Biochemical characterization of CopA, the Escherichia coli Cu(I)-translocating P-type ATPase.

Escherichia coli CopA is a copper ion-translocating P-type ATPase that confers copper resistance. CopA formed a phosphorylated intermediate with [gamma-(32)P]ATP. Phosphorylation was inhibited by vanadate and sensitive to KOH and hydroxylamine, consistent with acylphosphate formation on conserved Asp-523. Phosphorylation required a monovalent cation, either Cu(I) or Ag(I). Divalent cations Cu(II), Zn(II), or Co(II) could not substitute, signifying that the substrate of this copper-translocating P-type ATPase is Cu(I) and not Cu(II). CopA purified from dodecylmaltoside-solubilized membranes similarly exhibited Cu(I)/Ag(I)-stimulated ATPase activity, with a K(m) for ATP of 0.5 mm. CopA has two N-terminal Cys(X)(2)Cys sequences, Gly-Leu-Ser-Cys(14)-Gly-His-Cys(17), and Gly-Met-Ser-Cys(110)-Ala-Ser-Cys(113), and a Cys(479)-Pro-Cys(481) motif in membrane-spanning segment six. The requirement of these cysteine residues was investigated by the effect of mutations and deletions. Mutants with substitutions of the N-terminal cysteines or deletion of the first Cys-(X)(2)-Cys motif formed acylphosphate intermediates. From the copper dependence of phosphoenzyme formation, the mutants appear to have 2-3 fold higher affinity for Cu(I) than wild type CopA. In contrast, substitutions in Cys(479) or Cys(481) resulted in loss of copper resistance, transport and phosphoenzyme formation. These results imply that the cysteine residues of the Cys-Pro-Cys motif (but not the N-terminal cysteine residues) are required for CopA function.

achieved approximately the same final density in the absence of copper.
Strain Construction and Plasmids-Standard molecular and genetic techniques were used for strain and plasmid construction (19). The construct of copA deletion strain DC194 and Cys-(X) 2 -Cys motif mutants were described previously (18). To introduce mutations in the coding sequence for the cysteines in the Cys-Pro-Cys motif, a sequence corresponding to base pairs 892-1854 of the copA gene was cloned into plasmid pGEM-T Easy (Promega) by polymerase chain reaction (PCR) using forward primer 5Ј-CTCGGCCATATGCTGGAAGCG-3Ј and reverse primer 5Ј-GTAATCGCCGCGGTTTCAATG-3Ј. Site-directed mutagenesis was carried out using a QuickChange method (Stratagene). The inserts were excised by double digestion with NdeI and SacII and ligated into plasmid pGEXCopA 2 in which copA is inserted into the pGEX-6p-2 vector (Amersham Biosciences). The copA gene was excised from that plasmid by digestion with NcoI and SacII and ligated into pCopA 2 (1). A copA gene with deletion of codons 7-54 (corresponding to amino acid residues 3-18 and including the first Cys-(X) 2 -Cys sequence) was generated by PCR amplification of copA of using forward primer 5Ј-GTTCCATGGCAAAACGCGTGAAAGAAAG-3Ј and reverse primer 5Ј-ATGGCCGCCGGCGAAAACCATCACTGC-3Ј, with subsequent cloning into plasmid pGEM-T Easy. The insertion was excised by NcoI/NaeI double digestion and ligation into pCopA 2 , generating the plasmid pCopA 2 ⌬N1. Plasmid pCopA 2 ⌬N2, which has a deletion of copA codons 7-342 bp (corresponding to amino acid residues 3-114 and includes both Cys-(X) 2 -Cys sequences) was generated by PCR using forward primer 5Ј-GTTCCAATGGCAACCCGCGTACAAAATGCG-3Ј and the same reverse primer as with pCopA 2 ⌬N1. The PCR product was excised by NcoI/NaeI double digestion and ligation into plasmid pCopA 2 generating plasmid pCopA 2 ⌬N2. All mutations were verified by sequencing the entire insert. In pCopA 2 copA is in-frame with the sequence for the Myc epitope and six histidine codons, which are at the 3Ј-end of copA (1). The wild type and mutant pCopA 2 plasmid were introduced into strain DC194.
Preparation of Everted Membrane Vesicles-Cells were grown overnight at 37°C in 5 ml of LB, diluted 100-fold into prewarmed medium and allowed to grow to an optical density of 0.7 at 600 nm. Cells were induced with 0.0002% arabinose for 2.5 h at 30°C. Everted membrane vesicles were prepared as described previously (1). The membrane vesicles were suspended in a buffer consisting of 25 mM Tris-HCl, pH 7.0, containing 0.25 M sucrose, 0.2 M KCl, and 0.05 mM EDTA and stored at Ϫ70°C until use. Protein concentrations were determined using a bicinchoninic acid method (Sigma Chemical Co.).
Polyacrylamide Gel Electrophoresis and Immunoblotting-Samples were prepared by incubation in SDS sample buffer for 30 min at room temperature and separated by SDS-PAGE on 8% polyacrylamide gels. Immunoblotting was performed with antibody to the His 6 tag (Clontech) using an enhanced chemiluminescence assay (PerkinElmer Life Sciences) and exposed on x-ray film at room temperature, as described previously (20). To correct for differences in expression of the mutant proteins, the amount of CopA in everted membrane vesicles was normalized to that of the wild type by quantitative Western blot analysis. Each set of membranes was analyzed at four different concentrations, and a standard curve was constructed from the values obtained by densitometry.
64 Cu Transport Assays-64 Cu was obtained from the Division of Radiological Science, Mallinckrodt Institute of Radiology, Washington University Medical School, which is supported by NCI Research Resource Grant NIH 1 R24 CA86307. Transport assays were performed at room temperature as described previously (1). Unless otherwise noted, the reaction mixture (1 ml) contained 40 mM histidine (pH 6.8), 0.2 M KCl, 0.25 M sucrose, 1 mM dithiothreitol (DTT), 0.5 mg of membrane protein, 10 M 64 CuCl 2 (0.5-10 Ci/ml), and 5 mM Na 2 ATP. The reaction was initiated by addition of 5 mM MgCl 2 . At intervals, 0.1-ml samples were withdrawn and filtered through nitrocellulose filters (0.22-m pore size, Whatman). The filters were presoaked in a buffer consisting of 40 mM histidine, pH 6.8, 0.2 M KCl, 0.25 M sucrose, 10 mM MgSO 4 , and 20 mM CuCl 2 . Following filtration, the filters were washed with 5 ml of the same buffer, dried, and the radioactivity quantified by liquid scintillation counting. The values obtained with the assay mixture without membrane vesicles were subtracted from all time points.
Phosphoenzyme Formation with [␥- 32  To measure sensitivity to vanadate, 5 mM sodium orthovanadate was added to a final concentration of 50 M. The mixture was incubated at 4°C for 5 min, and the reaction was initiated by addition of 2.5 Ci of [␥-32 P]ATP at a final concentration of 50 M. After incubation on ice (or at room temperature for purified CopA) for the indicated times, 40 l of ice-cold 50% tricholoracetic acid was added to terminate the reaction. For pulse-chase analysis, 1 mM cold ATP was added 30 s after the start of the reaction, followed by incubation on ice for another 15 to 30 s. Following addition of tricholoracetic acid, the membrane vesicles were kept on ice for an additional 10 min and then harvested by centrifugation at 15,000 ϫ g for 10 min. The pellet protein was washed once with 0.5 ml of distilled water and once with 0.5 ml of a solution of 50 mM H 3 PO 4 /NaOH, pH 2.4. To examine alkali lability, the pelleted material was suspended in 0.2 ml of 0.5 M KOH and kept on ice for 5 min. Sensitivity to hydroxylamine was examined by suspending the pelleted material in 0.2 ml of 0.1 M sodium acetate, pH 5.6, followed by the addition of 0.2 ml of 0.25 M NH 2 OH, with an additional 10 min incubation at room temperature. The treated samples were again precipitated with tricholoracetic acid and dissolved in 30 l of 2ϫ concentrated SDS sample buffer diluted with an equal volume of 50 mM H 3 PO 4 /NaOH, pH 2.4, containing 5% SDS. Samples of 10 l were loaded on an acidic 8% polyacrylamide gel (21). Following electrophoresis, the gels were stained with Coomassie Blue, dried using a DryEase mini-gel drying system (Novex). Radioactivity was analyzed with a PhosphorImager (Molecular Dynamics).
Purification of CopA-Membrane vesicles (5 mg/ml) were solubilized in buffer D, consisting of 25 mM Tris-HCl, pH 7.0, 50 mM KCl, 1 M CuCl 2 , 5 mM MgCl 2 , and 10% glycerol, containing 2% dodecyl maltoside (DDM). The mixture was gently shaken at 4°C for 1 h, and the insoluble fraction was removed by centrifugation at 250,000 ϫ g for 1 h. The soluble fraction was loaded onto a Probond Ni 2ϩ affinity column (Invitrogen) pre-equilibrated with buffer D containing 0.1% DDM. The column was washed with 5 bed volumes of the same buffer, followed by 5 bed volumes of the same buffer containing 40 mM imidazole. The protein was eluted with the same buffer containing 0.1 M imidazole. Fractions of 0.5 ml were collected into tubes containing concentrated DTT and EDTA such that the final concentrations became 1 and 0.1 mM, respectively. The samples were analyzed for CopA by SDS-PAGE and immunoblotting. CopA-containing fractions were concentrated 10fold using Centricon concentrators (Millipore), and the imidazole was removed by gel filtration on a 5-cm column filled with 3 ml of Sephadex G-25 column pre-equilibrated with buffer D containing 0.1% DDM. All buffers were degassed under vacuum or bubbled with Argon before use. All steps were performed at 4°C.

Cu(I)-dependent Formation of an Acylphosphate Bond in
CopA-A signature property of the P-type ATPases is that the conserved aspartate residue in the DKTG motif accepts the ␥-phosphate from ATP during the catalytic cycle, forming a covalent acylphosphate intermediate (23). To investigate phosphoenzyme formation in CopA, everted membrane vesicles were prepared from cells of strain DC194 (⌬copA) expressing wild type copA on a plasmid. CopA was phosphorylated with [␥-32 P]ATP in a time-dependent manner, with maximum labeling within 30 s (Fig. 1A). Vesicles from the copA deletion strain DC194 did not exhibit a radioactive band at the corresponding position (Fig. 1, lane 1). The intermediate was sensitive to basic pH and hydroxylamine (Fig. 1B), which are considered to be reliable criteria for the presence of an acylphosphate bond. The label could be chased by 1 mM unlabeled ATP within 15 s, indicating the transient nature of the intermediate. In the presence of 50 M sodium orthovanadate, which mimics the transition state when bound in the active site of P-type AT-Pases, the reaction was entirely inhibited. On the other hand, phosphoenzyme formation was not inhibited by azide, an inhibitor of many ATPases.
Since CopA is a copper ion pump, acylphosphate formation would be expected to be stimulated by copper ion. In the absence of copper ion, no radioactive band was observed (Fig. 1C, lane 1). To date there has been no clear determination of the redox state of the substrate of copper ion pumps, that is, whether these ATPases transport Cu(II), Cu(I) or both. Our previous work showed that a reductant, DTT, was required for the transport of 64 Cu with membrane vesicles (1,18). To investigate whether reductant is also required for phosphoenzyme formation, several reductants were added to the assay. In agreement with the results of the transport assays, a significant radioactive band was observed only in the presence of Cu(II) and reductant: DTT, GSH, or cysteine (Fig. 1C). DTT and GSH were more effective than cysteine. For those reasons DTT was usually added as reductant in subsequent assays.
The requirement for reductant could be to reduce Cu(II) to Cu(I) or to reduce the cysteines in CopA or both. For that reason the effect of mono-and divalent soft metal ions on acylphosphate formation was examined (Fig. 1C). No phosphoenzyme formation was observed in the presence of the divalent cations Zn(II) or Co(II), whether or not DTT was present. In contrast, either Ag(I) or Cu(I) (added in the form of acetonitrile copper(I) hexafluorophosphate) stimulated phosphoenzyme formation even in the absence of DTT, showing that DTT is not necessary to maintain the enzyme in a reduced state. Moreover, bathocuproindisulfonate (BCDS), which chelates Cu(I) and not Cu(II), prevented labeling. These results demonstrate that CopA is a monovalent (but not a divalent) soft metal ion P-type ATPase.
Requirement of the N Terminus for CopA Activity-One characteristic that differentiates soft metal P-type ATPases from their hard metal homologues is the presence of one or more Cys-(X) 2 -Cys metal binding domains in the cytosolic N-terminal region. We have previously shown that none of the four cysteine residues in the Gly-Leu-Ser-Cys 14 -Gly-His-Cys 17 and Gly-Met-Ser-Cys 110 -Ala-Ser-Cys 113 sequences of CopA is required for either copper resistance or transport, although deletion of codons 8 -150 (⌬NCopA) lost both copper resistance and transport (18). To extend this observation, two N-terminally truncated CopAs were constructed: pCopA 2 ⌬N1, which is truncated immediately after the first Cys-(X) 2 -Cys motif, and pCopA 2 ⌬N2, which has both Cys-(X) 2 -Cys motifs deleted. Cells expressing ⌬N1 CopA retained partial resistance to CuSO 4 in vivo, while cells expressing ⌬N2 CopA were as sensitive as the CopA deletion strain (Fig. 2), even though the truncated proteins were made in amounts comparable to that of the wild type (Fig. 3A). These results suggest that at least a portion of the cytosolic N terminus of the protein is necessary, although whether that is for activity or proper folding cannot be deduced from these results.
CopA mutants with alanine substitutions in the first (C14A/ C17A) or second (C110A/113A) Cys-(X) 2 -Cys motifs or with all four cysteines substituted (C14A/C17A/C110A/C113A) formed an acylphosphate intermediate equivalent to wild type CopA (Figs. 1B and 3B). Indeed, deletion of the N terminus through the first Cys-(X) 2 -Cys motif did not appear to affect the ability of the ⌬N1 CopA to form a phosphorylated intermediate, just as a strain expressing this mutant copA conferred copper resistance in vivo. Consistent with the loss of resistance in cells bearing pCopA 2 ⌬N2, the ⌬N2 enzyme did not form a phosphorylated intermediate.
Requirement of an Intramembrane CPC Motif for CopA Activity-A distinctive feature of soft metal ATPases is a highly conserved Cys-Pro-Cys motif in TM6, the putative translocation domain (24). The Cys-Pro-Cys motif in CopA includes Cys 479 and Cys 481 . The two cysteines were individually mutated to C479A and C481A. Since the Enterococcus hirae CopB has a Cys-Pro-His sequence (6), a C481H mutant was also constructed. All three mutant proteins were expressed in amounts comparable to wild type CopA (Fig. 3A). The effect of the mutations on copper resistance in vivo (Fig. 4A) and ATPdriven 64 Cu transport in vitro were examined. None of the mutant genes conferred resistance to copper, and everted membrane vesicles from cells expressing those genes were unable to accumulate 64 Cu (Fig. 4). No phosphoenzyme intermediates were detected in the C479A, C481A, or C481H proteins (Fig. 3B).
Copper Dependence of Phosphoenzyme Formation-Forma-  (Fig. 5). Wild type CopA, the C14A/C17A/C110A/C113A quadruple mutant, and the ⌬N1 mutant each responded to increasing concentrations of copper (Fig. 5A). The bands from three separate 32 P-labeled gels were quantified by densitometry. The values were averaged and, for clarity, were normalized to the value at 50 M (Fig. 5B). The apparent K m for copper of wild type CopA was 1.5 Ϯ 0.5 M, which is similar to the K m of 3.9 M for the copper dependence of ATPase activity of the Achaeoglobius fulgidus CopA (25). The apparent K m of the quadruple mutant was 0.45 Ϯ 0.1 M, and the ⌬N1 exhibited an apparent K m of 0.84 Ϯ 0.2 M. Thus, removal of the first N-terminal Cys-(X) 2 -Cys motif or modification of all four N-terminal cysteines seemed to produce an apparent increase in affinity for Cu(II).
Solubilization and Purification of CopA-Membrane vesicles were prepared from arabinose-induced cells bearing plasmid pCopA2, which carries a copA gene in-frame with the sequence for the Myc epitope and six histidine codons under control of the arabinose promoter. The vesicles were solubilized with 2% dodecylmaltoside, and CopA was purified by metal chelate affinity chromatography (Fig. 6A). The presence of a small amount of copper during solubilization and purification was found to result in more active and stable CopA. A minor band was observed upon immunoblotting (Fig. 6B). Since the slightly smaller protein reacted with antibody to the histidine tag, there was probably some degradation at the N terminus of CopA. Although the purified protein exhibited ATPase activity, hydrolysis was inhibited 80 -90% by sodium azide. Since phosphoenzyme formation is not inhibited by sodium azide (Fig. 1B), it is likely that the azide-sensitive ATPase activity is due to minor contamination by a highly active ATPase.
The azide-insensitive ATPase activity of 130 nmol/mg/min was inhibited to 65 nmol/mg/min by BCS, a Cu(I) chelator, that may reflect residual copper remaining from purification. Neither Cu(II) nor Zn(II) stimulated ATPase activity. Compared with the rate in the presence of BCS, there was an ϳ4-fold activation by Cu(I) (280 nmol/mg/min) or Cu(II) plus DTT (295 nmol/mg/min). In other experiments, Ag(I)-stimulated activity was 4-to 5-fold higher than the rate with BCS. The Ag(I)stimulated activity exhibited a relatively broad pH profile, with approximately equal activity at pH 7 and 8, and about one- third lower at pH 6. These results indicate that the azideinsensitive ATPase activity is catalyzed by CopA in a reaction that requires either Cu(I) or Ag(I). Cu(I)-stimulated ATPase exhibited an apparent K m for ATP of 0.5 mM (Fig. 6C). The Cu(I)-stimulated V max of 0.19 mol/mg/min is similar to the V max for the Ag(I)-stimulated ATPase reported for the CopA homologue from A. fulgidus (25). DISCUSSION The distinctive feature of P-type ATPases that gave the superfamily its name is the formation of an acylphosphate on a conserved aspartate residues during the catalytic cycle (23). Phosphoenzyme formation has been shown for several coppertranslocating P-type ATPase homologues of CopA (25)(26)(27)(28). In this report we demonstrate Cu(I)-dependent phosphorylation of CopA. Phosphorylation was sensitive to alkali and hydroxylamine, consistent with acylphosphate formation on the conserved Asp 523 .
Whereas it is presumed that copper pumps are specific for Cu(I) and not Cu(II), this has not been proven. Since an Nterminal peptide containing the Cys-(X) 2 -Cys motifs of the Wilson disease protein has been shown to bind Cu(II) and other divalent cations (29), it is possible that these proteins could pump both Cu(I) and Cu(II). Phosphorylation of WND and E. hirae CopA were not copper-dependent. This could have been due to copper contamination in the buffers, since copper chelators were found to inhibit. Phosphorylation of MNK and the A. fulgidus CopA showed copper dependence, but it was not clear whether the activity was stimulated by Cu(I), Cu(II), or both. In both cases DTT was present in the phosphorylation assays, which would reduce some or all of the Cu(II) to Cu(I), but DTT might also have been necessary to reduce cysteine thiolates in the proteins. The data reported here clearly show that Cu(I) and not Cu(II) stimulates phosphorylation of CopA, and that DTT is not required. While phosphoenzyme formation was stimulated by Cu(II) in the presence of DTT, Cu(I) and Ag(I) were equally effective in the absence of DTT. Thus the requirement for DTT appears to be in reduction of copper rather than for maintenance of protein cysteine thiolates. There was no phosphorylation in the presence of Cu(II) or other divalent cations, so CopA is not a divalent cation pump.
The soft metal P-type ATPase are distinguished from their hard metal homologues by the presence of N-terminal metal binding motifs, usually Cys-(X) 2 -Cys, and a second cysteine motif, usually Cys-Pro-Cys, in TM6. The physiological role of these motifs and the function of the individual cysteine residues in the biochemical mechanism of the ATPases are open questions. In eukaryotes the N-terminal motifs may interact with metal chaperones or may be involved in copper-regulated trafficking of copper pumps from intracellular compartments to the plasma membrane. Most prokaryotes lack intracellular membranes or compartments, and the function of the Cys-(X) 2  first Cys(X) 2 Cys motif similarly was phosphorylated. In both there was a 2-3-fold decrease in the concentration of copper required for maximal phosphoenzyme formation. Does this apparently counterintuitive result imply that binding of copper to the N-terminal motifs decrease the affinity of the pump for copper, perhaps for a regulatory function? An alternative explanation is that in vitro the Cys-(X) 2 -Cys motifs reduce access of copper to the translocation pathway or reduce the local concentration of copper in that region. The mutations did not produce altered copper resistance, so it is questionable whether the in vitro results reflect an increased efficiency of the pump in vivo. While none of the four N-terminal cysteines are required for CopA activity under the conditions studied, their contributions to CopA function may be apparent only under different physiological conditions. Similar conclusions have been made about the Cys-(X) 2 -Cys motifs in other copper Ptype ATPases. For example, in MNK mutation or deletion of the first four of the six Cys-(X) 2 -Cys motifs had no apparent effect on copper-induced trafficking, their only identified function (15). Although none of the cysteine residues of the two Cys-(X) 2 -Cys motifs of CopA are obligatory, deletion of the N terminus to just before the first putative transmembrane region resulted in loss of function, but it is possible that the remainder of the protein does not fold properly.
In three CopA homologues the Cys-Pro-Cys motif that is distinctive in soft metal ion-translocating ATPases has been shown to be located in TM6 (30 -32), so it is reasonable to assume that it is in TM6 of CopA as well. The Cys-Pro-Cys motif has been proposed to be part of the translocation domain (24). To examine the requirement of the cysteine residues, Cys 479 and Cys 481 were changed to alanine residues. Cells expressing either mutant became sensitive to copper. Both mutant proteins were found in the membrane in normal amounts, but neither was phosphorylated by [␥-32 P]ATP. Since the CopB homologue from E. hirae has a Cys-Pro-His motif in the corresponding TM6 (6), we constructed a C481H mutant. Cells expressing the mutant CopA were also sensitive to copper, and the protein with a Cys-Pro-His motif did not form an acylphosphate intermediate. A reasonable interpretation of these results is that the Cys-Pro-Cys motif of CopA is essential for Cu(I) binding, and that mutants with Cys-Pro-His, Cys-Pro-Ala, or Ala-Pro-Cys no longer bind metal. This is consistent with the known catalytic cycle for hard metal ion-translocating P-type ATPases, where binding of metal is required for acylphosphate formation at the conserved aspartate residue (33). Thus, we propose that the Cys-Pro-Cys residues of CopA are required for binding of Cu(I) in the ion translocation pathway and hence for Cu(I) transport.