Advertisement

The Role of the Invariant His-1069 in Folding and Function of the Wilson's Disease Protein, the Human Copper-transporting ATPase ATP7B*

  • Ruslan Tsivkovskii
    Footnotes
    Affiliations
    Department of Biochemistry and Molecular Biology, Oregon Health and Science University, Portland, Oregon 97239
    Search for articles by this author
  • Roman G. Efremov
    Footnotes
    Affiliations
    Shemyakin-Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences, 117997 Moscow, Russia
    Search for articles by this author
  • Svetlana Lutsenko
    Correspondence
    To whom correspondence should be addressed. Tel.: 503-494-6953; Fax: 503-494-8393
    Affiliations
    Department of Biochemistry and Molecular Biology, Oregon Health and Science University, Portland, Oregon 97239

    Shemyakin-Ovchinnikov Institute of Bioorganic Chemistry, Russian Academy of Sciences, 117997 Moscow, Russia
    Search for articles by this author
  • Author Footnotes
    * This work was funded by National Institutes of Health Grant DK55719 (to S. L.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
    § Recipient of American Heart Association Postdoctoral Fellowship Grant 0120573Z.
    ‖ Recipient of financial support from the Science Support Foundation (Russia).
Open AccessPublished:January 27, 2003DOI:https://doi.org/10.1074/jbc.M300034200
      The copper-transporting ATPase ATP7B is essential for normal distribution of copper in human cells. Mutations inATP7B lead to Wilson's disease, a severe disorder with neurological and hepatic manifestations. One of the most common disease mutations, a H1069Q substitution, causes intracellular mislocalization of ATP7B (the Wilson's disease protein, WNDP). His-1069 is located in the nucleotide-binding domain of WNDP and is conserved in all copper-transporting ATPases from bacteria to mammals; however, the specific role of this His in the structure and function of WNDP remains unclear. We demonstrate that substitution of His-1069 for Gln, Ala, or Cys does not significantly alter the folding of the WNDP nucleotide-binding domain or the proteolytic resistance of the full-length WNDP. In contrast, the function of WNDP is markedly affected by the mutations. The ability to form an acylphosphate intermediate in the presence of ATP is entirely lost in all three mutants, suggesting that His-1069 is important for ATP-dependent phosphorylation. Other steps of the WNDP enzymatic cycle are less dependent on His-1069. The H1069C mutant shows normal phosphorylation in the presence of inorganic phosphate; it binds an ATP analogue, β,γ-imidoadenosine 5′-triphosphate (AMP-PNP), and copper and undergoes nucleotide-dependent conformational transitions similar to those of the wild-type WNDP. Although binding of AMP-PNP is not disrupted by the mutation, the apparent affinity for the nucleotide is decreased by 4-fold. We conclude that His-1069 is responsible for proper orientation of ATP in the catalytic site of WNDP prior to ATP hydrolysis.
      MNKP
      Menkes disease protein
      WNDP Wilson's disease protein
      AMP-PNP, β,γ-imidoadenosine 5′-triphosphate
      wt
      wild-type
      MES
      4-morpholineethanesulfonic acid
      BCS
      bathocuproine disulfonate
      ATP-BD
      ATP-binding domain
      CD
      circular dichroism
      Copper is a cofactor of important metabolic enzymes that are involved in a variety of physiological processes such as radical detoxification, oxidative phosphorylation, and iron metabolism. The key role in copper distribution in human cells belongs to the copper-transporting ATPases ATP7A and ATP7B (Menkes disease and Wilson's disease proteins, respectively). These large polytopic membrane proteins utilize the energy of ATP hydrolysis to transport copper from the cytosol into the lumen of the secretory compartment where copper can be incorporated into various copper-dependent enzymes (
      • Harris E.D.
      ). When intracellular copper exceeds a certain level, the copper-transporting ATPases traffic to the plasma membrane where they export excess copper out of the cell (
      • Petris M.J.
      • Mercer J.F.
      • Culvenor J.G.
      • Lockhart P.
      • Gleeson P.A.
      • Camakaris J.
      ,
      • Llanos R.M.
      • Mercer J.F.
      ). Mutations in ATP7A and ATP7Bresult in disruption of copper transport from the cytosol and severe pathologies in humans, i.e. Menkes disease and Wilson's disease, respectively.
      Menkes disease and Wilson's disease proteins (MNKP1 and WNDP) belong to a family of the P-type ATPases. The primary structure of MNKP and WNDP include several signature sequences (Fig. 1A) that are essential for enzymatic activity of all members of this family. The ATP-binding domain of WNDP and MNKP also contains a sequence SEHPL, which is conserved in all ATPases transporting transition metals (the P1-type ATPases or CPx-ATPases) but not in other P-type ATPases (Fig. 1). In the SEHPL sequence, His is an invariant residue (Fig. 1B, alignment), suggesting that this amino acid is essential for function or structure of the P1-type ATPases. Significantly, substitution of His-1069 in WNDP for Gln is the most frequent cause of Wilson's disease in northern European populations (
      • Thomas G.R.
      • Roberts E.A.
      • Walshe J.M.
      • Cox D.W.
      ), an observation that underscores the importance of His-1069.
      Figure thumbnail gr1
      Figure 1Schematic representation of the transmembrane organization of WNDP. A, the letters TGEA,TGDN, DKTG, and GDGVND indicate sequence motifs conserved in all P-type ATPases. The bold letter D in the DKTG motif marks the position of Asp-1027, an acceptor of Pi during catalysis. The SEHPL sequence is conserved in all P1-type ATPases and contains the invariant His-1069, marked by larger font. The CXXC motifs indicate copper-binding sites in the cytosolic copper-binding domain. B, the alignment of the ATP-binding domain segments of several P1-type ATPase. The alignment was generated using ClustalW (www.ebi.ac.uk/clustalw/). The protein data base accession numbers are given in parentheses for the following: atp7a_human, MNKP (Q04656); atp7b_human, WNDP (P35670); atu2_yeast, yeast copper-transporting ATPase CCC2 (P38995); atzn_ecoli, lead-, cadmium-, and zinc-transporting ATPase (P37617); cada_stauu, cadmium-transporting ATPase (P20021); copa_enthr, copper-importing ATPase A from Enterococcus hirae (P32113); and copb_enthr, copper-exporting ATPase B from E. hirae (P05425). The invariant His is in bold; the SEHPL-like sequences are underlined.
      Recent studies have shown that mutations of His-1069 to Gln or Ala result in decreased intracellular stability of WNDP and retention of the mutants in the endoplasmic reticulum (
      • Payne A.S.
      • Kelly E.J.
      • Gitlin J.D.
      ,
      • Huster D.
      • Hoppert M.
      • Zinke J.
      • Lutsenko S.
      • Lehmann C.
      • Mossner J.
      • Berr F.
      • Caca K.
      ); the latter effect can be overcome by lowering the temperature at which the cells are grown (
      • Payne A.S.
      • Kelly E.J.
      • Gitlin J.D.
      ). These results suggested a role for His-1069 in folding and stability of WNDP. In contrast, replacement of the equivalent His residue in MNKP, which is 60% homologous to WNDP, does not alter the protein steady-state levels and does not disrupt normal targeting of MNKP to the trans-Golgi network (
      • Petris M.J.
      • Voskoboinik I.
      • Cater M.
      • Smith K.
      • Kim B.E.
      • Llanos R.M.
      • Strausak D.
      • Camakaris J.
      • Mercer J.F.
      ), indicating that in MNKP the structural consequences of His replacement are minor. Thus, it remains unclear whether His-1069 plays an important role in the structural organization of WNDP.
      The effect of the H1069Q substitution on the transport function of WNDP was analyzed by the ability of the mutant protein do the following: (i) complement the growth defects of a Δccc2 yeast strain lacking the WNDP homologue Ccc2 (
      • Hung I.H.
      • Suzuki M.
      • Yamaguchi Y.
      • Yuan D.S.
      • Klausner R.D.
      • Gitlin J.D.
      ,
      • Iida M.
      • Terada K.
      • Sambongi Y.
      • Wakabayashi T.
      • Miura N.
      • Koyama K.
      • Futai M.
      • Sugiyama T.
      ); and (ii) decrease the toxic effects of copper on growth of fibroblasts in cell culture (
      • Payne A.S.
      • Kelly E.J.
      • Gitlin J.D.
      ). These experiments demonstrated that mutations of His-1069 impaired the transport function of WNDP. However, the specific role of this invariant histidine in copper transport by WNDP has not been explained.
      To elucidate the functional role of His-1069, we substituted His-1069 for various amino acid residues and characterized the effects of these mutations on WNDP folding and enzymatic activity. Our experiments indicate that His-1069 is important for a specific step in the WNDP catalytic cycle, phosphorylation from ATP, and is not critical for other steps of the enzymatic reaction or for overall protein folding. His-1069 appears to control, directly or indirectly, positioning of ATP in the active site of WNDP.

      EXPERIMENTAL PROCEDURES

      Generation of the His-1069 Mutants of the Full-length WNDP and Expression in SF9 Cells

      The H1069Q, H1069A, and H1069C mutants of WNDP were generated using polymerase chain reaction. The following forward and reverse primer pairs were used to introduce His → Gln, His → Ala, and His → Cys substitutions, respectively: 5′-AGCAGTGAACAACCCTTGGGCGTGG and 5′-ACGCCCAAGGGTTGTTCACTGCTGG; 5′-AGCAGTGAAGCTCCCTTGGGCGTGG and 5′-ACGCCCAAGGGAGCTTCACTGCTGG; and 5′-AGCAGTGAATGCCCCTTGGGCGTGG and 5′-ACGCCCAAGGGGCATTCACTGCTGG. The oligonucleotides: 5′-GGTATGGATTGTAATCGG-3′ and 5′-CGTCGACGCCTGCCTGAA-3′ were used as a forward and reverse flanking primers. Following PCR, the fragments containing mutations were digested with the ClaI and SalI restriction endonucleases and the ClaI-SalI fragments were exchanged with the corresponding fragment of the wild-type, full-length WNDP cDNA cloned into the pFastBacDual-WNDP plasmid. Generation of the pFastBacDual-WNDP plasmid has been described previously (
      • Tsivkovskii R.
      • Eisses J.F.
      • Kaplan J.H.
      • Lutsenko S.
      ). The presence of the mutations in the final construct and the absence of unwanted mutations were confirmed by automated DNA sequencing.
      The generated plasmids were then utilized to produce recombinant baculoviruses; the wild-type (wt) and mutant WNDPs were then expressed using virus-mediated infection of SF9 cells as described previously (
      • Tsivkovskii R.
      • Eisses J.F.
      • Kaplan J.H.
      • Lutsenko S.
      ). Expression of the WNDP mutants was verified by SDS-PAGE and Western blot analysis with anti-WNDP antibodies; the amount of expressed protein was determined by densitometry of the Coomassie-stained bands. The level of expression for mutant proteins was, on average, ∼50–70% of wt WNDP. The membrane preparations of insect cells expressing either wt or mutant WNDP were isolated according to Ref.
      • Tsivkovskii R.
      • Eisses J.F.
      • Kaplan J.H.
      • Lutsenko S.
      ; the membrane protein was stored at −80 °C in a buffer containing 25 mm imidazole, pH 7.4, and 250 mm sucrose until further use. Analysis of WNDP phosphorylation in the presence of [γ-32P]ATP was carried out as described in Ref.
      • Tsivkovskii R.
      • Eisses J.F.
      • Kaplan J.H.
      • Lutsenko S.
      .

      WNDP Phosphorylation from 32Pi Inorganic Phosphate

      50 μg of membrane protein preparation was resuspended in 200 μl of a buffer containing 50 mm MES-Tris (pH 7.0), 10 mm MgCl2, 20% Me2SO (Pi-buffer), and then 0.66 μl of 100 μCi of32Pi (specific activity, 6000 Ci/mmol) was added to the mixture. Following a 10 min incubation at room temperature, 50 μl of 1 mm NaH2PO4 in 50% (w/v) trichloroacetic acid was added to stop the reaction, and precipitated proteins were collected by centrifugation at 20,000 × g for 10 min and then rinsed with 1 ml of cold H2O. The protein was resuspended in the sample buffer (5 mm Tris-PO4, pH 5.8, 6.7 m urea, 0.4 m dithiothreitol, 5% SDS), and incorporation of32P into WNDP was analyzed on acidic SDS-PAGE as described (
      • Tsivkovskii R.
      • Eisses J.F.
      • Kaplan J.H.
      • Lutsenko S.
      ).
      To determine the effect of ATP and the non-hydrolyzable ATP analogue β,γ-imidoadenosine 5′-triphosphate (AMP-PNP) on phosphorylation of WNDP by 32Pi, 50 μg of membrane preparation containing either wt or mutant WNDP was pre-incubated in Pi-buffer with increasing concentrations of the nucleotide (10 μm–2 mm) for 10 min at room temperature, and then Pi-mediated phosphorylation was carried out as described above.
      To determine the effect of copper on phosphorylation from Pi, wt and mutant WNDPs were incubated at room temperature in Pi buffer containing 100 μmTris-(2-carboxyethyl)phosphine (TCEP) and 100 μmascorbate in the absence or presence of 250 μm copper chelator bathocuproine disulfonate (BCS) and then phosphorylated by Pi as described above. To reverse the effect of the chelator, 135 μm CuCl2 was added to the WNDP sample preincubated with 250 μm BCS. BCS binds copper with a stoichiometry of 2:1; thus this procedure generates 10 μm free copper in a solution. Following 10 min of incubation at room temperature, the Pi phosphorylation reaction was then carried out as described above.

      Generation of the ATP-binding Domain Mutants and Expression of Recombinant Proteins in Escherichia coli

      The expression plasmid pET28b-ATP-BD encoding the ATP-binding domain of WNDP (ATP-BD, amino acid residues Lys-1010 to Ly-1325) was described previously (
      • Tsivkovskii R.
      • MacArthur B.C.
      • Lutsenko S.
      ). To introduce various mutations of His-1069 into the recombinant ATP-BD, the Bsu36I-PstI fragments of the mutant full-length WNDP cDNA generated as described above were excised from the pFastBacDual-WNDP plasmid and cloned into the pET28b-ATP-BD plasmid digested with the Bsu36I and PstI restriction endonucleases. Expression in E. coli and purification of mutant ATP-BDs were performed according to (
      • Tsivkovskii R.
      • MacArthur B.C.
      • Lutsenko S.
      ). Protein expression and purity were verified by SDS-PAGE on 12% Laemmli gels.

      Folding of the Wild-type and Mutant ATP-BDs

      The effect of His-1069 substitutions on ATP-BD folding was analyzed using intrinsic Trp fluorescence as described in (
      • Tsivkovskii R.
      • MacArthur B.C.
      • Lutsenko S.
      ) and by circular dichroism spectroscopy (CD). The CD spectra in the region 180–260 nm were recorded in 50 mm NaH2PO4, 250 mm NaF buffer at a protein concentration of 0.2 mg/ml using an AVIV CD model 215 spectrometer. The protein concentration was verified by amino acid analysis of corresponding samples and further used to calculate the Δε values. The secondary structure contents in WNDP and its mutants were estimated using the set of 43 reference protein spectra and the following algorithms of CD spectra deconvolution: CONTIN, SELCON3, CDSSTR, and CDNN (
      • Sreerama N.
      • Woody R.W.
      ).

      Proteolityc Digestion of Full-length WNDP Expressed in Insect Cells

      10 μg of membrane preparations from insect cells expressing the wt WNDP or His mutants was resuspended in 20 μl of 25 mm imidazole, pH 7.4, and 250 mm sucrose buffer. Tosylphenylalanyl chloromethyl ketone (TPCK)-treated trypsin (Sigma) was added up to 0.83 μg/ml, and the reaction mixture was incubated at room temperature for 30 min. The proteolytic reaction was terminated by addition of the protease inhibitor 4-(2-aminoethyl)benzenesulfonyl fluoride (AEBSF) to 2 μm. The reaction mixture was then diluted to 100 μl with SDS-PAGE loading buffer (100 mm Tris, pH 6.8, 3.3% SDS, 2.6 murea, 5% β-mercaptoethanol), and 10 μl of the mixture was loaded onto a 7.5% Laemmli gel. Following separation and Western blotting, the WNDP proteolytic pattern was analyzed by immunostaining using anti-ATP-BD antibody as described previously (
      • Tsivkovskii R.
      • Eisses J.F.
      • Kaplan J.H.
      • Lutsenko S.
      ). To determine the effect of the nucleotide binding on proteolytic sensitivity of WNDP and the H1069C mutant, 10 μg of respective membrane preparations were incubated in 20 μl of 25 mm imidazole, pH 7.4, and 250 mm sucrose buffer with an increasing concentration of AMP-PNP (0–50 μm) for 10 min at room temperature. Trypsin was then added, and the proteolytic digestion and analysis of the fragments were performed as described above.

      Molecular Modeling

      The high resolution structure of the ATP-binding domain of SERCA1 Ca2+-ATPase (residues Ala-320 to Lys-758) was utilized as a template to generate a molecular model for the ATP-binding domain of WNDP (residues Met-996 to Arg-1322). The ATP-binding domain of WNDP is composed of segments Val-997 to Ala-1065, Asp-1185 to Ile-1236, and Phe-1240 to Ile1311, which share significant homology with Ca2+-ATPase. Segment Ser-1066 to Ile-1184 of ATP-BD is located between these homologous regions and has little similarity to the corresponding region of Ca2+-ATPase. The secondary structure predictions for WNDP and subsequent search in the data base of high-resolution structures using threading algorithms revealed that the predicted fold of the ATP-binding domain of WNDP matches unequivocally to the fold of the Ca2+-ATPase ATP-binding domain in both conserved and non-conserved regions. The alignment of sequences and secondary structures generated by the GenTHREADER program (
      • Jones D.T.
      ) served as a basis for the ATP-BD model. The model was built using Modeler software (
      • Sali A.
      • Overington J.P.
      ) and included thorough loop optimization. 25 models with different loop conformations were generated. The model with the highest quality score, as defined by the Profiles_3D program (
      • Lüthy R.
      • Bowie J.U
      • Eisenberg D.
      ), was selected as the best one. The above procedure was employed to build ATP-BD models using three structural templates, i.e. the high-resolution structures of Ca2+-ATPase in the E1 and E2 states (Protein Data Bank codes 1EUL (
      • Toyoshima C.
      • Nakasako M.
      • Nomura H.
      • Ogawa H.
      ) and 1IWO (
      • Toyoshima C.
      • Nomura H.
      ), respectively) as well as electron microscopy structure in the E2 state (Protein Data Bank code 1KJU(
      • Xu C.
      • Rice W.J.
      • He W.
      • Stokes D.L.
      )).

      RESULTS

      The Effect of the His-1069 Substitution on Folding of the ATP-binding Domain

      As shown in Fig.1, His-1069 is located in the ATP-binding domain of WNDP (ATP-BD) and could be essential for correct folding of this important functional domain. To test this hypothesis, we expressed and purified the wt ATP-BD and ATP-BDs in which the His residue was replaced with one of the following amino acid residues, Gln, Ala, and Cys. Analysis of protein folding using CD spectroscopy revealed that overall secondary structure of all mutants is similar to the wt ATP-BD structure (Fig. 2A, and Table I), indicating that the mutations do not cause marked misfolding of this domain.
      Figure thumbnail gr2
      Figure 2Folding properties of the wt and mutant ATP-binding domains. A, Circular dichroism spectra for wt ATP-BD (WT) and the H1069Q (HQ), H1069A (HA), and H1069C (HC) mutants were recorded at a protein concentration of 0.2 mg/ml in buffer containing 50 mm NaH2PO4 and 250 mmNaF, pH 7.0. B, the fluorescence emission spectra of wt ATP-BD (WT), the H1069Q mutant (HQ), andl-tryptophan (Trp) at equimolar concentrations.
      Table ISecondary structure composition of ATP-BD (WT) and the ATP-BD mutants with His-1069 substituted for Ala (HA), Cys (HC), or Gln (HQ)
      Proteinα-Helixβ-Sheetβ-TurnOther
      %
      WT39.2 ± 4.515.0 ± 2.419.5 ± 2.526.3 ± 4.6
      HA34.2 ± 6.816.3 ± 3.920.8 ± 2.928.8 ± 4.4
      HC36.3 ± 3.416.0 ± 2.919.8 ± 2.528.0 ± 3.7
      HQ44.5 ± 6.312.3 ± 3.219.3 ± 4.624.0 ± 5.5
      Values represent an average and standard deviation of the secondary structure predictions by four different algorithms described under “Experimental Procedures.”
      This conclusion was further verified using intrinsic Trp fluorescence. ATP-BD contains a single Trp residue, which has a buried position in the wt protein (11). Unfolding of ATP-BD results in a decrease of Trp fluorescence intensity and a red shift of the fluorescence maximum (
      • Tsivkovskii R.
      • MacArthur B.C.
      • Lutsenko S.
      ). Fig. 2B illustrates that even the bulkiest H1069Q substitution does not have a significant negative effect on intrinsic Trp fluorescence of ATP-BD, confirming that His-1069 is not critical for ATP-BD folding. This conclusion was at odds with recent cellular studies wherein decreased stability and mislocalization of the H1069Q mutant pointed to somewhat abnormal folding of the mutated WNDP (
      • Payne A.S.
      • Kelly E.J.
      • Gitlin J.D.
      ). Consequently, we hypothesized that mutations of His may alter domain-domain interactions within the full-length WNDP rather than folding of the individual domains. The disruption of this interaction may expose some protein regions to cellular proteases and thus decrease protein stability.
      To determine the effect of His substitution on folding of the full-length WNDP, the Gln, Cys, and Ala substitutions of His-1069 were introduced into the full-length WNDP, and the mutants were expressed in insect cells using baculovirus-mediated infection of SF9 cells. As shown in Fig.3A, the steady state protein levels were similar for all three mutants and not markedly different from the protein levels of the wt WNDP (∼50–70% of the wt). More importantly, we did not observe significant fragmentation of any of the WNDP mutants, suggesting that the expressed proteins are fairly stable (Fig. 3B). To further evaluate stability of the expressed proteins in vitro, the membrane fractions containing the wt and mutant WNDPs were prepared and subjected to limited tryptic digestion, and the proteolytic resistance of the mutants and wt WNDP was compared. Fig. 3B illustrates that the patterns of proteolytic fragments for the wt and mutant WNDPs are very similar. In some experiments we observed a slight difference in relative abundance of the produced fragments between the mutant and wt WNDPs, suggesting that the replacement of His made the contacts between different regions of the protein somewhat more flexible or exposed. There was also a subtle difference in the proteolytic resistance of the three mutants. Consistently, the H1069A mutant was less resistant than the H1069Q and H1069C mutants, whereas the latter mutant appeared to be the most resistant. Overall however, it is clear that the folding of WNDP is not grossly affected by the His-1069 mutations.
      Figure thumbnail gr3
      Figure 3Expression of the His-1069 mutants using baculovirus-mediated infection of insect cells and analysis of their proteolytic stability. A, Coomassie staining of the membrane fractions isolated from SF9 cells infected with empty virus (Mock), virus expressing wt WNDP (WT), or one of the following mutants: H1069Q (HQ), H1069C (HC), and H1069A (HA). 50 μg of total membrane protein is loaded per lane. B, left panel, Western blot analysis of 2 μg of membrane preparations containing either wt or mutant WNDPs. Right panel, 10 μg of membrane preparation containing indicated proteins were treated with trypsin as described under “Experimental Procedures” and analyzed by Western blotting

      The Catalytic Phosphorylation of WNDP by ATP Is Disrupted by Mutation of His-1069

      As we demonstrated previously, WNDP functions as a copper-dependent P-type ATPase,i.e. during its catalytic cycle WNDP transfers the γ-phosphate from ATP to the invariant Asp-1027, generating a fairly stable phosphorylated intermediate (
      • Tsivkovskii R.
      • Eisses J.F.
      • Kaplan J.H.
      • Lutsenko S.
      ). Consequently, to evaluate the effect of His-1069 mutations on WNDP function, we determined whether the mutants were able to form a catalytic phospho-intermediate. Cell membranes containing wt or mutant WNDP were incubated with γ-[32P]ATP, and incorporation of 32P into WNDP was monitored following separation of membrane proteins on an acidic polyacrylamide gel (
      • Sarkadi B.
      • Enyedi A.
      • Foldes-Papp Z.
      • Gardos G.
      ). The catalytically inactive D1027A mutant of WNDP was used in these experiments as a background control.
      As shown in Fig. 4A, all three mutants demonstrated markedly reduced phosphorylation activity, which did not exceed that of the inactive D1027A mutant. Catalytic phosphorylation of the mutants could not be detected even when sensitivity of the assay was increased by raising the specific radioactivity of [γ-32P]ATP 10-fold (data not shown). His-1069 is located in the ATP-binding domain of the protein, and one of the reasons for the observed effect could be a dramatic decrease in the affinity of mutant WNDPs for ATP. To explore this possibility, the ATP concentration in the phosphorylation assay was increased up to 1 mm, but no catalytic phosphorylation of any of the His-1069 mutants was detected. Similarly, changing other reaction conditions (temperature, pH of the reaction in the range of 5.0–8.5, buffer composition) did not lead to appearance of the acylphosphate intermediate for any of the mutant WNDPs (data not shown). Altogether, the results suggested that mutations of His-1069 had a strong negative effect on the catalytic function of WNDP. Further experiments were carried out to determine more precisely which step of the catalytic cycle is most affected.
      Figure thumbnail gr4
      Figure 4Catalytic phosphorylation of the WNDP and mutants using [γ-32P]ATP and32Pi. A and B, 50 μg of total membrane preparation of wild-type or mutant WNDP were incubated with 1 μm of [γ-32P]ATP (5 μCi; specific activity, 20 Ci/mmol) on ice at pH 7.0 for 4 min (for phosphorylation from ATP) (panel A) or with 80 nm of 32Pi (100 μCi; specific activity, 6000 Ci/mmol) at room temperature at pH 7.0 for 10 min (for phosphorylation from Pi) (panel B). The phosphorylated intermediate was detected by autoradiography following separation of samples on an acidic gel. C, the gels were then rehydrated and stained with Coomassie R250 (Protein). Abbreviations for the mutants are the same as in the legend to Fig. . The lane marked DA contains a sample with the D1027A mutant. D, the incorporation of 32P from inorganic phosphate normalized to the amount of WNDP determined by densitometry.

      His-1069 Is Not Essential for Phosphorylation of WNDP by Inorganic Phosphate

      The characteristic feature of the P-type ATPases is their ability to form a phosphorylated catalytic intermediate when incubated with inorganic 32P phosphate in the presence of magnesium. This reaction reflects the reversibility of the dephosphorylation step in the enzyme catalytic cycle. The conformational state of the P-type ATPases, which favors phosphorylation from Pi (the E2-state), is different from the one favoring phosphorylation from ATP (the E1-state). Consequently, we used phosphorylation from Pi to test whether the His-1069 mutations eliminated phosphorylation in general or specifically affected the ATP-dependent step of the catalytic cycle. This experiment produced an interesting result (Fig. 4, BD). The H1069Q and H1069A mutants demonstrated markedly reduced level of phosphorylation from Pi (∼5–10%). In contrast, the H1069C mutant was phosphorylated at levels close to the wt phosphorylation (80–90%). These results indicate that His-1069 is not essential for phosphorylation of WNDP by PI; however, the 1069 position plays an important role in the environment of the catalytic site.
      In the P-type ATPases, phosphorylation from Pi takes place when the exported ion is released from the transport site, allowing the protein to adopt the E2 conformation. Accordingly, addition of the transported ion to the reaction prevents formation of the E2 state and decreases the Pi-mediated phosphorylation. This property of the P-type ATPases was used to test whether the His mutations have an effect on the ability of WNDP to bind copper at the site(s) important for catalytic phosphorylation. As shown in Fig.5, addition of the copper chelator BCS to the wt WNDP stimulates phosphorylation of WNDP from Pi, whereas the addition of copper inhibits the phosphorylation reaction in agreement with the predicted properties of WNDP as a P-type ATPase. Importantly, the His-1069 mutants behaved in these reactions similarly to the wt WNDP, confirming that the His substitutions do not compromise the ability of mutants to bind copper.
      Figure thumbnail gr5
      Figure 5Effect of copper on phosphorylation of the WNDP mutants in the presence of 32Pi inorganic phosphate. Phosphorylation was measured under standard conditions or following treatment of the membrane preparations with the copper chelator BCS with or without subsequent addition of copper. The comparison of the wt WNDP (WT) and two mutant WNDPs (HC and HQ) are shown; each lane contains 50 μg of membrane protein. The HA mutant was also tested and showed a similar response.

      Nucleotide Binding Properties and Conformational Transitions of the H1069C Mutant

      Further studies were carried out with the H1069C mutant, which showed close to normal phosphorylation from Pi and normal response to copper additions but could not be phosphorylated from ATP. One explanation for this phenotype is that binding of ATP is disrupted by the mutation. Alternatively, the H1069C mutant can bind ATP but fails to undergo conformational transitions in response to nucleotide binding. For WNDP, direct measurements of ATP binding are still technically difficult. Consequently, to examine the above possibilities, we carried out limited proteolysis of WNDP in the presence of increasing concentrations of a non-hydrolyzable ATP analogue, AMP-PNP. As shown in Fig.6A, AMP-PNP alters the proteolytic pattern for both the wt WNDP and the H1069C mutant, making several sites less accessible to proteolytic digestion. Moreover, the same protein regions of the wt WNDP and the mutant became resistant to proteolysis, suggesting that the major structural changes upon binding of AMP-PNP are similar for both proteins. Thus, the lack of ATP-dependent phosphorylation in the case of the H1069C mutant is unlikely due to lack of ATP binding.
      Figure thumbnail gr6
      Figure 6Effect of AMP-PNP or ATP on proteolytic sensitivity (A) and on32Pi-mediated phosphorylation (Band C) of the wt WNDP and the H1069C mutant. A, 10 μg of membrane protein was incubated with increasing concentrations (0–50 μm) of AMP-PNP and then proteolyzed with trypsin as described under “Experimental Procedures.” The fragments were separated on a 7.5% Laemmli gel and visualized by immunostaining using antibody directed against the ATP-binding domain of WNDP. B and C, membrane preparations containing the wt WNDP (WT) and the H1069C (HC) mutant were incubated with increasing concentrations of AMP-PNP (panel B) or 200 μm ATP or AMP-PNP (panel C) for 10 min at room temperature, and then32Pi-mediated phosphorylation was carried out. Following autoradiography, the intensity of the 32P-labeled bands was quantified by densitometry.
      The proteolysis experiments provide only a rough estimate of the ability of WNDP and the His mutant to bind the nucleotide. It is quite possible that mutations of His-1069 alter the positioning of ATP within the catalytic site rather than disrupt the nucleotide binding. If this is the case, one may expect to see quantitative differences in the nucleotide-binding characteristics of wt WNDP and the H1069C mutant. To characterize ATP binding more quantitatively, we incubated WNDP and the H1069C mutant with increasing concentrations of the non-hydrolyzable ATP analogue (AMP-PNP) and then measured phosphorylation from Pi. Binding of the nucleotide was expected to stabilize the enzyme in the E1-like state and therefore decrease efficiency of the Pi-mediated phosphorylation. As shown in Fig.6B, the addition of AMP-PNP inhibits incorporation of32Pi for both the wt WNDP and H1069C mutant; however, the efficiency of inhibition is different. The apparentKi values for the wt WNDP and the H1069C mutant are equal to 116 ± 19 μm and 533 ± 166 μm, respectively, suggesting that the H1069C mutant binds AMP-PNP less effectively.
      Finally, we compared the effects of ATP and AMP-PNP on phosphorylation by 32Pi. ATP is expected to allow enzyme turnover, producing an additional decrease in 32P incorporation compared with AMP-PNP. As shown in Fig. 6C, the addition of ATP to the wt WNDP indeed leads to a larger inhibition of the 32Pi-incorporation than a non-hydrolyzable analogue. In the case of the H1069C mutant, ATP and AMP-PNP have a similar effect on the Pi-mediated phosphorylation, suggesting that ATP binds to the mutant but is not hydrolyzed.

      Spatial Location of His-1069 in the ATP-binding Domain

      To better visualize the position of His-1069 in the ATP-binding domain, we generated a molecular model of the ATP-binding domain using the recently published high resolution structure of Ca2+-ATPase in the E2-conformation (
      • Toyoshima C.
      • Nomura H.
      ). In this conformation, the nucleotide-binding region (the N-domain) and the phosphorylation region containing the DKTG and GDGXXD motifs (the P-domain) come fairly close together in contrast to the Ca2+-bound form in which these two subdomains are located far apart (
      • Toyoshima C.
      • Nakasako M.
      • Nomura H.
      • Ogawa H.
      ). Although the E2-state structure cannot be used to predict the exact position of ATP prior to hydrolysis, the model based on this structure provides a useful guide to relative distances and mutual positions of the residues in the ATP-binding domain. The generated model is shown in Fig.7. It revealed that the spatial location of the SEH1069PL sequence is equivalent to that of the438GEAT441 segment of Ca2+-ATPase. The 438GEAT441 segment is a part of the nucleotide-binding region of Ca2+-ATPase and was shown to be in direct proximity to bound ATP (20). Furthermore, even in the E2 state His-1069 is only 18Å away from the Asp residues of the GDGVND motif that are known to be close to the βγ-phosphates of the ATP-magnesium complex during catalysis (
      • Pedersen P.A.
      • Jorgensen J.R.
      • Jorgensen P.L.
      ).
      Figure thumbnail gr7
      Figure 7A molecular model of the ATP-binding domain of WNDP. The model is based on the high-resolution structure of Ca2+-ATPase in the E2-state (Protein Data Bank accession number 1IWO). The molecule is shown in a ribbon representation. The SEHPL motif with the His-1069 side chain and the DKTG and GDGVND motifs are shown in green. An ATP molecule (inpink) at the top is shown to illustrate relative distances between these three motifs. D1027 and arrow indicate the position of Asp-1027, an acceptor of γ-phosphate during ATP-dependent phosphorylation.

      DISCUSSION

      WNDP has a central role in human copper homeostasis; however, the molecular mechanism of the WNDP-mediated copper transport remains poorly understood. Until recently, the functional characterization of this protein has been hindered by the lack of an expression system suitable for biochemical investigations. Recent development of functional expression of WNDP in insect cells (
      • Tsivkovskii R.
      • Eisses J.F.
      • Kaplan J.H.
      • Lutsenko S.
      ) opened the door to analysis of the molecular mechanism of WNDP and supplied necessary tools for understanding how disease-causing mutations affect folding, function, and regulation of WNDP.
      To provide detailed characterization of the Wilson's disease mutations, in this work we expanded our earlier analysis of the enzymatic properties of WNDP. We demonstrated that WNDP can be phosphorylated in the presence of inorganic phosphate and magnesium and that this reaction is facilitated upon removal of copper by BCS and is inhibited by copper addition. We also found that binding of nucleotides induces conformational changes in WNDP. These properties confirmed that the overall catalytic cycle of WNDP resembles those of a typical P-type ATPase. The developed tools were then utilized to characterize the role of His-1069 in the structure and function of WNDP.
      His-1069 is particularly interesting because it is a site of the most frequent mutation in Wilson's disease patients and an invariant residue in the P1-type ATPases. His-1069 is located in the ATP-binding domain of WNDP, a protein region that is most conserved in all P-type ATPases. The ATP-binding domains of several P-type ATPases have been extensively characterized, and chemical modification, mutagenesis, and NMR studies have yielded a detailed map of residues important for nucleotide binding in Ca2+-ATPase (
      • Abu-Abed M.
      • Mal T.K.
      • Kainosho M.
      • MacLennan D.H.
      • Ikura M.
      ). However, this information has only limited value for understanding the role of His-1069, because this residue lies in the region of the ATP-binding domain (the so called N-subdomain), the sequence of which is unique for the P1-type ATPases. No equivalent histidine is present in the ATP-binding domain of Ca2+-ATPase or other well characterized P2-type ATPases. Conversely, several residues known to play an important role in the binding of ATP in Ca2+-ATPase and other P2-ATPases are absent in WNDP and its homologues. This raises an interesting possibility that, despite overall similarity of the catalytic cycles, the P1-type ATPases and P2-type ATPases have distinct subsets of amino acid residues involved in ATP-coordination and possibly catalysis. The results of our experiments support this hypothesis.
      The first important result of this work is the demonstration that folding of either the isolated ATP-binding domain or the full-length WNDP is not significantly altered by mutations of His-1069. These results argue against a significant role for this residue in the overall structure of WNDP and seem to be at odds with recent studies showing decreased intracellular stability of the His-1069 mutant (
      • Payne A.S.
      • Kelly E.J.
      • Gitlin J.D.
      ). Two considerations may help to reconcile this apparent discrepancy. First, although we observed only a slight increase in proteolytic sensitivity of the His-1069 mutants compared with the wt WNDP, it seems likely that in a cell even small changes in the folding are sufficient to decrease protein stability over time. Second, in the cellular studies the negative effect of the H1069Q mutation on intracellular localization of WNDP was eliminated at room temperature (
      • Payne A.S.
      • Kelly E.J.
      • Gitlin J.D.
      ), suggesting that, at temperatures lower than 37 °C, the folding of WNDP was sufficient to pass quality control in the endoplasmic reticulum. For our studies, we produce WNDP in the insect cells, which grow at 27 °C. Therefore, it seems likely that folding of the WNDP mutants in insect cells is better than in mammalian cells. This unexpected benefit of the protein expression in SF9 cells helped us to marginalize the effect of the His mutations on the WNDP structure and dissect the functional role of His-1069.
      Functional analysis of the His-1069 mutants revealed that His is not essential for the dephosphorylation step of the catalytic cycle as evidenced by the normal level of Pi-mediated phosphorylation of the H1069C mutant. Mutations of His-1069 do not disrupt copper binding to the site(s) that control catalytic phosphorylation. Similarly, ATP binding by the mutant, although diminished, is not critically affected. This last result is consistent with the lack of significant structural changes in ATP-BD, where ATP coordination is likely to be provided by several amino acid residues. In striking contrast, phosphorylation from ATP is markedly decreased in the mutants, suggesting that His-1069 plays an important role during this step of the catalytic cycle.
      To better understand the role of His-1069 in ATP-dependent phosphorylation, it is useful to compare the results of our work with recent studies on bacterial homologues of WNDP. The histidine residues equivalent to His-1069 have been mutated in two bacterial P1-type ATPases, a copper-transporting ATPase CopB (
      • Bissig K.D.
      • Wunderli-Ye H.
      • Duda P.W.
      • Solioz M.
      ) and a zinc and cadmium-transporting ATPase, ZntA (
      • Okkeri J.
      • Bencomo E.
      • Pietila M.
      • Haltia T.
      ). In CopB, the His → Gln mutation drastically reduced the ATPase activity and decreased phosphorylation from ATP by 80% (
      • Bissig K.D.
      • Wunderli-Ye H.
      • Duda P.W.
      • Solioz M.
      ). In a study of ZntA, introduction of His → Gln mutation also decreased the ATPase activity but less than in the case of CopA (37% of activity remained). The level of phosphorylation from ATP and Pi in ZntA mutants was reduced, and dephosphorylation was slowed down (
      • Okkeri J.
      • Bencomo E.
      • Pietila M.
      • Haltia T.
      ). In this earlier work, the multiple consequences of the His substitution made it difficult to assign a specific role to the histidine. Nevertheless, these studies are very helpful. Together with our results, they indicate that His strongly affects the environment in close proximity to the phosphorylated Asp but is not essential for the chemistry of ATP hydrolysis, because ATP-hydrolysis is not abolished in the bacterial ATPase mutants. The mild decrease in the apparent affinity for AMP-PNP and the drastic effect on phosphorylation from ATP observed in our experiments further suggest that His-1069 plays a key role in orienting ATP with respect to catalytic aspartate, thus allowing phosphorylation to occur.
      The resolution of our methods is currently insufficient to distinguish between direct involvement of His-1069 in coordination of ATP and a more indirect role in which His-1069 interacts with other residues in WNDP and helps to organize the catalytic site by forming specific protein-protein contacts. It seems likely that, prior to transfer of the γ-phosphate, His-1069 directly orients ATP with respect to catalytic Asp, whereas after the phosphorylation step the role of His-1069 became less direct. This conclusion is consistent with the predicted location of His-1069 in the ATP-binding domain (Fig. 7) and is in agreement with our experimental data showing normal PI phosphorylation by the H1069C mutant. The negative effect of the H1069A and H1069Q mutations on Piphosphorylation most likely stems from altered protein-protein contacts within WNDP. It is interesting that H1069A substitution has a stronger negative effect on structure and function of WNDP than either H1069Q or H1069C mutations, suggesting that the presence of a residue that can form a hydrogen bond is tolerated better than the presence of the small and neutral Ala.
      In summary, we characterized the role of the invariant His-1069 in the structure and function of WNDP. We conclude that His-1069 is important for catalytic activity of WNDP and is likely to be involved in positioning of ATP prior to transfer of the γ-phosphate. Because His in the SEHPL motif is an invariant residue in all P1-type ATPases, it is plausible that this histidine has a similar role in all these transporters. We also demonstrated that, at lower temperatures, the folding of the common Wilson's disease mutant H1069Q is close to normal but that this improvement of protein folding does not restore catalytic activity of WNDP, an important conclusion for future development of the corrective therapies for Wilson's disease.

      Acknowledgments

      We thank Dr. Jens Peter Andersen and Dr. Michel Green for helpful discussion and for pointing out the similarities in the predicted spatial location of His-1069 and Glu-441 of SERCA1 Ca2+-ATPase. We also thank Kerry Maddox (Shriners Hospital for Children, Portland, OR) for performing amino acid analysis, Ms. Tina Purnat for help with preparation of figures, and Drs. D. Huster and J. H. Kaplan and Ms. Mee Min for critical reading of the manuscript.

      REFERENCES

        • Harris E.D.
        Annu. Rev. Nutr. 2000; 20: 291-310
        • Petris M.J.
        • Mercer J.F.
        • Culvenor J.G.
        • Lockhart P.
        • Gleeson P.A.
        • Camakaris J.
        EMBO J. 1996; 15: 6084-6095
        • Llanos R.M.
        • Mercer J.F.
        DNA Cell Biol. 2002; 21: 259-270
        • Thomas G.R.
        • Roberts E.A.
        • Walshe J.M.
        • Cox D.W.
        Am. J. Hum. Genet. 1995; 56: 1315-1319
        • Payne A.S.
        • Kelly E.J.
        • Gitlin J.D.
        Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 10854-10859
        • Huster D.
        • Hoppert M.
        • Zinke J.
        • Lutsenko S.
        • Lehmann C.
        • Mossner J.
        • Berr F.
        • Caca K.
        Gastroenterology. 2003; 124: 335-345
        • Petris M.J.
        • Voskoboinik I.
        • Cater M.
        • Smith K.
        • Kim B.E.
        • Llanos R.M.
        • Strausak D.
        • Camakaris J.
        • Mercer J.F.
        J. Biol. Chem. 2002; 277: 46736-46742
        • Hung I.H.
        • Suzuki M.
        • Yamaguchi Y.
        • Yuan D.S.
        • Klausner R.D.
        • Gitlin J.D.
        J. Biol. Chem. 1997; 272: 21461-21466
        • Iida M.
        • Terada K.
        • Sambongi Y.
        • Wakabayashi T.
        • Miura N.
        • Koyama K.
        • Futai M.
        • Sugiyama T.
        FEBS Lett. 1998; 428: 281-285
        • Tsivkovskii R.
        • Eisses J.F.
        • Kaplan J.H.
        • Lutsenko S.
        J. Biol. Chem. 2002; 277: 976-983
        • Tsivkovskii R.
        • MacArthur B.C.
        • Lutsenko S.
        J. Biol. Chem. 2001; 276: 2234-2242
        • Sreerama N.
        • Woody R.W.
        Anal. Biochem. 2000; 287: 252-260
        • Jones D.T.
        J. Mol. Biol. 1999; 287: 797-815
        • Sali A.
        • Overington J.P.
        Protein Sci. 1994; 3: 1582-1596
        • Lüthy R.
        • Bowie J.U
        • Eisenberg D.
        Nature. 1992; 356: 83-85
        • Toyoshima C.
        • Nakasako M.
        • Nomura H.
        • Ogawa H.
        Nature. 2000; 405: 647-655
        • Toyoshima C.
        • Nomura H.
        Nature. 2002; 418: 605-611
        • Xu C.
        • Rice W.J.
        • He W.
        • Stokes D.L.
        J. Mol. Biol. 2002; 316: 201-211
        • Sarkadi B.
        • Enyedi A.
        • Foldes-Papp Z.
        • Gardos G.
        J. Biol. Chem. 1986; 261: 9552-9557
        • Abu-Abed M.
        • Mal T.K.
        • Kainosho M.
        • MacLennan D.H.
        • Ikura M.
        Biochemistry. 2002; 41: 1156-1164
        • Pedersen P.A.
        • Jorgensen J.R.
        • Jorgensen P.L.
        J. Biol. Chem. 2000; 275: 37588-37595
        • Bissig K.D.
        • Wunderli-Ye H.
        • Duda P.W.
        • Solioz M.
        Biochem. J. 2001; 357: 217-223
        • Okkeri J.
        • Bencomo E.
        • Pietila M.
        • Haltia T.
        Eur. J. Biochem. 2002; 269: 1579-1586