Mutation of a single conserved tryptophan in multidrug resistance protein 1 (MRP1/ABCC1) results in loss of drug resistance and selective loss of organic anion transport.

Multidrug resistance protein 1 (MRP1/ABCC1) belongs to the ATP-binding cassette transporter superfamily and is capable of conferring resistance to a broad range of chemotherapeutic agents and transporting structurally diverse conjugated organic anions. In this study, we found that substitution of a highly conserved tryptophan at position 1246 with cysteine (W1246C-MRP1) in the putative last transmembrane segment (TM17) of MRP1 eliminated 17beta-estradiol 17-(beta-d-glucuronide) (E(2)17betaG) transport by membrane vesicles prepared from transiently transfected human embryonic kidney cells while leaving the capacity for leukotriene C(4)- and verapamil-stimulated glutathione transport intact. In addition, in contrast to wild-type MRP1, leukotriene C(4) transport by the W1246C-MRP1 protein was no longer inhibitable by E(2)17betaG, indicating that the mutant protein had lost the ability to bind the glucuronide. A similar phenotype was observed when Trp(1246) was replaced with Ala, Phe, and Tyr. Confocal microscopy of cells expressing Trp(1246) mutant MRP1 molecules fused at the C terminus with green fluorescent protein showed that they were correctly routed to the plasma membrane. In addition to the loss of E(2)17betaG transport, HeLa cells stably transfected with W1246C-MRP1 cDNA were not resistant to the Vinca alkaloid vincristine and accumulated levels of [(3)H]vincristine comparable to those in vector control-transfected cells. Cells expressing W1246C-MRP1 were also not resistant to cationic anthracyclines (doxorubicin, daunorubicin) or the electroneutral epipodophyllotoxin VP-16. In contrast, resistance to sodium arsenite was only partially diminished, and resistance to potassium antimony tartrate remained comparable to that of cells expressing wild-type MRP1. This suggests that the structural determinants required for transport of heavy metal oxyanions differ from those for chemotherapeutic agents. Our results provide the first example of a tryptophan residue being so critically important for substrate specificity in a eukaryotic ATP-binding cassette transporter.

The 190-kDa multidrug resistance protein 1 (MRP1) 1 is a member of a branch of the ATP-binding cassette (ABC) superfamily of transport proteins designated ABCC (1)(2)(3). When overexpressed in tumor cells, MRP1 (gene symbol ABCC1) confers resistance to anticancer drugs and other xenobiotics of remarkably diverse structures and charge. Thus, the resistance spectrum associated with MRP1 expression extends from cationic and neutral natural product drugs (e.g. vincristine, doxorubicin, and VP-16) to the antimetabolite methotrexate to arsenical and antimonial oxyanions (4). MRP1 is also a primary active transporter of conjugated organic anions that include GSH-, glucuronide-, and sulfate-conjugated derivatives of both endo-and xenobiotics, suggesting a role for MRP1 in the disposition and elimination of these compounds (2).
The ability of MRP1 to confer drug resistance and transport conjugated organic anions is shared by at least two other proteins belonging to the ABCC subfamily, MRP2 (ABCC2) and MRP3 (ABCC3) (5,6). These three proteins share a common five-domain structure that distinguishes them from most other ABC transporters, which more typically have four structural domains. Thus, MRP1, -2, and -3 have a third membranespanning domain (MSD) with an extracytosolic N terminus that precedes a core structure consisting of two tandemly arranged units containing an MSD and a nucleotide-binding domain (see Fig. 1A) (7,8). The drug resistance profiles and the specificity and relative affinities of the three proteins for organic anions are similar, but not identical (5, 8 -10). For example, cisplatin resistance is associated with elevated expression of MRP2, but not MRP1 or MRP3 (8,11). Moreover, MRP3 is a relatively poor transporter of GSH-conjugated organic anions compared with MRP1 and MRP2 (9). Such differences in substrate specificity are not unexpected given that the amino acid sequences of MRP2 and MRP3 are only 48 and 58% identical, respectively, to that of MRP1. Two of the best characterized substrates of MRP1 are the GSH-conjugated arachidonic acid derivative leukotriene C 4 (LTC 4 ) and the glucuronidated estrogen 17␤-estradiol 17-(␤-Dglucuronide) (E 2 17␤G) (5,12). MRP1 has also been shown in vitro to transport certain unconjugated xenobiotics, but only in the presence of GSH (13)(14)(15)(16). Conversely, transport of GSH itself by MRP1 is markedly stimulated by a variety of structurally diverse heterocyclic molecules that may or may not be cotransported with this tripeptide (16,17). The mechanism by which MRP1-mediated transport and/or cotransport of its conjugated and unconjugated substrates occurs is presently un-clear. Nevertheless, it has been proposed that MRP1 contains a bipartite binding site(s) that can accommodate the structural and physical diversity of its substrates (3,13,18).
The specific amino acids involved in the recognition, binding, and transport of MRP1 substrates are largely unknown. In previous studies, the differing abilities of human MRP1 and its murine ortholog to confer resistance to anthracycline drugs have allowed us to identify a nonconserved glutamic acid residue at position 1089 in putative transmembrane segment 14 (TM14) in the third C-proximal MSD (MSD3) of human MRP1 that is essential for its ability to confer resistance to this class of drugs (19). Anthracyclines such as doxorubicin and daunorubicin exist predominantly as cations at physiological pH, and consequently, it seems reasonable for a negatively charged membrane-embedded amino acid to be involved in conferring resistance to these drugs (20). In this study, we have identified a highly conserved tryptophan residue in predicted TM17, the last TM segment in MSD3, that is critical not only for enabling the transport of the conjugated organic anion E 2 17␤G, but also for the ability of MRP1 to confer resistance to both cationic and electroneutral natural product drugs.
Transient and Stable Transfections of MRP1 Expression Vectors-For transient transfections, wild-type and mutant pcDNA3.1(Ϫ)-MRP1 expression vectors were transfected into SV40-transformed human embryonic kidney cells (HEK293T). Briefly, ϳ3.9 ϫ 10 6 cells were seeded in 175-cm 2 flasks, and 24 h later, DNA (16 g) was added using FuGENE TM -6 (Roche Molecular Biochemicals, Laval, Quebec, Canada) according to the manufacturer's instructions. After 48 -60 h, the HEK293T cells were harvested, and inside-out membrane vesicles were prepared as described previously (13). Empty vector and vector containing the wild-type cDNAs were included as controls in all experiments. Levels of wild-type and mutant MRP1 proteins were determined by immunoblotting as described below. For stable transfections, ϳ1.5 ϫ 10 5 HeLa cells were seeded in each well of a 6-well plate, and 24 h later, DNA (1 g) was added with FuGENE TM -6. After 48 h, the cells were subcultured 1:12, and the medium was replaced with fresh medium containing G418 (1000 g/ml; Geneticin, Life Technologies, Inc.). After ϳ14 days, cell colonies were removed individually using cloning cylinders and subcultured in 600 g/ml G418. Levels of MRP1 protein in G418-resistant cell populations were then determined by immunodot blotting, and the proportion of cells expressing MRP1 was determined by flow cytometry, followed by cloning by limiting dilution as described below.
Measurement of Protein Levels in Transfected Cells-The levels of wild-type and mutant MRP1 proteins were determined by immunoblot analysis of membrane protein fractions from transfected cells essentially as described (22). Proteins were resolved on a 6 -7% polyacrylamide gel and electrotransferred to a nylon membrane. Blots were blocked with 4% (w/v) skim milk powder for 1 h, followed by incubation with primary antibody. Wild-type and mutant MRP1 proteins were detected with the human MRP1-specific murine mAb QCRL-1 (diluted 1:10,000), which recognizes a linear epitope consisting of amino acids 918 -924 (23). After washing, blots were incubated with horseradish peroxidase-conjugated goat anti-mouse antibody (Pierce, Edmonton, Alberta, Canada), followed by application of Renaissance ® chemiluminescence blotting substrate (PerkinElmer Life Sciences). Relative levels of protein expression were estimated by densitometric analysis using a ChemiImager TM 4000 (Alpha Innotech, San Leandro, CA).
Confocal Microscopy-HEK293T cells were seeded at 3.5 ϫ 10 5 cells/ well in a 6-well plate on coverslips coated with 0.1% gelatin in 2 ml of Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and transiently transfected with the MRP1-GFP cDNA constructs (1 g of DNA/well) using FuGENE TM -6. Forty-eight hours later, the coverslips were washed once with PBS and fixed with 4% paraformaldehyde for 10 min at room temperature. After washing twice with PBS, the cells were permeabilized by adding 0.2% Triton X-100 in PBS. The coverslips were then incubated in RNase A (10 g/ml in 0.1% Triton X-100 and 1% bovine serum albumin in PBS) for 60 min at room temperature. The coverslips were washed again, and cell nuclei were stained with 1 ml of propidium iodide (2 g/ml in PBS) for 45 min in the dark. Finally, the coverslips were placed on slides containing 1 drop of Antifade solution (Molecular Probes, Inc., Eugene, OR), and cells were examined using a Meridian InSight Plus confocal microscope equipped with an air-cooled argon laser. Images obtained at 488 nm excitation were pseudo-colored and overlaid using Maxim DL software.
Flow Cytometry and Cloning by Limiting Dilution of Stably Transfected HeLa Cells-To determine the proportion of stably transfected cells expressing MRP1, populations of HeLa cells selected in G418 for Ͼ3 weeks were analyzed by flow cytometry with the MRP1-specific mAb QCRL-3, which recognizes a conformation-dependent epitope within the first nucleotide-binding domain of MRP1 (22). Briefly, cells were washed twice with cold PBS and fixed with 1% paraformaldehyde in PBS for 30 min at 4°C. After washing with blocking solution (1% bovine serum albumin and 0.1% Triton X-100 in PBS), cells were incubated with mAb QCRL-3 (diluted 1:100) for 1 h. The cells were washed twice with blocking solution and incubated for 30 min with fluorescein-conjugated goat anti-mouse IgG (H ϩ L) F(abЈ) 2 fragment (Pierce) diluted 1:100 in blocking solution. After two washes with blocking solution, cells were resuspended in 1% paraformaldehyde in PBS and analyzed on a Coulter Epic flow cytometer. Several populations containing the highest level of MRP1 with the largest proportion of MRP1-positive cells were cloned by limiting dilution and expanded to obtain stable cell lines with Ͼ90% of the cells expressing wild-type or mutant MRP1.
MRP1-mediated Transport by Inside-out Membrane Vesicles-In-side-out membrane vesicles were prepared from transiently transfected HEK293T cells or stably transfected HeLa cells, and ATP-dependent transport of 3 H-labeled substrates by the membrane vesicles was measured using a rapid filtration technique as described previously (13). Briefly, LTC 4 transport assays were performed at 23°C in a 50-l reaction containing 50 nM LTC 4 (40 nCi), 4 mM AMP or ATP, 10 mM MgCl 2 , 10 mM creatine phosphate, 100 g/ml creatine kinase, and 2-4 g of vesicle protein in transport buffer (50 mM Tris-HCl, 250 mM sucrose, pH 7.4). Uptake was stopped at selected times by rapid dilution in ice-cold buffer, and then the reaction was filtered through glass-fiber filters (type A/E) that had been presoaked in transport buffer. Radioactivity was quantitated by liquid scintillation counting. All data were corrected for the amount of [ 3 H]LTC 4 that remained bound to the filter, which was usually Ͻ10% of the total radioactivity. Transport in the presence of AMP was subtracted from transport in the presence of ATP to determine ATP-dependent LTC 4 uptake. All transport assays were carried out in triplicate, and results are expressed as means Ϯ S.D. Uptake of [ 3 H]E 2 17␤G was measured in a similar fashion, except that membrane vesicles (2-4 g of protein) were incubated at 37°C in a total reaction volume of 50 l containing E 2 17␤G (400 nM, 40 nCi) and the components as described for LTC 4 transport. [ 3 H]GSH uptake was also measured by rapid filtration with membrane vesicles (20 g of protein) incubated at 37°C in a 60-l reaction volume with 100 M [ 3 H]GSH (300 nCi/reaction) (17). To minimize GSH catabolism by ␥-glutamyltranspeptidase during transport, membranes were preincubated in 0.5 mM acivicin for 10 min at 37°C prior to measuring [ 3 H]GSH uptake in the presence of verapamil (100 M) (16).

Chemosensitivity Testing and 3 H-Labeled Drug Accumulation in Stably Transfected HeLa Cell Lines-
The relative drug resistance of the HeLa cell lines stably transfected with wild-type and mutant MRP1 cDNAs was determined using a tetrazolium-based microtiter plate assay (24). Cells were seeded in 96-well plates (1 ϫ 10 4 cells/well), incubated at 37°C for 24 h before the addition of drug, and then incubated for a further 72 h before the addition of 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (2 mg/ml of PBS). After 3 h at 37°C, isopropyl alcohol and 1 N HCl (24:1) were added to solubilize the formazan crystals, and the absorbance at 570 nm was determined using an EL X 800 microplate reader (Bio-Tek Instruments, Inc.). IC 50 values were obtained from the dose-response curves using GraphPAD Prism ® software. All assays were carried out in quadruplicate and reported as means Ϯ S.D. Relative resistance was calculated as the IC 50 of the wild-type or mutant MRP1-transfected HeLa cell line divided by the IC 50 of the HeLa cell line transfected with empty pcDNA3.1(Ϫ) vector alone.
[ 3 H]Vincristine accumulation in intact cells was measured as described previously (25). Briefly, HeLa cells (1.25 ϫ 10 6 /ml) were incubated at 37°C in the presence of 1 M [ 3 H]vincristine (1 Ci/ml) in RPMI 1640 medium supplemented with 5 mM HEPES (pH 7.0), 10 mM glucose, and 5% fetal bovine serum. Aliquots of suspended cells were removed at 0 and 60 min, and accumulation of [ 3 H]vincristine was stopped by rapid dilution into ice-cold PBS. Cells were washed twice with PBS, and the cell pellets were solubilized in 1% SDS. Cell-associated radioactivity was determined by liquid scintillation counting.

Identification of Trp 1246 as a Functionally Important Amino
Acid in MRP1-During the course of a mutational analysis of proline residues in MRP1, we generated an MRP1 mutant in which proline at position 196 was replaced with alanine (P196A). When stably expressed in HeLa cells, the ability of this mutant to transport organic anions was reduced, and it no longer conferred resistance to natural product drugs. In contrast, a second independently derived P196A mutant HeLa cell line displayed a phenotype similar to that of cells expressing wild-type MRP1. 2 Using three sets of overlapping primers, the integrated MRP1-transfected cDNAs of the wild-type MRP1 and the first P196A mutant cell lines were amplified by polymerase chain reaction. The DNA fragments obtained were of the size expected, and when sequenced, it was discovered that, in addition to the P196A mutation, the transfected copies of MRP1 cDNA in the first cell mutant line with an altered phenotype contained a G 3 C nucleotide mutation, resulting in substitution of cysteine for tryptophan at position 1246. This finding, together with the highly conserved nature of Trp 1246 among ABC transporters belonging to the ABCC subfamily (Fig. 1B), prompted us to initiate a more extensive analysis of the structural and functional importance of this amino acid.
A W1246C Substitution Causes Complete Loss of E 2 17␤G Transport Activity in Membrane Vesicles from Transiently Transfected HEK293T Cells-As a first step to confirming the importance of Trp 1246 , this amino acid was substituted with cysteine in pcDNA3.1(Ϫ)-MRP1 K by site-directed mutagenesis, and the resulting construct was then transiently transfected into HEK293T cells. Control transfections with the wild-type construct and the empty pcDNA3.1(Ϫ) vector were carried out at the same time. Membrane vesicles were prepared, and protein expression levels were determined by immunoblotting. For the experiments shown in Fig. 2, the levels of W1246C-MRP1 in the transient transfectants were ϳ70% those of wild-type MRP1 ( Fig. 2A). A time course of LTC 4 uptake was performed, and the levels of uptake by the W1246C-MRP1 mutant relative to wild-type MRP1 were 2 K. Ito, R. G. Deeley, and S. P. C. Cole, unpublished observations. observed to be essentially proportional to their relative levels of expression (Fig. 2B). Subsequent kinetic analyses of LTC 4 transport by the W1246C-MRP1 mutant indicated that its affinity for this substrate was reduced compared with wild-type MRP1 (K m ϭ 189 nM versus 79 nM), but the V max values for the two proteins were similar (40.6 versus 35.1 pmol/mg/min). 3 These results indicate that substitution of Trp 1246 with Cys reduces the affinity of MRP1 for this substrate, but that the transport capacities of the mutant and wild-type proteins are comparable. In contrast, uptake of E 2 17␤G by the W1246C-MRP1 mutant was dramatically reduced and was comparable to that of the vector-transfected controls (Fig. 2C). We have previously shown that E 2 17␤G is a competitive inhibitor of MRP1-mediated LTC 4 transport, with K i ϳ 25 M (26). Thus, the ability of E 2 17␤G to inhibit LTC 4 transport by the W1246C-MRP1 mutant was examined, and the results are shown in Fig. 2D. E 2 17␤G (25 M) inhibited LTC 4 transport by wild-type MRP1 by 54%, as expected. In contrast, E 2 17␤G had no effect on LTC 4 transport by W1246C-MRP1, indicating that the loss of E 2 17␤G transport by the mutant protein is associated with a loss of binding of this substrate.
Conservative and Nonconservative Substitutions of MRP1 Trp 1246 Cause Loss of E 2 17␤G Transport Activity-To determine whether the loss of E 2 17␤G transport activity was related specifically to a Cys substitution at position 1246, several different amino acids were introduced in place of Trp 1246 . These included substitution with a nonpolar non-aromatic amino acid (Ala; W1246A-MRP1) as well as conservative substitutions with polar (Tyr; W1246Y-MRP1) and nonpolar (Phe; W1246F-MRP1) aromatic amino acids. Membrane vesicles were prepared from HEK293T cells transiently transfected with these constructs, and the ATP-dependent transport of LTC 4 and E 2 17␤G was determined (Fig. 3). As shown in Fig. 3A, the levels of expression of the mutant proteins were comparable or slightly higher than those of wild-type MRP1. The LTC 4 transport levels of the W1246A-MRP1 mutant (Fig. 3B) and the W1246Y-MRP1 and W1246F-MRP1 mutants (Fig. 3C) were similar to those of wild-type MRP1 and the W1246C-MRP1 mutant. In contrast, like the W1246C mutant, the W1246A mutant did not transport E 2 17␤G (Fig. 3D). E 2 17␤G transport by the W1246Y and W1246F mutants was also extremely low (ϳ10% of wild-type MRP1) (Fig. 3E).
Verapamil-stimulated Transport of GSH Remains Intact after Substitution of Trp 1246 -To determine whether the Trp 1246 substitution might affect the transport of other MRP1 substrates, GSH uptake was measured in membrane vesicles pre-3 K. Ito, R. G. Deeley, and S. P. C. Cole, manuscript in preparation. pared from HEK293T cells transfected with wild-type and Trp 1246 mutant MRP1 cDNAs. We have shown previously that MRP1 mediates low level ATP-dependent GSH transport that is stimulated by the calcium channel blocker verapamil and certain of its dithiane analogs (17,27). Consequently, GSH transport by wild-type MRP1 and the Trp 1246 mutant proteins was measured in the presence of 100 M verapamil. As shown in Fig. 4, all four Trp 1246 mutant MRP1 proteins retained the ability to mediate verapamil-stimulated GSH uptake at levels proportional to the expressed protein.
Trp 1246 Mutant MRP1 Molecules Are Correctly Routed to the Plasma Membrane-To ensure that the loss of transport activity in the Trp 1246 MRP1 mutants was not caused by impaired trafficking of the mutant molecules to the plasma membrane, GFP-tagged constructs of wild-type MRP1 and mutants W1246C and W1246A were generated and transiently transfected into HEK293T cells. When viewed under the confocal microscope, both the wild-type and mutant MRP1 proteins were observed to localize strongly to the plasma membrane, indicating that in the transiently transfected cells, the mutants were correctly routed to the cell surface (Fig. 5). Studies with HEK293T and HeLa cells transfected with wild-type MRP1-GFP cDNA showed that fusion of GFP to the C terminus of wild-type MRP1 did not affect the LTC 4 and E 2 17␤G transport activities or the drug resistance-conferring properties of the protein. 4 Drug Resistance Is Lost and Accumulation Is Restored in Stably Transfected Trp 1246 Mutant MRP1 HeLa Cells-In addition to its ability to transport organic anions such as GSH, E 2 17␤G, and LTC 4 , MRP1 confers resistance to natural product anticancer drugs by reducing drug accumulation in cells in which it is overexpressed. To examine whether Trp 1246 plays a role in conferring drug resistance, stably transfected cell lines were established by transfection of HeLa cells with pcDNA3.1(Ϫ)-MRP1 K and pcDNA3.1(Ϫ)-W1246C-MRP1. After G418 selection and cloning by limiting dilution, stable cell lines were obtained, and uniformity of protein expression levels was established by flow cytometric and immunoblot analyses. The transfected HeLa cell lines were then tested for vincristine resistance using a tetrazolium salt-based chemosensitivity assay. As expected, HeLa cells expressing wild-type MRP1 displayed ϳ8-fold resistance to this drug (Fig. 6A). In contrast, the IC 50 of the W1246C-MRP1 mutant for vincristine was similar to that of the vector control-transfected cell line. When accumulation of [ 3 H]vincristine was measured, steady-state concentrations of the drug in the W1246C-MRP1-transfected cells were comparable to those in vector control-transfected cells (33.83 Ϯ 1.94 versus 34.25 Ϯ 2.24 pmol/10 6 cells/h) (Fig. 6B). However, in wild-type MRP1-transfected cells, vincristine accumulation was reduced to 19.44 Ϯ 1.11 pmol/10 6 cells/h (ϳ57% of vector control-transfected cells), consistent with results obtained in previous studies (4).
The drug resistance phenotype of the W1246C-MRP1-transfected cells was further characterized by determining the sensitivity of these cells to the cationic anthracyclines doxorubicin and daunorubicin as well as the electroneutral epipodophyllotoxin VP-16 (etoposide). As shown in Fig. 7, the IC 50 values of the W1246C-MRP1 mutant cells for all three drugs were similar to those of the vector control-transfected cells. Thus, in addition to eliminating the ability to bind and transport the conjugated estrogen E 2 17␤G, the W1246C substitution in MRP1 results in loss of resistance to all classes of natural product drugs tested.
Resistance to Potassium Antimony Tartrate Is Retained, but Resistance to Sodium Arsenite Is Reduced, in W1246C-MRP1expressing Cells-We have previously shown that in addition to conferring resistance to anticancer drugs, both human and murine MRP1 can confer low level resistance to arsenical and antimonial oxyanions (4,28). When tested for sensitivity to potassium antimony tartrate, the IC 50 of the W1246C mutant HeLa cells was comparable to that of cells expressing wild-type MRP1 (40 versus 30 g/ml) (Fig. 8A). Thus, both cell lines showed a similar level of resistance (6 -8-fold) to this heavy metal oxyanion. In contrast, W1246C mutant cells displayed reduced resistance to sodium arsenite (Fig. 8B). Thus, the IC 50 of W1246C mutant HeLa cells for sodium arsenite was 1 g/ml compared with 3.5 g/ml for cells expressing wild-type MRP1 and 0.6 g/ml for vector control-transfected cells. DISCUSSION MRP1 was first cloned and identified as a protein capable of effluxing cationic and electroneutral chemotherapeutic agents and subsequently demonstrated to be a transporter of GSH-, glucuronide-, and sulfate-conjugated organic anions (1, 2, 28 -31). Earlier studies from our group and others demonstrating competitive inhibition between structurally diverse substrates indicated that MRP1 contains mutually exclusive substratebinding sites and TM structures that allow the recognition and passage of cationic, anionic, and electroneutral molecules. Thus, among ABC transporters, MRP1 and its related proteins have a remarkably broad capacity to transport substrates of striking physical and chemical diversity. However, the molecular basis for the breadth of this transport capacity is largely unknown.
Previous studies have shown that portions of the first 280 amino acids of MRP1 encompassing the N-terminal MSD (MSD1) and the cytoplasmic loop connecting it to the second MSD (MSD2) are important for its expression in mammalian cell membranes as well as its ability to transport at least some of its substrates (32,33). In the course of a mutational analysis of this region of MRP1, we derived a mutant in which Pro 196 was replaced with Ala by site-directed mutagenesis (34). When stably expressed in HeLa cells, the ability of this MRP1 mutant to transport organic anions was diminished, and in particular, E 2 17␤G transport was not detectable. Moreover, this mutant protein failed to confer drug resistance. In contrast, a second independently derived transfected cell line also expressing an MRP1 protein with the same P196A mutation displayed a phenotype similar to that of cells expressing wild-type MRP1. 2 It was subsequently discovered that the transfected MRP1 in the first P196A mutant cell line had acquired a second nonengineered mutation causing the substitution of Cys for Trp at position 1246. Consequently, we generated a single W1246C mutant of MRP1 so that its phenotype could be compared with that of the double P196A/W1246C mutant.
When transiently transfected into HEK293T cells, W1246C-MRP1 could be expressed at levels comparable to those of wild-type MRP1 and also transported LTC 4 with a similar efficiency. Moreover, verapamil-stimulated GSH transport was still intact. However, as observed for the double P196A/ W1246C mutant, transport of E 2 17␤G by W1246C-MRP1 membrane vesicles was not detectably different from that obtained with vesicles from vector control-transfected cells, and although LTC 4 transport appeared unchanged in the W1246C mutants, it could no longer be inhibited by E 2 17␤G. Taken together, these results allowed us to conclude that the altered phenotype of the P196A/W1246C-transfected cells was attributable to the single tryptophan substitution at position 1246 in MSD3 and did not require the additional proline substitution at position 196 in the cytoplasmic loop between MSD1 and MSD2. Additional mutants in which Trp 1246 was substituted with Ala, Phe, and Tyr displayed transport characteristics similar to those of W1246C-MRP1, indicating that replacement of the tryptophan residue rather than the introduced cysteine was responsible for the phenotype.
To examine the drug resistance phenotype of the W1246C mutation, stably transfected HeLa cells were generated. In addition to the selective loss of E 2 17␤G transport, we found that W1246C-MRP1-expressing cells were no longer resistant to natural product chemotherapeutic agents, including the electroneutral VP-16 and the cationic vincristine and anthracyclines. Consistent with this loss of drug resistance, vincristine accumulation in intact cells expressing W1246C-MRP1 was comparable to that in vector control-transfected cells, which was ϳ2-fold higher than vincristine accumulation in HeLa cells expressing wild-type MRP1. On the other hand, the W1246C-MRP1-expressing cells were still resistant to antimony tartrate and partially resistant to sodium arsenite. Taken together, these observations suggest that the structural determinants in MRP1 necessary for recognition and transport of these heavy metal oxyanions differ from those for the natural product chemotherapeutic agents.
Current topological models of MRP1 are based on predictions from computer-based algorithms for which there are limited supporting biochemical data. Although similar in many respects, some models differ in their placement of several TM segments in the MSDs (7,35). However, with the exception of the PredictProtein model, which predicts only four TM helices in MSD3, all of the algorithms place Trp 1246 in the last TM segment close to the cytoplasmic face of the membrane (Fig.  1A). This tryptophan residue is extremely well conserved among MRP (ABCC) subfamily members, including MRP2, -3, -4, and -6 and murine MRP1. It is also found in the cystic fibrosis transmembrane conductance regulator, the sulfonylurea receptor SUR1, the Saccharomyces cerevisiae cadmium resistance protein Ycf1, Leishmania tarentolae PGPA, and the Arabidopsis thaliana conjugate transporter AtMRP3 (Fig. 1B). However, it is not conserved in human P-glycoprotein (MDR1) or other members of the ABCB subfamily to which this multidrug resistance protein belongs. The last TM helix in MSD3 of MRP1 has a highly amphipathic character, with Trp 1246 and other amino acids with hydrogen-bonding side chains densely clustered on one side of the ␣-helix (Fig. 1C) and predominantly in the region predicted to be in the inner leaflet of the plasma membrane (Fig. 1A). The last predicted TM helix of other ABCC subfamily members in which Trp 1246 is conserved such as MRP2 and MRP3 has a similar amphipathic character, lending support to the idea that the extensive hydrogen-bonding capacity of putative TM17 may play an important role in the substrate-binding and transport properties of these proteins. Interestingly, the topologically comparable TM segment in P-glycoprotein (TM12) has also been demonstrated to contain key determinants of substrate specificity, but TM12 of this ABC transporter is significantly less amphipathic than TM17 of MRP1, -2, and -3 (36 -38). This difference may be relevant to the apparent differences in the substrate specificity and transport mechanisms of MRP1 and P-glycoprotein (2, 3). MRP1 Trp 1246 may function to either maintain or form part of a substrate-binding pocket, and/or it may interact directly with the natural product drugs and E 2 17␤G. The facts that the Trp 1246 mutants were expressed at comparable levels and correctly routed to the plasma membrane in the transiently transfected cells and that their LTC 4 -and verapamil-stimulated GSH transport activities remained intact indicate that the mutations did not perturb the global structural integrity of the protein. The inability of the variously substituted Trp 1246 MRP1 mutants to transport E 2 17␤G suggests that tryptophan possesses specific physical and/or chemical properties that are essential for recognition and transport of some MRP1 substrates. Thus, the absence of E 2 17␤G transport activity and drug resistance in the W1246C-MRP1 mutant can be explained by the fact that although cysteine can participate in hydrogen bonding, it is considerably smaller than tryptophan and lacks aromaticity. However, the introduction of a phenylalanine or tyrosine residue at position 1246 also resulted in the loss of E 2 17␤G transport, indicating that aromatic interactions and general hydrogen-bonding capability are not sufficient to maintain the architecture of the binding site in MRP1 for this substrate and that specific tertiary interactions involving the indole residue are required. To our knowledge, our data represent the first example of a tryptophan residue being critically important for substrate specificity in a eukaryotic ABC transporter. It will be of interest to determine whether the analogous tryptophan residue in other members of the ABCC subfamily is also important for their transport functions.
Our data clearly demonstrate that Trp 1246 is critical for the recognition and transport of E 2 17␤G by MRP1. However, although the human and murine orthologs of MRP1 differ markedly in their ability to transport this conjugated estrogen, this amino acid is conserved in the two proteins, suggesting that other, nonconserved residues are important as well (28). Thus, it seems probable that several conserved and nonconserved determinants within the MSDs and possibly cytosolic regions of MRP1 come together to form a multipartite binding pocket for this organic anion and, furthermore, that substitution of just one amino acid can be sufficient to abrogate E 2 17␤G recognition and transport activity as well as the ability to confer drug resistance. Our recent finding that wild-type MRP1, but not W1246A-MRP1, transports the O-glucuronide of the tobaccospecific carcinogen 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol indicates that Trp 1246 is important for the transport of other glucuronide conjugates as well. 5 It has been proposed (and there is considerable evidence to support the notion) that substrates for P-glycoprotein are taken up from the inner leaflet of the plasma membrane (39). Whether or not this is also true for every MRP1 substrate is not known, but may depend on the physical characteristics of the substrate, such as its hydrophobicity, aromaticity, and charge, as well as certain aspects of its chemical structure. The fact that a hydrogen-bonding aromatic amino acid close to the cytosolic face of the protein is of such critical importance for the binding and transport of E 2 17␤G suggests that this organic anion may be taken up by MRP1 from the cytoplasm. Why LTC 4 transport is not affected by the Trp 1246 substitutions is unclear. It is possible that steric complementarity is more important for interaction of MRP1 with E 2 17␤G than with LTC 4 . In support of this idea are our earlier findings that a change in the site of glucuronidation on the D-ring of 16␣,17␤estriol from the 17␤-position to the 16␣-position increased the K i for E 2 17␤G transport ϳ30-fold (26). The LTC 4 molecule is significantly less rigid and more lipophilic than E 2 17␤G. These physical properties of LTC 4 may make its recognition and binding less affected by the changes in the architecture of a multipartite substrate-binding pocket caused by replacement of the Trp 1246 residue as well as possibly favoring its uptake from the membrane leaflet.
As mentioned previously, we have recently shown that the ability of MRP1 to confer resistance to cationic anthracyclines is lost when Glu 1089 is substituted with a glutamine residue (19). This finding, together with our present data, indicates that recognition and transport of these drugs depend not only on the presence of a membrane-embedded negatively charged amino acid in putative TM14, but also on the presence of the bulky indole Trp 1246 in TM17. Thus, these data raise the possibility that interactions between TM14 and TM17 are required for MRP1 to confer drug resistance. The comparable interaction in P-glycoprotein would be between TM9 and TM12. However, such an interaction has not been reported, providing further evidence of significant differences between these two drug resistance proteins with respect to the mechanisms by which they recognize and transport xenobiotics.
As has been proposed for other multidrug transporters (20), MRP1 substrate molecules may penetrate the hydrophobic core of the protein, where they form a number of van der Waals and stacking interactions with the surrounding hydrophobic and aromatic residues. Our data suggest that for a broad range of MRP1 substrates, aromatic stacking interactions, hydrogenbonding interactions, and possibly -bonding interactions with Trp 1246 (and possibly other amino acids in TM17) are also important, as are electrostatic interactions with negatively charged amino acids such as Glu 1089 in the case of cationic drugs (19). Whether or not anionic substrates have electrostatic interactions with positively charged residues in MRP1 is not yet known. However, the predicted MSDs of MRP1 and related proteins contain a significant number of membraneembedded arginine and lysine residues. Indeed, it has recently been reported that mutants of the related protein MRP2 in which several basic residues in MSD2 and MSD3 have been replaced by other amino acids show decreased transport activity (40,41). Thus, additional molecular and pharmacological studies are anticipated to continue to provide important information on the structural features that determine how MRP1 binds and transports both its charged and electroneutral substrates.