Introduction

Top Abstract Introduction Materials and methods Results Discussion References

Multidrug resistance, one of the major obstacles encountered in cancer chemotherapy, is often acquired by overexpression of ATP-binding cassette transmembrane transporters such as P-glycoprotein (P-gp) and multidrug resistance-associated protein (MRP) (Gottesman et al., 1996; Lautier et al., 1996; Loe et al., 1996). Although the overexpression of P-gp and MRP provides almost the same spectrum of resistance to antitumor drugs, the substrate specificity of these transporters determined in isolated plasma membrane vesicles differ markedly. In general, P-gp accepts amphipathic cationic or neutral compounds as a substrate, whereas MRP acts as a glutathione S-conjugate export pump (GS-X pump) (Ishikawa and Akimura, 1996; Lautier et al., 1996; Loe et al., 1996; Müller et al., 1996, 1997; Keppler and König, 1997). In addition, MRP can extrude several organic anions including glucuronide and sulfate conjugates (Loe et al., 1996; Lautier et al., 1996; Keppler et al., 1997). Cumulative evidence indicates the presence of GS-X pump/MRP family members (Kool et al., 1997); for example, the recently cloned canalicular multispecific organic anion transporter (cMOAT), which is predominantly expressed on the canalicular membrane of hepatocytes under physiological conditions, exhibits GS-X pump activity (Oude Elferink et al., 1995; Yamazaki et al., 1996; Meier, 1996; Müller et al., 1996; 1997; Keppler et al., 1997; Paulusma and Elferink, 1997). By comparing the transport properties across the bile canalicular membrane between normal and mutant rats (e.g., TR and Eisai hyperbilirubinemic rats) whose cMOAT function is hereditarily defective, it has been established that the substrate specificity of cMOAT is similar to that of MRP in that cMOAT accepts the previously described conjugated metabolites as substrates (Oude Elferink et al., 1995; Yamazaki et al., 1996; Meier, 1996; Müller et al., 1996, 1997; Keppler and König, 1997; Paulusma and Oude Elferink, 1997). This observation has been further confirmed by examining the function of cloned cMOAT (Büchler et al., 1996; Paulusma et al., 1996; Taniguchi et al., 1996; Ito et al., 1997; 1998; Borst et al., 1997; Madon et al., 1997; Evers et al., 1998; Koike et al., 1997; van Aubel et al., 1998). Moreover, we found that several antitumor drugs are extruded via cMOAT, including methotrexate, irinotecan [7-ethyl-10-[4-(1-piperidino)-1-pipertidino)-1-piperidino] carbonyloxy camptothecin) (CPT-11)] and its metabolites (Chu et al., 1997a,b; Masuda et al., 1997).

CPT-11 is a derivative of camptothecin (CPT) with substantial anticancer activity in a broad spectrum of human tumor cells (Slichenmyer et al., 1993). CPT-11 is a prodrug that can be converted by carboxyl esterase to its active metabolite, SN-38, which is able to inhibit topoisomerase I (Slichenmyer et al., 1993). The -hydroxy--lactone ring in CPT-11 and SN-38 is in equilibrium with its carboxylate form, and the equilibrium reaction favors the production of the carboxylate form at physiological pH and the lactone form at pH values below physiological (Fassberg and Stella, 1992). We found that cMOAT is responsible for the biliary excretion of four CPT-11-related compounds with anionic charges (the carboxylate forms of CPT-11 and SN-38 along with the lactone and carboxylate forms of SN38-Glu) in rats (Chu et al., 1997a,b). Due to the similar substrate specificity of MRP and cMOAT, it is possible that CPT-11 and its metabolites can be substrates for MRP and/or other GS-X pump family members. Therefore, the overexpression of such pumps may confer on the tumor cells resistance to these drugs. In addition, several reports have been published in which overlapping substrate specificity between P-gp and MRP is described (Gottesman et al., 1996). In the present study, we investigated these hypotheses by examining the uptake of CPT-11-related compounds into membrane vesicles isolated from human epidermoid carcinoma KB-3-1 derived cell lines; these include KB-C2 and C-A500, which overexpress P-gp (Akiyama et al., 1985) and MRP (Sumizawa et al., 1994), respectively, and KCP-4, a cisplatin-resistant cell line with a higher transport activity of the unidentified GS-X pump (Fujii et al., 1994; Chuman et al., 1996). In addition, the results of the transport studies were compared with the resistance of these cell lines to CPT-11 and SN-38 to examine the role of these transporters in the acquisition of resistance.

	Materials and Methods

Top Abstract Introduction Materials and methods Results Discussion References

Materials. CPT-11, SN-38, and SN38-Glu were kindly provided by Daiichi Pharmaceutical Co. Ltd. (Tokyo, Japan) and Yakult Honsha Co. Ltd. (Tokyo, Japan). The lactone and carboxylate forms of CPT-11, SN-38, and SN38-Glu were produced by dissolving the compounds in 50 mM phosphate buffer at pH 3.0 or 9.0, respectively, and leaving the solutions overnight. Under these experimental conditions, the conversion of lactones into carboxylates or carboxylates into lactones is virtually complete (>99%) as determined by the high-performance liquid chromatography (HPLC) method described below. CPT, used as an internal standard, was obtained from Sigma (St. Louis, MO). ATP, AMP, creatine phosphate, and creatine phosphokinase were purchased from Sigma. All other chemicals were commercial products and of analytical grade.

Cells. Human epidermoid carcinoma KB-3-1 cells, multidrug resistant KB-C2, which overexpress P-gp (Akiyama et al., 1985) and C-A500, which overexpress MRP (Sumizawa et al., 1994), as well as KCP-4, a cisplatin-resistant cell line (Chuman et al., 1996), were used in this study. KB-3-1 cells were cultured in minimal essential medium, and supplemented with 10% fetal bovine serum. The other KB-3-1-derived cells were selected from KB-3-1 cells and cultured in selection medium containing colchicine (2 µg/ml) for KB-C2 (Akiyama et al., 1985); cepharanthine (1 µg/ml), doxorubicin (0.5 µg/ml), and mezerein (0.065 µg/ml) for C-A500 (Sumizawa et al., 1994), and cisplatin (7 µg/ml) for KCP-4 (Fujii et al., 1994; Chuman et al., 1996).

Isolation of Membrane Vesicles. Membrane vesicles were prepared from above KB-derived cells by the nitrogen cavitation method (Ito et al., 1998). Cell monolayers (10⁷-10⁸ cells) were washed once and scrapped into phosphate-buffered saline. The cells were washed by centrifugation (4,000g, 10 min) at 4°C in phosphate-buffered saline and then in buffer A consisting of 250 mM sucrose, 0.2 mM CaCl₂, and 10 mM Tris/HCl (pH 7.4) and equilibrated at 4°C under a nitrogen pressure of 900 psi for 15 min. EDTA was added to the cell lysate to give a final concentration of 1 mM. The lysed cell suspension was then diluted with 4 volumes of buffer B containing 250 mM sucrose and 10 mM Tris/HCl (pH 7.4) and centrifuged (1,000g, 10 min) at 4°C. The supernatant was layered onto 35% sucrose cushion containing 10 mM Tris/HCl and 1 mM EDTA (pH 7.4) and centrifuged for 30 min at 16,000g at 4°C. The interface was collected, diluted 5-fold with buffer B, and then centrifuged for 45 min at 100,000g. The vesicle pellet was resuspended in buffer B using a 25-gauge needle. Vesicles were stored at 100°C until use. The enrichment of the isolated membrane vesicles assessed by alkaline phosphatase activity was 4.44 ± 0.26, 5.55 ± 0.24, 5.12 ± 0.35, and 5.24 ± 0.40 (mean ± S.E. of two preparations determined in duplicate) for KB-3-1, KB-C2, C-A500, and KCP-4, respectively (K Ueda, H Suzuki, S Akiyama, and Y Sugiyama, submitted). The sidedness of the membrane vesicles was also determined by monitoring this enzyme activity in the presence and absence of Triton X-100 (0.2%). The fraction of inside-out membrane vesicles was 42.9 ± 2.3, 39.2 ± 3.3, 37.4 ± 5.9, and 43.9 ± 4.1% (mean ± S.E. of two preparations determined in duplicate) for KB-3-1, KB-C2, C-A500, and KCP-4, respectively (K Ueda, H Suzuki, S Akiyama, and Y Sugiyama, submitted).

Uptake by Membrane Vesicles. The uptake of the carboxylate and lactone forms of CPT-11 and its metabolites was studied as described previously (Chu et al., 1997a,b). The uptake study was performed at 37°C in medium (20 µl) containing 250 mM sucrose, 10 mM Tris/HCl (pH 7.4), 10 mM MgCl₂, 5 mM ATP, 10 mM creatine phosphate, and 100 µg/ml creatine phosphokinase. For the control experiments, ATP was replaced by AMP. The final concentration of membrane vesicles was adjusted to 0.25 mg/ml. In the present study, we examined the uptake of ligands into membrane vesicles at 2 min because our previous studies have suggested that the uptake of CPT-11 and its metabolites (anionic form) was significantly stimulated by ATP and shows linearity for at least 2 min in canalicular membrane vesicles (CMVs) from both humans and rats (Chu et al., 1997a, 1998). Transport was terminated by adding 1 ml ice-cold stop solution followed by immediate filtration through a 0.45-µm filter (HAWP 02500; Millipore Corporation, Bedford, MA), and subsequently washing twice with 5-ml ice-cold stop solution. The stop solution consisted of 10 mM Tris/HCl (pH 7.4), 250 mM sucrose, and 100 mM NaCl. ATP-dependent uptake was determined as the difference in uptake in the presence and absence of ATP.

Resistance to CPT-11 and SN-38. The resistance of KB-3-1-derived cells to CPT-11 and SN-38 was determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) colorimetric assay performed in 96-well plates (Carmichael et al., 1987). Cells (2 × 10³ for KB-3-1 and KB-C-2 and 5 × 10³ for C-A500 and KCP-4) were seeded to each well and cultured with 50 µl of phenol red-free culture medium. After overnight incubation at 37°C in an atmosphere of 5% CO₂, CPT-11 and SN-38 were added to the cells and incubated for 4 days. Then, MTT was added to each well and incubated for 4 h. The resulting formazan was dissolved in 100 µl dimethyl sulfoxide. Plates were placed on a plate shaker for 5 min and the absorbance at 595 nm was immediately read using a microplate reader (model 3550; Bio-Rad, Hercules, CA).

Data Analysis. The Eadie-Hofstee plot was used to express the concentration dependence of ATP-dependent uptake. ATP-dependent uptake was obtained by subtracting the uptake in the presence of AMP from that in the presence of ATP. Kinetic parameters for the ATP-dependent uptake were obtained by fitting using the following equation:

(1)

where V_o is the initial uptake rate of substrate (pmol/min/mg protein), S is the substrate concentration in the medium (µM), K_m is the Michaelis constant (µM), and V_max is the maximum uptake rate (pmol/min/mg protein). The uptake data were fitted to eq. 1 by a nonlinear least-squares method using the MULTI program (Yamaoka et al., 1981) to obtain estimates of kinetic parameters. The input data were weighted as the reciprocal of the square of the observed values, and the algorithm used for the fitting was the Damping Gauss Newton Method (Yamaoka et al., 1981).

(2)

Statistical Methods. The results are shown as means ± S.E. for the number of determinations. Student's t test was used to determine the significant difference between the means of two groups, with P < .01 and P < .05 as the minimum levels of significance.

	Results

Top Abstract Introduction Materials and methods Results Discussion References

Uptake of Carboxylate and Lactone Forms of SN38-Glu by Membrane Vesicles from KB-Derived Cells. Uptake of the carboxylate form of SN38-Glu (50 µM) into membrane vesicles from KB-3-1, KB-C2, C-A500, and KCP-4 cells is shown in Fig. 1A. The uptake was significantly stimulated by ATP in C-A500 and KB-C2 vesicles (Fig. 1A), although the transport activity in KB-C2 was much lower than that in C-A500 (Fig. 1A). By contrast, no significant ATP-dependence was found in KB-3-1 and KCP-4 vesicles (Fig. 1A). Similar results were obtained for the uptake of the lactone form of SN38-Glu (50 µM) (Fig. 1B), although the transport activity for the lactone form of SN38-Glu was lower than that of its corresponding carboxylate form in C-A500 vesicles (Fig. 1, A and B).

View larger version (25K): [in this window] [in a new window]   Fig. 1.   Uptake of the carboxylate and lactone forms of SN38-Glu and the carboxylate form of SN-38 by membrane vesicles. Membrane vesicles from KB-3-1, KB-C2, C-A500 and KCP-4 cells (5 µg protein) were incubated with 50 µM the carboxylate form of SN38-Glu (A), the lactone form of SN38-Glu (B), and the carboxylate form of SN-38 (C) in the presence of 5 mM ATP (closed column) or AMP (open column) and ATP-regenerating system for 2 min. Data are mean ± S.E. of five different experiments. **P < .01, significantly different from that in the absence of ATP.

Uptake of Carboxylate Forms of SN-38 and CPT-11 by Membrane Vesicles from KB-Derived Cells. For the carboxylate form of SN-38 (50 µM), ATP-dependent uptake was observed only in C-A500 vesicles (Fig. 1C). No significant ATP dependence was found in KB-3-1, KB-C2, or KCP-4 vesicles (Fig. 1C). Kinetic analysis revealed that the ATP-dependent uptake of the carboxylate form of SN-38 by C-A500 vesicles consists of a single saturable component with a K_m of 17 µM and a V_max of 1.18 nmol/min/mg protein (Fig. 2).

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Eadie-Hofstee plot for ATP-dependent uptake of the carboxylate form of SN-38 by membrane vesicles isolated from C-A500 cells. Membrane vesicles (5 µg protein) were incubated with varying concentrations of the carboxylate form of SN-38 in the presence of 5 mM ATP or AMP and ATP-regenerating system for 2 min. ATP-dependent uptake was obtained by subtracting the uptake in the presence of AMP from that in the presence of ATP. Kinetic parameters (Km = 17 µM, Vmax = 1.12 nmol/min/mg protein) were obtained by fitting using eq. 1. Data are mean ± S.E. of three different experiments. Solid line represents the fitted line.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Uptake of the carboxylate form of CPT-11 by membrane vesicles. Membrane vesicles from KB-3-1, KB-C2, C-A500, and KCP-4 cells (5 µg protein) were incubated with 5 (A) or 50 µM (B) of the carboxylate form of CPT-11 in the presence of 5 mM ATP (closed column) or AMP (open column) and ATP-regenerating system for 2 min. Data are mean ± S.E. of three different experiments. *P < .05, significantly different from that in the absence of ATP.

Resistance of KB-Derived Cells to CPT-11 and SN-38. The resistance of KB-derived cells to CPT-11 and SN-38 was determined by MTT assay. As shown in Table 1, the resistance of these cells to CPT-11 and SN-38 was increased 6.3- and 6.8-fold for KB-C2 cells compared with KB-3-1 cells, respectively, whereas the corresponding figures for C-A500 cells were 12- and 27-fold, respectively, and for KCP-4 cells 2.3- and 20-fold, respectively.

	Discussion

Top Abstract Introduction Materials and methods Results Discussion References

Although the overexpression of transport proteins on tumor cells has been reported to be one of the mechanisms for the acquisition of resistance to CPT-11, a promising anticancer agent in clinical situations, no detailed mechanism has been proposed (Minato et al., 1990; Takigawa et al., 1992; Hasegawa et al., 1995). Because our previous studies suggested that CPT-11 and its metabolites are substrates for cMOAT (Chu et al., 1997a,b), cMOAT and/or related transporters might be involved in the active efflux of these drugs from tumor cells. To provide a deeper insight into the mechanism of resistance to CPT-11 acquired by the overexpression of efflux transporters and to examine the substrate specificity of P-gp and GS-X family members, uptake studies involving membrane vesicles isolated from KB-derived cell lines were performed.

For the carboxylate and lactone forms of SN38-Glu, marked ATP-dependent uptake was observed in C-A500 vesicles (Fig. 1, A and B), suggesting that the transporter, which is predominantly responsible for the efflux of this conjugated metabolite, is MRP. According to Northern blot analysis, we found that C-A500 cells also overexpressed cMOAT, although its expression level was much lower than that of MRP (K Ueda, H Suzuki, S Akiyama, and Y Sugiyama, submitted). Because our previous study with isolated rat bile CMVs also indicated that SN38-Glu is a high-affinity substrate for cMOAT (K_ms for the lactone and carboxylate forms are 2.30 and 0.96 µM, respectively) (Chu et al., 1997b), it is possible that cMOAT may also be involved in the ATP-dependent transport of SN38-Glu in C-A500. As shown in Fig. 1, A and B, the ATP-dependent uptake of the carboxylate form of SN38-Glu into C-A500 and KB-C2 vesicles is higher than that of its lactone form, which is consistent with our previous findings in rat CMVs (Chu et al., 1997b). These results can be explained if we consider the previous hypothesis that divalent anions are better substrates for cMOAT and MRP (Oude Elferink et al., 1995) and the fact that the carboxylate form of SN38-Glu has one additional anionic charge compared with its lactone form.

Significant ATP-dependent uptake of the carboxylate and lactone forms of SN38-Glu was also observed in KB-C2 vesicles where P-gp is overexpressed, although the transport activity in KB-C2 was much lower compared with that in C-A500 (Fig. 1, A and B). The suggestion that SN38-Glu is also a substrate for P-gp is supported by the finding that some glucuronide conjugates can be substrates for P-gp. Vore et al. (1996) reported that estradiol 17--D-glucuronide (E₂17G), a typical substrate for MRP and cMOAT, may also be transported via P-gp, based on the finding that the ATP-dependent uptake of E₂17G into rat CMVs was inhibited by C219, a monoclonal antibody against P-gp. In addition, we have also reported that the primary active transporter(s), which is also expressed in Eisai hyperbilirubinemic rat CMVs, is responsible for the ATP-dependent uptake of SN38-Glu in rat CMVs with low affinity (K_ms for the lactone and carboxylate forms were 189 and 75 µM, respectively) (Chu et al., 1997b). In contrast, no significant ATP-dependent uptake of SN38-Glu was found in KCP-4 vesicles (Fig. 1, A and B), although glutathione conjugates (LTC₄ and DNP-SG) are taken up into the same vesicle preparations (Fujii et al., 1994; Chuman et al., 1996). These results, along with our recent observation that no ATP-dependent uptake of E₂17G is detectable in KCP-4 vesicles (K Ueda, H Suzuki, S Akiyama, and Y Sugiyama, submitted), indicate that the unidentified GS-X pump expressed on KCP-4 may exhibit a different substrate specificity from that of MRP, particularly as far as recognition of the glucuronide moiety is concerned.

As shown in Fig. 1C, ATP-dependent uptake for the carboxylate form of SN-38 was observed only in C-A500 vesicles, suggesting that MRP is involved in the active efflux of this compound. Although cMOAT was also expressed in C-A500, its expression level was considerably lower than that of MRP (K Ueda, H Suzuki, S Akiyama, and Y Sugiyama, submitted). Moreover, no ATP-dependent uptake in KB-3-1 vesicles was found in which cMOAT was also detected. By comparing the transport affinity of the carboxylate form of SN-38 in C-A500 vesicles (K_m = 17 µM) with its ATP-dependent uptake in rat and human CMVs, we found that the K_m of this compound was 69 and 180 µM for rat and human cMOAT, respectively (Chu et al., 1997b, 1998). This comparison suggested that the affinity of the carboxylate form of SN-38 in C-A500, which overexpresses MRP, is approximately 10-fold higher than that for cMOAT from the same species. In addition, this compound was not transported via a GS-X pump expressed on KCP-4 cells (Fig. 1C). Taking these results into consideration, we speculate that MRP is predominantly responsible for the ATP-dependent uptake of the carboxylate form of SN-38 in C-A500, whereas the contribution of cMOAT to its active efflux is minor.

The transport properties determined in membrane vesicles need to be discussed in relation to the resistance of tumor cells to antitumor drugs. The resistance of C-A500 cells to SN-38 determined by MTT assay was 27-fold higher than that of KB-3-1 cells (Table 1). Taking this together with the finding that SN-38 is a good substrate for MRP, suggests that MRP is responsible for the resistance to this drug. In contrast, the results with KB-C2 vesicles (Fig. 1C) indicate that P-gp may play only a very minor role in the resistance to SN-38. Indeed, the resistance of KB-C2 cells to SN-38 (6.8-fold higher than that of KB-3-1 cells) was much lower than that of C-A500 cells (Table 1). Our observations are further supported by the finding of Hoki et al. (1997) that the transfection of a plasmid containing cDNA for wild-type P-gp does not confer any resistance to SN-38 in NIH/3T3 cells.

Although no significant ATP-dependent uptake of the carboxylate form of SN-38 was found in KCP-4 vesicles (Fig. 1C), MTT assay revealed that KCP-4 cells also exhibited a similar resistance to SN-38 (20-fold higher than that of KB-3-1 cells) as that of C-A500 cells (Table 1). Although we cannot ignore the possibility that the lactone form of SN-38, whose antitumor activity is much more potent than its carboxylate form, is extruded via a GS-X pump expressed on KCP-4 cells, this may not be plausible because it is believed that the GS-X pump can accept only anionic compounds as substrates (Lautier et al., 1996; Loe et al., 1996). The resistance of KB-C2 and KCP-4 cells to SN-38 may be accounted for other mechanism(s), such as alterations in the expression level of topoisomerase and/or mutations in the topoisomerase gene (Slichenmyer et al., 1993; Tanizawa et al., 1993).

The involvement of P-gp in the extrusion of CPT-11 is supported by the finding that the uptake of CPT-11 into KB-C2 vesicles is stimulated by ATP at a substrate concentration of 5 µM. In addition, no ATP-dependent uptake of the carboxylate form of CPT-11 was observed in membrane vesicles from C-A500 and KCP-4, as well as KB-3-1, at both 5 and 50 µM. This suggested that active efflux of the carboxylate form of CPT-11 is not related to MRP or the GS-X pump. These results should be discussed in relation to our recent studies in which we reported that the uptake of the carboxylate form of CPT-11 into rat CMVs consists of a high (K_m = 3.4 µM and V_max = 115 pmol/min/mg protein) and a low (K_m = 236 µM and V_max = 1.99 nmol/min/mg protein) affinity component, the latter being attributed to cMOAT (Chu et al., 1997b). Our recent finding that the high-affinity component was inhibited by several drugs such as PSC-833, cyclosporin A, and verapamil (Sugiyama et al., 1998) suggests that P-gp is involved in the efflux of this compound. If we consider the transport activity under linear conditions (V_max/K_m), the contribution of P-gp to the uptake of CPT-11 by CMVs is approximately four times higher than that of cMOAT (Chu et al., 1997b), suggesting that CPT-11 is preferentially extruded via P-gp rather than a member of the GS-X pump family. These observations are consistent with the present finding that the ATP-dependent uptake of CPT-11 mediated by MRP is much less marked than that mediated by P-gp (Fig. 3).

The observation of ATP-dependent uptake of CPT-11 in KB-C2 vesicles (Fig. 3) is in good agreement with the fact that KB-C2 exhibited a 6-fold resistance to this compound compared with KB-3-1 cells (Table 1). It has also been reported that a bladder cancer cell line (KK47/ADM) and etoposide-resistant small-cell lung cancer cell lines (H69/VP and SBC-3/ETP), both of which overexpress P-gp, show cross-resistance to CPT-11 (Minato et al., 1990; Takigawa et al., 1992; Hasegawa et al., 1995). Irrespective of the fact that CPT-11 is per se not a good substrate for MRP (Fig. 3), the MTT assay revealed that C-A500 cells acquired resistance to CPT-11 (12-fold higher than that of KB-3-1 cells). This discrepancy may be accounted for by considering that the active metabolite formed intracellularly (SN-38) may be efficiently extruded via MRP (Fig. 1C). Alternatively, the expression level of carboxyesterase, an enzyme responsible for the conversion of CPT-11 to SN-38 (Slichenmyer et al., 1993), may be down-regulated in C-A500 cells.

Our study focused on four compounds related to CPT-11 with anionic moiety, because the uptake of the lactone form of SN-38 cannot be determined due to its very low solubility in transport medium and the limit of detection of the HPLC assay (Chu et al., 1997a, 1998). In addition, our previous rat CMV studies indicated that no ATP-dependent uptake was observed for the lactone form of CPT-11. Moreover, it did not inhibit the ATP-dependent uptake of 2,4-dinitrophenyl-S-glutathione, a typical substrate for cMOAT (Chu et al., 1997a). Thus, it is possible that the efflux of the lactone form of CPT-11 may not be related to the primary active transport protein expressed in KB-derived cells.

In conclusion, the results of the present study suggest that MRP and P-gp are involved in the active efflux of SN-38 and CPT-11, respectively, from human KB-derived cells and confer drug resistance. In addition, a difference in the substrate specificity among GS-X pump members was demonstrated in that SN-38 and its glucuronide can be a substrate for MRP, but not for a GS-X pump expressed on cisplatin-resistant KCP-4 cells.