Identification of a Nonconserved Amino Acid Residue in Multidrug Resistance Protein 1 Important for Determining Substrate Specificity EVIDENCE FOR FUNCTIONAL INTERACTION BETWEEN TRANSMEMBRANE HELICES 14 AND 17*

Murine multidrug resistance protein 1 (mrp1), differs from its human ortholog (MRP1) in that it fails to confer anthracycline resistance and transports the MRP1 substrate, 17 (cid:1) -estradiol 17-( (cid:1) - D -glucuronide) (E 2 17 (cid:1) G), very poorly. By mutating variant residues in mrp1 to those present in MRP1, we identified Glu 1089 of MRP1 as being critical for anthracycline resistance. However, Glu 1089 mutations had no effect on E 2 17 (cid:1) G transport. We have now identified a nonconserved amino acid within the highly conserved COOH-proximal transmembrane helix of MRP1/mrp1 that is important for transport of the conjugated estrogen. Converting Ala 1239 in mrp1 to Thr, as in the corresponding position (1242) in MRP1, increased E 2 17 (cid:1) G transport 3-fold. Any mutation of mrp1 Ala 1239 , including substitution with Thr, decreased resistance to vincristine and VP-16 without double mutation, pGEM-MRP1-(2730–4832) containing mutation MRP1E1089Q was digested with Stu I and Xba I to yield a 4.1-kilobase pair fragment comprised of nucleotides 2730–3758 of MRP1E1089Q attached to the vector fragment, and this fragment was then ligated to a 1.1-kilobase pair Stu I- Xba I fragment encompassing nucleotides 3758–4832 of MRP1T1242A.Allmutations were confirmed by sequencing using DNA Thermo Sequenase and Cy5.5 and Cy5.0 dye terminator/primer (Amersham Pharmacia Biotech). DNA fragments containing the desired mutations were transferred into pCEBV7-mrp1 or pCEBV7-MRP1, after which the entire mutated inserts and the cloning sites were verified by DNA sequencing.

Murine multidrug resistance protein 1 (mrp1), differs from its human ortholog (MRP1) in that it fails to confer anthracycline resistance and transports the MRP1 substrate, 17␤-estradiol 17-(␤-D-glucuronide) (E 2 17␤G), very poorly. By mutating variant residues in mrp1 to those present in MRP1, we identified Glu 1089 of MRP1 as being critical for anthracycline resistance. However, Glu 1089 mutations had no effect on E 2 17␤G transport. We have now identified a nonconserved amino acid within the highly conserved COOH-proximal transmembrane helix of MRP1/mrp1 that is important for transport of the conjugated estrogen. Converting Ala 1239 in mrp1 to Thr, as in the corresponding position (1242) in MRP1, increased E 2 17␤G transport 3-fold. Any mutation of mrp1 Ala 1239 , including substitution with Thr, decreased resistance to vincristine and VP-16 without altering anthracycline resistance. However, introduction of a second murine to human mutation, Q1086E, which alone selectively increases anthracycline resistance, into mrp1A1239T restored resistance to both vincristine and VP-16. To confirm the importance of MRP1 Thr 1242 for E 2 17␤G transport and drug resistance, we mutated this residue to Ala, Cys, Ser, Leu, and Lys. These mutations decreased E 2 17␤G transport 2-fold. Conversion to Asp eliminated transport of the estrogen conjugate and also decreased leukotriene C 4 transport ϳ2-fold. The mutations also reduced the ability of MRP1 to confer resistance to all drugs tested. As with mrp1, introduction of a second mutation based on the murine sequence to create MRP1E1089Q/ T1242A restored resistance to vincristine and VP-16, but not anthracyclines, without affecting transport of leukotriene C 4 and E 2 17␤G. These results demonstrate the important role of Thr 1242 for E 2 17␤G transport. They also reveal a highly specific functional relationship between nonconserved amino acids in TM helices 14 and 17 of both mrp1 and MRP1 that enables both proteins to confer similar levels of resistance to vincristine and VP-16.
Human multidrug resistance protein 1 (MRP1), 1 a 190-kDa glycoprotein, is a member of the ATP-binding cassette (ABC) transporter superfamily (1)(2)(3). The predicted structure of MRP1 differs from that of a typical eukaryotic ABC transporter such as P-glycoprotein (P-gp). It contains a P-gp-like core region, which consists of two membrane-spanning domains (MSDs), each containing six transmembrane (TM) ␣-helices, and two nucleotide binding domains. However, MRP1 contains a third MSD predicted to consist of five TM ␣-helices with an extracellular NH 2 terminus and a cytoplasmic linker connecting the additional MSD with the core region (3)(4)(5)(6)(7).
Like P-gp, MRP1 confers resistance to many commonly used, structurally diverse natural product chemotherapeutic agents including anthracyclines, Vinca alkaloids, and epipodophyllotoxins (7)(8)(9)(10). However, several lines of evidence suggest that MRP1 and P-gp confer resistance to these drugs by different mechanisms. In vitro studies using P-gp-enriched membrane vesicles or purified reconstituted protein demonstrate that P-gp directly binds and transports its drug substrates in an ATP-dependent manner (11,12). Under similar conditions, MRP1 will only transport unmodified amphipathic drugs, such as vincristine and daunorubicin, in the presence of GSH in addition to ATP (13)(14)(15)(16)(17). In some cases, GSH appears to be co-transported with these compounds (15).
Despite the very broad substrate specificity of MRP1 and the relatively high level of primary structure conservation with mrp1 (87% identity), the murine and human proteins display some striking functional differences (26,27). Although mrp1 confers resistance to Vinca alkaloids and epipodophyllotoxins as efficiently as MRP1, cells overexpressing the murine protein do not show any detectable resistance to anthracyclines. In addition, despite the fact that MRP1 and mrp1 transport LTC 4 with similar efficiency, the murine protein transports E 2 17␤G very poorly (26). Studies using hybrid murine/human proteins revealed that the COOH-terminal third of MRP1 contains nonconserved amino acids that contribute to its ability to confer anthracycline resistance and to transport E 2 17␤G efficiently (28).
In a previous study, we demonstrated that mutating Glu 1089 in TM14 of MRP1 to Gln, as present in mrp1, essentially eliminated the ability of the protein to confer anthracycline resistance without affecting transport of E 2 17␤G (29). However, transport of the estrogen conjugate was increased in a mrp1/MRP1 hybrid containing a relatively highly conserved region of the human protein extending from amino acid 1188 to the COOH terminus (28). In the present study, we examined the consequences of replacing amino acid residues in this region that differ between the human and murine proteins. By doing so, we have identified a Thr residue at position 1242 in predicted TM17 of MRP1 that appears to account for the enhanced E 2 17␤G transport observed with the hybrid protein. In mrp1, the corresponding amino acid is Ala 1239 . Reciprocal mutation of these residues in the human and murine proteins decreased or increased E 2 17␤G transport, respectively. The nonconserved A1239/T1242 residues are close to a highly conserved tryptophan residue (W1246/1243 in MRP1/mrp1) present in human MRP2 to -4 and MRP6 and several more distantly related members of the ABCC branch of the ABC superfamily (30). We have shown recently that mutation of Trp 1246 , which has only a minor effect on transport of LTC 4 , eliminates the ability of MRP1 to transport E 2 17␤G. In addition, MRP1 W1246 mutant proteins fail to confer drug resistance (30). Unexpectedly, reciprocal exchange of the nonconserved Ala 1239 /Thr 1242 residues in mrp1 and MRP1 also decreased the ability of both proteins to confer resistance to vincristine and VP-16. This alteration in specificity was not expected because wild-type mrp1 and MRP1 display no differences in their ability to confer resistance to these drugs (26). However, we also show that a second reciprocal exchange of nonconserved amino acids between the TM17 mutant human and murine proteins, at the previously identified residue in TM14 shown to be critical for anthracycline resistance (29), can in both cases restore the ability to confer resistance to vincristine and VP-16. This observation provides compelling evidence that the two pairs of nonconserved residues, Glu 1089 and Thr 1242 in MRP1 and Gln 1086 and Ala 1239 in mrp1, must act in concert to enable the proteins to confer resistance to these, and possibly other, chemotherapeutic drugs. Site-directed Mutagenesis and Generation of Expression Vectors-To generate the hybrid protein mrp1/MRP1-(1354 -1531), a new ClaI site was introduced into pBluescript-mrp1 by site-directed mutagenesis at position 4049. Digestion of this construct with ClaI and NotI yielded a 7.0-kilobase pair fragment composed of nucleotides 1-4049 of mrp1 attached to the vector. This fragment was ligated to a 0.8-kilobase pair ClaI-NotI fragment containing nucleotides 4058 -4832 of MRP1. The resulting construct was excised using EcoRV and NotI and ligated into pCEBV7 (31) digested with PvuII and NotI to give construct pCEBV7-mrp1/MRP1-(1354 -1531).

Materials
All mutations were generated using the Transformer TM site-directed mutagenesis kit (CLONTECH, Palo Alto, CA). Templates were prepared as described previously (29). Mutagenesis was then performed according to the manufacturer's instructions using a selection primer 5Ј-GAG AGT GCA CGA TAT CCG GTG TG-3Ј that mutates a unique NdeI site in the vector to an EcoRV restriction site. Oligonucleotides bearing mismatched bases at the residues to be mutated (  All mutations were confirmed by sequencing using DNA Thermo Sequenase and Cy5.5 and Cy5.0 dye terminator/primer (Amersham Pharmacia Biotech). DNA fragments containing the desired mutations were transferred into pCEBV7-mrp1 or pCEBV7-MRP1, after which the entire mutated inserts and the cloning sites were verified by DNA sequencing.
Cell Lines and Tissue Culture-Stable transfection of HEK293 cells with the pCEBV7 vector containing the wild type MRP1 cDNAs or wild type mrp1 cDNAs has been described previously (26,29). All of the mutated MRP1 or mrp1 constructs were analyzed as stably transfected HEK293 cells grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 100 g/ml hygromycin B (Roche Molecular Biochemicals). Briefly, HEK293 cells were transfected with pCEBV7 vectors containing mutant MRP1 or mrp1 cDNAs using Fugene6™ (Roche Molecular Biochemicals) according to the manufacturer's instructions. After ϳ48 h, the transfected cells were supplemented with fresh medium containing 100 g/ml hygromycin B. Approximately 3 weeks post-transfection, the hygromycin B-resistant cells were cloned by limiting dilution, and the resulting cell lines were tested for expression of the mutant proteins.
Determination of Protein Levels in Transfected Cells-Plasma membrane vesicles were prepared as described previously (26,29). After determination of protein levels by Bradford assay (Bio-Rad), 1.25, 2.5, and 5 g of total membrane protein from transfectants expressing wild type MRP1/mrp1 were analyzed together with comparable amounts of membrane protein from cells expressing various mutant proteins by SDS-polyacrylamide gel electrophoresis (7.5% gel). Proteins were subsequently transferred to Immobilon-P polyvinylidene difluoride membranes (Millipore, Bedford, MA) by electroblotting. The mrp1 proteins were identified using the monoclonal antibody (mAb), MRPr1, which cross-reacts with the murine and human proteins (26). The human proteins were detected with mAb QCRL-1 (32). Antibody binding was detected with horseradish peroxidase-conjugated goat anti-rat/antimouse IgG (Pierce), followed by enhanced chemiluminescence detection and X-Omat™ Blue XB-1 films (PerkinElmer Life Sciences). Densitometry of the film images was performed using a ChemiImager™ 4000 (Alpha Innotech Corporation, San Leandro, CA). The relative protein expression levels were calculated by dividing the integrated densitometry values obtained for 1.25, 2.5, and 5 g of total membrane protein from transfectants expressing the mutant proteins by the integrated densitometry values obtained for the comparable amounts of total membrane proteins from transfectants expressing wild type proteins. Each comparison was performed at least three times in independent experi-ments. The results were then pooled, and the mean values were used for normalization purposes.
LTC 4 and E 2 17␤G Transport by Membrane Vesicles-Plasma membrane vesicles were prepared as described previously, and ATP-dependent transport of [ 3 H]LTC 4 into the inside-out membrane vesicles was measured by a rapid filtration technique (16,18). Briefly, vesicles (10 g of protein) were incubated at 23°C in 100 l of transport buffer (50 mM Tris-HCl, 250 mM sucrose, 0.02% sodium azide, pH 7.4) containing ATP or AMP (4 mM), 10 mM MgCl 2 , and [ 3 H]LTC 4 (50 nM, 100 nCi). At the indicated times, 20-l aliquots were removed and added to 1 ml of ice-cold transport buffer, followed by filtration under vacuum through glass fiber filters (Type A/E, Gelman Sciences, Dorval, Quebec, Canada). Filters were immediately washed twice with 5 ml of cold transport buffer and then dried before the bound radioactivity was determined by scintillation counting. All data were corrected for the amount of [ 3 H]LTC 4 that remained bound to the filter in the absence of vesicle protein (usually Ͻ5% of the total radioactivity). [ 3 H]LTC 4 uptake was expressed relative to the total protein concentration in each reaction. ATP-dependent [ 3 H]E 2 17␤G (400 nM, 120 nCi) uptake was measured as described for [ 3 H]LTC 4 except that 20 g of vesicle protein was used, and the reaction was carried out at 37°C.
K m and V max values of ATP-dependent [ 3 H]LTC 4 uptake by membrane vesicles (2.5 g of protein) were measured at various LTC 4 concentrations (0.01 to 2 M) for 1 min at 23°C in 25 l of transport buffer containing 4 mM ATP and 10 mM MgCl 2 , followed by nonlinear regression analyses. Kinetic parameters of ATP-dependent [ 3 H]E 2 17␤G (0.1-16 M) uptake were determined as described for [ 3 H]LTC 4 except that 5 g of vesicle protein was used, and the reaction was carried out at 37°C.
Confocal Microscopy-Confocal microscopy was carried out as described previously (29). Briefly, ϳ5 ϫ 10 5 stably transfected HEK293 cells were seeded in each well of a six-well tissue culture dish on coverslips. When the cells had grown to confluence, they were washed once in phosphate-buffered saline and then fixed with 2% paraformaldehyde in phosphate-buffered saline, followed by permeabilization using digitonin (0.25 mg/ml in phosphate-buffered saline). MRP1 proteins were detected with the monoclonal antibody MRPm6, which reacts with an epitope close to the COOH terminus of MRP1 (amino acids 1511-1520) (33). Antibody binding was detected with Alexa Fluor 488 antimouse IgG (H ϩ L) F(abЈ) 2 fragment. Nuclei were stained with propidium iodide. Localization of MRP1 in the transfected cells was determined using a Meridian Insight confocal microscope (filter, 620/40 nm for propidium iodide; 530/30 nm for Fluor 488).
Chemosensitivity Testing-Drug resistance was determined using the colorimetric 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay as described previously (8,26,29). Briefly, cells were seeded at 5 ϫ 10 3 cells/well in 100 l of culture medium in 96-well tissue culture plates. The following day, various concentrations of drug diluted in culture medium were added to cells (100 l/well). After incubation for an additional 96 h, 100 l of medium was removed from each well, and the MTT reagent (25 l/well, 2 mg/ml) (Sigma) was added. After 3 h, the formazan was solubilized by mixing with 1 N HCl/isopropyl alcohol (1:24) (100 l/well). Color density was determined using the ELX 800 UV spectrophotometer (570 nm). Mean values of quadruplicate determinations Ϯ S.D. were plotted using GraphPad software. IC 50 values were obtained from the best fit of the data to a sigmoidal curve. Relative resistance is expressed as the ratio of the IC 50 value of cells transfected with MRP1 and mrp1 expression vectors to that of cells transfected with empty vector. Resistance factors were determined in three or more independent experiments.

Transport of [ 3 H]LTC 4 and [ 3 H]E 2 17␤G by Wild Type and Mutant
Murine mrp1-Previous studies demonstrated that the murine/human hybrid protein in which amino acids 1185-1528 of mrp1 were replaced with the corresponding region of MRP1 (mrp1/MRP1-(1188 -1531)) transported E 2 17␤G more efficiently than wild type murine protein with no detectable change in the efficiency of LTC 4 transport (28). To further localize residues in this region responsible for enhancing E 2 17␤G transport activity, amino acids 1351-1528 of mrp1 were replaced with the corresponding part of MRP1. In addition, 9 residues between amino acids 1185 and 1350 of mrp1 that differ from the human protein were substituted with the corresponding amino acid(s) from MRP1 by single, double, or triple mutations (Figs. 1A and 2E). Episomal pCEBV7 expression vectors containing mutated forms of full-length mrp1 cDNAs were transfected into HEK293 cells, and populations of transfected cells were selected in hygromycin B. The resulting stably transfected cell populations were cloned by limiting dilution. Populations derived from clones that expressed high levels of mrp1 mutant proteins were used in all subsequent studies. The levels of mutant proteins relative to wild type mrp1 in previously characterized HEK transfectants were determined by immunoblotting and densitometry as described under "Experimental Procedures" (Figs. 1A and 2E). Under the conditions used for immunoblotting, no endogenous MRP1 was detected in cells transfected with empty vector (data not shown). ATP-dependent transport of [ 3 H]LTC 4 and [ 3 H]E 2 17␤G was examined using plasma membrane vesicles from transfected HEK cells (Figs. 1 and 2). The levels of LTC 4 uptake by vesicles prepared from HEK transfectants expressing wild type and mutant mrp1 were proportional to the relative expression levels of all mutant proteins tested and comparable with that of wild type mrp1 (Figs. 1B and 2A and B). Only two mutations increased the ability of the protein to transport E 2 17␤G, a double mutation in which Ile 1237 and Ala 1239 were replaced with Val and Thr (I1237V/A1239T), and a single conversion of Ala 1239 to Thr (A1239T) (Figs. 1C and 2C). Substitution of Ala 1239 with Thr increased the ability of murine mrp1 to transport E 2 17␤G ϳ3-fold, as did the double mutation I1237V/ A1239T. However, when A1239 was mutated to Cys (A1239C) or Ser (A1239S), neither of the mutations had any effect on E 2 17␤G transport activity (Fig. 2D). These results indicated a highly specific requirement for a Thr residue at position 1239 of mrp1 for efficient transport of the conjugated estrogen.
Mutational Analysis of Thr 1242 in Predicted TM17 of Human MRP1-Based on the findings with the mrp1A1239T mutation, we investigated how specific the requirement was for Thr at position 1242 of the human protein for its ability to transport E 2 17␤G. Thus, Thr 1242 was mutated to Ala, Cys, Ser, Leu, Lys, and Asp. Stably transfected cell populations expressing high levels of MRP1 mutant proteins were isolated as described for mrp1 mutations, and the levels of protein expression were determined by immunoblotting (Fig. 3A). The ability of the mutant proteins to transport LTC 4 and E 2 17␤G was then determined by vesicle transport assays (Figs. 3 and 4). After normalization for differences in expression levels, substitution of Thr 1242 with Ala, Cys, Ser, Leu, and Lys decreased the ability of MRP1 to transport E 2 17␤G by more than 2-fold relative to the wild type protein with no significant effect on LTC 4  transport. However, conversion of Thr 1242 to a negatively charged amino acid, Asp, essentially eliminated transport of E 2 17␤G and also decreased LTC 4 transport ϳ2-fold (Figs. 3 and  4). These data support those obtained from studies of mrp1A1239 mutations and confirm the importance of Thr 1242 in human MRP1 for efficient transport of E 2 17␤G.

FIG. 2. Time course of ATP-dependent [ 3 H]LTC 4 and [ 3 H]E 2 17␤G uptake by membrane vesicles prepared from
Comparison of the Trafficking of Mutant Human Proteins with Wild Type MRP1 in Transfected HEK293 Cells-To determine whether the decrease in E 2 17␤G transport observed following mutation of Thr 1242 , particularly the influence of substituting Thr with an acidic amino acid such as Asp, might be attributable to changes in trafficking of the protein, we compared the subcellular localization of wild type and mutant human proteins by confocal microscopy. As shown in Fig. 5, cells expressing mutations MRP1T1242A, MRP1T1242C, MRP1T1242D, and MRP1T1242K showed a pattern of strong plasma membrane staining indistinguishable from that of cells expressing wild type MRP1, indicating that the trafficking was unaffected.

Kinetic Parameters of [ 3 H]LTC 4 and [ 3 H]E 2 17␤G
Transport-We have shown that interconversion of Ala 1239 in mrp1 and Thr 1242 in MRP1 affected the ability of the proteins to transport E 2 17␤G with no apparent effect on LTC 4 transport.
To determine the influence of these mutations on transport more precisely, we compared K m and V max values for the wild type murine and human proteins with those of mutant mrp1A1239T and MRP1T1242A (Fig. 6). For wild type mrp1 and mrp1A1239T, the K m values and normalized V max for LTC 4 were essentially identical ( Fig. 6A and Table I). For E 2 17␤G transport, K m values were 2.0 and 1.3 M for wild type mrp1 and mrp1A1239T, respectively, and the normalized V max value for mrp1A1239T was ϳ2-fold higher than that for wild type mrp1 (Table I).
The kinetic parameters of ATP-dependent LTC 4 and E 2 17␤G transport were also examined for the wild type and mutant human proteins (Fig. 6, C and D). The K m and normalized V max for LTC 4 transport obtained with vesicles containing wild type MRP1 or mutant MRP1T1242A were essentially identical (Table I). Consistent with the results obtained with the mrp1 mutant, substitution of Thr 1242 with Ala in MRP1 decreased the normalized V max value for the mutant ϳ2-fold relative to wild type MRP1. Apparent K m values for wild type protein and MRP1T1242A were 0.5 and 0.7 M, respectively (Table I). Thus, the mrp1A1239T and MRP1T1242A mutations increased or decreased the V max /K m ratio for E 2 17␤G 3-fold, respectively ( Table I).
Effect of Mutation mrp1A1239T and MRP1T1242A on the Inhibition of mrp1/MRP1-mediated LTC 4 Transport by E 2 17␤G-As an alternative means of assessing the effects of the TM17 mutations on the interaction between E 2 17␤G and the human and murine proteins, we examined the ability of the conjugated estrogen to inhibit transport of LTC 4 . IC 50 values for wild type and mutant murine and human proteins were obtained from the best fit of the inhibition data to a sigmoidal curve (Fig. 7). For the wild type murine protein, the IC 50 value for E 2 17␤G was 127 M compared with 27 M for mutation mrp1A1239T. In MRP1, substitution of Thr 1242 with Ala increased the IC 50 value from 15 to 134 M. These results are independent of protein expression levels and provide strong evidence that the increase or decrease in E 2 17␤G transport by mrp1A1239T and MRP1T1242A, respectively, is at least partially attributable to changes in the affinity of the proteins for this substrate.
Resistance Profiles of Wild Type and TM17 Mutant Murine/ Human Proteins-In addition to the differences in E 2 17␤G transport activity, the hybrid protein mrp1/MRP1-(1188 -1531) has been shown to increase the resistance to anthracyclines ϳ2-fold relative to wild type mrp1 (28). Consequently, we ex- FIG. 5. Confocal microscopy of HEK transfectants expressing wild type and mutant MRP1 proteins. Cells were grown and stained for immunofluorescence detection of MRP1 as described under "Experimental Procedures." MRP1 was detected using mAb MRPm6, which reacts with a cytoplasmic continuous epitope near the COOH terminus MRP1 (amino acids 1511-1520) (31). The location of MRP1 is indicated in green. Nuclei were stained with propidium iodide and are shown in red. Transfectants tested were expressing wild type or mutant MRP1 as indicated. D) uptake by membrane vesicles prepared from HEK293 cells transfected with wild type or mutant proteins was measured at various LTC 4 concentrations (0.01-2 M) for 1 min at 23°C or at various E 2 17␤G concentrations (0.1-16 M) for 1 min at 37°C in transport buffer, as described under "Experimental Procedures." Values shown are means Ϯ S.D. of triplicate determinations in a single experiment. Data were plotted as V 0 versus [S] to confirm that concentration range selected was appropriate to observe both zero-order and first-order rate kinetics. The transfectants tested were HEK mrp1 (f), HEK mrp1A1239T (OE) (A and B), HEK MRP1 (), and HEK MRP1T1242A (q) (C and D). Kinetics parameters for LTC 4 or E 2 17␤G transport were determined by nonlinear regression analysis of the combined data using GraphPad software and are shown in Table I. amined the drug resistance profiles of transfectants expressing mrp1 mutations A1239T, A1239S, and A1239C. The results are summarized as relative resistance factors in Table II. None of the mutations increased resistance to either doxorubicin or epirubicin. However, they all decreased the ability of the murine protein to confer vincristine resistance by 6 -7-fold and reduced the resistance to VP-16 by 40 -45% (Table II).

FIG. 6. Kinetics of ATP-dependent [ 3 H]LTC 4 and [ 3 H]E 2 17␤G transport by wild type and mutant proteins. The initial rate of ATP-dependent [ 3 H]LTC 4 (A and C) and [ 3 H]E 2 17␤G (B and
When comparable studies were carried out with the human mutant proteins, MRP1T1242A, MRP1T1242C, MRP1T1242S, MRP1T1242L, MRP1T1242D, and MRP1T1242K, we found that these mutations also decreased the ability of the protein to confer vincristine and VP-16 resistance by ϳ40% and reduced the ability of MRP1 to confer resistance to doxorubicin or epirubicin by 2-3-fold (Table III). Thus, in both the human and murine proteins, Thr 1242 and Ala 1239 , respectively, are involved in conferring resistance to vincristine and VP-16 and, in MRP1, also resistance to anthracyclines.
Effect of Double Mutations in TM14 and TM17 of mrp1 and MRP1 on Transport Activity and Drug Resistance Profiles-We demonstrated previously that converting Gln 1086 in mrp1 to Glu, as it is in MRP1, increased the relative resistance to doxorubicin 4 -5-fold. However, this mutation had no influence on the ability of the protein to transport LTC 4 and E 2 17␤G (29). Given the observed effect of the mrp1A1239T and MRP1T1242A mutations on drug resistance, including, in the case of MRP1T1242A, resistance to anthracyclines, we exam-ined whether the mutations in TM helices 14 and 17 had any combined effect on transport or drug specificity. To do so, we made a double mutation in mrp1, in which Gln 1086 was replaced with Glu and Ala 1239 was substituted with Thr (Q1086E/A1239T). The double mutant was stably expressed in HEK293 cells, and the transport of both LTC 4 and E 2 17␤G by membrane vesicles was examined (Fig. 8, A, C, and D). The levels of uptake of the two substrates by vesicles containing either mrp1A1239T or mrp1Q1086E/A1239T were proportional to the relative expression levels of the proteins (Fig. 8, A, C, and  D). Thus, in the murine protein, mutation Q1086E has no significant effect on the ability of mutant mrp1A1239T to transport either LTC 4 or E 2 17␤G.
We had also found previously that substitution of Glu 1089 in MRP1 with Gln essentially eliminated resistance to anthracyclines without affecting either LTC 4 or E 2 17␤G transport. However, unlike the reciprocal mutation in mrp1, which had no detectable effect on vincristine and VP-16 resistance, replacement of Glu 1089 with uncharged amino acids decreased the capacity of the protein to mediate vincristine and VP-16 resistance by ϳ60 and ϳ40%, respectively (29). Consequently, a double mutation of MRP1 was also made, in which Glu 1089 was replaced with Gln and Thr 1242 was substituted with Ala (E1089Q/T1242A). Stably transfected cell populations expressing high levels of the mutant protein were isolated, and the levels of protein expression were determined by immunoblotting (Fig. 8B). As in the case of mrp1, the second mutation in TM14 of MRP1 had no additional effect on transport of either LTC 4 or E 2 17␤G over and above that observed with the single TM17 mutation (Fig. 8, E and F). Trafficking of the double mutation was also comparable with that of wild type MRP1 (Fig. 5).
In contrast to the lack of effect of the second reciprocal mutation in TM14 on transport of E 2 17␤G by either mrp1A1239T or MRP1T1242A, introduction of the TM14 mutations into both proteins markedly affected their ability to confer drug resistance. Mutation mrp1Q1086E/A1239T increased the capacity of the mutant protein mrp1A1239T to confer resistance to both vincristine and VP-16 (Table II). After normalizing for differences in expression levels, the double mutation increased resistance to doxorubicin and epirubicin 5-fold (Table II), as observed previously with the single mutation mrp1Q1086E (29). Similarly, introduction of the E1089Q mutation into MRP1T1242A increased resistance to vincristine and VP-16 despite the fact that the single TM14 mutation in the human protein has been shown to decrease resistance to both of these drugs (29). However, the E1089Q mutation eliminated the ability of MRP1T1242A to confer resistance to anthracyclines (Table III), as observed previously when this amino acid substitution was introduced into wild type MRP1 as a single mutation (29). These results indicate a highly specific requirement for the combination of Gln 1086 and Ala 1239 in mrp1 and Glu 1089 and Thr 1242 in MRP1 for the ability of both proteins to confer resistance to vincristine and VP-16.  4

and E 2 17␤G uptake by vesicles from HEK cells transfected with vectors encoding wild type and mutant proteins
The kinetic parameters of LTC 4 and E 2 17␤G uptake were determined as described in the legend to Fig. 6. The normalized V max values were obtained by adjusting determined V max values to compensate for differences in the relative levels of the wild type and mutant proteins.

DISCUSSION
Human MRP1, like P-gp, confers resistance to many hydrophobic natural product chemotherapeutic agents including Vinca alkaloids, epipodophyllotoxins, and anthracyclines (7-10). However, unlike P-gp, MRP1 is able to transport many relatively hydrophilic organic anion conjugates including the two well characterized possible physiological substrates, LTC 4 and E 2 17␤G (16,18,19). We have recently shown that Glu 1089 in predicted TM14 of human MRP1 is critical for the ability of the protein to confer anthracycline resistance and is also involved in mediating vincristine and VP-16 resistance. In contrast, mutation of this or the corresponding residue in mrp1 has no effect on transport of either LTC 4 or E 2 17␤G (29). Amino acid(s) that selectively affect transport of organic anion conjugates without altering the drug resistance profile conferred by mrp1 or MRP1 have not been identified. However, we have shown recently that mutation of a highly conserved tryptophan residue at position 1246 in TM17 of MRP1 that eliminates the ability of the protein to confer drug resistance also abolishes E 2 17␤G transport but leaves LTC 4 transport and verapamilstimulated GSH transport relatively intact (30).
Earlier comparisons of the substrate specificity of mrp1 and MRP1 indicated that the efficiency of LTC 4 transport by the two proteins was similar but that mrp1 was a very poor transporter of E 2 17␤G (26). Previous studies with mrp1/MRP1 hy-brids suggested that one of the most highly conserved regions of the two proteins including TM16 and TM17 extending to the COOH terminus contains variant residues important for the transport of this substrate (28). We have now examined the consequences of converting all nonconserved amino acid residues in mrp1 in the region between amino acids 1185 and 1528 to the corresponding amino acid present in MRP1 to identify specific residue(s) involved in the efficient transport of E 2 17␤G. We have also determined whether the same residue(s) influence the drug resistance profile of the protein.
Replacement of the variant residues in the cytoplasmic COOH-terminal region of mrp1 with those present in the human protein had no significant effect on its ability to transport either LTC 4 or E 2 17␤G. In contrast, substitution of Ala 1239 in TM17 with Thr, as it is in MRP1, increased the ability of the protein to transport E 2 17␤G 2-3-fold without affecting transport of LTC 4 . Consistent with the findings obtained with mrp1 mutations, the reciprocal mutation in MRP1 confirmed the importance of Thr 1242 for E 2 17␤G transport. Replacement of Thr 1242 with Ala, Ser, Cys, Leu, and Lys decreased the ability of the protein to transport E 2 17␤G 2-3-fold without significantly affecting LTC 4 transport. However, introduction of an acidic residue at this location not only abolished E 2 17␤G transport but also reduced LTC 4 transport 2-fold. Of all mutants tested, only the mrp1A1239T mutation increased E 2 17␤G The resistance of HEK293 cells transfected with expression vectors encoding wild type and mutant mrp1 relative to that of cells transfected with empty vector was determined using a tetrazolium salt-based microtiter plate assay. Data were analyzed as described under "Experimental Procedures." The relative resistance factor was obtained by dividing the IC 50 values for wild type/mutant mrp1-transfected cells by the IC 50 value for control transfectants. The values shown represent the mean Ϯ S.D. of relative resistance values determined from 3-6 independent experiments. Resistance factors normalized for differences in the levels of mutant proteins expressed in the transfectant populations used are shown in parentheses.  The resistance of HEK293 cells transfected with expression vectors encoding wild type and mutant MRP1 relative to that of cells transfected with empty vector was determined as described in Table II transport. In contrast, mutation of Ala 1239 to Cys, Ser, and Thr decreased the ability of mrp1 to confer vincristine and VP-16 resistance. Since wild type mrp1 does not confer resistance to anthracyclines, the effect of these mutations on resistance to this class of drugs could not be assessed. Replacement of the comparable amino acid residue in MRP1, Thr 1242 , with Ala, Cys, Ser, Leu, Lys, or Asp also decreased the ability of the protein to confer resistance to vincristine and VP-16 and, in this case, the anthracyclines tested. The effects of the reciprocal Ala/Thr substitutions on drug resistance of both mrp1 and MRP1 were unexpected, because exchanging MSD3 and the COOH-terminal regions of the two proteins had no effect on resistance to vincristine and VP-16 (28). The consequences of mutating mrp1A1239 and MRP1T1242, with the exception of the T1242D mutation, are qualitatively similar to those we observed following mutation of the highly conserved Trp 1246 , which abolished drug resistance and E 2 17␤G transport (30). Together, our results are also consistent with a direct involvement of residues in TM17 in substrate binding, as suggested by recent photoaffinity labeling studies in which cross-linking to a proteolytic fragment containing TM16 and TM17 of MRP1 was observed (34). The topologically comparable transmembrane helix in P-gp, due to the five additional NH 2 -proximal TMs in MRP1/mrp1, is TM12, which has been demonstrated to play an important role in substrate specificity and drug binding. Triple substitution of the nonconserved residues, Leu 975 , Val 981 , and Phe 983 , with Ala collectively abrogates drug transport, drug-stimulated ATP hydrolysis, and photoaffinity labeling with the drug analogue, [ 125 I]iodoarylazidoprazosin, while having minimal effect on [␣-32 P]8-azido-ATP labeling and basal ATPase activity of the protein (35). Use of the thiol-reactive substrate, dibromobimane, together with cysteine-scanning mutagenesis has confirmed that hydrophobic residues Leu 975 , Val 982 , and Ala 985 in TM12 are important for the interaction of substrates with P-gp (36,37). In addition, mutation F978A in TM12 of Pgp decreases resistance to colchicine and doxorubicin with little effect on resistance to vinblastine or actinomycin D (38).
The mechanism by which various substitutions of the threonine residue at position 1242 in MRP1 and the alanine residue at position 1239 in mrp1 affect E 2 17␤G transport and drug resistance is presently unclear. It has been proposed that MRP1 binds its substrates via hydrogen bond formation with the electron donor (or hydrogen bond acceptor) groups of the substrate (39). The helical wheel projection for MRP1 TM17 reveals a highly amphipathic character with amino acid residues possessing hydrogen bond donor and acceptor side chains arrayed preferentially on one side of the helix and amino acid residues with non-hydrogen-bonding side chains on the other side (30,39). However, substitution of Thr 1242 with either Cys or Ser, which retain hydrogen bonding capability, decreased the ability of the protein to transport E 2 17␤G and to confer drug resistance. Their effect was similar to that of the MRP1T1242A mutation, suggesting that the size of the side chain, in addition to hydrogen bonding capability, is also critical for retaining substrate specificity and transport capacity. That the exchange of Ala and Thr affects the affinity of the mouse and human proteins for E 2 17␤G is supported by the results of experiments in which we examined the ability of the estrogen conjugate to inhibit LTC 4 transport. We have shown previously that E 2 17␤G competitively inhibits LTC 4 transport by MRP1 with a K i of 22 M (18), which is in reasonable agreement with the IC 50 value of 15 M determined in this study. In contrast, the IC 50 for LTC 4 transport by MRP1 T1242A was ϳ130 M, a value similar to that obtained for wild type mrp1, while the IC 50 of mrp1A1239T decreased to ϳ30 M.
It has been proposed that the transport of anionic/cationic substrates by MRP1 is facilitated by cationic/anionic acid residues present in the transmembrane helices (39). We have shown previously that an acidic amino acid at position 1089 in TM14 of MRP1 is critical for the ability of the protein to confer resistance to anthracyclines, which are cationic at physiological pH, and that substitution with a positively charged amino acid essentially eliminates drug resistance (29). Similarly, the replacement of basic residues (Lys 332 , Lys 483 , Arg 1210 , and Arg 1257 ) in the predicted transmembrane domains in the related human protein, MRP2, with Ala has been reported recently to decrease the ability of the protein to transport an organic anion substrate, glutathione-methylfluorescein (40). It has also been reported recently that the charged amino acids in transmembrane helices of rat mrp2 may play an important role in the recognition and/or transport of its conjugated substrates. Substitution of Lys 325 with Met or Arg 586 with Leu markedly decreases the ability of rat mrp2 to transport the glutathione conjugates, 2,4-dinitrophenyl-S-glutathione and LTC 4 (41). However, we found in this study that mutation of Thr 1242 to Lys did not enhance LTC 4 transport and actually decreased transport of E 2 17␤G. In addition, mutation of Thr 1242 to either a negatively or positively charged amino acid similarly decreased the ability of the protein to confer resistance to all of the drugs tested including the cationic anthracyclines. Taken together, these findings indicate both a strong restriction on the size of the residue side chain at position 1242 and a re-

FIG. 8. Time course of ATP-dependent [ 3 H]LTC 4 and [ 3 H]E 2 17␤G uptake by membrane vesicles prepared from
HEK293 stable transfectants expressing wild type mrp1/MRP1 or TM14/17 mutant proteins. The relative expression levels of wild type and mutant mrp1/MRP1 proteins in the membrane vesicles were determined as described in the legend to Fig. 1 (A and B). The murine proteins were determined by immunoblotting with mAb MRPr1, and the human proteins were detected with mAb QCRL-1 as described under "Experimental Procedures." Similar results were obtained from at least three replicate experiments. [ 3 H]LTC 4 (C and E) and [ 3 H]E 2 17␤G (D and F) uptake was determined as described in the legend to Fig. 2. Transfectants tested were HEK mrp1 (f), HEK mrp1A1239T (OE), and HEK mrp1Q1086E/A1239T (q) (C and D) and HEK MRP1 (Ⅺ), HEK MRP1T1242A (‚), and HEK MRP1E1089Q/T1242A (E) (E and F). quirement for hydrogen bonding capacity for the protein to function as an efficient transporter of relatively bulky amphipathic, sterically rigid drugs and E 2 17␤G, compared with the smaller more structurally flexible substrate, LTC 4 . Only substitution of Thr 1242 with Asp affected transport of both E 2 17␤G transport and, albeit to a lesser extent, LTC 4. One possible explanation for this observation is that the interaction of LTC 4 with other residues in a common substrate-binding pocket of the protein, such as Trp 1246 (30), may be affected by charge repulsion from the aspartate side chain.
Having found previously that mutation of Glu 1089 in TM14 of MRP1 to Gln not only markedly reduced that ability of the protein to confer resistance to anthracyclines but also, to a lesser extent, to vincristine and VP-16 (29), we investigated the effect of combining this mutation, and the reciprocal mutation in the murine protein, with the A1239T and T1242A mutations of mrp1 and MRP1, respectively. We observed that a reciprocal exchange of the naturally occurring residue in TM14 of mrp1 and MRP1 introduced as a second mutation completely rescued the effect of the TM17 mutations on vincristine and VP-16 resistance. Previously, we have assumed that the similar levels of resistance that wild type mrp1 and MRP1 confer to these drugs reflected the high degree of structural conservation between the two proteins. Our current findings indicate that this is not the case and that compensatory mutations in nonconserved residues have occurred, which have allowed both proteins to retain comparable abilities to transport some shared substrates. They also suggest that Glu 1089 and Thr 1242 in MRP1 and Gln 1086 and Ala 1239 of mrp1 may form direct contacts for interaction with vincristine and VP-16 in substrate binding pockets on the human and murine proteins, respectively. The fact that creation of these double mutants did not alter the LTC 4 and E 2 17␤G transport activity of the single mutations in TM17 or change the anthracycline resistance observed previously with the single TM14 mutations provides strong supporting evidence for a model in which different substrates interact with different, but partially shared sets of determinants in a common binding pocket on the protein (15, 16, 18, 21, 22, 28 -30).