Functional properties of the copper-transporting ATPase ATP7B (the Wilson's disease protein) expressed in insect cells.

Copper-transporting ATPase ATP7B is essential for normal distribution of copper in human cells. Mutations in the ATP7B gene lead to copper accumulation in a number of tissues and to a severe multisystem disorder, known as Wilson's disease. Primary sequence analysis suggests that the copper-transporting ATPase ATP7B or the Wilson's disease protein (WNDP) belongs to the large family of cation-transporting P-type ATPases, however, the detailed characterization of its enzymatic properties has been lacking. Here, we developed a baculovirus-mediated expression system for WNDP, which permits direct and quantitative analysis of catalytic properties of this protein. Using this system, we provide experimental evidence that WNDP has functional properties characteristic of a P-type ATPase. It forms a phosphorylated intermediate, which is sensitive to hydroxylamine, basic pH, and treatments with ATP or ADP. ATP stimulates phosphorylation with an apparent K(m) of 0.95 +/- 0.25 microm; ADP promotes dephosphorylation with an apparent K(m) of 3.2 +/- 0.7 microm. Replacement of Asp(1027) with Ala in a conserved sequence motif DKTG abolishes phosphorylation in agreement with the proposed role of this residue as an acceptor of phosphate during the catalytic cycle. Catalytic phosphorylation of WNDP is inhibited by the copper chelator bathocuproine; copper reactivates the bathocuproine-treated WNDP in a specific and cooperative fashion confirming that copper is required for formation of the acylphosphate intermediate. These studies establish the key catalytic properties of the ATP7B copper-transporting ATPase and provide a foundation for quantitative analysis of its function in normal and diseased cells.

P-type ATPases form a large family of membrane proteins found in all living organisms from bacteria to man (1)(2)(3). The major function of these proteins is to couple the energy of ATP hydrolysis with the uphill transport of a variety of cations across cell membranes. The P-type ATPases share several highly conserved sequence motifs, most of which are located in the ATP-hydrolyzing domain of these proteins (Fig. 1). For several members of the family, it has been shown that the invariant Asp residue in the DKTG motif accepts ␥-phosphate from ATP during the catalytic cycle, forming a covalent acylphosphate intermediate (see for example, Refs. 4 -8). The formation of a phosphorylated intermediate is a signature property of these transporters, which also gave the name to this family.
More than 150 members of the P-type ATPase family have been identified. Current classification divides the P-type ATPases into 5 subfamilies based on their sequence, cation specificity, membrane topology, and presence of various regulatory domains (2). Among these diverse molecular machines, ATPases involved in the transport of copper form one of the most intriguing and least characterized groups. The first mammalian copper-transporting ATPases were identified as the result of a search for genes affected in patients with disorders of copper metabolism, i.e. Menkes disease and Wilson's disease (9 -12). Menkes disease is caused by various mutations or deletions in the gene ATP7A and is associated with overall copper deficiency due to impaired export of copper from intestinal cells. Wilson's disease, in contrast, is caused by copper accumulation predominantly in the liver, brain, and kidneys as a result of mutations in the ATP7B gene. The Menkes disease protein (MNKP) 1 and Wilson's disease protein (WNDP) are homologous and both contain sequence motifs characteristic of the P-type ATPases.
Both WNDP and MNKP, as well as their numerous homologues from other organisms belong to the P 1 -subfamily of cation-transporting P-type ATPases. This family is characterized by their selectivity toward soft and transition metals and by the presence of metal-binding sites in the N-terminal region of the protein, 8 transmembrane segments (6 segments before, and 2 segments after the ATP-binding domain, Fig. 1), the conserved CP(C/H) motif in the transmembrane segment 6, and SEHPL sequence located in the ATP-binding domain downstream of the putative phosphorylation site DKTG (3,13). Very few, predominantly bacterial, members of the P 1 -subfamily have been functionally characterized (see, for example, Refs. 7, 14, and 15).
The ability of MNKP and WNDP to transport copper has been demonstrated (8,(17)(18)(19) and recent studies from Voskoboinik et al. (8,20) confirmed that copper transport by these proteins is ATP-dependent. Further characterization of MNKP indicated that this protein formed a phosphorylated intermediate in a manner similar to other P-type ATPases (8). The enzymatic properties of WNDP have not yet been described.
Earlier we isolated and provided detailed biochemical characterization of two major functional domains of WNDP, the copper-binding domain and the ATP-binding domain (21,22).
In the present work we evaluate functional characteristics of the full-length WNDP utilizing baculovirus-infected insect cells, a system previously optimized for expression of Na,K-ATPase, a P 2 -type ATPase (23). Using this approach we were able to obtain a high yield expression of the full-length WNDP and analyze its major functional properties.

EXPERIMENTAL PROCEDURES
Generation of Recombinant Baculovirus-The 4.4-kb fragment containing the full-length WNDP cDNA was excised from the pMT2-WNDP plasmid using digestion with the restriction nuclease SalI and partial digestion with EcoRI (partial digestion with EcoRI was required due to the presence of another EcoRI site inside the WNDP cDNA sequence). The obtained fragment was cloned into the pFastBacDual vector (Invitrogen, Carlsbad CA) digested with the same enzymes and was placed under the control of the polH promoter. The resultant plasmid pFast-BacDual-WNDP (pWNDP) was then utilized to generate the recombinant WNDP-expressing baculovirus using the commercially available Bac-to-Bac kit (Invitrogen) and previously described protocols (23). DH10 Bac cells were transformed with pWNDP and allowed to generate bacmids via a transposition mechanism as previously described (24). The WNDP bacmids were then used to transfect Spodoptera frugiperda 9 (Sf9) cells and produce baculovirus expressing WNDP. Baculovirus was amplified as described in the Bac-to-Bac TM manual and in Ref. 23.
The D1027A mutant of WNDP was generated by polymerase chain reaction. The oligonucleotides, 5Ј-GTTTGCCAAGACTGGCACCAT-3Ј and 5Ј-TGGTGCCAGTCTTGGCAAACATC-3Ј, were used as forward and reverse primers, respectively, to introduce the mutation and the 5Ј-TGTGCATTGCCTGCCCCT-3Ј and 5Ј-ACCACAGCCAGAACCTT-C-3Ј oligonucleotides were used as a forward and reverse flanking primers. Following PCR, the fragment containing the D1027A mutation was digested with the XcmI and Bsu36I restriction endonucleases and the XcmI-Bsu36I fragment was exchanged with the corresponding fragment of the wild-type WNDP cloned into the pFastBacDual plasmid. The presence of the D1027A mutation in the final construct and the absence of the unwanted mutations were confirmed by automated DNA sequencing.
WNDP Expression in Insect Cells and Preparation of Membrane Fractions-Sf9 cells (Invitrogen) were maintained at 27°C in 150-ml suspension cultures in the Ex-Cell TM 420 growth medium (JRH Biosciences Inc., Lenexa, KS) and were split every 2-3 days with fresh medium to maintain cell densities between 0.5 ϫ 10 6 and 4 ϫ 10 6 cells/ml. Cells were infected with recombinant virus as previously described (23,25), and harvested 3 days post-infection. To obtain a total membrane preparation, the cells were centrifuged at 500 ϫ g for 10 min and cell pellet was frozen at Ϫ20°C and then thawed to facilitate lysis. Cells from a 50-ml culture were pelleted and resuspended in 4 ml of homogenizing buffer (HB): 25 mM imidazole, pH 7.4, 0.25 M sucrose, 1 mM dithiothreitol (1 tablet of Roche complete protease inhibitor mixture was added per 50 ml of buffer solution).
Cell were lysed by a 20-stock homogenization in a Dounce homogenizer, and then centrifuged for 10 min at 500 ϫ g. The pellet was discarded and the soluble fraction was subjected to an additional centrifugation for 30 min at 20,000 ϫ g to sediment cell membranes. The pelleted cell membranes were then resuspended in 0.5 ml of HB and stored frozen at Ϫ80°C until further use. Protein concentration was determined by the method of Lowry (26). To analyze protein expression 1 g of total membrane protein was separated by a 7.5% Laemmli gel (27), transferred to an Immobilon-P membrane using 10 mM CAPS, pH 11, 10% methanol, and stained with polyclonal antibody a-ABD (1: 20,000 dilution) directed against the central region of WNDP (22). To test subcellular localization of WNDP, the total membrane preparation was fractionated using a discontinuous five-step sucrose gradient and analyzed by Western blotting, as previously described (23,25).
In several experiments, a baculovirus construct expressing both the ␣ and ␤ subunits of the Na,K-ATPase was used as a control. The construct was generated using the pFastBacDual vector, in which the ␣ subunit of Na,K-ATPase was placed under the control of the p10 promoter while the ␤ subunit was expressed from the polH promoter. The expression of Na,K-ATPase was carried out using previously described protocols (23,25). Analysis of protein expression was performed by Western blotting using a mouse anti-Na,K-ATPase ␣-1 antibody (Affinity Bioreagents, Inc., Golden, CO).
Phosphorylation of WNDP from [␥-32 P]ATP-50 g of total membrane protein was resuspended in 200 l of the assay buffer: 20 mM bis-Tris propane, pH 6.0, 200 mM KCl, 5 mM MgCl 2 . Radioactive [␥-32 P]ATP (5 Ci, specific activity 20 mCi/mol) was added to a final concentration of 1 M and the reaction mixture was incubated on ice for 4 min or for various time periods in the case of kinetic measurements. All additional treatments were done as described in the figure legends. The reaction was stopped by addition of 50 l of ice-cold 1 mM NaH 2 PO 4 in 50% trichloroacetic acid and then centrifuged for 10 min at 20,000 ϫ g. The protein pellet was washed once with ice-cold water and resuspended in 40 l of sample buffer (5 mM Tris-PO 4 , pH 5.8, 6.7 M urea, 0.4 M dithiothreitol, 5% SDS) and loaded on the acidic 7.5% polyacryalmide gel (28). After electrophoresis, the gels were fixed in 10% acetic acid for 10 min and dried on a blotting paper. The dried gels were exposed either overnight to the Molecular Imaging screen CS (Bio-Rad) or for several hours at Ϫ80°C to the Kodak BioMax MS film and the intensity of the bands was quantified using a Bio-Rad Molecular Imager or Bio-Rad densitometer, respectively. Then, the dried gels were re-hydrated, stained with Coomassie, and the amount of protein in the WNDPrelated bands was determined by a second round of densitometry. The 32 P incorporation into WNDP was then normalized to the WNDP protein levels. Phosphorylation of Na,K-ATPase and analysis of 32 P incorporation was carried out identically, with the exception that the phosphorylation buffer contained 150 mM NaCl instead of KCl.
Analysis of Endogenous WNDP in HepG2 Cells and WNDP, Transiently Expressed in COS-7 Cells-Human hepatoma (HepG2) cells were grown in minimum essential medium (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (Invitrogen), 0.1 mM minimal essential medium, nonessential amino acids, and 1 mM sodium pyruvate (basal medium) in a 37°C, 5% CO 2 , 95% air incubator. COS-7 cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 100 units/ml of penicillin-streptomycin (all from Invitrogen) at 37°C, 5% CO 2 . For transfection with the pMT2-WNDP plasmid, COS-7 cells at 50 -70% confluency were rinsed once with phosphate-buffered saline and treated for 45 min under growth conditions with DNA/DEAE-dextran solution: 2 g of plasmid DNA, 250 g of DEAE-dextran (Amersham Bioscience, Inc., Peapack, NJ) per 1 ml of phosphate-buffered saline. The cells were then incubated in growth medium containing 57 g/ml chloroquine (ICN, Costa Mesa, CA) for 3 h under growth conditions. Next, the cells were treated with 10% dimethyl sulfoxide in phosphate-buffered saline for 1.5 min and then rinsed with phosphate-buffered saline. Growth medium was added and the cells were grown under normal conditions for 24 -36 h. To compare protein expression, cells grown on a 15-cm plate (HepG2 at 90 -95% confluence and COS-7 cell 24 h after transfection) were scraped in 4 ml of HB buffer using rubber "policeman" and thoroughly resuspended. Cells were then homogenized and membrane fractions were prepared as described above for the insect cells.
Treatment with Bathocuproine Disulfonate (BCS) and Copper Reactivation Experiments-Membrane preparations from insect cells expressing WNDP or Na,K-ATPase were resuspended in the respective phosphorylation buffers to a protein concentration of 0.25 mg/ml and then incubated on ice with increasing concentrations of the copper chelator BCS for 15 min. BCS was then removed by sedimentation of the reaction mixture at 20,000 ϫ g for 5 min, the pellets were resuspended in the respective phosphorylation buffer, and analysis of phosphorylation from ATP was carried out as described above.
For metal reactivation experiments, ascorbate alone or ascorbate with tris-(2-carboxyethyl)phosphine hydrochloride (TCEP, Sigma) were added to the membrane protein to a final concentration of 100 M each before addition of BCS. Following a 15-min incubation of the membrane preparation with 100 M BCS, the chelator was removed by centrifugation as described above. The membrane pellets were resuspended in the phosphorylation buffer containing either ascorbate or ascorbate with TCEP and increasing concentrations of CuCl 2 (or ZnCl 2 and CdCl 2 at 4 or 20 M) were added to the mixture. Following 10 min incubation on ice the radioactive ATP was added and the phosphorylation reaction and analysis of 32 P incorporation were carried out as described above. The data were analyzed by a nonlinear regression using SigmaPlot software. The highest r 2 values were obtained using either sigmoidal function, where n was 2.6 Ϯ 0.6 (r 2 ϭ 0.98).

Characterization of the WNDP Expression in Insect Cells-
The ATP7B gene encoding human copper-transporting ATPase WNDP was characterized several years ago (29), however, analysis of the WNDP enzymatic properties has been hindered by very low levels of endogenous protein (ϳ0.005% of total cell membrane protein), and by insufficient levels of expression of WNDP following transfection of mammalian cells. Currently, the most frequently used system for analysis of WNDP and its disease-related mutants utilizes a yeast strain, in which the endogenous copper-transporting ATPase CCC2 has been deleted (30). Deletion of CCC2 disrupts copper delivery to the secretory compartment and renders the copper-dependent ferroxidase Fet3 inactive, while expression of WNDP restores the Fet3 activity to some extent (18,31). This elegant assay permits rapid screening of mutants (32), however, the indirect nature of the assay prevents characterization of the molecular mechanism of WNDP. The recently reported expression of WNDP in Chinese hamster ovary cells does permit direct analysis of copper transport (20), however, the low level of protein expression makes other measurements difficult.
To overcome these problems we constructed a donor plasmid for baculovirus-mediated expression of WNDP (Fig. 1B) and tested expression of WNDP in Sf9 cells. As shown in Fig. 2A, infection of Sf9 cells with baculovirus containing WNDP cDNA leads to a time-dependent appearance of the 165-kDa protein, which reacts with the WNDP-specific antibody on a Western blot. We also detected an additional 140-kDa band, which is likely to represent a proteolytic fragment of WNDP, since the amount of the 140-kDa band can be minimized using a protease inhibitors mixture.
Comparison of WNDP expressed in Sf9 cells with endogenous WNDP present in hepatocytes (HepG2) and with WNDP expressed in COS-7 cells demonstrated that the yield of WNDP per mg of membrane protein in Sf9 cells was about 20-fold higher than in COS cells and about 400-fold higher than in HepG2 cells (Fig. 2B). WNDP represents ϳ2% of the membrane protein in Sf9 cells and this amount of WNDP is sufficiently high to allow its visualization by standard Coomassie staining (Fig. 2C). The apparent molecular mass of the heterologously expressed WNDP corresponds to that of endogenous protein in HepG2 cells (Fig. 2B).
In addition to the high level of protein expression, another important advantage of expression in insect cells is the relative ease with which cell membranes can be fractionated. Centrifugation using a 5-step sucrose gradient permits separation of endoplasmic reticulum, Golgi, and plasma membrane (23,25). Therefore, we utilized this procedure to evaluate the subcellular distribution of the expressed WNDP. As shown in Fig. 3, the major portion of WNDP is found in the Golgi fraction in agreement with its known primary localization in mammalian cells (33)(34)(35). This distribution was different from the distribution of Na,K-ATPase, which was more abundant at the plasma membrane, the primary location of this transporter. A significant portion of WNDP was also seen in endoplasmic reticulum, its site of synthesis, where protein is likely to accumulate due to high levels of expression. Our later studies confirmed that WNDP protein in all fractions is functional, consequently unfractionated membrane preparations were used for most of the experiments described below.
Analysis of Functional Properties of WNDP-The P-type ATPases are known to bind ATP and form an acylphosphate intermediate during their catalytic cycle. The characteristic properties of this intermediate are its sensitivity to treatment with hydroxylamine at neutral pH, sensitivity to basic pH, the transient nature of the intermediate, and reversibility of the reac-tion. To test whether WNDP can form an acylphosphate intermediate, the cell membranes were incubated with 1 M [␥-32 P]ATP and incorporation of 32 P into protein was monitored using an acidic gel electrophoresis (28).
As shown in Fig. 4, A and B, WNDP incorporates ␥-phosphate from ATP, while no radioactive band was present in the lane with a control membrane fraction obtained from the cells mock-infected with the virus lacking WNDP (compare lanes 1 and 2). Further characterization of phosphorylated WNDP revealed that phosphorylation is hydroxylamine sensitive (Fig. 4,  A and B, lane 4), unstable at pH 7 (Fig. 4, A and B, lane 5), and transient, as indicated by the effect of cold ATP (Fig. 4, A and  B, lane 3). ADP at 1 mM concentration apparently reverses the phosphorylation (Fig. 4, A and B, compare lanes 6 and 7), in agreement with the known properties of other P-type ATPases.
WNDP was originally assigned to the family of P-type ATPases because of the presence in its structure of several signature motifs, including the DKTG sequence with an invariant aspartate residue 1027. Substitution of this aspartate with alanine abolishes phosphorylation of WNDP (Fig. 4C), suggesting that Asp 1027 is indeed a target of catalytic phosphorylation. It should be noted that the D1027A substitution does not have FIG. 1. A, schematic representation of the membrane-bound coppertransporting ATPase ATP7B. Gray blocks with CXXC motif indicate the repetitive metal-binding sites in the N-terminal copper-binding region of the protein; TGE, DKTG, TGDN, and MGDGVND are sequences conserved in all P-type ATPases. CPC, a sequence motif characteristic for P 1 -type ATPases, is believed to form an intramembrane metalbinding site. SEHP is another sequence motif conserved in P 1 -type ATPases. B, the donor plasmid pFastBacDual-WNDP used for generation of recombinant baculovirus.

Functional Expression of Copper-ATPase ATP7B
a negative effect on protein expression in insect cells (see Fig.  4), and in COS-7 cells the expression level of mutant was consistently higher than the level of the wild-type protein (not shown).
Finally, analysis of the rate of formation of phosphorylated WNDP revealed that t1 ⁄2 for this reaction is about 30 s and that maximal phosphorylation is achieved in 3-4 min (Fig. 5). Therefore, in subsequent experiments the time of the reaction was set at 4 min.
ATP Dependence of Phosphorylation-We have previously isolated the ATP-binding domain of WNDP and characterized its nucleotide-binding properties (22), however, the nucleotide binding characteristics of the full-length copper-transporting ATPase remained unknown. Consequently, we utilized the phosphorylation assay to assess the apparent affinity of the full-length WNDP for ATP. As shown in Fig. 6A, phosphorylation of WNDP is ATP-dependent with a K m for ATP equal to 0.95 Ϯ 0.25 mol. This value is in agreement with values reported earlier for the ATP-dependent phosphorylation of other P-type ATPases (for example, a K m of 0.5 M was observed for Na,K-ATPase, Ref. 36). Interestingly, the K m for ATP observed in our experiments was significantly lower than the K m for ATP determined for the homologous Menkes disease ATPase (17 Ϯ 7 M) using an ATP-dependent copper transport assay (8).
As predicted for a P-type ATPase, the hydrolysis of ATP by WNDP is reversible and the phosphorylated intermediate formed from ATP is sensitive to the addition of ADP (see Fig. 4 above). To characterize the reverse reaction in more detail, we determined the dependence of dephosphorylation on the concentration of ADP. As shown in Fig. 6B, ADP robustly dephosphorylates WNDP with an apparent K m of 3.2 Ϯ 0.7 mol. The ease with which ADP dephosphorylates WNDP suggests that following phosphorylation from ATP, WNDP is stabilized in a conformation particularly suitable for ADP binding.
The Effect of Copper on Catalytic Activity of WNDP-The presence of a transported cation in the reaction buffer is required for efficient catalytic phosphorylation of various P-type ATPases, therefore we analyzed whether copper can stimulate phosphorylation of WNDP. From the studies shown above, it appears that addition of copper is not required to generate a phosphorylated intermediate of WNDP. Furthermore, additions of copper to the reaction mixture did not stimulate phosphorylation of WNDP. 2 We hypothesized that trace amounts of copper present in the cell growth medium and in buffers (about 1-1.5 M) 2 were sufficient to fully activate WNDP. This idea is not unreasonable, since intracellular concentrations of copper are expected to be extremely low (in yeast Ͻ10 Ϫ18 M (37)). We therefore tested whether chelation of copper in the medium with the copper chelators, bicinchoninic acid and bathocuproine disulfonate (BCS), would have an effect on catalytic phosphorylation of WNDP.
Addition of either bicinchoninic acid or BCS to the reaction mixture in vitro or addition of 50 M BCS to the WNDPexpressing cells 12 h prior to harvesting markedly decreased the level of WNDP phosphorylation, supporting our hypothesis 2 (see Fig. 7 for details on the in vitro experiments). To verify that the effect of chelator is specific and is due to copper removal we compared phosphorylation of WNDP and Na,K-ATPase, a copper-independent P-type ATPase. As shown in Fig. 7, BCS had essentially no inhibitory effect on phosphorylation of Na,K-ATPase, while phosphorylation of WNDP was markedly decreased in the presence of BCS (K m for BCS is 50 M).
These results pointed to the important role of copper in catalytic phosphorylation of WNDP and implied that the inhibiting effect of BCS can be reversed by additions of copper in vitro. Our earlier experiments, as well as studies from other groups, demonstrated that, in vitro, copper could be bound to the copper-binding domains of WNDP and MNKP (21, 38 -40) as well as to the full-length MNKP (8) in the presence of ascorbate or dithiothreitol. However, additions of 0.1-50 M copper to the BCS-treated WNDP in the presence of ascorbate and dithiothreitol, or addition of copper in a complex with Atox1, a protein which is believed to act as an endogenous copper donor for WNDP (41), did not restore phosphorylation of WNDP. 2 It appeared that following treatment with BCS, the copper-binding site(s) on WNDP became unavailable, perhaps due to rapid oxidation. This was somewhat puzzling, because reducing reagents were present in the buffers at all times, and no rapid oxidation of copper-coordinating Cys residues was previously observed for the purified N-terminal domain of WNDP. 2 The intramembrane portion of the P 1 -type ATPases contains . Following incubations the reaction was stopped by acid precipitation and all samples were analyzed by acidic gel electrophoresis (28). The gel was then dried and exposed to BioMax film for several hours at Ϫ80°C or overnight at room temperature. B, the amount of incorporated phosphate was calculated by densitometry and normalized to the amount of the WNDP protein. The gels were dried and exposed for 12 h to the Bio-Max film (left); the intensity of 32 P-labeled bands was quantified using densitometry (right panel).

FIG. 6. The effects of ATP (A) and ADP (B) on formation of the acylphosphate intermediate. A, total membrane fractions (50 g)
from the mock-infected cells or from cells expressing WNDP were incubated with increasing concentration of [␥-32 P]ATP for 4 min on ice. The reaction was stopped and the samples were analyzed by acidic gel electrophoresis. The gels were dried and exposed to the film (left); the intensity of the 32 P-labeled bands was quantified using densitometry (right panel). B, the membrane fractions (50 g) from the mock or the WNDP-expressing cells were incubated with 1 M [␥-32 P]ATP for 4 min. ATP was removed by centrifugation, samples were resuspended in the same phosphorylation buffer, and various concentrations of ADP were added to the preparation. Following 4 min incubation on ice, the reaction was stopped and the samples were analyzed by acidic gel electrophoresis. The gels were then dried and exposed to the film (left); the intensity of the 32 P-labeled bands was quantified using densitometry (right panel). The representative experiments are shown. a highly conserved sequence motif CP(C/H) (see Fig. 1), which was shown to be critical for the transport function of these proteins (18,43). This motif is likely to serve as an intramembrane metal-coordination site. We speculated that the removal of copper by BCS led to rapid oxidation of Cys residues in the CPC motif. Such inactivation due to oxidation of the CPC cysteines was recently observed for the purified bacterial P 1type ATPase, ZntA, and it was reversed by subsequent treatment with dithiothreitol (44). The presence of 50 M dithiothreitol in our buffers was apparently insufficient to prevent oxidation. Consequently, we modified the reactivation protocol and included in the buffer a highly efficient cistine-reducing reagent TCEP in addition to ascorbate.
As shown in Fig. 8, in the presence of TCEP we observed a significant copper-dependent reactivation of the BCS-treated WNDP. Glutathione, which was previously shown to markedly facilitate the metal transport activity of bacterial ATPase, ZntA (44,45), did not substitute for TCEP. 2 Apparently, the reduced state of cysteines, rather than the form in which copper is supplied to WNDP, is critical for the copper-dependent reactivation of this enzyme. Analysis of the data using nonlinear regression revealed that a sigmoidal function provides the best fit for our experimental data (Fig. 8A). This result suggests that the effect of copper on the WNDP activity was cooperative and that more than one copper-binding site has to be occupied to activate phosphorylation of WNDP. The EC 50 for copper in these experiments was 1.5 Ϯ 0.6 M.
Finally we tested whether copper-dependent reactivation of the BCS-treated WNDP is metal-specific. In our earlier experiments we demonstrated that copper binds to the soluble Nterminal domain of WNDP, while zinc and cadmium bind less efficiently, if at all (21). It was interesting to determine whether or not the metal dependence of phosphorylation follows a similar pattern. As shown in Fig. 8B, only copper but not zinc or cadmium was able to stimulate phosphorylation of WNDP, suggesting that the sites at the N-terminal domain and the site(s) essential for the metal-dependent phosphorylation have similar metal specificity. DISCUSSION In this paper we describe a heterologous expression system suitable for direct analysis of the functional properties of the Wilson's disease protein, a human copper-transporting ATPase ATP7B. The baculovirus-mediated infection of Sf9 cells permitted for the first time measurements of the catalytic properties of native, membrane-bound WNDP. Using this system in combination with characterization of isolated domains of WNDP (21,22), we can now thoroughly analyze the functional properties of WNDP and identify specific effects of the Wilson's disease-causing mutations on various steps of the WNDP catalytic cycle. The results described in this paper provide strong evidence that the mammalian copper-transporting ATPase ATP7B has catalytic properties characteristic for the members of the P-type family.
We demonstrate that during its catalytic cycle WNDP forms  Fig. 4. Top, autoradiogram of a typical gel. Bottom, the results of densitometry for five independent experiments; the mean Ϯ S.D. are represented by filled circles and vertical lines, respectively. The solid line is the theoretical sigmoidal curve generated using SigmaPlot software, the r 2 factor is 0.99. B, WNDP was treated with BCS as described above and then various metals at two different concentrations were added to the reaction mixture. Following phosphorylation under standard conditions, equal amounts of protein (50 g) were run on a gel and incorporation of 32 P was analyzed by densitometry. a transient acylphosphate intermediate that can be reversed by excess ADP, the product of the reaction. Copper, an ion transported by WNDP, is required for catalytic phosphorylation, as evidenced by the specific inhibitory effect of copper chelators and by the ability of copper to reactive the BCS-treated enzyme. Other metals, such as cadmium and zinc, do not stimulate the WNDP phosphorylation suggesting that the effect of copper is specific.
While this article was in preparation, Voskoboinik and colleagues (8) published their work on characterization of the phosphorylated intermediate formed by the human coppertransporting ATPase ATP7A, the Menkes disease protein, MNKP. Since MNKP and WNDP are homologous, it is interesting to compare their properties.
Both MNKP and WNDP are phosphorylated in an ATP-dependent manner and the sensitivity to hydroxylamine and ADP is similar for these proteins. However, phosphorylation of WNDP appears somewhat slower than phosphorylation of MNKP and there is also a significant difference in the K m for ATP (17 M Ϯ 7 for MNKP and 0.95 Ϯ 0.25 M for WNDP). The 17-fold difference in apparent affinity between these two ATPases could be a true reflection of their functional properties. Alternatively, the difference could be a result of different experimental protocols: we characterized the ATP dependence of the formation of acylphosphate intermediate, while Voskoboinik and co-workers (8) measured the ATP dependence of copper transport.
Another interesting difference is the effect of copper on formation of the acylphosphate intermediate. Voskoboinik and colleagues (8) reported increased incorporation of 32 P into MNKP upon addition of 2-5 M copper to the protein. In contrast, WNDP appeared fully active at concentrations of copper below 1 M and additions of copper to WNDP either in vivo or in vitro did not facilitate formation of the acylphosphate intermediate. Therefore, it seems possible that WNDP and MNKP have different affinities for copper that leads to their different in vitro response to copper additions. This conclusion is supported by the 10-fold difference in the K i for BCS (5 M for MNKP versus 50 M for WNDP).
To observe the effect of copper on catalytic phosphorylation, WNDP has to be pretreated with copper chelator. After this treatment the activity of WNDP decreases dramatically; additions of copper specifically re-activate the enzyme and the effect of copper is cooperative (see Fig. 8). The apparent cooperative response is very interesting, since it suggests that at least two copper-binding sites should be occupied before activation occurs.
WNDP has multiple copper-binding sites: six of them are located at the N-terminal portion of the molecule, while the CPC motif (Fig. 1) is believed to bind copper in the membrane during the copper translocation step (the actual number of the copper-binding sites in the membrane is currently unknown). The observed cooperativity could be either due to interactions between different intramembrane copper-binding sites or due to cross-talk between the intramembrane copper-binding site and the site(s) at the N-terminal portion of WNDP. The latter explanation is very appealing since previous studies demonstrated that the deletion/mutations of the 5th and/or 6th Nterminal metal-binding sites inhibited transport activity of WNDP (31,47).
In their recent important studies, Voskoboinick and colleagues (8) demonstrated that MNKP with mutations in all 6 N-terminal metal-binding sites could undergo copper-dependent catalytic phosphorylation. In these experiments the effect of copper was not cooperative, suggesting that the authors indeed monitored occupation of a single, most likely intramem-brane copper-binding site. Curiously, the wild-type MNKP also showed a noncooperative "single-site" response to copper. Thus, it is possible that the mechanism of copper-dependent activation of phosphorylation is somewhat different for MNKP and WNDP. At the same time, one cannot exclude the possibility that the lack of cooperativity for the wild-type MNKP was due to partial oxidation of the metal-binding sites, which made them unavailable for copper binding.
The role of the N-terminal metal-binding domain for the function of the P 1 -type ATPase has been the subject of several recent investigations (16,42,(45)(46)(47)(48). The emerging consensus is that the metal-binding sites at the N-terminal domain are not essential for the catalytic step. Whether the N-terminal domain regulates enzyme activity and/or delivery of copper to the intramembrane sites remains uncertain, because the results obtained in different expression systems are not entirely consistent. Although in this study we did not directly investigate the role of the N-terminal domain, several unexpected results made us consider the likely involvement of the N terminus in regulation of the WNDP activity. These results are related to the BCS effect and they are the following.
Given the fact that copper concentration in the phosphorylation buffer was only 1 M, it was surprising that as high as 100 M BCS (with K m for copper (I) close to 10 Ϫ20 M, (37)) was required to completely inhibit catalytic phosphorylation of WNDP. Also, it was unclear why the pretreatment with BCS was "remembered" by WNDP, i.e. why the activity of the BCStreated protein remained markedly lower than that of the control sample even after BCS was removed and the protein was resuspended in a standard phosphorylation buffer. Finally, it was unclear why addition of micromolar concentrations of copper were needed to activate the BCS-pretreated WNDP, while the nontreated protein had maximum activity without any addition of copper to the buffer.
The likely explanation for these observations is that BCS, in addition to simple chelation of copper in a solution, has a direct effect on WNDP. This effect can be best described as a marked decrease of the WNDP affinity for copper. Such a decrease could be due to tight binding of BCS to WNDP that renders the protein in the low affinity state. Alternatively, BCS may strip tightly bound copper from the site(s) important for the highaffinity state of WNDP thus down-regulating the enzyme activity.
Earlier, we demonstrated that copper binding to the N-terminal domain of WNDP alters the domain-domain interactions within WNDP, affecting the conformation of the ATP-binding domain, and we proposed that the N-terminal domain could regulate the WNDP activity (22). Given our current results, it is tempting to speculate that copper binding to the N-terminal regulatory sites within the cell generates the "high-affinity state" of WNDP, which has high affinity for copper and allows enzyme to be active even at submicromolar copper concentrations. Stripping of copper from the N-terminal domain with BCS would then generate the low affinity state of WNDP with catalytic activity only at micromolar copper concentrations. Why copper additions in vitro do not restore the high affinity state of WNDP is currently unclear. Oxidation of cysteine residues, an incorrect coordination environment for copper upon in vitro loading, or lack of copper chaperone are among the possibilities that we are currently considering.
More experiments are clearly needed to test the possible regulatory role of the N-terminal domain, however, the following consideration adds credibility to this hypothesis. It is interesting that in vitro, copper activates phosphorylation of MNKP or the BCS-treated WNDP when supplied at micromolar concentrations. This result is consistent with the idea that the metal-binding sites in the membrane-translocation pathway should have a fairly low affinity for copper to ensure its efficient release during the transport step. At the same time the observed low affinity is at odds with the fact, that the concentration of free copper in a cell is extremely low (10 Ϫ19 M in yeast, Ref. 37), suggesting that MNKP and WNDP would be inactive under normal intracellular conditions. Therefore, one must conclude that in a cell, ATPases rely on a specific mechanism, which facilitates copper binding to the intramembrane sites either by regulating their affinity, or by providing direct delivery of copper to these sites. It seems that the role of such a mechanism could be best performed by the cytosolic copperbinding site(s) at the N-terminal portion of these proteins (Fig.  1). The functional expression system described in this paper provides the necessary experimental background for careful testing of these hypotheses.
In summary, we developed a direct approach for characterization of the functional properties of the full-length human copper-transporting ATPase ATP7B in a native membrane environment, and have shown that the protein displays the characteristics of a P-type ATPase. The described system can be utilized for careful analysis of the molecular mechanism of copper transport by WNDP, and for characterization of various WNDP mutants. The ability to measure apparent nucleotide affinities is particularly useful for analysis of the Wilson's disease-causing mutations located in the ATP-binding domain of this protein.