ATPase Activity of Purified Multidrug Resistance-associated Protein*

Human multidrug resistance protein (MRP) was expressed at high levels in stably transfected baby hamster kidney (BHK-21) cells. These cells exhibited a pattern of cross-resistance to several different drugs typical of an MRP-mediated phenotype despite the addition of 10 histidine residues at the C terminus to facilitate purification. Consistent with this functional evidence of the presence of MRP at the surface of these transfectants, strong signals were detected by immunoblotting and immunofluorescence using a specific monoclonal antibody to MRP. There was intense uniform staining of the cell surface as well as weaker staining of intracellular membranes. MRP-containing membranes were solubilized in 1%N-dodecyl-β-d-maltoside in the presence of 0.4% sheep brain phospholipids. Two sequential affinity purification steps on Ni-NTA agarose and wheat germ agglutinin agarose provided substantial enrichment, and contaminating bands were not detected. ATPase activity of the purified protein was assayed in the presence of the phospholipids, which had been maintained throughout all purification steps. ATP was hydrolyzed in proportion to the amount of purified protein assayed, and typical Michaelis-Menten behavior was exhibited, yielding estimations of K m of ∼3.0 mm and V max of 0.46 μmol mg−1 min−1. This activity was moderately stimulated by the drugs that others have shown to be transported by MRP-containing membrane vesicles. This stimulation was enhanced by reduced glutathione as is its drug transport, and oxidized glutathione, itself a substrate for transport, caused a strong stimulation. These data describe the first purification of MRP and provide the first direct evidence that the molecule possesses drug-stimulated ATPase activity.

The multidrug resistance-associated protein (MRP) 1 was discovered in multidrug-resistant tumor cells that did not express any of the structurally related P-glycoproteins (1). Several such cell lines have been described (2)(3)(4)(5)(6)(7), although drug-resistant tumor cells frequently express both MRP and P-glycoprotein. MRP function has since been characterized using cells in which it is either endogenously or heterologously expressed and in membrane vesicles containing the protein, isolated from these cells (8 -18). Results of these studies are consistent with the notion that the protein brings about the active export of a number of anti-tumor drugs provided that intracellular glutathione is present (12,19). Significantly, the protein has also been found to transport conjugated natural substrates of physiological importance, including glutathione-conjugated leukotrienes (9, 11, 13, 19 -21) and glucuronide-conjugated steroids (14). More recently, cDNAs have been cloned for several socalled multispecific organic anion transporters, and their sequences indicate that they are homologs of MRP (22,23). The sequences of these molecules indicate that they all contain two nucleotide binding domains, and their transport functions require ATP. Therefore, they may all be transport ATPases. However, although ATP has been shown to bind to MRP (24), ATPase activity has not yet been demonstrated for any of the MRP or multispecific organic anion transporters, indeed none of the proteins have yet been purified. To fill these gaps in our understanding of MRP, we have cloned its cDNA in a plasmid (pNUT; Ref. 25) enabling high level stable expression in mammalian cells, generating sufficient protein for purification to homogeneity and assay of ATPase activity in a lipid environment in the absence and presence of compounds that it is able to transport. AGA TAT GCC AGC). These PCR-amplified fragments were cloned into pBluescript, digested with the restriction endonucleases indicated in Fig. 1, and ligated together to generate a complete MRP cDNA (from nucleotide 197 to nucleotide 4824) (1). This MRP cDNA was sequenced completely. To facilitate the purification of MRP, we inserted sequence coding for 10 histidine residues before the stop codon. Two primers were used in the PCR reaction to insert the 10 His residues: MRPm (CCT GTT TGC GGT GAT CTC CA) and MRP/His/3Ј (TGA TAT CTA GTG ATG GTG ATG GTG ATG GTG ATG GTG ATG CAC CAA GCC GGC GTC TTT GGC). The sequences of this fragment generated by PCR were fully verified after its insertion in the MRP cDNA. This modified fulllength MRP cDNA with 10 C-terminal histidine residues was inserted into the pNUT expression vector, which we have employed previously (26) and designated pNUT-MRP/His.
Cell Culture and Stable Transfection of MRP-Baby hamster kidney (BHK-21) cells were cultured at 37°C in 5% CO 2 . Subconfluent cells were transfected with pNUT-MRP/His in the presence of 20 mM HEPES (pH 7.05), 137 mM NaCl, 5 mM KCl, 0.7 mM Na 2 HPO 4 , 6 mM dextrose, and 125 mM CaCl 2 (26). The cells were shocked 5 h later with 25% glycerol. After a further 24 h, 500 M methotrexate was added to the medium. Cells continued to grow in the selective medium for about 10 days. Surviving individual colonies were picked and amplified in the selective medium.
MRP Protein Detection-To detect MRP expression in BHK cells by immunofluorescence, cells grown on coverslips were fixed in 70% cold methanol for 10 min. The cells were subsequently permeabilized in 1% Triton X-100 in phosphate-buffered saline at room temperature for 5 min and blocked with 0.5% goat serum. The permeabilized cells were incubated with the rat monoclonal antibody MRPr1 (27) at room temperature for 60 min. After rinsing several times, the cells were incubated with goat anti-rat antibody-fluorochrome conjugates.
To detect MRP protein in whole cell lysates by immunoblotting, cultures grown up from individual colonies were lysed with 1% SDS, and DNA was sheared by cycling through a 30-gauge needle before SDS-PAGE (6% polyacrylamide gel). The sheared cell lysate was mixed with sample buffer and subjected to SDS-PAGE. Amounts of lysate protein loaded are indicated in the figure legends. The samples (after electrophoresis) were electroblotted to nitrocellulose membrane and probed with the rat anti-MRP monoclonal antibody MRPr1. The secondary antibody was anti-rat Ig conjugated with horseradish peroxidase. Chemiluminescent film detection was performed according to the manufacturer's recommendations.
Chemosensitivity Assay-The resistance of MRP/His expressing cells to anticancer agents was tested using the MTT assay (28). Briefly, the parental BHK cells or the MRP-expressing BHK cells were plated in a volume of 200 l at 2.5 ϫ 10 4 cells per well in 96-well plates. After incubation at 37°C overnight, the media were replaced with fresh media containing different concentrations of drugs. The cells were incubated for a further 4 days. At the end of drug exposure, media containing 0.5 mg/ml MTT were added to the wells, and the plates were incubated for an additional 4 h. The media were removed, and the plates were dried overnight. 200 l of isopropanol:1 N HCl (24:1) were added to solubilize the formazan crystals at 37°C for 1 h, and then the absorbance at 570 nm was determined. IC 50 was defined as being the dose of drug that reduced this absorbance to 50% of control values.
Purification of MRP Protein-A crude membrane fraction was prepared from BHK cells expressing MRP/His by homogenizing them in a cold (4°C) hypotonic buffer A containing 10 mM KCl, 1.5 mM MgCl 2 , and 10 mM Tris-HCl (pH 7.4). Nuclei were removed by centrifugation at 300 ϫ g for 15 min. Membranes were collected at 31,000 ϫ g for 30 min. The pellet was resuspended in buffer B (0.6 ml/15-cm-diameter dish) containing 20 mM Tris/HCl (pH 7.9), 300 mM NaCl, 5 mM imidazole, 1% (w/v) n-dodecyl-␤-D-maltoside (DDM), 20% glycerol, and 0.4% sheep brain lipid. The sample was sonicated, and the insoluble material was removed by centrifugation at 10,000 ϫ g for 15 min. The supernatant was applied onto a His.Bind Resin column, which had been pre-equilibrated with buffer B. The column was then washed with 6 column volumes of buffer B containing 0.1% DDM and 25 mM imidazole. The column was washed once more with 6 column volumes of buffer B containing 0.1% DDM and 40 mM imidazole. The protein was eluted with 2 column volumes of buffer B containing 0.1% DDM and 300 mM imidazole. The eluate was diluted with an equal volume of buffer C containing 50 mM Tris-HCl (pH 7.4), 2 mM MgCl 2 , 20% glycerol, 0.4% sheep brain lipid, and 0.2% ␤-mercaptoethanol. The diluted sample was mixed with buffer C pre-equilibrated wheat germ lectin Sepharose 6MB and shaken for 2 h. The wheat germ lectin Sepharose 6MB beads were packed into a small column and washed with 10 column volumes of buffer C. The bound proteins were eluted with 2 column volumes of buffer containing 50 mM Tris-HCl (pH 7.4), 20% glycerol, 0.4% sheep brain lipid, and 0.3 M N-acetylglucosamine.
Assay of ATPase Activity-ATPase activity was assayed employing the methods of Gibson et al. (29) with some modifications. Briefly, the reaction buffer (40 mM Tris-HCl (pH 7.4), 1 mM dithiothreitol, 10 mM MgCl 2 ), [␥-32 P]ATP (1 Ci), and ATP (2 mM or as otherwise indicated in the figure legends) was placed in an ice bath and mixed with transported substrates where indicated. The purified MRP protein was added last, mixed well, transferred to a 37°C water bath, and incubated for 1 h. The reaction was stopped by adding 20 reaction volumes of ice-cold stop buffer containing 50 mM Tris-HCl (pH 8.0), 2 mM KH 2 PO 4 , 9 mM ATP, 5.6 mM MgSO 4 , and 8.2% HClO 4. The released P i was precipitated with 21 reaction volumes of ice-cold precipitation solution containing 3% (w/v) ammonium molybdate, 67 mM HCl, 1% (v/v) triethylamine, and 1% (v/v) bromine water. After incubation on ice for 20 min, the precipitate was separated from supernatant by centrifugation in a microfuge for 5 min. The supernatant was transferred to a scintillation vial, and the pellet was then washed once with 1 ml of 1 N HCl. The pellet was dissolved in 0.6 ml of 1 N NaOH and transferred to another scintillation vial. The amounts of 32 P radioactivity in supernatant and pellet were counted separately, and specific ATPase activity was calculated.

RESULTS
Sequence of Expressed MRP cDNA-Sequencing on both strands of the full-length cDNA generated from Hela cell RNA as described under "Experimental Procedures" and Fig. 1 indicated identity with the human MRP sequence originally determined by Cole et al. (1) except for single nucleotide substitutions at nine positions. These differences and the resulting amino acid changes are as follows: Thr 434 to Gly (L80V), Cys 546 to Thr (T117M), Cys 1304 to Gly (L370V), Thr 1880 to Cys (L582L), Thr 2250 to Cys (L685S), Ala 2283 to Gly (D696G), Cys 4040 to Gly and Gly 4041 to Cys (R1282A), and Gly 4732 to Ala (S1512S). The changes underlined were also found by Zaman et al. (10) in a cDNA synthesized with RNA from a human tumor cell line as template for reverse transcription. L685S and R1282A changes were also found by Stride et al. (31). The locations of the amino acid changes with respect to the putative domain structure and membrane topology are as follows. The first two, L80V and T117M fall within the 220-residue N- The amplified fragments were cloned into pBluescript. BamHI, NcoI, HindIII, and EcoRI indicate the locations of those restriction endonuclease sites. Fragments were derived from the above subclones by digesting the DNAs with these enzymes and then ligated together to generate full-length MRP cDNA, to which codons for 10 histidine residues were added before transfer to the pNUT expression plasmid.
terminal extension, which distinguishes MRP and some other organic ion transporters from members of the superfamily such as P-glycoprotein and CFTR; however, no specific function has yet been assigned to this putative domain (30). L370V represents a conservative hydrophobic replacement in TM2 of the currently favored topology model (31). L685S is immediately C-terminal of the Walker A lysine residue of NBF1; interestingly, a hydroxyl amino acid is frequently found at this position in the P-loop of other ATPases (32). Similarly, an acidic residue is sometimes present in these other molecules at the position comparable with the D696G change. It is difficult to speculate on the possible significance of the R1282A substitution because positively charged residues usually are present at this position in the hydrophilic region between the last TM and NBF2. In fact, some of these nucleotide substitutions could just reflect PCR or other cloning artifacts or polymorphisms among the individuals from which the RNA originated. Regardless of their basis, the potential consequences of amino acid changes were of primary concern in the present context. Therefore, it was important to demonstrate that the cDNA imparted drug resistance on cells into which it was transfected. Functional testing was also essential to detect any possible deleterious effects of the C-terminally added 10 histidine residues.
MRP Expression in Stable BHK Transfectants-When examined by immunofluorescence ( Fig. 2A), stable transfectants of BHK cells stained strongly with an MRP-specific monoclonal antibody (MRPr1) (27). Greatest intensity is observed at the cell surface but also in association with intracellular membranes including the perinuclear region, which may represent the Golgi apparatus as well as some more diffuse reticular structures typical of the endoplasmic reticulum. Untransfected BHK cells were essentially unstained (Fig. 2B). Western blots (Fig. 2C) of lysates of the same clonal MRP-expressing BHK cells stained in Fig. 2A revealed a single strong MRPr1 antibody-reactive band with an apparent molecular mass of approximately 190 kDa, similar to that seen in other MRP-expressing cells (24,33). MRP was not detected in a lysate of the parental BHK cells nor was their endogenous gene amplified or induced during the methotrexate selection because it was also not detected in BHK cell transfectants similarly selected for CFTR expression. Its expression was readily detected with the CFTR-specific antibody M3A7 (34) as shown in Fig. 2D. Pglycoprotein expression could not be detected in similar blots of any of the three cell lines used in Fig. 2 (not shown).
To determine whether MRP is functional in these BHK cells, their survival when grown in the continued presence of several cancer drugs was measured employing the MTT assay (28). As indicated in Table I, the MRP-expressing cells were approximately 5-fold more resistant to colchicine than untransfected BHK cells, 5-to-10-fold in the case of the anthracyclines, doxorubicin and daunomycin, and more than 20-fold more resistant to vincristine. Although not as highly resistant as tumor cells actively selected from survival in these agents (2,35), these BHK cells do exhibit a cross-resistance pattern typical of MRP overexpressing variants of several different cell types (8,10,20,36).
Purification of MRP-A crude microsomal membrane fraction containing a large proportion of the human MRP expressed in BHK cells was employed as starting material. To solubilize the protein from these membranes, the non-ionic detergent, n-dodecyl-␤-D-maltoside, was effective in either the presence or absence of added phospholipids. Because of the requirement of P-glycoprotein ATPase activity for phospholipids (37,38), we included 0.4% sheep brain lipids during DDM solublization and maintained its presence throughout each step of the purification. Fig. 3 illustrates the fractionation of total membranes and MRP during the two-step purification. Under the conditions of loading and washing of Ni-NTA agarose beads, which were arrived at empirically (see "Experimental Procedures"), the MRP/His was quantitatively bound and, upon elution with elevated imidazole (lane 5), was already very highly enriched and readily detectable by Coomassie Blue staining. However, when the gel was loaded to allow detection at this level, the presence of multiple additional bands was observed. The Nlinked oligosaccharide chains demonstrated to be present on MRP (24) made the use of the lectin-affinity chromatography an obvious choice for further purification. When the MRPcontaining material eluted from the Ni-NTA agarose was applied to wheat germ lectin Sepharose, not all MRP was bound. Some was detectable in both the flow-through (lane 6) and wash (lane 7). However, a larger proportion did bind specifically via N-acetylglucosamine-containing determinants and was eluted by a high concentration of this monosaccharide (lane 8). Importantly, most of the contaminating bands in the nickel bead eluate could be detected in the flow-through and wash from wheat germ lectin beads and not in the N-acetylglucosamine eluate (lanes 6 -8). Hence, these two successive affinity purification steps yield a very highly enriched preparation of the MRP glycoprotein in a lipid matrix, which might be expected to provide an approximation of its native environment. We have not quantitated purification factors or yields at each step, but the entire procedure produces, from one 150-cm 2 plate of confluent cells, approximately 2 g of purified MRP. Aside from the major MRP monomer band of approximately 190 kDa, the only additional staining band present in the final preparation has an apparent molecular weight approaching that of a homodimer, although its size has not been more  accurately determined. This band clearly contains MRP as indicated by its detection with the specific monoclonal antibody as well as by staining (Fig. 3). Although it is more evident once the protein has been purified to at least the Ni-NTA column stage, a small amount was detected even in the membranes before solubilization. A similar large immunoreactive band is observable in immunoblots of other drug-resistant cells expressing high levels of MRP (11,33). ATPase Activity of Purified MRP-To determine if purified MRP eluted from wheat germ lectin beads in the absence of detergent but in the presence of 0.4% sheep brain lipid-possessed intrinsic ATPase activity, the release of 32 P from [␥-32 P]ATP was assayed. Fig. 4A indicates that indeed ATP was hydrolyzed in direct proportion to the amount of purified MRP present, at least over a range of 50 -250 ng, indicating that it is active as an ATPase.
The dependence of the rate of hydrolysis on ATP concentration shown in Fig. 4B exhibited typical Michaelis-Menten behavior and, when expressed as a Lineweaver-Burk plot in Fig.  4C, showed a linear relationship, yielding a K m for ATP of approximately 3 mM and a V max of 0.46 mol mg Ϫ1 min Ϫ1 . This maximal specific activity is of the same order as the basalspecific ATPase activity of P-glycoprotein in the absence of drugs (37)(38)(39).
Effect of Transported Substrates on MRP ATPase-MRPcontaining membrane vesicles have been shown to transport glutathione and glucuronide conjugates of a range of hydrophobic compounds including some of the cancer drugs to which MRP overexpressing cells are resistant (9, 11, 13, 14, 19 -21, 40). Although there is some controversy over whether conjugation of substrate is essential to transport by MRP (41), it seems that the presence of a conjugating agent such as glutathione is necessary (12,19). We first tested the influence of the cysteinyl leukotriene, LTC 4 , one of the best characterized MRP substrates for transport (9,11,13,42), on ATPase activity (Fig. 5). 1 nM LTC 4 caused an increase in specific activity of approximately 25%, and this enhancement increased further to as much as 60% at 10 M. Hence, while MRP is clearly not de-pendent on such substrates for its ATPase activity, there is significant stimulation.
We next tested the influence of the cancer drugs to which the MRP-transfected cells exhibited cross-resistance. Fig. 6A shows that the anthracyclines, daunomycin and doxorubicin, also both cause moderate augmentation of MRP ATPase, with doxorubicin having a greater influence at higher concentrations. The vinka alkaloids, vinblastine and vincristine, had similar stimulatory effects, with the latter eliciting its activation at lower concentrations. Increasing concentrations of colchicine caused a smooth rise in ATPase-specific activity. Hence, while the response of MRP ATPase to these drugs is moderate, it is quite similar in magnitude to that of P-glycoprotein ATPase (37)(38)(39).
Because glutathione-conjugated compounds seem to be the preferred substrates for transport by MRP (9,11,13,42), we also examined the influence of reduced and oxidized glutathione in the presence and absence of doxorubicin (Table II). Reduced glutathione alone had a minor stimulatory effect whereas the oxidized form, which has itself been reported to be transported by MRP (15), caused a much stronger stimulation. Interestingly, the addition of GSH together with doxorubicin seemed to increase activity more than when GSSG was combined with the anthracycline. Interpretation of this observation in the presence of only purified MRP and no conjugating or other enzymes is not obvious at this stage. Nevertheless, it does seem that the presence of glutathione increases the drug activation and that oxidized glutathione alone is capable of substantial activation. These observations support the conclusion that several substrates for transport by MRP enhance but are not required for ATPase activity of the protein. DISCUSSION We have developed stable mammalian cell transfectants capable of high level expression of functional MRP designed to facilitate its purification. This enabled its separation from other membrane proteins by two successive affinity chromatography steps in the continuous presence of membrane lipids. The purified glycoprotein hydrolyzed ATP with a specific activity of approximately 0.5 mol mg Ϫ1 min Ϫ1 and a K m for ATP of about 3 mM. Activity could be stimulated by a maximum of 2-fold by compounds that have been previously shown to be transported by MRP-containing membrane vesicles.
The stable BHK cell lines, in which integrated MRP sequences were highly amplified because of strong selection of the dihydrofolate reductase-containing pNUT (25) plasmid with methotrexate, seemed to produce the mature glycoprotein at higher levels than some other heterologous expression systems, surprisingly, even including the baculovirus-insect cell system (18). This could reflect the fact that this amplification at the DNA level may to some extent mimic the situation in MRP overexpressing cells selected for resistance to drugs which it transports. Transcription from the metallothionein gene promoter in the pNUT plasmid also occurs at a high level. Pulsechase experiments also revealed that the core-glycosylated form present in the endoplasmic reticulum matured with the addition of complex oligosaccharide chains in a relatively efficient fashion (data not shown). The 190-kDa band representing this mature form of the protein is formed on passage through the Golgi complex, but our immunocytochemical observations (Fig. 2) indicated that a large amount of the protein is present at the cell surface in the steady state, indicating that most of it had proceeded beyond the Golgi. Much of the protein has been reported to be in intracellular membranes in some cells and tissues, and one report (43)  resistance correlated with the conversion of a lower molecular mass band to the mature 190-kDa band (43). It was the principal form of the molecule in the transfected BHK cells, although a larger band, either a dimer or a monomer in stable association (resistant to SDS and DTT) with another molecule, is also present.
The protein with the addition of 10 histidine residues at the C terminus was apparently active because the cells exhibited a cross-resistance pattern similar to that reported for several different MRP-expressing cell types, although the resistance to vincristine was higher than in some of the others. Because these cells had never been subjected to any selection pressure by this drug or the anthracyclines, it seems extremely likely that resistance is solely due to their expression of MRP.
Since the objective was to purify a fully functional molecule, we solubilized MRP-containing membranes in a non-ionic detergent in the presence of excess lipid. Lipid was present throughout the purification, and in the final step, elution from the lectin beads, detergent was omitted. DDM had been used previously by Loo and Clarke (44) for the purification of Pglycoprotein by Ni-NTA affinity chromatography. They stated that phospholipids interfered with that step. While perhaps not optimal, the data in Fig. 3 suggest that the binding to and elution of MRP from Ni-NTA agarose was quite efficient. It is possible that the presence of phospholipid micelles with which other hydrophobic proteins might associate contributed to the significant contamination that remained following this step. However, if that were the only explanation, one might have expected the contaminants to also carry through the wheat germ lectin affinity step, and they do not. Although not optimized for yield, the purification provides abundant amounts for enzymatic assays and is sufficient for many but not all other purposes; 100 g was obtained in a preparation using routine cell culture methods and scale up by at least one order of magnitude is immediately feasible.
Human MRP purified in this way has ATPase activity that is at least superficially similar to that of P-glycoprotein in that the V max is in the mol mg Ϫ1 min Ϫ1 range and the K m is in the mM or physiological range. It is important to note that this K m for ATP is much higher than that determined in very rigorous transport assays with several different substrates in MRPcontaining membrane vesicles (13,14). We have no explanation of this apparent discrepancy at this stage, but as mentioned further below there is still no compelling evidence that MRP, P-glycoprotein, or other structurally related members of that class of molecules function mechanistically as classical transport ATPases with obligatory tight coupling between ATP hydrolysis and translocation of solute. However, there are also several possible technical reasons for the discrepancy. The membrane vesicles in which the transport assays have been performed contain many proteins in addition to MRP, which could conceivably increase its affinity for ATP. On the other hand, despite our attempt to maintain a membranous environment, the purified molecule could well have lost a higher ATP Ci of [␥-32 P]ATP, respectively, and transferred to a 37°C water bath to start the reaction. To stop the reaction, the samples were transferred to an ice bath and 20 volumes of cold stopping solution plus 21 volumes of cold precipitation solution added. Protein-free 0.4% sheep brain lipid was used as negative control. The amount of P i released and intact ATP was determined as described under "Experimental Procedures." The amount of free P i detected in the negative control was subtracted from each reaction. Each point shows the mean Ϯ S.D. of triplicate experiments. In some cases the standard deviation bars are smaller than the symbols. B, effects of increasing ATP concentration on MRP ATPase activity. 192 ng of purified MRP protein in 0.4% sheep brain lipid was utilized in each reaction. The reaction conditions were the same as in A, except for the varying ATP. C, Lineweaver-Burk. The data derived from B were converted to 1/V and 1/S and plotted. The linear fit of the data illustrated was obtained. V max and K m values for ATP were estimated to be 463 nmol mg Ϫ1 min Ϫ1 and 2950 M, respectively. Similar V max and K m values were also derived from an Eadie-Hofstee plot (plot of V against V/S, not shown). affinity, which it may process in its truly native state. It will be necessary to characterize MRP ATPase much more extensively in terms of inhibitor sensitivity and other properties. In vitro mutagenesis should readily reveal the contribution of each of the two nucleotide binding domains to activity.
The extent of ATPase activation by drugs to which MRP overexpressing cells are resistant in the presence or absence of GSH or by GSSG is modest and similar to that of P-glycoprotein by hydrophobic drugs. In the case of P-glycoprotein, this has been considered by some investigators to reflect high basal activity (38), possibly due to activation by lipids. We have not yet attempted to distinguish between the possibilities of artifactually elevated basal activity and low inherent sensitivity to transported substrates in the case of MRP ATPase.
Nevertheless, from previous work and the experiments described in this paper, it would seem that MRP, like P-glycoprotein, now qualifies as a transport ATPase, at least according to the loose criteria that it contributes to the transport of conjugated hydrophobic xenobiotics and metabolites and has ATPase activity that is influenced by these substances. However, in comparison with prototypic transport ATPases such as sodium and calcium pumps, for example, dependence of ATP hydrolysis on solute that is able to be translocated is much less complete, and it has not yet been shown that MRP alone is sufficient for transport. Demonstration of the latter property has been difficult even for P-glycoprotein, which has been much more extensively studied (45). However, given the hydrophilicity imparted upon the hydrophobic substrates by conjugation with either a cysteinyl peptide or a glucuronide moiety, it may be easier to perform appropriate transport assays with large proteoliposomes containing only purified MRP than in the case of P-glycoprotein.