion of drugs (Dietrich et al., 2003; Sarkadi et al., 1992). Especially, MDR1 expressed in the lining epithelium of gastrointestinal tract forms a major chemical barrier against xenobiotic agents including commonly prescribed anti-cancer chemotherapeutic agents.

Paclitaxel, a taxane derivative, is used widely in chemotherapy for cancers of the breast, lung, head and neck, and ovaries (Spratlin and Sawyer, 2007). Either alone or in combination with other drugs (Baker et al., 2002; Gill et al., 1999), paclitaxel elicits a robust and significant tumor response rate. However, in multi-drug resistant cancer cells, it is difficult to maintain the effective target concentration of paclitaxel, because MDR1 efficiently pumps it out of the cancer cell (Takahashi et al., 2006). In addition, as a typical substrate of MDR1, paclitaxel requires parenteral administration because of its poor oral bioavailability.

During last decades, active search for a potent MDR1 modulator has been performed to increase the efficacy of anti-cancer chemotherapy by reversing the multi-drug resistance phenotype. A continual search for potency and selectivity produced the third generation MDR1 inhibitors, including GF120918 (elacridar) and XR9576 (tariquidar), which may have clinically acceptable pharmacologic properties (Bardelmeijer et al., 2004; Kruijtzer et al., 2002; Kühnle et al., 2009; Walker et al., 2004). In the present study, we compared the action of HM30181, a recently introduced MDR1 inhibitor (Paek et al., 2007), on MDR1 and MRPs with the action of other MDR1 inhibitors, using in vitro and in vivo assay systems. We also tested the effect of HM30181 on the oral bioavailability and therapeutic effectiveness of paclitaxel. Our results show that HM30181 inhibits MDR1 potently and selectively, and that in combination with paclitaxel, may provide an orally active anti-tumor regimen.

2. Materials and methods

2.1. Chemicals and animals

HM30181, GF12098, XR9676 and cyclosporin A were synthesized by the Research Center, Hanmi Pharm. Co. (Hwaseong, Korea). Chemical structure of HM30181 is presented in Supplementary Fig. 1. Paclitaxel was purchased from Sigma (St. Louis, MO) and [3H] paclitaxel was from Moravek Biochemicals Inc. (Brea, CA). Sulfasala- zine, N-ethylmaleimide glutathione (NEM-GS), glutathione (GSH), bebzmarone, verapamil and Hoechst 33342 were included in the ATPase assay kits from Solvo Biothechnology (Budapest, Hungary). All the other chemicals used in the experiments were of the highest quality available from Sigma. Male Sprague–Dawley rats and athymic nude mice were purchased from Charles River Korea Co. (Gyeonggi- do, Korea). This study was approved by the Committees for the Care and Use of Laboratory Animals, Yonsei University College of Medicine (Seoul, Korea) and the Hanmi Research Center (Hwaseong, Korea), and the animal studies were carried out in accordance with the Declaration of Helsinki.

2.2. ATPase assay

The inhibitory activities of HM30181 and other MDR1 and MRP inhibitors were measured using the ATPase assay kit (Solvo) according to the manufacturer’s instructions. Briefly, the purified membrane vesicles were diluted to 0.1 μg/μL with assay mix (50 mM Mops–Tris, pH 7.0; 50 mM KCl; 5 mM Na-azide; 2 mM DTT; 0.1 mM EGTA–Tris, pH 7.0; 1 mM ouabain in distilled water) and a 40-μL volume of diluted membrane suspension was loaded into each well of a 96-well microplate. One μL of the tested compound dissolved in dimethylsulfoxide (DMSO) with or without Na-orthovanadate (600 mM) was added to the membrane suspension. The same volume of DMSO was added to the control wells. The mixtures were pre- incubated at 37 °C for 10 min and the reaction was started by addition of 10 μL Mg-ATP (200 mM). After a 10-min incubation at 37 °C, the inorganic phosphate (Pi) released was determined colorimetrically. The absorbance at 600 nm was read in a microplate reader, and the concentration of liberated Pi was calculated from the calibration curve (Sarkadi et al., 1992). For the inhibition experiment, ATPase assays were performed in the presence of each activator substance (0.05 μM paclitaxel for MDR1, 10 μM sulfasalazine for breast cancer resistance protein [BCRP, ABCG2], NEM-GS 3 mM with GSH 2 mM for MRP1 [ABCC1], sulfasalazine 150 μM with GSH 2 mM for MRP2 [ABCC2], and bebzmarone 50 μM with GSH 2 mM for MRP3 [ABCC3]). The IC50 value was calculated using the Prism 4 software package (http://www. graphpad.com/).

2.3. Generation of 293FRT-MDR1 cells, and the rhodamine transport assay

HEK293 (human embryonic kidney) cells were maintained in DMEM with 10% fetal bovine serum at 37 °C. HEK293 cells that stably express MDR1 (293FRT-MDR1) were generated using the Flp-In system (Invitrogen; Carlsbad, CA, USA). This system creates isogenic cell lines with a fixed number of Flp recombination target (FRT) sites. The HEK293-FRT cells that stably express the single FRT site were purchased from Invitrogen and maintained in a zeocin (100 μg/mL)- containing medium. The 293FRT-MDR1 cells were generated by co- transfecting pcDNA5/FRT-MDR1 and pOG44 (Flp recombinase ex- pression plasmid from Invitrogen) using Lipofectamine 2000 into the HEK293-FRT cells. The pcDNA5/FRT-MDR1 plasmid was constructed from pcDNA3.1-MDR1 using the Nhe I and Xho I restriction sites. The pcDNA3.1-MDR1 plasmid had been previously constructed using a PCR-amplified full-length human MDR1 cDNA and the EcoR I and Xho I restriction sites. After transfection, the 293FRT-MDR1 cells were selected and maintained in a hygromycin B (50 μg/mL)-containing medium.

Cell-based rhodamine transport assays in the presence or absence of MDR1 inhibitors were performed in 293FRT-MDR1 cells. The 293FRT-MDR1 and 293FRT-Mock cells cultured in a 60-mm culture dish were washed with 1× phosphate-buffered saline (PBS) and dissociated with 0.05% Trypsin-EDTA buffer. Cells (5 × 105 cells) were resuspended in 1× PBS and incubated with rhodamine123 (5 μM) for 80 min at 37 °C. After washing out extracellular rhodamine123 by centrifugation (400 g for 5 min), the cells were treated with graded concentrations (10−11 – 10−6 M) of MDR1 inhibitors and incubated for 5 min at 37 °C to allow rhodamine 123 efflux. The residual intracellular rhodamine123 was measured with a fluorescence- activated cell sorter (FACS) system (FACSCalibur equipped with the CellQuest software package; BD Biosciences, Mansfield, MA).

2.4. Transcellular transport assay

Madin–Darby canine kidney (MDCK) cells were maintained in DMEM with 10% fetal bovine serum at 37 °C. MDCK cells that stably express MDR1 (MDCK–FRT–MDR1) were generated using the Flp-In system as described above. First, MDCK-FRT cells that stably express the FRT site were generated by transfection with pFRT/lacZeo using Lipofectamine 2000 (Invitrogen) and zeocin selection (100 μg/mL). The MDCK–FRT–MDR1 cells were then generated by co-transfection with MDR1-pcDNA5/FRT and pOG44 plasmids, followed by hygromycin B selection (50 μg/mL).

The [3H]paclitaxel transport assay was performed as described previously, with a minor modification (Evers et al., 1998; Taub et al., 2005). Briefly, the MDCK–FRT–MDR1 cells were seeded onto polycarbonate cell culture membranes (Transwell 3414, 3 μm pore size, 1.33 mm diameter; Costar, Cambridge, MA) at a density of 5 × 104 cells/well and cultured for 5–6 days with daily medium replacement. Integrity of the confluent monolayers was determined from measure- ments of transepithelial electrical resistance (TEER) taken with an epithelial voltohmmeter (Millicell-ERS; Millipore, Billerica, MA). When TEER values exceeded 200 ohm×cm2, MDCK monolayers were cultured in a medium containing sodium butyrate (10 mM) for 24 h and used for experiments. For transport assays, cells were washed three times and both sides of monolayers were pre-incubated at 37 °C with the Krebs–Henseleit buffer (KHB; 142 mM NaCl,
23.8 mM Na2CO3, 4.83 mM KCl, 0.96 mM KH2PO4, 1.20 mM MgSO4,
12.5 mM HEPES, 5 mM glucose, and 1.53 mM CaCl2, pH 7.3). The assay was initiated by the replacement of buffer at either apical (200 μL) or basal side (800 μL) with KHB containing [3H]paclitaxel (1 μM) and graded concentrations (10−11 – 10−6 M) of HM30181A. At 0, 30, 60, 90, 120, 150 and 180 min, 100-μL samples of buffer were removed from the receiver compartment and replaced with preheated buffer (100 μL). Radioactivity of the medium was measured in a liquid scintillation counter (LS 6000SE, Beckman Instruments, Inc., Fullerton, CA) after addition of 2 mL of scintillation cocktail. The cumulative amount of [3H]paclitaxel was plotted and analyzed using the Prism 4 software package.

2.5. Measurements of paclitaxel concentration in rat plasma

Male Sprague–Dawley rats aged 8 weeks were housed under standard husbandry conditions (22 ± 2 °C, 12-h light and 12-h dark cycle, and 50–60% relative humidity). Rat plasma samples were collected after intravenous (i.v.) or oral (p.o.) administration of paclitaxel with or without HM30181. Plasma paclitaxel concentra- tions were measured using liquid chromatography with tandem mass spectrometry (LC-MS/MS) as previously reported (Paek et al., 2006). Briefly, a 100-μL volume of each blank, calibration standard, and sample plasma was dispensed into a 2-mL eppendorf tube and mixed with the internal standard (HM30059, 50 μL) and extraction reagent (methyl-tert-butyl ether containing 0.01% perchloric acid, 1 mL). After extraction, the organic layer was reconstituted in the mobile phase (50 μL) and a 10-μL volume was injected into the LC-MS/MS system. This system included a Waters 2795HT pump and autosampler (Waters, Milford, MA). The separation was performed on an X-terra RP18 column (Waters) using a mixture of 0.1% trifluoroacetic acid (pH 3.6) in 80% methanol at a flow rate of 0.2 mL/min. The eluent was introduced directly into the ElectroSpray Ionization source of a tandem quadrupole mass spectrometer (Quattro Micro; Waters). The cone energy of paclitaxel was 26 V and the protonated molecular ions of paclitaxel were fragmented at a collision energy of 18 V by collision- activated dissociation. Detection of paclitaxel was monitored by the ion transition of m/z 854.16 → 286.12. Peak areas for all components were automatically integrated using the Masslynx software package (version 4.0, Waters).

2.6. Measurements of paclitaxel anti-tumor activity

Measurements of the in vivo anti-tumor efficacy of i.v. paclitaxel versus oral combination of paclitaxel and HM30181 were performed using the human colon cancer cell (HT-29) xenograft model (Radulovic et al., 1991). Athymic male nude (nu/nu) mice aged 6 weeks were maintained under pathogen-limited conditions. The HT-29 human colonic adenocarcinoma cell line was grown in RPMI 1640 (Gibco, Grand Island, NY) supplemented with 10% colostrum-free bovine serum at 37 °C in a humidified 5% CO2 atmosphere. Tumor cells growing exponentially were harvested by brief incubation with 0.25% trypsin- EDTA solution (Gibco). Xenografts were initiated in 5 nude mice by subcutaneous (s.c.) injection of 1 × 107cells into the right flank. The tumors resulting after 3 weeks were aseptically dissected and mechan- ically minced, and 10-mm3 pieces were transplanted s.c into 28 animals using a trocar needle. Fifteen days after transplantation, tumors had grown to a volume of approximately 450 mm3 and the animals were divided into four experimental groups: 1) control, 2) paclitaxel 20 mg/ kg, i.v., 3) paclitaxel 20 mg/kg and HM30181 10 mg/kg, p.o., and 4) paclitaxel 40 mg/kg and HM30181 20 mg/kg, p.o. Drugs were admin- istered on days 15, 19, 23, and 27. The tumor volume was calculated as (length×width×height×π)/6.

3. Results

3.1. Inhibition of MDR1-mediated transport by HM30181

Inhibition of MDR1-mediated xenobiotic transport by HM30181 was first determined with the ATPase assay using MDR1-enriched membrane vesicles. The ATPase activity of MDR1 was triggered with 0.05 μM paclitaxel, a concentration 5-fold higher than the EC50 (Supplementary Fig. 2). Dose–response curves for MDR1 inhibition and IC50 values (Fig. 1A and B) showed the very potent inhibitory effect of HM30181 on MDR1-mediated paclitaxel transport in membrane vesicles. The IC50 value of HM30181 (IC50 = 0.63 nM) was approximately 220-fold, 50-fold, and 7.7-fold lower than those of cyclosporine A, XR9576, and GF120918, respectively.

We also tested the inhibitory effects of MDR1 inhibitors in intact cells by measuring rhodamine123 transport with the FACS system (Fig. 1C and D). To decrease the variability between cells in MDR1 expression, 293FRT-MDR1 cells that stably express a single copy of human MDR1 cDNA with the cytomegalovirus promoter were generated as described in Materials and methods. The rhodamine 123 transport assay again showed HM30181 to be a potent inhibitor of MDR1. However, in this cell-based assay the relative potencies of the four inhibitors differed from those obtained with the ATPase assay using membrane vesicles. The IC50 value of HM30181 (IC50= 11.5 nM) was 23.8-fold and 4.6-fold lower than those of cyclosporine A and XR9576, respectively, but 3.1- fold higher than that of GF120918 (Fig. 1D). In addition, this value was 18.3-fold higher than that in the ATPase assay.

Next, the effect of HM30181 on the MDR1-mediated paclitaxel transport was analyzed in MDR1-overexpressed MDCK monolayers. As shown in Fig. 2, the basal-to-apical paclitaxel transport (B to A, 5996 ± 324 pmol/mg protein at 0 nM HM30181) was 17-fold higher than that of apical-to-basal transport (A to B, 353 ± 14 pmol/mg protein at 0 nM HM30181) due to the apical localization of MDR1 in the polarized MDCK monolayers (Evers et al., 1998; Taub et al., 2005). Consistent with its direct effect on MDR1, HM30181 strongly inhibited the basal-to-apical transport of paclitaxel in MDCK mono- layers with an IC50 value of 35.4 nM.

3.2. Effects of HM30181 on other ABC transporters

The specificities of MDR1 inhibitors were determined using membrane vesicles enriched in other ABC transporters, including BCRP, MRP1, MRP2, and MRP3. The ATPase activity of each transporter was initiated with its corresponding substrate (see Materials and methods). In BCRP and MRP2 assays, Hoechst 33342 and benzmarone, respectively, were used as control inhibitory agents. ATPase assays of xenobiotic transporters revealed that HM30181 selectively inhibits MDR1. HM30181 did not inhibit MRP1, MRP2, and MRP3 (Fig. 3) and partially inhibited BCRP only at very high concentrations with an IC50 value greater than 3.7 μM. This value was 20-fold and 15-fold higher than those of XR9576 and GF120918, respectively (Table 1).

3.3. Effects of HM30181 on the oral bioavailability of paclitaxel in rats

Previously, a preliminary result that HM30181 could increase the oral bioavailability of paclitaxel has been reported when developing the assay system of HM30181 (Paek et al., 2006). In the present study, the effects of HM30181 on the paclitaxel oral bioavailability were more thoroughly investigated using appropriate i.v. controls and increased dose of HM30181. The median lethal dose (LD50) of i.v. paclitaxel in male Sprague–Dawley rats is 8.3 mg/kg (95% confidence interval, 7.11–9.58 mg/kg) (Kim et al., 2001). Therefore, i.v. 6 mg/kg paclitaxel was chosen for a standard single dose of pharmacokinetic evaluation. In a preliminary experiment with a small number of animals, it was found that 20 mg/kg p.o. paclitaxel+ 10 mg/kg p.o. HM30181 showed comparable pharmacokinetics to that of i.v. 6 mg/ kg paclitaxel. Therefore, these doses were selected for the standard pharmacokinetic study with increased number of animals. Pharma- cokinetic parameters of paclitaxel were determined in male Sprague– Dawley rats after both intravenous and oral administration of the drug, with or without HM30181, to reveal the effects of the inhibitor on oral bioavailability. The average plasma concentration–time curves for paclitaxel are shown in Fig. 4 and the pharmacokinetic parameters are summarized in Table 2. The area under the plasma concentration– time curve (AUC) and apparent elimination half-life (t1/2) were based on a non-compartmental analysis (WinNonlin; Pharsight Corp, Moun- tain View, CA), and the peak concentration (Cmax) and the time to peak concentration (Tmax) were determined by visual inspection from the experimental data.

Fig. 1. Inhibitory effects of HM30181 and other modulators on MDR1. (A,B) The inhibitory activities of MDR1 modulators were measured using the ATPase assay as described in Materials and methods. The assay began with the activation of the MDR1 ATPase with 0.05-μM paclitaxel, a concentration 5-fold higher than the EC50. Dose–response curves and the IC50 values for each compound are presented in panels A and B, respectively. (C,D) Inhibitory effects of MDR1 inhibitors were also tested in intact cells by measuring rhodamine123 transport in 293FRT-MDR1 cells with the FACS system. Dose–response curves and the IC50 values for each compound are presented in panels C and D, respectively (each trial n = 4). CI: confidence interval.

Treatment with HM30181 significantly increased the oral bio- availability of paclitaxel. The AUC for control i.v. paclitaxel (6 mg/kg) was 2728.1 ± 281.7 ng h/mL. The AUC value for oral paclitaxel-alone (20 mg/kg) was 308.5 ± 160.6 ng h/mL, indicating that only 3.4% of orally administered paclitaxel was absorbed. Co-administration of HM30181 (10 mg/kg) increased the AUC of oral paclitaxel (20 mg/kg)

to 3756.5 ± 865.9 ng h/mL, which corresponds to an oral bioavail- ability of 41.3%. In addition, the Cmax value increased from 127.2 ± 72.9 ng/mL to 1253.7 ± 269.8 with HM30181, while Tmax and t1/2 did not change significantly.

Fig. 2. Inhibition of the MDR1-mediated transcellular paclitaxel transport by HM30181. Transcellular transport of paclitaxel in MDCK cells that overexpress MDR1 was measured at graded concentrations of HM30181. The basal-to-apical (B to A) paclitaxel transport was 17-fold higher than the apical-to-basal (A to B) transport due to the apical orientation of MDR1 in the polarized MDCK monolayers. HM30181 exerted potent inhibition on the B to A paclitaxel transport (IC50 = 35.4 nM, each trial n =3) and no effects on the A to B transport.

3.4. Effects of oral co-administration of HM30181 and paclitaxel on tumor growth in nude mice

In vivo anti-tumor efficacy of the oral combination of paclitaxel and HM30181 was measured using the human colon cancer cell (HT-29) xenograft model (Radulovic et al., 1991) (Fig. 5). Intravenous paclitaxel was used as a reference treatment and drugs were administered on days 15, 19, 23, and 27 after establishing the tumor xenograft. The maximum tolerable dose (MTD) of i.v. paclitaxel in nude mice is 20 mg/kg (Kim et al., 2001). Therefore, this dose was chosen for the evaluation of anti-tumor efficacy of i.v. and p.o paclitaxel. Because oral bioavailability of paclitaxel is estimated to be around 40% when HM30181 was co- administered (Fig. 4), p.o. 40 mg/kg paclitaxel was also evaluated. Mice tolerated the i.v. injection with 20 mg/kg paclitaxel, with an average maximum weight loss of 4.34%, and the oral administration of 20 and 40 mg/kg paclitaxel in the combination regimens, with average maxi- mum weight losses of 0% and 4.87%, respectively. Against the HT-29 tumor cell line in nude mice, the combination regimens containing 20 and 40 mg/kg oral paclitaxel inhibited tumor growth by 74.0% and 94.2%, respectively. These values compared favorably with, or exceeded, the effectiveness of i.v. paclitaxel injection (20 mg/kg), which inhibited tumor growth by 76.5 %. Furthermore, the oral combination regimen containing 40 mg/kg paclitaxel and 20 mg/kg HM30181 induced a remission of tumor growth until day 47 (Table 3).

4. Discussion

Development of an orally effective regimen offers many benefits for anti-tumor chemotherapy. Convenience for the patient, improved efficiency of its use in health care facilities (time, personnel, and equipment), and drug economy all argue in favor of oral therapy (DeMario and Ratain, 1998; Liu et al., 1997). In addition, the oral route facilitates the use of more chronic treatment regimens, which result in prolonged exposure to the drug (Malingré et al., 2001). A selective and potent MDR1 inhibitor will pave the way to oral chemotherapy for drugs such as the taxanes (paclitaxel and docetaxel), which activate MDR1 to excrete them from the intestinal epithelia. This phenomenon greatly restricts the oral bioavailability of taxanes and other MDR1 substrates (Sparreboom et al., 1997). In fact, MDR1 inhibition with cyclosprin A had been shown to increase the oral bioavailability of docetaxel in human (Malingré et al., 2001). However, the intrinsic effect of cyclosporine A on its primary target, immunosuppression by the absorbed cyclosporine A, hampered its wide use in the clinical setting. This led many researchers to look for a highly potent inhibitor of MDR1 that has a minimal systemic effect. In a recent study, it has been shown that GF120918 showed the most potent effect in increasing the oral bioavailability of paclitaxel in mice among several newer generation MDR1 inhibitors, including PSC833, LY335979, and R101933 (Bardelmeijer et al., 2004).

The present study demonstrates that HM30181 has a potent and
selective inhibitory effect on MDR1. In the ATPase assay using MDR1- enriched vesicles, HM30181 showed the highest potency of several MDR1 inhibitors, including cyclosporin A, XR9576, and GF120918. In addition, HM30181 effectively blocked transepithelial transport of paclitaxel in MDCK monolayers. In the ATPase assay, HM30181 showed high specificity for MDR1, did not inhibit MRP1, MRP2, or MRP3, and very weakly inhibited BCRP. Importantly, MDR1 inhibition with HM30181 greatly increased the oral bioavailability of paclitaxel. HM30181 has a low water solubility (1 g in 4556 L, at pH 3.8), and consequently a low oral bioavailability (6.3 % in rats; AUC0 − 48h with 3 mg/kg i.v., 618.7 ± 43.7 ng h/mL; AUC0 − 48h with 10 mg/kg p.o.,
128.9 ± 22.4 ng h/mL). This property would be advantageous to the inhibition of MDR1 in the gastrointestinal tract. Because low systemic absorption will minimize the possible adverse reactions of HM30181, a high dose of HM30181 can be used to fully inhibit MDR1 in the gastrointestinal tract. For this reason, HM30181 was developed primarily as an oral adjuvant, to increase the oral bioavailability of MDR1 substrates, rather than an MDR antagonist at the level of the target cancer cell. An interesting and puzzling observation in this study is an 18.3-fold difference in IC50 values for HM30181 in the ATPase assay, using membrane vesicles (IC50 = 0.63 nM), and in the cell-based assay, using rhodamine 123 efflux (IC50 = 11.5 nM). This did not result simply from the difference in the assay systems, because both assays produced similar IC50 values for GF120918 (Fig. 1). A plausible explanation for the difference in the HM30181 IC50 values is that HM30181 may bind to the cytosolic face of MDR1. The ATPase assay offers the test drug easy access to the cytosolic face of the MDR1-enriched membrane vesicles, while in the rhodamine efflux assay, the relatively insoluble HM30181 must penetrate the cell membrane to inhibit MDR1 at the cytosolic face. Further studies of the HM30181 binding site and inhibitory mechanisms, to clarify the above question, are currently in progress.

Among the newer MDR1 inhibitors tested in this study, HM30181 showed the highest selectivity for MDR1, while XR9576 and GF120918 inhibited BCRP-mediated transport by 50% at 180–250 nanomolar concentrations. On the other hand, more than 3.7 μM of HM30181 was needed to achieve similar efficacy (Fig. 3A). At micromolar concentrations, XR9576 and  also exerted a weak inhibitory effect on MRP1, which HM30181 did not. In addition to the low systemic absorption, the high selectivity of HM30181 can be advantageous to prevent untoward effects in clinical applications. On the other hand, this suggests that HM30181 would be a narrow spectrum agent effective only in increasing the oral bioavailability of pure MDR1 substrates. In this regard, it will be interesting to evaluate the effect of HM30181 on the oral bioavailability of agents that can be transported by multiple transporters, such as irinotecan, a substrate for both MDR1 and BCRP (Oostendorp et al., 2009; Su et al., 2006).

In the xenograft model in nude mice, oral co-administration of paclitaxel and HM30181 effectively inhibited tumor growth (Fig. 5). The combination of 20 mg/kg paclitaxel and 10 mg/kg HM30181 given orally suppressed growth by approximately 75%, which is comparable to the result of 20 mg/kg paclitaxel injected i.v. Interestingly, at the dose of 20 mg/kg paclitaxel, oral administration did not induce a weight loss, while i.v. injection did (4.3%), which suggests that oral administration is more tolerable. Importantly, the combination of 40 mg/kg paclitaxel and 20 mg/kg HM30181 given orally inhibited tumor growth by 94% and induced a remission of tumor growth until day 47, a result superior to that of 20 mg/kg paclitaxel by intravenous injection. The most plausible explanation about this finding is that orally absorbed paclitaxel may produce a safer pharmacokinetic such as avoiding an extremely high peak plasma concentration, and hence is more suitable for increasing the dose and efficacy of anti-tumor chemotherapy. The MTD of i.v. paclitaxel in nude mice is 20 mg/kg (Kim et al., 2001). In the present study, the dose with twice the amount of MTD (40 mg/kg paclitaxel) was tolerable when administered orally. The oral administration of 40 mg/kg paclitaxel induced a 4.9% weight loss, similar to that of i.v. 20 mg/kg paclitaxel. Another possibility of higher effect when paclitaxel was administered with HM30181 is that systemically absorbed HM30181 may further increase paclitaxel effectiveness by preventing drug efflux from the target cells, although oral bioavailability of HM30181 is low. In fact, it has been shown recently that oral administration with 16 mg/kg HM30181 in C3H mice improved the brain penetration of intraperitoneally administered paclitaxel in an independent study (Joo et al., 2008).

In conclusion, our results suggest that HM30181 is a potent and selective inhibitor of MDR1, and that HM30181 and paclitaxel in combination can provide orally effective chemotherapy for clinical application. We are awaiting human trials that will validate the effectiveness of HM30181 on the oral bioavailability and therapeutic efficacy of paclitaxel in clinical settings.