Charged Amino Acids in the Sixth Transmembrane Helix of Multidrug Resistance Protein 1 (MRP1/ABCC1) Are Critical Determinants of Transport Activity*

The multidrug resistance protein, MRP1 (ABCC1), is an ATP-binding cassette transporter that confers resistance to chemotherapeutic agents. MRP1 also mediates transport of organic anions such as leukotriene C4 (LTC4), 17β-estradiol 17-(β-d-glucuronide) (E217βG), estrone 3-sulfate, methotrexate (MTX), and GSH. We replaced three charged amino acids, Lys332, His335, and Asp336, predicted to be in the sixth transmembrane (TM6) helix of MRP1 with neutral and oppositely charged amino acids and determined the effect on substrate specificity and transport activity. All mutants were expressed in transfected human embryonic kidney cells at levels comparable with wild-type MRP1, and confocal microscopy showed that they were correctly routed to the plasma membrane. Vesicular transport studies revealed that the MRP1-Lys332 mutants had lost the ability to transport LTC4, and GSH transport was reduced; whereas E217βG, estrone 3-sulfate, and MTX transport were unaffected. E217βG transport was not inhibited by LTC4 and could not be photolabeled with [3H]LTC4, indicating that the MRP1-Lys332 mutants no longer bound this substrate. Substitutions of MRP1-His335 also selectively diminished LTC4 transport and photolabeling but to a lesser extent. Kinetic analyses showed that V max(LTC4) of these mutants was decreased butK m was unchanged. In contrast to the selective loss of LTC4 transport in the Lys332 and His335 mutants, the MRP1-Asp336 mutants no longer transported LTC4, E217βG, estrone 3-sulfate, or GSH, and transport of MTX was reduced by >50%. Lys332, His335, and Asp336 of TM6 are predicted to be in the outer leaflet of the membrane and are all capable of forming intrahelical and interhelical ion pairs and hydrogen bonds. The importance of Lys332 and His335 in determining substrate specificity and of Asp336 in overall transport activity suggests that such interactions are critical for the binding and transport of LTC4 and other substrates of MRP1.

Multidrug resistance is responsible for significantly limiting the effectiveness of many anti-cancer drugs. In experimental model systems, multidrug resistance in vivo and in vitro is characterized by cross resistance to a broad range of cytotoxic drugs that may have little structural similarity to one another and may exert their cytotoxic effects through different cellular pathways. Multidrug resistance in human tumor cells is often associated with enhanced ATP-dependent drug efflux attributed to elevated expression of some members of the ATPbinding cassette (ABC) 1 superfamily of transporter proteins. These include MRP1 (gene designation ABCC1), P-glycoprotein (gene ABCB1), and BCRP or MXR (gene ABCG2) (1)(2)(3)(4)(5)(6)(7).
The 190-kDa multidrug resistance protein, MRP1, is frequently overexpressed in drug-resistant tumor cell lines and is also expressed in a wide variety of human tumors. The xenobiotics transported by MRP1 range from complex heterocyclic natural products and chemotherapeutic agents such as vincristine, doxorubicin, and the folate antagonist methotrexate to arsenical and antimonial oxyanions (3,8). MRP1 has also been shown to be an efficient ATP-dependent transporter of various conjugated organic anions, including a mediator of inflammation, the cysteinyl leukotriene LTC 4 , the cholestatic glucuronide conjugated estrogen E 2 17␤G, and the sulfate conjugate estrone 3-sulfate (8,9). To transport unconjugated drugs such as vincristine, MRP1 requires the presence of reduced GSH, or analog, which appears to be co-transported with the drug (10 -13).
MRP1 belongs to subfamily C of the ABC transporter superfamily, which, in addition to MRP2-7, includes the cystic fibrosis transmembrane conductance regulator (CFTR) and the sulfonylurea receptors, SUR1 and SUR2 (8,14). Some members of the ABCC subfamily (MRP4, MRP5, CFTR, ABCC11, and ABCC12) have a typical four domain ABC transporter structure with two membrane-spanning domains (MSDs) and two nucleotide binding domains (8,14,15). However, the predicted topologies of the remaining members of this subfamily (MRP1, MRP2, MRP3, MRP6, and MRP7 as well as the K ϩ channel regulators, SUR1 and SUR2), together in some cases with biochemical studies, indicate that they contain an additional NH 2 -terminal MSD with five TM segments and an extracytosolic NH 2 terminus (16 -19). Thus, these ABCC transporters are predicted to have 17 TM segments, assumed to be ␣-helices, distributed among three MSDs: (MSD1, TM1-5; MSD2, TM6 -11; and MSD3, TM12-17) (Fig. 1A). Overall, MRP2, MRP3, and MRP6 are the most closely related to MRP1 with respect to their structure and ability to transport conjugated organic anions. Like MRP1, MRP2 and MRP3 can also confer resistance to a variety of cytotoxic drugs (9, 19 -25).
It has previously been noted that MRP1 contains a signifi-cantly greater number of potentially charged amino acids in its predicted TM ␣-helices than P-glycoprotein (26). Since many of the molecules recognized and transported by MRP1 contain a relatively large hydrophobic domain as well as a hydrophilic domain with at least one anionic or cationic charge at physiological pH, it seems reasonable to suggest that charged amino acids in the TM segments of this protein might well play a role in determining its substrate specificity and transport activity (26 -30). Hydropathy analyses predict that the first TM helix of MSD2, TM6, spans amino acids 320 -340 and contains three potentially charged amino acids (Fig. 1). In the present study, we have mutated TM6 Lys 332 , His 335 , and Asp 336 , which are predicted to cluster in the exoplasmic leaflet of the membrane, and show that replacing them individually with a neutral or oppositely charged amino acid markedly affects the transport activity of MRP1. Importantly, we show that substitutions of Lys 332 and to a lesser extent His 335 selectively decrease the binding and transport of LTC 4 and GSH while leaving the transport of other organic anions unchanged. In contrast, substitutions of Asp 336 show no such selectivity, since mutation of this amino acid markedly reduces or eliminates transport of all MRP1 substrates and thus appears critical for the overall activity of the protein.
Generation of MRP1-GFP Fusion Constructs-The 238-amino acid coding sequence of jellyfish GFP with Ser 65 3 Thr was cloned in frame to the 3Ј-end of MRP1 to generate the wild-type construct pcDNA3.1(Ϫ)-MRP1 K -GFP as described (31). Mutant MRP1-GFP fusion proteins were generated by replacing the 1-kb BamHI/Bsu36I fragment in the pcDNA3.1(Ϫ)-MRP1 K -GFP construct with the comparable fragment containing the desired altered cDNA sequence.
Transfections with MRP1 Expression Vectors-Constructs containing wild-type and TM6 mutant pcDNA3.1(Ϫ)-MRP1 expression vectors were transfected into SV40-transformed human embryonic kidney cells (HEK293T). Briefly, ϳ7 ϫ 10 6 cells were seeded in 150-mm plates, and 24 h later, DNA (16 g) was added using FuGENE™6 (Roche Diagnostics, Laval, Canada) according to the manufacturer's instructions. After 72 h, the HEK293T cells were harvested, and inside-out membrane vesicles were prepared as described previously (32). Empty vector and vector containing the wild-type cDNAs were included as controls in all experiments. Levels of wild-type and TM6 mutant MRP1 proteins were determined by immunoblotting as described below.
Measurement of MRP1 Protein Levels in Transfected Cells-The expression levels of wild-type and TM6 mutant MRP1 proteins were determined by immunoblot analysis of membrane protein fractions from transfected cells essentially as described (33). Proteins were resolved on a 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 the human MRP1-specific murine monoclonal antibody QCRL-1 (diluted 1:10,000), which recognizes a linear epitope consisting of amino acids 918 -924 (34). After washing, blots were incubated with horseradish peroxidase-conjugated goat antimouse antibody (Pierce) followed by application of Renaissance ® chemiluminescence blotting substrate (PerkinElmer Life Sciences). Relative levels of MRP1 expression were estimated by densitometric analysis using a ChemiImager™ 4000 (Alpha Innotech, San Leandro, CA).
Confocal Microscopy-HEK293T cells were seeded at 3.5 ϫ 10 5 cells/ well in a six-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™6 as before. Forty-eight h 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 Ϫ1 in 0.1% Triton X-100 plus 1% bovine serum albumin in PBS) for 60 min at room temperature. The coverslips were washed again, and cell nuclei were stained in 1 ml of propidium iodide (2 g ml Ϫ1 in PBS) for 45 min in the dark. Finally, the coverslips were placed on slides containing one drop of Antifade Solution (Molecular Probes, Inc., Eugene, OR), and cells were examined using a Meridien InSight Plus confocal microscope equipped with an air-cooled argon laser. Images obtained with 488-nm excitation were pseudocolored and overlaid using Maxim DL software.

MRP1-mediated Transport of 3 H-Labeled
Substrates by Inside-out Membrane Vesicles-Preparation of inside-out membrane vesicles from transfected HEK293T cells has been described previously (32), and ATP-dependent uptake of 3 H-labeled substrates by the membrane vesicles was measured using a rapid filtration technique also as described previously (10). Briefly, LTC 4 transport assays were performed at 23°C in a 50-l reaction containing 50 nM LTC 4 (50 nM; 40 nCi), 4 mM AMP or ATP, 10 mM MgCl 2 , creatine phosphate (10 mM), creatine kinase (100 g ml Ϫ1 ), and 2 g of vesicle protein. Uptake was stopped at selected times by rapid dilution in ice-cold buffer, and then the incubation mixture was filtered through glass fiber (Type A/E) filters 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 ATPdependent LTC 4 uptake. All transport assays were carried out in triplicate, and results were expressed as means Ϯ S.D.
Uptake of E 2 17␤G was measured in a similar fashion except that membrane vesicles (2 g of protein) were incubated at 37°C in a total reaction volume of 50 l containing [ 3 H]E 2 17␤G (400 nM; 40 nCi) and components as described for [ 3 H]LTC 4 transport. ATP-dependent transport of estrone 3-sulfate into membrane vesicles was measured as described above for [ 3 H]E 2 17␤G except that the initial substrate concentration was 300 nM (200 nCi) [ 3 H]estrone 3-sulfate, and the reaction volume was 60 l containing 3 mM GSH and 10 mM dithiothreitol. GSH uptake was also measured by rapid filtration with membrane vesicles (20 g of protein) incubated at 37°C for 20 min in a 60-l reaction volume with 100 M [ 3 H]GSH (300 nCi/reaction). 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 apigenin (30 M) (32). MTX uptake was also measured as described previously (35). Assays were carried out at 37°C in a 60-l reaction volume containing [ 3 H]MTX (100 M; 200 nCi), membrane vesicles (10 g of protein), and other components as above. Uptake was stopped after 20 min by rapid dilution in ice-cold buffer and processed as before.
Kinetic Analysis of ATP-dependent [ 3 H]LTC 4 Transport-K m and V max values of ATP-dependent LTC 4 transport by membrane vesicles (2.5 g of protein) were determined by measuring uptake at eight different [ 3 H]LTC 4 concentrations (10 -1000 nM) for 1 min at 23°C in 50 l of transport buffer containing components as described above. Data were analyzed using Graph Pad Prism™ software and kinetic parameters determined by nonlinear regression analyses and Michaelis-Menten analysis. 4 -Wild-type and TM6 mutant MRP1 membrane proteins were photolabeled with [ 3 H]LTC 4 essentially as described (10). Briefly, vesicles prepared from HEK293T cells transfected with wild-type and TM6 mutant MRP1 cDNAs (75 g of protein in 50 l) were incubated with [ 3 H]LTC 4 (0.5 Ci; 200 nM) at room temperature for 30 min and then frozen in liquid nitrogen. Samples were then alternately irradiated at 302 nm for 1 min using a CL-1000 ultraviolet cross-linker (DiaMed, Mississauga, Canada) and snapfrozen in liquid N 2 10 times. Radiolabeled proteins (75 g) were resolved by SDS-PAGE. The gel was fixed (isopropyl alcohol/water/acetic acid; 25:65:10) and then soaked in Amplify (Amersham Biosciences) for 30 min. After drying for 2 h, the gel was exposed to Kodak X-Omat film for 1 week at Ϫ70°C.

Expression of MRP1 in Transfected HEK293T Cells Is Not Affected by Substitutions of Charged Amino Acids in TM6 -
Site-directed mutagenesis was performed on the three charged amino acids found in predicted TM6 of MRP1. An alignment of the amino acid sequence of MRP1 TM6 and the corresponding sequences in related ABC-C proteins revealed that Asp 336 is the most highly conserved of the three amino acids, whereas His 335 is the least conserved (Fig. 1B). An ␣-helical wheel projection of putative TM6 shows that it is highly amphipathic and, furthermore, that all three charged amino acids are on the same face of the helix within an arc of ϳ100°and cluster in the outer leaflet of the membrane (Fig. 1C). Lys 332 was replaced with a nonpolar neutral amino acid (Leu) to eliminate charge and with Asp to introduce the opposite charge. Similarly, Asp 336 was replaced with Leu as well as Arg. Histidine, although not positively charged at physiological pH, has a strong basic character. Consequently, in addition to substituting His 335 with Leu, it was replaced with a polar neutral amino acid (Gln) and an acidic Glu residue.
Expression of the wild-type and TM6 mutant MRP1 proteins was confirmed by immunoblot analysis. In all cases, the MRP1specific monoclonal antibody QCRL-1 detected a single band of 190 kDa in membrane vesicles prepared from transfected HEK293T cells (Fig. 3A). Densitometric analysis showed that all mutants were expressed at levels comparable with that of wild-type MRP1 (80 -100%). The GFP fusion proteins of wildtype MRP1 and the TM6 MRP1 mutants were also expressed at similar levels (data not shown).
All TM6 Mutant MRP1 Proteins Are Correctly Routed to Plasma Membrane of Transfected HEK293T Cells-To determine whether the substitutions of Lys 332 , His 335 , and Asp 336 impaired trafficking of MRP1 to the plasma membrane, the subcellular localization of the MRP1 TM6 mutants was compared with wild-type MRP1 by confocal laser-scanning fluorescence microscopy. For these experiments, GFP-tagged constructs encoding wild-type MRP1 and TM6 mutant MRP1 proteins were generated and transfected into HEK293 cells. When viewed under the confocal microscope, both wild-type and mutant MRP1 proteins exhibited an exclusively plasma membrane localization, confirming that the mutant proteins were correctly routed to the cell surface. Representative confocal micrographs of cells expressing GFP-tagged wild-type MRP1 and mutants K332D, H335E, and D336R are shown in Fig. 2.
[ 3 H]LTC 4 Uptake Is Reduced or Eliminated by Substitution of Charged Amino Acids in TM6 of MRP1-To determine the effect of the TM6 mutations on MRP1 transport activity, a time course of ATP-dependent [ 3 H]LTC 4 uptake was performed using inside-out membrane vesicles prepared from transfected HEK293T cells that expressed comparable levels of MRP1 proteins (Fig. 3A). In the case of the MRP1-Lys 332 mutants K332D and K332L (Fig. 3B) and the MRP1-Asp 336 mutants D336L and D336R (Fig. 3D), LTC 4 uptake was reduced to levels that were indistinguishable from those observed with vesicles prepared from empty vector-transfected control cells. LTC 4 uptake by the MRP1-His 335 mutants was also reduced but to a lesser degree. Thus H335E and H335L and H335Q all transported LTC 4 at levels that were ϳ50 -60% of wild-type MRP1 uptake levels (Fig. 3C). These results show that LTC 4 transport activity of MRP1 is adversely affected by both charged and neutral changes to all three charged amino acids of TM6.   uptake by these mutants (Fig. 5). The Lys 332 mutants (Fig. 5A) and His 335 mutants (Fig. 5B) exhibited the same levels of E 2 17␤G uptake as wild-type MRP1. In contrast, substitution of Asp 336 with either Leu or Arg led to a complete loss of E 2 17␤G uptake. Thus, as shown in Fig. 5C, MRP1-Asp 336 mutants D336L and D336R exhibited the same E 2 17␤G uptake activity as the empty vector control. These observations indicate that substitutions of Lys 332 and His 335 selectively reduce LTC 4 transport but leave E 2 17␤G uptake intact, whereas replacing Asp 336 eliminates MRP1-mediated transport of both conjugated organic anions altogether.
Reciprocal and competitive inhibition of LTC 4 and E 2 17␤G transport has been reported previously (10,36). Consequently, to determine whether E 2 17␤G transport by the TM6 mutant MRP1 proteins that either no longer transport LTC 4 or exhibit reduced LTC 4 transport activity could still be inhibited by this cysteinyl leukotriene, [ 3 H]E 2 17␤G uptake by the Lys 332 and His 335 mutants in the presence of LTC 4 was examined. As shown in Fig. 6A, LTC 4 at a concentration of 2 M completely inhibited [ 3 H]E 2 17␤G uptake by wild-type MRP1 as expected. In contrast, LTC 4 had very little effect (Ͻ15%) on E 2 17␤G uptake by MRP1 mutants K332D and K332L, indicating that loss of LTC 4 transport in these mutants is associated with a loss of binding of this substrate. On the other hand, LTC 4 was still able to inhibit E 2 17␤G uptake by MRP1 mutants H335E, H335L, and H335Q, which is consistent with only a partial reduction in LTC 4 transport activity observed with these mutants (Fig. 6B).

FIG. 4. Kinetics of [ 3 H]LTC 4 uptake by wild-type MRP1 and MRP1 TM6 mutants H335E, H335L, and H335Q. A, shown is the Michaelis-Menten plot of the initial rate of ATP-dependent [ 3 H]LTC 4 uptake by membrane vesicles prepared from HEK293T cells transfected
with wild-type MRP1 (f) or MRP1 His 335 mutants H335E (Ⅺ), H335L (ƒ), and H335Q (q). Uptake was measured at various LTC 4 concentrations (0.01-1 M) for 1 min at 23°C in transport buffer. B, shown are Lineweaver-Burk plots for the LTC 4 uptake data shown in A. All values are means Ϯ S.D. of triplicate determinations in a single experiment. Relative MRP1 protein expression levels in the membrane vesicles were as shown in Fig. 3A. Similar results were obtained in at least one additional independent experiment. In previous studies, we have shown that MRP1 exhibits low levels of ATP-dependent GSH transport that can be markedly stimulated by a variety of compounds including the bioflavone apigenin (37,38). Similarly, estrone 3-sulfate by itself is a poor substrate for MRP1, but uptake of this conjugated estrogen is increased ϳ5-fold in the presence of physiological concentrations of GSH (30,39). Thus, apigenin and GSH were included in membrane vesicle transport assays to allow more accurate measurements of [ 3 H]GSH uptake and [ 3 H]estrone 3-sulfate uptake, respectively.

Photolabeling of Wild-type and TM6 Mutant MRP1 Proteins with [ 3 H]LTC 4 -To
When membrane vesicles from cells expressing the Lys 332 , His 335 , and Asp 336 MRP1 mutants were examined, it was found that apigenin-stimulated GSH uptake activity was reduced in all cases. The relative levels of [ 3 H]GSH uptake by the MRP1 K332D and K332L mutants were less than 15% of wild-type MRP1 (Fig. 7A)   was also reduced, but only by ϳ60% (Fig. 7B). Substitution of Asp 336 with Leu (D336L) reduced GSH uptake to ϳ25% of wild-type MRP1 levels, whereas substitution of this negatively charged residue with a positively charged residue (D336R) further reduced uptake of this tripeptide to just above basal levels observed with vesicles from the empty vector control (Fig. 7C).
In contrast to [ 3 H]GSH uptake, both neutral and negatively charged substitutions of Lys 332 and His 335 had essentially no effect on GSH-stimulated uptake of [ 3 H]estrone 3-sulfate (Fig.  7, D and E). This suggests that these mutants are still capable of binding GSH although GSH transport is not detectable. On the other hand, as observed for LTC 4 , E 2 17␤G, and GSH uptake, both Leu and Arg substitutions of Asp 336 eliminated estrone 3-sulfate transport by MRP1 (Fig. 7F).
[ 3 H]MTX Uptake in TM6 Mutant MRP1-enriched Membrane Vesicles-Several groups have reported previously that MRP1 mediates ATP-dependent uptake of MTX (22,40,41). In the membrane vesicles prepared from the transfected HEK cells used in this study, we found that MTX uptake by wild-type MRP1 was ϳ6.6 Ϯ 0.6 nmol mg of protein Ϫ1 20 min Ϫ1 compared with 1.1 nmol Ϯ 0.4 mg of protein Ϫ1 20 min Ϫ1 observed with control vesicles prepared from cells transfected with the pcDNA3.1(Ϫ) vector alone. Substitution of Lys 332 with a neutral (K332L) or negatively charged (K332D) amino acid had no effect on MTX uptake by MRP1. Similarly, MTX uptake by the H335E, H335L, and H335Q MRP1 mutants was comparable with wild-type MRP1. In contrast, MTX uptake was reduced by ϳ50% when Asp 336 was substituted with Leu (D336L) and by ϳ65% when substituted with Arg (D336R) (not shown).

DISCUSSION
Two of the best characterized substrates of MRP1 are the physiological metabolites LTC 4 and E 2 17␤G, which show reciprocal competitive inhibition of each other's transport. Consequently, it has been suggested that these two conjugated organic anions bind to the same or mutually exclusive binding sites on the protein (10,19,36,42). However, it is clear that the binding sites for these two substrates are not identical, since substitutions of single amino acids in TM17 can eliminate or reduce E 2 17␤G transport while leaving LTC 4 transport essentially intact (31,43). In the present study, we have now identified two basic amino acids, Lys 332 and His 335 , in the highly amphipathic TM6 of MRP1 that play a critical and selective role in the binding and transport of LTC 4 . Thus, replacing Lys 332 with either Leu or Asp eliminated the ability of MRP1 to transport LTC 4 (and markedly reduced GSH transport) without affecting the transport of other organic anions including E 2 17␤G, estrone 3-sulfate, and MTX. The MRP1-Lys 332 mutants also could not be photolabeled with LTC 4 , and E 2 17␤G transport by these mutants could no longer be inhibited by LTC 4 . Moreover, since the loss of LTC 4 transport and binding was the same whether Lys 332 was replaced with a neutral or negatively charged amino acid, it may be concluded that the loss of the positive charge at position 332 rather than the introduction of a neutral or negative charge is responsible for the phenotype observed.
Despite the fact that the MRP1-Lys 332 mutant proteins no longer transported LTC 4 , they still retained the ability to transport GSH, albeit at a much reduced level. In addition, GSH-stimulated estrone 3-sulfate transport by these mutants remained comparable with that observed with wild-type MRP1. One possible explanation for these findings is that MRP1 may contain two binding sites for GSH, one of which is involved in the stimulation of estrone 3-sulfate transport and the other of which is involved in the transport of GSH itself. If indeed more than one site exists, our results indicate that only the site for GSH transport is affected by the Lys 332 mutations (44). MRP1-Lys 332 is quite well conserved among ABCC subfamily members, including MRP2 and MRP3, and the murine, canine, Saccharomyces cerevisiae, and Leishmania tarentolae orthologs of MRP1, which are known to transport organic anions. However, mutation of Lys 329 in human MRP2 (which is analogous to MRP1-Lys 332 ) had no effect on transport of glutathioneconjugated methylfluorescein, suggesting that this amino acid is not critical for the transport of all GSH-conjugated substrates (29). On the other hand, Ito et al. (28) showed that the analogous Lys 325 in rat Mrp2 was important for LTC 4 transport as well as for transport of GSH-conjugated 2,4dinitrophenyl.
Substitution of MRP1-His 335 also diminished GSH and LTC 4 transport by MRP1, but the effect on LTC 4 transport was considerably less than observed with the MRP1-Lys 332 mutants. The retention of some activity allowed us to determine the kinetic parameters of LTC 4 transport by the His 335 mutants, which revealed that their apparent affinity for this substrate was similar to that of wild-type MRP1. However, their V max values were significantly decreased, as was photolabeling by [ 3 H]LTC 4 . This suggests that these mutations caused a reduction in both the efficiency of binding and translocation of this substrate, although why the reduction in photolabeling was not associated with a change in K m is not clear. MRP1-His 335 is relatively poorly conserved among ABCC family members and is present only in the canine, mouse, and yeast orthologs of MRP1 (Fig. 1B). MRP2, MRP3, and MRP6 have a much lower LTC 4 transport efficiency than MRP1, but whether this might be related to the absence of a His residue in the analogous position of MRP1-His 335 in these other transporters is not known.
Substitutions of MRP1-Asp 336 had a more dramatic effect on MRP1 transport activity than substitutions of either Lys 332 or His 335 in that the D336L and D336R mutants did not transport four of the five organic anion substrates tested, and transport of the fifth, MTX, was reduced by at least 50%. Thus mutations of MRP1-Asp 336 caused a nonselective overall loss of MRP1 transport activity, suggesting that this residue plays a more global and essential structural role in assembling or maintaining MRP1 in a transport-competent state. Among the ABCC proteins, MRP1-Asp 336 is the most conserved of the three charged residues examined that are all clustered on the same face of TM6 in the outer leaflet of the membrane. The functional importance of this acidic residue may also be conserved, since the analogous Asp 329 in rat Mrp2 has been demonstrated to be important for transport of the GSH-conjugated substrates, 2,4-dinitrophenyl S-glutathione and LTC 4 , and, to a lesser extent, E 2 17␤G and other conjugated organic anions (28). Interestingly, a missense mutation of the analogous acidic residue in CFTR, E92K, has been associated with a benign cystic fibrosis phenotype (45), and in vitro studies indicate that it is not critical for the chloride conducting-activity of CFTR (46).
The exact roles of TM6 Lys 332 , His 335 , and Asp 336 in the mechanism of substrate binding and translocation by MRP1 are unknown, although it is interesting to note that residues in the analogous TM1 of CFTR form part of the chloride channel lining of this protein (46,47). It is possible that the selective loss or reduction of LTC 4 and GSH transport by mutations of MRP1-Lys 332 and His 335 occurs because the side chains of these amino acids are required to form ionic or hydrogen bonding interactions directly with the glutathione moiety of these substrates. Additionally, side chain interactions between Lys 332 and His 335 and other amino acids, either in the same or a different TM helix of MRP1, may be required to maintain the architecture of the LTC 4 and GSH binding pocket(s). Thus mutations in these basic residues may cause a shift in TM helix packing that makes the LTC 4 (and GSH) binding site and/or translocation pathway less accessible to substrate. On the other hand, substitutions of Asp 336 may perturb normal packing of the TM helices of MRP1 or impede their movement in response to substrate translocation to such a degree that binding and transport of all substrates is markedly reduced or eliminated altogether. Further investigations of MRP1 helixhelix interactions such as those described for several TM segments of CFTR (48,49) may help to distinguish among these possibilities.
There are numerous examples of proteins where side chains of charged amino acids are known to form ion pairs or salt bridges with oppositely charged amino acids and thus provide protein stability and/or promote interactions with ligands. In addition, the introduction of salt bridges at the x and x ϩ 4 positions in short peptides and on protein surfaces can facilitate ␣-helix formation (50). Lys 332 and Asp 336 are located just four amino acids apart and thus could potentially form an intrahelical stabilizing salt bridge with one another. However, if this were their sole interaction, then it might be expected that mutation of either amino acid would result in the same phenotype, and clearly this was not the case. It is thus highly likely that these charged residues form additional ion pairs and/or hydrogen bonds with neighboring polar or charged amino acids in other TM helices of MRP1.
Ion pair interactions between TM helices have been reported for a variety of transporters and ion channels. For example, it has been shown that a salt bridge can occur between oppositely charged amino acids in TM2 and TM11 of the rat vesicular monoamine transporter, and when disrupted, high affinity recognition of its substrate serotonin is lost (51). Similarly, mutation of Lys 131 or His 338 that make up a salt bridge between TM2 and TM8 in the rat vesicular acetylcholine transporter results in loss of substrate binding and transport (52). Whether or not the three MRP1 TM6 amino acids studied here form ion pairs with oppositely charged residues or form hydrogen bonds with polar amino acids in other TM segments that are essential for MRP1 transport activity remains to be determined.
In summary, we have identified three potentially charged residues in the region of TM6 predicted to be in the outer leaflet of the plasma membrane that are critical for the transport activity of MRP1. Mutations of two of them, Lys 332 and His 335 , selectively affect the recognition and transport of GSH and the GSH conjugate LTC 4 . We previously showed that the naturally occurring mutation of MRP1-Arg 433 to Ser predicted to be located in the fourth cytoplasmic loop in close proximity to the membrane interface of TM8 results in a 2-fold decrease in LTC 4 and estrone 3-sulfate transport efficiency. Like the present findings with the Lys 332 and His 335 mutants, E 2 17␤G transport was unaffected by the Arg 433 3 Ser mutation (30). Thus, potentially charged amino acids both in the TM segments of MSD2 and in at least one of the cytoplasmic loops connecting them appear particularly important for MRP1-mediated transport of LTC 4 . Additional charged residues in other regions of MRP1 are currently being investigated to determine their participation, if any, in the substrate specificity and transport activity of MRP1.