| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 278, Issue 46, 46052-46063, November 14, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
¶||¶**¶
From the
Division of Cancer Biology and Genetics, Cancer Research Institute, and the Departments of ¶Pathology & Molecular Medicine, Biochemistry, and Anatomy & Cell Biology, Queen's University, Kingston, Ontario K7L 3N6, Canada
Received for publication, July 31, 2003 , and in revised form, September 3, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-estradiol 17-(-D-glucuronide) (E217G) but not cysteinyl leukotriene 4 (LTC4), demonstrating the importance of multiple residues in this helix in determining substrate specificity. In contrast, mutations of Asp1084 that eliminate the carboxylate side chain markedly decreased resistance to all drugs tested, as well as transport of both E217G and LTC4, despite the fact that LTC4 binding was unaffected. We show that these mutations prevent the ATP-dependent transition of the protein from a high to low affinity substrate binding state and drastically diminish ADP trapping at nucleotide binding domain 2. Based on results presented here and crystal structures of prokaryotic ATP binding cassette transporters, Asp1084 may be critical for interaction between the cytoplasmic loop connecting TM13 and TM14 and a region of nucleotide binding domain 2 between the conserved Walker A and ABC signature motifs. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
In addition to conferring drug resistance, transport studies using inside-out membrane vesicles prepared from MRP1-transfected cells have shown that MRP1 is capable of transporting glutathione-, glucuronate- and sulfate-conjugated organic anions such as the cysteinyl leukotriene 4 (LTC4), 17-estradiol 17-(-D-glucuronide) (E217G), and estrone-3-sulfate (15-23). MRP1-mediated LTC4 and E217G transport has further been demonstrated in reconstituted proteoliposomes with immunoaffinity-purified native MRP1, clearly proving that MRP1 alone is sufficient for transport of conjugated organic anions (24). LTC4 is a high affinity endogenous glutathione S-conjugate substrate for MRP1 and plays important roles in the inflammatory response (15, 16, 19). In mMRP1-/- mice, failure to express mMRP1 leads to decreased LTC4 secretion from leukotriene-synthesizing cells and impaired response to an inflammatory stimulus (25). Studies by Robbiani et al. (26) found that mMRP1-mediated efflux of LTC4 was also involved in regulating the migration of dendritic cells to lymph nodes. Whether E217G is a physiological substrate for MRP1 is not yet known, but the protein actively transports this cholestatic conjugated estrogen in vitro with a Km of 1-3 µM (13, 17, 19).
MRP1, or ABCC1, is a member of the ABCC branch of the ATP binding cassette (ABC) superfamily. This branch also includes MRP2 to -6 (ABCC2 to -6), MRP7 to -9 (ABCC10 to -12), as well as the sulfonylurea receptors SUR1 (ABCC8) and SUR2 (ABCC9), and the cystic fibrosis transmembrane conductance regulator (CFTR/ABCC7) (27-31). The predicted topologies of MRP1 to -9 all contain a typical ABC transporter core region, composed of two hydrophobic membrane-spanning domains (MSDs), each with six transmembrane (TM) 
-helices, and two hydrophilic nucleotide-binding domains (NBDs). In addition, MRP1, -2, -3, -6, and -7 have a third, NH2-terminal MSD (MSD1), which consists of five TMs with an extracellular NH2 terminus (Fig. 1) (27, 31-36).
|
G efficiently (38-40). In addition, these studies provided evidence for cross-talk between residues in the two TM helices (39). TM17 of MRP1 is topologically equivalent to TM12 of P-gp, which has been shown to be intimately involved in substrate binding and determination of substrate specificity (41). Similarly, TM 14 corresponds to TM 9 of P-gp, but the involvement of residues in this helix in determining substrate specificity or overall activity has only recently been identified (42, 43). In addition to Glu1089, TM14 of MRP1 contains two charged residues, Asp1084 and Lys1092, as well as four amino acid residues with polar side chains (Fig. 1). We have now characterized the role of all of these residues in determining the substrate specificity and/or overall activity of MRP1. These studies demonstrate that Lys1092, Ser1097, and Asn1100, which are all predicted to be on the same face of the helix, are important for defining the substrate specificity of MRP1 and that Asp1084 is critical for its overall activity. We also show that proteins in which Asp1084 has been mutated to non-carboxylic amino acids retain the ability to bind LTC4 with high affinity but are unable to shift into a low affinity binding state in the presence of ATP and vanadate. Finally, we demonstrate that the inability of the Asp1084 mutant protein to enter a transition state is attributable to a defect in the ability of NBD2 to hydrolyze ATP, as evidenced by a loss of vanadate-dependent trapping of ADP at this NBD.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
G (44 Ci/mmol) from PerkinElmer Life Sciences, and 8-azido-[-32P]ATP (11.8 Ci/mmol) from Affinity Labeling Technologies Inc. (Lexington, KY). Doxorubicin HCl, etoposide (VP-16), and vincristine sulfate were obtained from Sigma. Site-directed Mutagenesis—All mutations were generated using the TransformerTM site-directed mutagenesis kit (Clontech, Palo Alto, CA). Templates were prepared as described previously (37-39). Mutagenesis was then performed according to the instructions from the manufacturer using a selection primer, 5'-GAG AGT GCA CGA TAT CCG GTG TG-3', that mutates a unique NdeI site in the vector to an EcoRV restriction site. Oligonucleotides bearing mismatched bases at the residues to be mutated (underlined) were synthesized by ACGT Corp. (Toronto, Ontario, Canada). They are as follows: T1082A (5'-C TCC AAG GAG CTC GAC GCA GTG GAC TCC-3'), D1084N (5'-CTG GAC ACA GTG AAT TCC ATG ATC CCG-3'), D1084A (5'-CTG GAC ACA GTG AAA TCG ATG ATC CCG-3'), D1084E (5'-CTG GAC ACA GTG GAC TCG ATG ATC CCG-3'), D1084V (5'-CTG GAC ACA GTG GTA TCG ATG ATC CCG-3'), S1085A (5'-CTG GAC ACA GTC GAC GCC ATG ATC CCG G-3'), K1092M (5'-C CCG GAG GTC ATC ATG ATG TTC ATG GGC-3'), K1092A (5'-C CCG GAG GTC ATC GCG ATG TTC ATG GGC-3'), K1092E (5'-C CCG GAG GTC ATC GAG ATG TTC ATG GGC-3'), K1092R (5'-C CCG GAG GTC ATC AGG ATG TTC ATG GGC-3'), S1097A (5'-G ATG TTC ATG GGC GCC CTG TTC AAC-3'), N1100A (5'-TTC ATG GGC TCG CTC TTC GCC GTC ATT GGT G-3'), N1100S (5'-TTC ATG GGC TCG CTC TTC AGT GTC ATT GGT G-3').
After confirming all mutations by DNA Thermo Sequenase Cy5.5 and Cy5.0 dye terminator cycle sequencing (Amersham Biosciences) according to the instructions from the manufacturer, DNA fragments containing the desired mutations were transferred into pCEBV7-MRP1. After reconstructing the mutations into the full-length clones, the integrity of the entire mutated inserts and cloning sites was verified by DNA sequencing (ACGT Corp.).
Mutation D1084R was generated using the QuikChangeTM site-directed mutagenesis kit (Stratagene, La Jolla, CA). Templates were prepared as described previously (44). Mutagenesis was then performed according to the instructions from the manufacturer. Oligonucleotide bearing mismatched bases at the residues to be mutated (underlined) was synthesized by ACGT Corp. It is as follows: D1084R (5'-CTG GAC ACA GTG CGG TCC ATG ATC CCG-3'). After confirming the mutation by DNA sequencing (ACGT Corp.), DNA fragments containing the desired mutations were transferred into pcDNA3.1-MRP1.
Cell Lines and Tissue Culture—Stable transfection of HEK293 cells with the pCEBV7 vector containing the wild type and mutant MRP1 cDNAs has been described previously (13, 37-39). Briefly, HEK293 cells were transfected with pCEBV7 vectors containing mutant MRP1 using FuGENE 6 (Roche Molecular Biochemicals) according to the instructions from the manufacturer. After 
48 h, the transfected cells were supplemented with fresh medium containing 100 µg/ml hygromycin B. Approximately 3 weeks after transfection, the hygromycin B-resistant cells were cloned by limiting dilution and the resulting cell lines were tested for high level expression of the mutant proteins.
Confocal Microscopy—Confocal microscopy was carried out as described previously (37-39). Briefly, 5 x 105 stably-transfected HEK293 cells were seeded in each well of a 6-well tissue culture dish on coverslips. When the cells had grown to confluence, they were washed once in PBS and then fixed with 2% paraformaldehyde in PBS, followed by permeabilization using digitonin (0.25 mg/ml in PBS). MRP1 proteins were detected with the monoclonal antibody MRPm6. Antibody binding was detected with Alexa Fluor 488 anti-mouse IgG (H+L) (Fab')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 and 530/30 nm for Fluor 488).
Determination of Protein Levels in Transfected Cells—Plasma membrane vesicles were prepared by centrifugation through sucrose, as described previously (13, 16). After determination of protein levels by Bradford assay (Bio-Rad), total membrane protein (0.5, 1.0, and 1.25 µg) from transfectants expressing wild type MRP1 and various mutant proteins were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using a 7.5% gel. Proteins were subsequently transferred to Immobilon-P polyvinylidene difluoride membranes (Millipore, Bedford, MA) by electroblotting. The proteins were detected with mAb MRPm6 (Alexis Biochemicals, San Diego, CA). Antibody binding was detected with horseradish peroxidase-conjugated goat anti-mouse IgG (Pierce), followed by enhanced chemiluminescence detection and X-OmatTM Blue XB-1 films (PerkinElmer Life Sciences). Densitometry of the film images was performed using a Chemi-ImagerTM 4000 (Alpha Innotech Corp., San Leandro, CA). The relative protein expression levels were calculated by dividing the integrated densitometry values obtained for 0.5, 1.0, and 1.25 µg of total membrane proteins from transfectants expressing the mutant proteins by the integrated densitometry values obtained for the comparable amounts of total membrane proteins from transfectants expressing the wild type protein. Each comparison was performed at least three times in independent experiments. The results were then pooled and the mean values used for normalization purposes.
Expression of the NH2- and the COOH-proximal Half-molecules of MRP1 in SF21 Cells—The construction of the dual expression vector pFASTBAC Dual (Invitrogen) encoding the NH2- and the COOH-proximal half-molecules of MRP1 has been described previously (45). To generate MRP1D1084N-pFASTBAC Dual vector, pCEBV7-MRP1 containing mutation D1084N was digested with BstEII, and an 600-bp fragment comprising nucleotides 3369-3913 of MRP1D1084N was isolated. This fragment was then ligated to an 9.4-kb fragment isolated from BstEII-digested MRP1-pFASTBAC Dual vector.
Recombinant bacmids and baculoviruses were generated as described previously (45). The conditions used for viral infection were also similar to those described previously (45). Plasma membrane vesicles were prepared, and the expression levels of wild type and mutant protein in infected cells were determined, as described previously (13, 16). The NH2-proximal proteins were detected with mAb MRPr1 (Alexis Biochemicals), and the COOH-proximal proteins were detected with mAb MRPm6 (Alexis Biochemicals).
LTC4, E217G, and GSH Transport by Membrane Vesicles—Plasma membrane vesicles were prepared as described previously, and ATP-dependent transport of [3H]LTC4 into the inside-out membrane vesicles was measured by a rapid filtration technique (16, 17). Briefly, vesicles (10 µg of total proteins) were incubated at 23 °C in 100 µl 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 MgCl2 and [3H]LTC4 (50 nM, 200 nCi). At the indicated times, 20-µl 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 4 ml of cold transport buffer. The bound radioactivity was determined by scintillation counting. All data were corrected for the amount of [3H]LTC4 that remained bound to the filter in the absence of vesicle protein (usually <5% of the total radioactivity). [3H]LTC4 uptake was expressed relative to the total protein concentration in each reaction. ATP-dependent uptake of [3H]E217G (400 nM, 120 nCi) was measured as described for [3H]LTC4 except that the reaction was carried out at 37 °C.
Km and Vmax values of ATP-dependent [3H]LTC4 uptake by membrane vesicles (2.5 µg of total proteins) were measured at various LTC4 concentrations (0.01-1 µM) for 1 min at 23 °C. in 25 µl of transport buffer containing 4 mM ATP and 10 mM MgCl2, followed by non-linear regression analyses. Kinetic parameters of ATP-dependent [3H]E217G (0.1-16 µM) uptake were determined as described for [3H]LTC4 except that the reaction was carried out at 37 °C.
GSH uptake was also measured by rapid filtration with membrane vesicles (20 µg of total proteins) incubated at 37 °C for 20 min in a 60-µl reaction volume with [3H]GSH (100 µM, 300 nCi). 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 [3H]GSH uptake in the presence of verapamil (100 µM).
Photoaffinity Labeling of the Wild Type and Mutant Proteins with [3H]LTC4—Wild-type and mutant MRP1 membrane proteins were photolabeled with [3H]LTC4 essentially as described previously (16, 46, 47). Briefly, membrane vesicles (50 µg of total proteins in 35 µl of transport buffer) were incubated with [3H]LTC4 (0.3 µCi, 200 nM) at room temperature for 10 min, frozen in liquid nitrogen, and UV-irradiated. Proteins were analyzed on a 5-15% gradient gel by SDS-PAGE. The gel was then fixed, treated with Amplify (Amersham Biosciences), dried, and exposed to film for 1 week at -70 °C.
Photolabeling of MRP1 by 8-Azido-[-32P]ATP and Orthovanadate-induced Trapping of 8-Azido-[-32P]ATP by MRP1—Wild-type and mutant MRP1 membrane proteins were photolabeled with 8-azido-[-32P]ATP essentially as described previously (45). Briefly, membrane vesicles (20 µg of total proteins in 10 µl of transport buffer containing 5 mM MgCl2) were incubated with 8-azido-[-32P]ATP (1 µCi) on ice for 5 min, and UV-irradiated. The reactions were stopped by the addition of 400 µl of ice-cold stop buffer (50 mM Tris-Cl, pH 7.2, 0.1 mM EGTA, 5 mM MgCl2), and the membranes were centrifuged at 14,000 rpm for 15 min at 4 °C. The pellets were resuspended in 20 µl of stop buffer and were then analyzed on a 5-15% gradient gel by SDS-PAGE. The gel was then dried and exposed to film for 2 h at room temperature.
Orthovanadate-induced trapping of 8-azido-[-32P]ATP by MRP1 was determined as described previously (45). Briefly, membrane proteins (20 µg of total proteins) were incubated in transport buffer (10 µl) containing 5 mM MgCl2, 1 mM sodium orthovanadate, and 8-azido-[-32P]ATP (1 µCi) at 37 °C for 15 min. The reactions were stopped by the addition of 400 µl of ice-cold stop buffer, and the membranes were centrifuged at 14,000 rpm for 15 min at 4 °C. The pellets were washed again and resuspended in 20 µl of stop buffer. The samples were transferred to a 96-well plate and UV-irradiated. Proteins were analyzed on a 5-15% gradient gel by SDS-PAGE. The gel was then dried and exposed to film for 6 h at room temperature.
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 (10, 11, 13). Briefly, cells were seeded at 7 x 103 cells/well in 100 µl of culture medium in 96-well tissue culture plate. The following day, various concentrations of drug diluted in culture medium were added to cells (100 µl/well). After incubation for an additional 72 h, 100 µl of medium was removed from each well and the MTT reagent (25 µl/well, 2 mg/ml) (Sigma) was added. After 3 h, the formazan was solubilized by mixing with HCl/isopropanol (1:24) (100 µl/well). Color density was determined using the ELX 800 UV spectrophotometer (Filter4, 570 nm). Mean values of quadruplicate determinations (±S.D.) were plotted using GraphPad software. IC50 values were obtained from the best fit of the data to a sigmoidal curve. Relative resistance is expressed as the ratio of the IC50 value of cells transfected with MRP1 expression vectors compared with cells transfected with empty vector. Resistance was determined in three or more independent experiments.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
Transport of [3H]LTC4 and [3H]E217G by Wild Type and Mutant MRP1—To determine whether any of the TM14 mutations altered the efficiency with which the protein transported LTC4 and E217G, we examined ATP-dependent uptake of these compounds by membrane vesicles prepared from HEK transfectants expressing each of the mutant proteins (Fig. 3). The levels of LTC4 uptake by vesicles prepared from HEK transfectants expressing either wild type MRP1 or mutations T1082A, S1085A, K1092M, S1097A, and N1100A were proportional to the relative expression levels of the wild type and mutant proteins. The only mutation that affected LTC4 transport was replacement of Asp1084 by Asn, which almost completely eliminated transport (Fig. 3, A-C).
|
G was unaffected by the T1082A, S1085A and K1092M mutations (Fig. 3, D-F), but substitution of Ser1097 with Ala and conversion of Asp1084 to Asn both dramatically decreased transport. Replacement of Asn1100 with Ala also decreased transport efficiency by 30-40%. Thus, mutations S1097A and N1100A only affected transport of the estrogen conjugate, whereas mutation D1084N decreased the ability of MRP1 to transport both LTC4 and E217G.
Resistance Profiles of Wild Type and Mutant Human Proteins—The drug resistance profiles of transfectants expressing mutant proteins were determined using MTT assays. The results are presented as relative drug resistance factors in Table I. Mutations T1082A, S1085A, and N1100A had no significant effect on drug resistance, but conversion of one polar residue, Ser1097, to Ala caused a 2-3-fold decrease in resistance to vincristine, VP-16, and doxorubicin. Similarly, mutation D1084N also affected the drug resistance profile of MRP1. The D1084N mutant protein essentially lost its ability to confer resistance to vincristine and doxorubicin, and retained only 30% of the ability of the wild type protein to increase VP-16 resistance. In contrast, although substitution of Lys1092 with Met had no detectable effect on resistance to the electroneutral drug VP-16, the mutation increased resistance to the cationic drugs, vincristine and doxorubicin. Thus, both Asp1084 and Ser1097 influence the level of resistance to all three classes of drugs tested, whereas Lys1092 only affects resistance to the cationic drugs vincristine and doxorubicin.
|
G, and drugs such as vincristine and VP-16, is a requirement for GSH, which may be co-transported with the unmodified drug (48, 49). Previously, we have reported that MRP1 exhibits low levels of ATP-dependent GSH transport that can be stimulated by compounds such as verapamil (50). Consequently, we examined the effects of TM14 mutations on verapamil-stimulated GSH transport to determine whether those that affected drug resistance also influenced transport of GSH (Fig. 4). As observed when examining LTC4 transport, only replacement of Asp1084 by Asn influenced the ability of MRP1 to transport GSH, and this mutation essentially eliminated transport activity. Other mutations, including mutations S1097A and K1092M that altered the drug resistance profile of MRP1, had no effect. These findings suggest that the effects of mutations S1097A and K1092M on MRP1-mediated drug resistance appear to result primarily from changes in the ability to interact with the drug substrate rather than GSH.
|
G Transport by Wild Type and Mutant MRP1—Because mutation D1084N affected the ability of the protein to transport both LTC4 and E217G whereas mutations S1097A and N1100A selectively decreased the transport of only E217G, we compared their effect on the kinetic parameters of transport of both substrates (Fig. 5). For E217G transport, the normalized Vmax values for mutations D1084N, S1097A, and N1100A were lower than that for wild type MRP1 (3207 pmol/mg/min for wild type MRP1, versus 151, 2279, and 1852 pmol/mg/min for mutations D1084N, S1097A and N1100A, respectively) (Fig. 5A and Table II). The apparent Km value for mutation N1100A (1.9 µM) was comparable with that of the wild type protein (1.6 µM). However, replacement of Asp1084 with Asn decreased the Km value 3-fold (0.5 µM). On the other hand, mutation S1097A significantly increased the Km value (28.4 µM) (Fig. 5A and Table II). Thus, although mutations D1084N, S1097A, and N1100A all decreased the conjugated estrogen transport, the three mutations had different effects on the apparent Km values for E217G uptake.
|
|
8- and 2-fold, respectively (30 pmol/mg/min, and 77 nM for mutation D1084N) (Fig. 5B and Table II). Thus, similar to the results obtained with E217G as a substrate, the D1084N mutation appeared to increase the affinity of MRP1 for substrate while decreasing the Vmax.
Effect of Mutation D1084N on Photolabeling of MRP1 with [3H]LTC4 and 8-Azido-[-32P]ATP—Because transport studies indicated that mutation D1084N markedly decreased the Vmax for LTC4 transport while modestly decreasing the apparent Km, we confirmed the ability of the mutant protein to bind [3H]LTC4 by photolabeling studies. Based on densitometry of several preparations of membrane proteins, substitution of Asp1084 with Asn had no significant effect on photolabeling with [3H]LTC4 (Fig. 6B). In addition, we examined photolabeling in the presence of ATP and vanadate to determine whether the mutant was altered in its ability to form a low affinity transition state. As we have shown previously, the presence of ATP significantly decreased the photolabeling of the wild type protein and this effect was further enhanced in the presence of vanadate (46). In contrast, the photolabeling of mutant MRP1D1084N with LTC4 was unaffected by ATP alone or ATP plus vanadate (Fig. 6B). To investigate how mutation D1084N abrogated the effect of ATP on the binding of LTC4, we examined the ability of the wild type and mutant proteins to be photolabeled with 8-azido-[-32P]ATP both at 4 °C to minimize hydrolysis and under vanadate-induced ADP trapping conditions at 37 °C. As shown in Fig. 6C, mutation D1084N had no marked effect on the binding of the ATP analog at 4 °C. However, the D1084N mutation virtually abolished the vanadate-dependent trapping of 8-azido-[-32P]ADP by MRP1 at 37 °C (Fig. 6D), suggesting that it substantially decreased ATP hydrolysis by MRP1.
|
|
G by Wild Type MRP1—Because mutation D1084N dramatically affected all of the MRP1 functions tested, we also mutated Asp1084 to Ala, Glu, Arg, and Val. In addition, because of the effect of the K1092M mutation on the drug resistance profile of MRP1, Lys1092 was converted to Ala, Glu, and Arg. Replacement of Asn1100 with Ala selectively decreased the transport of E217G. This residue is not conserved between the human and rodent proteins. In both rat and mouse MRP1, the comparable residue is Ser. We have shown previously that both rodent proteins transport E217G far less efficiently than human MRP1 and that non-conserved amino acid residues located within a region between amino acids 959 and 1197 of human MRP1 are partially responsible for the higher efficiency with which the human protein transports E217G (51, 52). To date only one such residue, Glu1089, has been identified (37). Consequently, we also mutated Asn1100 to Ser, as it is in the rodent proteins. These mutations were then stably expressed in HEK293 cells. The expression levels of these mutant proteins in stably transfected HEK293 cells were similar to that of wild type MRP1 (Fig. 8A).
|
G are shown in Fig. 8 (B and C). Mutations K1092A, K1092R, and K1092E had no effect on transport of either LTC4 or E217G. Mutation N1100S, like N1100A, decreased only E217G transport. In contrast, replacement of Asp1084 with Ala, Arg, and Val dramatically decreased the ability of MRP1 to transport both LTC4 and E217G, whereas conversion to Glu also resulted in an approximate 50% decrease in transport of both substrates. We also examined the ability of the mutants D1084E and D1084R to bind [3H]LTC4 by photolabeling studies. As shown in Fig. 8E, the relative normalized densitometry values of photolabeling of D1084E and D1084R with [3H]LTC4 were 120 and 100, respectively. Thus, similar to the results obtained with photolabeling of mutant D1084N with LTC4, substitution of Asp1084 by Arg or Glu had no significant effect on the photolabeling. These results further confirmed that Asn1100 is involved in the efficient transport of E217G, and that Asp1084 is important for transporting conjugated organic anions by MRP1. Effects of Mutations on Drug Resistance Profile of MRP1— The capacity of these mutant proteins to confer drug resistance was also examined by MTT assay (Table III). Mutation N1100S had no effect on the drug profile of MRP1. The mutation K1092R also showed the same phenotype as wild type MRP1, whereas mutations K1092A and K1092E, like K1092M, increased the ability of MRP1 to confer resistance to vincristine and doxorubicin but not to VP-16. Substitution of Asp1084 by Ala and Val, as observed with mutation D1084N, significantly reduced resistance to vincristine, VP-16, and doxorubicin. Interestingly, replacement of Asp1084 with a conserved residue, Glu, resulted in a mutant protein with decreased ability to confer resistance to vincristine and doxorubicin, but unaltered ability to confer VP-16 resistance, (Table III). These findings suggest that size together with charge in the side chain of residue at position 1084, and positive charge in the side chain of residue at position 1092 are important for determining the drug resistance profile of MRP1.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
Other than Glu1089, the only other negatively charged amino acid in TM14 is Asp1084. Rather than being involved in determining substrate specificity, the carboxylate side of Asp1084 was critical for the overall activity of MRP1. Replacement of Asp1084 with various neutral amino acid residues Asn, Ala, or Val dramatically decreased all of the MRP1 functions tested. In contrast, conservative substitution of Asp1084 with Glu resulted in a mutant protein which retained 50% of the ability of the wild type protein to transport LTC4 and E217G, and to confer vincristine and doxorubicin resistance, while having no effect on VP-16 resistance. Thus, reducing the length of the side chain by only a single methylene group had a readily detectable effect on transport of two structurally unrelated conjugated anions and on the ability to confer resistance to two structurally unrelated cationic drugs. This result, combined with the effect of mutations that eliminate the negative charge of the side chain, suggested that the presence and position of the carboxyl group may be critical for a common step in the transport process, rather than interaction with specific substrates.
In addition, we also demonstrated the important role of Lys1092, Ser1097, and Asn1100 in defining the substrate specificity of MRP1. Elimination of a positive charge in the side chain of the residue at position 1092 drastically increased resistance to the cationic drugs vincristine and doxorubicin, but had no effect on resistance to the electroneutral drug VP-16, or the transport of organic anionic conjugates. On the other hand, mutations S1097A, N1100A, and N1100S reduced E217G but not LTC4 or GSH transport activity. Mutation S1097A also reduced resistance to all three classes of drugs tested. Examination of a helical wheel projection of TM14 of MRP1 indicates that it is highly amphipathic with all polar residues, except for Asp1084, which influenced the overall activity of MRP1, located on one face of the helix (Fig. 9).
|
G transport without affecting transport of LTC4, suggesting that the hydrogen bonding capability of Ser1097 may be involved specifically in the interaction of the conjugated steroid with the protein. Substitution of Ser1097 with Ala also dramatically decreased resistance to all three drugs tested. A large body of evidence has suggested that unmodified drugs are transported by MRP1 via a co-transport mechanism with reduced glutathione (48, 49). However, in the present study, we did not observed any effect of mutation S1097A on verapamil-stimulated GSH transport by MRP1. These findings suggest that Ser1097 might be involved in the interaction of MRP1 with the drug substrate rather than GSH. In addition, in contrast to the effect of mutation S1097A on E217G transport, mutation N1100A, which only affected the transport of E217G, decreased the Vmax without significantly affecting Km, suggesting that the mutation affects a step in the transport process subsequent to initial binding of this substrate. Asn1100 is not conserved among the human and rodent proteins. In rat and mouse MRP1, the comparable residue is Ser. As indicated above, although both mouse and rat MRP1 transport LTC4 with efficiencies similar to that of the human protein, they are relatively poor transporters of E217G (13, 52). We recently showed that conversion of Ser1101 in TM14 of rat MRP1 to Asn, as it is in human MRP1, increased the ability of rat protein to transport the glucuronidated estrogen (52). Consistent with these results, we found that replacement of Asn1100 in human MRP1 with either Ala or Ser, as in the mouse and rat proteins, decreased E217G transport efficiency by 30%. Taken together, these findings confirm the importance of Asn1100 in the relatively efficient transport of E217G by the human protein.
More recently, we reported that positively charged residues Lys322, predicted to be in TM6, and Arg1249, predicted to be close to a TM-cytoplasm interface, are critical for the capacity of the human protein to transport LTC4 (55, 57).2 In the present study, we found that mutations of Lys1092, predicted to be in the middle of TM14 of MRP1, had no effect on the transport of LTC4 and E217G. In contrast, elimination of the positive charge by converting Lys1092 to various neutral residues, Met and Ala, or a negatively charged residue, Glu, increased resistance to cationic drugs vincristine and doxorubicin but not to the electroneutral drug VP-16. One possible explanation for this observation is that Lys1092 is predicted to be adjacent to Glu1089 on the same face of the helix. As mentioned before, Glu1089 plays an important role in the interaction of drugs, particularly cationic drugs, with MRP1. Replacement of Glu1089 with a positive charged residue, Lys, virtually abolished resistance to both doxorubicin and vincristine, whereas the mutation still retained 50% of the wild type proteins ability to confer VP-16 resistance (37). Thus, elimination of the positive charge in the side chain of Lys1092 may facilitate the interaction of cationic drugs with an acidic amino acid residue at position 1089.
The fact that elimination of the negative charge at position 1084 had a major effect on the overall activity of MRP1 raised the possibility that such mutations may have affected folding and trafficking of the protein. However, confocal microscopy revealed no evidence of a trafficking defect that might be indicative of misfolding. Photolabeling studies with [3H]LTC4 also showed that the replacement of Asp1084 by Asn and Glu had no effect on binding of the substrate to the protein. This observation strongly suggests that the mutations had not resulted in a major perturbation of the structure of the protein and, furthermore, that Asp1084 was not involved in the binding of LTC4. This differs from results we obtained recently following mutational analyses of Asp336 which is predicted to be in TM6. Mutations of Asp336 abolished the transport of LTC4, but they also eliminated LTC4 binding (55). In the present study, kinetic analyses of the residual transport activity of the D1084N mutant protein yielded a lower Km for both LTC4 and E217G than the wild type protein, suggesting that the affinity for both substrates was actually increased. However, we did not observe a significant increase in photolabeling of the mutant protein by LTC4. This may be attributable to the fact that the initial photolabeling experiments, in contrast to transport assays were carried out in the absence of ATP. We have shown previously that in the presence of ATP, LTC4 photolabeling of the wild type protein is decreased and that this decrease requires that NBD2 be competent to trap ADP (46). Consistent with this suggestion, we found that LTC4 binding by the D1084N mutation did not decrease in the presence of ATP even with the addition of vanadate. This finding together with the effect of mutation D1084N on the overall activity of MRP1 raised the interesting possibility that the mutation might affect the binding and/or hydrolysis of ATP, so that substrate could not be translocated and/or released from the mutant protein. We detected no change in binding of ATP at 4 °C by either NBD of the mutant protein, but the conformational change that results in decreased LTC4 binding does not occur at this temperature. At 37 °C, the D1084N mutation drastically decreased trapping of 8-azido-ADP at NBD2 when photolabeling was carried out in the absence or presence of vanadate. These findings indicate that the mutation D1084N decreased the ability of NBD2 to hydrolyze ATP. In recent studies we have determined that ATP binding by NBD2, rather than hydrolysis, is sufficient to result in a conformational change that decreases the affinity of MRP1 for LTC4 and that this persists while NBD2 is occupied by either ATP or ADP (58). Consequently, the primary effect of the mutation may be at the level of ATP binding by NBD2, a process that we and others have shown to be strongly stimulated by ATP binding to NBD1 (44, 59).
Asp1084 of MRP1 is predicted by a number of topology algorithms and by modeling of the MRP1 MSD structures, using the crystal structure of MsbA as a template, to be very close to the cytoplasm-membrane interface of TM14. Such analyses also predict that the TM14 helix is continuous with a helix that extends to Pro1068 in the cytoplasmic loop connecting with TM13 (Fig. 9B). The crystal structure of the comparable cytoplasmic loop in MsbA, which is a lipid A transporter from Escherichia coli, indicates that the loop forms a U-like configuration consisting of three -helices with the middle helix in close contact with a region between the Walker A motif and ABC signature of the NBD (60, 61). The cytoplasmic loop between TM13 and TM14 is also predicted to be extensively helical. If its location is similar to that of the corresponding region of MsbA, its position makes it a likely candidate for transduction of conformational changes that occur between NBD2 and MSD3 following nucleotide binding and hydrolysis. Because Asp1084 is highly conserved among different MRPs, as well as SUR1, SUR2, and CFTR (Fig. 1C), it is possible that this residue is particularly important for the transduction process.
|
FOOTNOTES |
|---|
|| Recipient of a Canadian Institutes of Health Research postdoctoral award.
** Canada Research Chair in Cancer Biology and Senior Scientist of Cancer Care Ontario.
Stauffer Research Professor of Queen's University and Director, Division of Cancer Biology and Genetics, Cancer Research Inst., Queen's University. To whom correspondence should be addressed: Cancer Research Inst., Suite 300, 10 Stuart St., Kingston, Ontario K7L 3N6, Canada. Tel.: 613-533-2979; Fax: 613-533-6830; E-mail: deeleyr{at}post.queensu.ca.
1 The abbreviations used are: MRP, multidrug resistance protein; P-gp, P-glycoprotein; CFTR, cystic fibrosis transmembrane conductance regulator; MSD, membrane-spanning domain; TM, transmembrane; NBD, nucleotide binding domain; mAb, monoclonal antibody; E217G, 17-estradiol 17-(-D-glucuronide); LTC4, leukotriene C4; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; VP-16, etoposide; HEK, human embryonic kidney; ABC, ATP binding cassette; PBS, phosphate-buffered saline.
2 D. Situ, manuscript in preparation.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
), HEKMRP1 (
), HEKMRP1T1082A (
), HEKMRP1D1084N (
), HEKMRP1S1085A (•), HEKMRP1K1092M (
), HEKMRP1S1097A (
), HEKMRPN1100A (
). The uptake of LTC4 and E217
) and HEKMRP1 (
)). The normalized transport values were obtained by adjusting experimentally determined values (1-min time point) to compensate for differences in the relative levels of the wild type and mutant proteins and are shown in panels C and F as gray bars. Data shown in panels A, B, D, and E have not been normalized to compensate for differences in expression levels. Values are mean ± S.D. of three or more independent experiments.
