|Home | About | Journals | Submit | Contact Us | Français|
The ctpA (ccoI) gene product, a putative inner membrane copper-translocating P1B-type ATPase present in many bacteria, has been shown to be involved only in the cbb3 assembly in Rhodobacter capsulatus and Bradyrhizobium japonicum. ctpA was disrupted in Rubrivivax gelatinosus, and the mutants showed a drastic decrease in both cbb3 and caa3 oxidase activities. Inactivation of ctpA results also in a decrease in the amount of the nitrous oxide reductase, NosZ. This pleiotropic phenotype could be partially rescued by excess copper in the medium, indicating that CtpA is likely a copper transporter that supplies copper-requiring proteins in the membrane with this metal. Although CtpA shares significant sequence homologies with the homeostasis copper efflux P1B-type ATPases including the bacterial CopA and the human ATP7A and ATP7B, disruption of ctpA did not result in any sensitivity to excess copper. This indicates that the CtpA is not crucial for copper tolerance but is involved in the assembly of membrane and periplasmic copper enzymes in this bacterium. The potential roles of CtpA in bacteria in comparison with CopA are discussed.
Copper is an essential cellular component, and it is required for a broad range of enzymes involved in numerous metabolic pathways including respiration and free radical scavenging (1). Maintenance of copper homeostasis and its delivery to target proteins are critical and require a variety of membrane copper transporters and chaperones. In bacteria, multiple copper enzymes are located within the periplasm or the inner membrane and are involved in oxidative reactions such as anaerobic and aerobic respiration. In the anaerobic denitrification pathway, the last step is catalyzed by the multicopper nitrous oxide reductase (NosZ)2 that generates N2 from N2O. NosZ is a dimeric enzyme with a binuclear CuA center and a tetranuclear CuZ center (Cu-S cluster) in the catalytic site (2). In the aerobic electron transport chain, the cytochrome c oxidases (Cox) are multi-heme-copper complexes catalyzing the reduction of O2 to water. Subunit II of the aa3 Cox contains the binuclear CuA center, which is the primary acceptor of electrons from reduced cytochrome c. Subunit I contains one low-spin heme a and a binuclear metal center formed by a high-spin heme a3 and CuB (3). The heme-copper cbb3 Cox are present only in eubacteria and are different from the class of aa3 Cox by their cofactor composition. The cbb3 enzyme lacks the CuA-containing subunit but instead features two membrane-associated heme subunits, the monoheme cytochrome CcoO and diheme cytochrome CcoP, in addition to the membrane-integral catalytic subunit I CcoN containing a low-spin heme b and a high-spin b3 in CuB-b3 binuclear center (4).
Insertion of copper in the redox active metal centers is essential for the formation of a functional enzyme complex. A number of copper-binding proteins were shown to be required for the transport and delivery of this metal to copper enzymes in bacteria. The maturation of the NosZ requires several proteins. In Pseudomonas stutzeri it was suggested that the ABC-type transporter NosDFY is involved in the CuZ center maturation (2) and that the outer membrane protein NosA and the NosL copper-soluble chaperone are involved in CuA insertion (5, 6). Curiously, and despite the occurrence of aa3- and cbb3-type Cox in many bacteria, very few data are available on their biogenesis and on the proteins involved in copper delivery. The biogenesis of cbb3 was mainly studied in Rhodobacter capsulatus and in Bradyrhizobium japonicum. In both strains, the ccoGHIS (fixGHIS) gene cluster was shown to be required for the formation of a functional cbb3 (7, 8). In Rhodobacter capsulatus it was suggested that CcoS is involved in the maturation of the CcoN CuB-b3 center. SenC, a homologue of Sco1, is also required for the assembly of an active cbb3 complex in R. capsulatus (9). CcoI, a member of P1B-type ATPase, is required for a normal amount of active cbb3 complex in both R. capsulatus and B. japonicum (8, 10).
Only trace amounts of copper are needed to sustain life. Excess copper is extremely toxic for cell viability. Most, if not all, cells have mechanisms to treat an excess of copper. Most free-living bacteria seem to possess the copper efflux ATPase CopA or homologues (11,–13). This copper ATPase belongs to the P1B-type ATPases subgroup of the P-type ATPases superfamily. They are characterized by conserved motifs including one to six CXXC motif metal binding domains at their N termini, the DKTGT motif, the ATP binding domain, and a conserved intramembranous CPC motif involved in copper transport. In Escherichia coli and other bacteria, copper tolerance requires the induction of copA and the cusCFBA system (12). The disruption of copA results in sensitivity to copper (14).
Rubrivivax gelatinosus is a purple β-proteobacterium that grows by photosynthesis or aerobic respiration using the cbb3 Cox and the bd-type quinol oxidase. A coxBAC operon encoding a caa3 Cox is also present in the genome of this bacterium. Genetic analyses showed that caa3 is not active in the wild type background. However, a mutant (Res2) that expresses only caa3 was isolated and characterized (15). A ccoI gene homologue encoding a putative P1B-type ATPase was identified in the vicinity of the cbb3-encoding operon. In this work, we examined the role of ccoI in the assembly of active cbb3 and caa3 Cox. Analyses of the ccoI-disrupted strains indicate that CcoI is a copper translocating pump required not only for cbb3 but also for caa3 and for NosZ assembly. Excess copper can partially restore the assembly of these complexes in the membrane. We have, therefore, renamed the gene ctpA and the gene product CtpA for copper transport protein. Given the resemblance between CtpA and CopA, we also investigated the involvement of CtpA in copper tolerance in this bacterium.
E. coli was grown at 37 °C in LB medium. R. gelatinosus was grown at 30 °C in the dark and aerobically (high oxygenation: 250-ml flasks containing 20 ml of medium), semi-aerobically (250-ml flasks containing 120 ml of medium), or micro-aerobically (low oxygenation: 50-ml flasks filled with 50 ml of medium) or in light (photosynthesis) in filled and sealed tubes on malate growth medium (16). Antibiotics were used at the following concentrations: kanamycin (Km), 50 μg/ml; ampicillin, 50 μg/ml; streptomycin, 50 μg/ml; spectinomycin, 50 μg/ml; trimethoprim (Tp), 50 μg/ml. The malate copper concentration was 1.6 μm, and exogenous copper CuSO4 was added to growth the medium at different concentrations. Bacterial strains and plasmids are listed in Table 1.
Standard methods were performed according to Sambrook et al. (17) unless indicated otherwise. The sequences reported in this paper were deposited in GenBankTM with accession numbers AY648960 (ctpA) and GQ900543 for (nosZ).
To clone the ctpA gene from genomic DNA of R. gelatinosus, a 2.5-kb fragment was amplified by PCR using the primers ol-342 and ol-328 (Table 2). The fragment was cloned into the PCR cloning vector pGEM-T to give pBS40. The ctpA gene was inactivated by the insertion of the Km and Tp cassettes at the PflMI and PshaI sites within the ctpA coding sequence. Briefly, plasmid pBS40 was subjected to restriction enzyme digestion and treated with Klenow polymerase before ligation with the 1.2-kb EcoRI blunt-digested Km cassette or the 0.9-kb XbaI-digested Tp cassette. The resulting recombinant plasmids were designated pBS41 (ctpA::Km) and pBS42 (ctpA::Tp).
A 1.56-kb ccoO fragment was cloned into pGEM-T using the primers 303–314 to give pBS11. ccoO gene was inactivated by the insertion of the Km cassette at the BsmBI site within the ccoO coding sequence. To that aim, pBS11 was subjected to restriction enzyme digestion and treated with Klenow polymerase before ligation with the 1.2-kb EcoRI blunt-digested kanamycin. The resulting recombinant plasmid was designated pBS21.
The nosZ primers designed according to nosZ from Ralstonia eutropha (NosZ-For1 and NosZ-Rev2) were used to clone the nosZ gene from the chromosome and from a genomic DNA library of R. gelatinosus. The 1.4-kb fragment obtained by PCR was cloned into the pDrive plasmid to give pBS70. The nosZ gene was inactivated by the insertion of a 0.9-kb SmaI-digested Tp cassette in the unique MscI site of nosZ. The resulting recombinant plasmid was designated pBS71.
Transformation of R. gelatinosus cells was carried out by electroporation as previously described (16). Transformants were selected on malate plates supplemented with the appropriate antibiotic either under aerobic conditions or anaerobic photosynthesis in sealed bags with H2- and CO2-producing gas packs (Anaerocult Merck). After transformant selection, template genomic DNA was prepared from the transformants, and confirmation of the antibiotic resistance marker presence at the desired locus was performed by PCR.
Peptide mixtures generated by standard in-gel digestion with trypsin (Gold, Promega) were SpeedVac-treated for 10 min, then analyzed with the QTOF Premier mass spectrometer coupled to the Acquity nano-ultraperformance liquid chromatography equipped with a trapping column (Symetry C18, 180 μm × 20 mm, 5-μm particle size) and an analytical column (BEH130 C18, 75 μm × 100 mm, 1.7-μm particle size) (Waters). The aqueous solvent (buffer A) was 0.1% formic acid in water, and the organic phase (buffer B) was 0.1% formic acid in acetonitrile. A 2–40% B gradient was set for 25 min. For exact mass measurements, glufibrinopeptide reference (m/z = 785.8426) was continuously supplied during nano-liquid chromatography-MS/MS analyses using the lockspray device. Peptide mass measurements were corrected during data-processing, and peak lists were generated by PLGS (ProteinLynx Global Server, Waters). Processed data were submitted to Mascot searching using the following parameters: data bank, NCBInr; taxonomy, bacteria; peptide tolerance, 20 ppm; fragment tolerance, 0.1 Da; digest, reagent trypsin; variable modification, oxidation (methionine); fixed modification, carbamido-methylation (cysteine). Validation criteria for protein identification were two peptides with a Mascot individual ion score of >43.
EPR spectra were recorded on a Bruker ESP300e X-band spectrometer fitted with an Oxford Instrument He-cryostat and temperature control system.
The membranes were prepared as previously described (16). To assay for cbb3 activity, wild type (wt), and ctpA− strains were grown semi-aerobically. For the caa3 activity, Res2 and Res2-ctpA− mutants were grown aerobically. Blue Native polyacrylamide gel electrophoresis and in-gel Cox activity assays were performed as described in Schagger and von Jagow (18) and Lemaire and Dujardin (19). 2 mg of membrane proteins were solubilized for 30 min on ice in 0.1 m Tris-HCl buffer, pH 8, containing 1% (w/v) n-dodecyl β-d-maltoside, 10% (v/v) glycerol, and 0.3 m NaCl. The supernatant was cleared by centrifugation at 70,000 rpm for 15 min. 4.5 μl of loading buffer (5% (w/v) Coomassie blue G250 (Serva), and 500 mm n-aminocaproic acid), 5% (w/v) blue G250, and 500 mm n-aminocaproic acid were added to the supernatant. BN electrophoresis was performed with modifications; the gel buffer contained 75 mm Bis-Tris and 0.75 m n-aminocaproic acid, pH 7.0, and 0.03% (w/v) of n-dodecyl β-d-maltoside was added in the cathode buffer. The Cox were revealed with 3,3′-diaminobenzidine tetrahydrochloride (DAB) (19).
The artificial electron donor TMPD can be oxidized by the Cox to form a blue indophenol compound (20). Cox activity was measured spectrophotometrically in whole cells by following the increase in the absorbance (A562 nm) at room temperature on a Carry 500 spectrophotometer. Cells were grown to reach the late exponential growth phase. The reaction was started upon the addition of 20 μl of 50 mm TMPD to 3 × 109 cells (1 A680) in a 2-ml final volume.
The Cox activity was detected by staining colonies with a 1:1 mixture of 35 mm α-naphthol dissolved in ethanol and 30 mm Nadi. Colonies with Cox activity turn blue after the addition of 10 μl of the substrate. Staining of membranes for covalently bound heme with 3,3′,5,5′-tetramethylbenzidine was achieved as described in Ouchane et al. (21).
In most cbb3-containing bacteria, the maturation genes ccoI, ccoS, ccoG, and ccoH are often located downstream of the ccoNOQP operon (22). In β-proteobacteria, in addition to the absence of the ccoH gene, homologues of ccoI and ccoS are located upstream of the structural operon. In other bacteria, ccoI and ccoS homologues are, however, not linked to the ccoNOQP operon; this is the case, for instance, of Thiobacillus denitrificans, Helicobacter pylori, and Campylobacter hominis (Fig. 1). In the β-proteobacterium R. gelatinosus, the ctpA (ccoI) gene is located upstream of the ccoNOQP operon. It encodes a 768-amino acid protein of ~81 kDa. It is highly homologous to P1B-type ATP-dependent metal ion transporters. R. gelatinosus CtpA contains only one copper binding domain in its N terminus. All the structural signatures common to the P1B-type ATPases are present within the sequence. Protein hydropathy analyses suggest eight putative transmembrane-spanning helices with a CPC motif in the sixth helix of the protein. The R. gelatinosus CtpA protein shares 56% amino acid identity with the putative P1B-type ATPases of β-proteobacteria Methylibium petroleiphilum (Mpe_A2479) and Leptothrix cholodnii (Lcho_2781). The identity is lower when CtpA is compared with the α-proteobacteria B. japonicum FixI (34% identity) or to R. capsulatus CcoI (35% identity). When compared with the copper efflux pumps CopA, CtpA displays 32% identity with CopA from E. coli and 34% identity with CopA from the closest species M. petroleiphilum and L. cholodnii. It also exhibits up to 29% identity with the human ATP7A and ATP7B sequences involved in Menkes and Wilson diseases (23).
Because of its significant primary and secondary structure homologies to the copper efflux P1B-type ATPase pump CopA, we asked whether CtpA is important for copper tolerance in this bacterium. For this purpose, a ctpA− deletion strain was constructed. The ctpA gene is not essential; however, the mutant was affected in its aerobic growth, with a doubling time of 3 h 30 min under semi-aerobic conditions compared with 2 h 15 min for the wild type and 4 h for the cbb3 null mutant. To check the effect of excess CuSO4 on R. gelatinosus growth, ctpA− mutant and the wild type were grown in liquid and on solidified malate medium (containing 1.6 μm CuSO4) supplemented with 0.1, 0.5, 1.0, and 1.5 mm CuSO4, and their ability to grow in the presence of increasing concentrations of CuSO4 was evaluated. As shown in supplemental Fig. S1, when the ctpA− deletion strain was grown on solidified medium containing 1.6 μm CuSO4, the mutant was less pigmented than the wild type; however, this phenotype was suppressed with increasing concentrations of CuSO4. The disrupted strain did not exhibit any increased sensitivity or resistance compared with the wild type strain even with a high copper concentration (supplemental Fig. S1). This indicates that CtpA protein is dispensable for copper tolerance in this bacterium in contrast to CopA and suggests that other genes contribute to copper tolerance in R. gelatinosus.
Considering the slow semi-aerobic growth of the ctpA− mutant, we first checked the ability of the mutant to respire with the Cox cbb3 oxidase. Colonies were tested with Nadi as an artificial electron donor, a method that has been used to detect cytochrome c oxidase-dependent respiration (24). The wild type colonies show cbb3 oxidase activity and reach complete blue coloration in ~45 s; in contrast, it takes more than 20 min to obtain even a faint blue coloration in the ctpA−, indicating that the mutant may contain a very low amount of cbb3 compared with the wild type. In the ccoO− control strain, the colonies never turn blue because of the absence of cbb3 oxidase. We further characterized these strains by in-gel Cox activity assays of solubilized membranes separated on a BN-PAGE (Fig. 2). A DAB-positive band (brownish color) was revealed in the wild type native membranes. This band appear within 10 min and was absent from the cbb3− strain regardless of length of exposure. This band corresponds to cbb3 oxidase in the wild type. In the ctpA− mutant, only traces of this band could be seen after 1 h. These results show the importance of CtpA and its requirement in the assembly of an active cbb3 complex in the membranes. Several other high molecular mass complexes were also visible at the top of the gel and will be presented and discussed below.
The decrease in cbb3 oxidase activity in the ctpA− mutant may be associated with a lower amount of protein in the membrane or with the lack of functionality of the enzyme because of the absence of copper. To assess the impact of ctpA disruption on the amount of cbb3 oxidase in the membranes, staining of wild type and the ctpA− mutant membranes for covalently bound heme was performed to check the presence of cytochrome c subunits CcoP (32 kDa) and CcoO (23 kDa) of cbb3. As shown on the heme-staining gel (Fig. 3), for membranes containing comparable amounts of the reaction center-attached tetraheme cytochrome PufC, both CcoP and CcoO are drastically reduced in the ctpA− membranes compared with the wild type. This indicates that inactivation of ctpA results in a significantly reduced amount of cbb3 within the membrane. This may be the consequence either of degradation of the oxidase subunits or to a regulatory negative feedback on cbb3 expression in the absence of copper (see “Discussion”).
Considering the high homology of CtpA to other P1B-type ATPase copper transporters and that the cbb3 Cox contains a copper atom in the binuclear center, we tested if the ctpA− phenotype and the cbb3 activity could be rescued in the ctpA− mutant by an excess of exogenous copper in the medium. As shown in supplemental Fig. S1, when the ctpA− deletion strain was grown on plates containing increasing concentrations of CuSO4, the pigmentation and growth phenotype of the mutant was suppressed, suggesting that copper may restore Cox activity in the mutant. To ascertain whether this suppression is concomitant to a rescue of the cbb3 Cox activity, cells were grown in liquid medium with increasing concentrations of copper, and Cox activity (TMPD) was measured using whole cells of the wild type and the ctpA− mutant. The activity assay (Fig. 4) indicated that the ctpA− mutant displayed a 90% decrease in cbb3 activity relative to that observed in the wild type cells grown in malate medium. This enzymatic activity increased in the mutant with the addition of copper to reach 65% that of the wild type activity in the presence of 250 μm copper. An increase of the cbb3 activity was also observed for the wild type. As a control, the Cox activity was also assayed in the cbb3-deficient ccoO− mutant strain grown under the same conditions. The ccoO− mutant did not exhibit any Cox activity regardless of copper concentration.
Interestingly, and in contrast to copper, the addition of other divalent cations such as cobalt and nickel to the growth medium of the ctpA− mutant had no suppressing effect on the growth of the mutant (data not shown). This suggests that these metals did not restore the cbb3 oxidase activity in the mutant. It appears, thus, that exogenous copper specifically allows the assembly of a functional cbb3 oxidase in the ctpA− mutant. Additional evidence confirming the rescue of the ctpA− phenotype by excess copper is provided by the analysis of in-gel Cox activity in solubilized membranes from the wild type and the mutant. An in-gel Cox activity assay (DAB) revealed an increase of cbb3 oxidase activity in the wild type grown with 20 μm CuSO4 (Fig. 2). Two DAB-positive bands (brownish color) corresponding to cbb3 oxidase were revealed in the wild type native membranes. Similarly, the addition of 20 μm CuSO4 to the ctpA− mutant results in a substantial increase of the cbb3 Cox activity. Both bands corresponding to cbb3 oxidase are revealed in the mutant. These results show that addition of exogenous copper to the growth medium results in partial suppression of the cbb3 oxidase activity phenotype of ctpA− strain and accounts for the improved growth ability of the ctpA− mutant on copper. Together, these data support the role of CtpA as a putative membrane copper transporter required for the assembly of a functional cbb3 oxidase.
CcoI has been shown to be involved only in cbb3 oxidase biogenesis in R. capsulatus and B. japonicum (7, 8). Nonetheless, inactivation of fixGHIS had no effect on the assembly of the aa3 type Cox in B. japonicum (7). In R. capsulatus, the inability to rescue the ccoI− mutant with excess copper suggested that CcoI is specific to the cbb3 oxidase and that CcoI is not an all-purpose copper transporter in these bacteria (8). The usual occurrence of the ccoI gene in the vicinity of the ccoNOQP operon corroborates this assumption. However, more detailed genomic surveys have challenged this view as in certain bacterial species, the ccoI homologue is not linked to the ccoNOQP operon (22) as shown previously (Fig. 1). Interestingly, when membranes from the wild type and ctpA− mutant were analyzed by EPR, the spectra showed a drastic decrease of copper content in the ctpA− membranes (Fig. 5). These data raise the question of the specificity of CtpA to cbb3 and prompted us to question whether CtpA is required for the assembly of other membrane copper proteins. We took advantage of the Res2 suppressor strain that expresses only a functional caa3 and not cbb3 to consider the involvement of CtpA in the assembly of an active caa3 Cox in R. gelatinosus. The ctpA gene was inactivated in this strain. The Res2-ctpA− mutant was affected in its aerobic growth when compared with the Res2 strain. The doubling time of Res2 was 10 h, whereas in the Res2-ctpA− mutant the doubling time was 22 h. Because of the low TMPD oxidase activity of caa3 in whole cells, we analyzed the caa3 activity in membranes of the Res2 and Res2-ctpA− mutant by a BN-PAGE in-gel Cox assay. A DAB-positive complex of about 240 kDa, corresponding to an active caa3, was detected in Res2 membranes (Fig. 6). The caa3 complex was substantially diminished in the Res2-ctpA− mutant strain. This shows that CtpA is also required for the assembly of a functional caa3 Cox in R. gelatinosus.
Copper supplementation (20 μm CuSO4) resulted in an increase of the caa3 oxidase activity in the Res2 strain. It also partially restored caa3 Cox activity in the Res2-ctpA− mutant (Fig. 6). These data reinforce the role of CtpA as a putative membrane copper transporter required not only for biogenesis of the Cox cbb3 but also for the assembly of an active caa3 Cox.
The BN-PAGE comparison of the solubilized membranes from the wild type and the ctpA− mutant revealed, in addition to the drastic reduction in cbb3 activity, a decrease in the amount of a high molecular blue-violet band observed in the wild type (Fig. 7). The blue-violet color suggests that this band may correspond to a copper-containing complex, in agreement with a “multi-purpose” copper transporter role for CtpA in this bacterium.
To identify the protein composition of this band in the membrane fraction from the wild type and from the ctpA− mutant, the bands were excised from BN-PAGE, immediately destained, and subjected to in-gel digestion with trypsin before mass spectrometry analysis (nano-liquid chromatography-MS/MS). Although most of the identified proteins were the same in wild type and ctpA− mutant bands, the unambiguously identified NosZ was found specifically in the wild type membranes and not in the ctpA− mutant membranes. Three peaks corresponding to the 127–140, 376–398, and 549–559 amino acidic sequences of the R. eutropha NosZ (accession number NP_942887.1) were detected and efficiently fragmented; the average mass accuracy was 15 ppm, and the Mascot individual score was more than 60 for each ion. These results demonstrate that R. gelatinosus expresses a membrane-associated copper-containing N2O reductase (NosZ) and suggests that CtpA is also required for the assembly of this enzyme.
The four peptides detected by mass spectrometry match the theoretical masses of tryptic peptides derived from the NosZ amino acid sequences of R. eutropha. To confirm the presence of a nos gene cluster in R. gelatinosus, primers were designed according to the R. eutropha nosZ sequence and used to screen a R. gelatinosus genomic library. A plasmid containing an 8.5-kb DNA fragment was isolated. Sequencing of this fragment revealed the presence of a nosZ gene and five related nos genes presumably involved in the assembly of NosZ. The gene organization and sequence of the R. gelatinosus nos cluster (nosXC-nosZRDF) is highly homologous to the nos cluster from R. eutropha. Secondary structure analyses predict that NosZ is a soluble protein, yet the protein co-purifies with the membrane fraction, suggesting that a fraction of NosZ can bind to the membrane.
A strain where the nosZ gene was disrupted by the insertion of an antibiotic cassette was constructed. Membranes from wild type, ctpA−, and nosZ− mutants were solubilized and analyzed by BN-PAGE (supplemental Fig. S2). Like in the ctpA− mutant, the blue-violet band present in the wild type was not present in the nosZ− strain. An in-gel Cox activity assay revealed the presence of an active cbb3 oxidase in both the wild type and the nosZ− mutant but not in ctpA− strain (supplemental Fig. S2). These data confirm the presence of NosZ in the solubilized membrane fraction of R. gelatinosus and the requirement of CtpA for the assembly of this multicopper periplasmic enzyme and the Cox cbb3. Furthermore, the data reveal for the first time that R. gelatinosus possesses a N2O reductase and may be a denitrifying photosynthetic bacterium.
As for cbb3 and caa3 Cox, we predicted that addition of exogenous copper would restore the assembly of NosZ in the ctpA− mutant. To test this, the ctpA− mutant was grown in 20 and 100 μm excess copper, and membranes were analyzed on BN-PAGE. The addition of copper to the ctpA− mutant restored the cbb3 activity and increased the intensity of the blue-violet band that contains NosZ (Fig. 8). This indicates that NosZ can also be rescued in the ctpA− mutant by exogenous copper. We cannot exclude at this stage that copper may also induce the expression of other proteins or complexes that co-migrate with the NosZ band. These results confirm the role of ctpA in the assembly of NosZ and support the hypothesis that in R. gelatinosus, CtpA is a putative copper transporter that delivers and sustains different membranous and membrane-associated complexes with copper.
Copper is required as a catalytic cofactor for many enzymes in the periplasm and the cytoplasmic membrane of bacteria. Nevertheless, and in contrast to copper tolerance and efflux, the mechanisms involved in copper uptake transport across membranes and delivery to cupro-proteins in bacteria are yet poorly understood (11, 12). So far only two copper uptake systems involved in the assembly of the membrane copper-requiring enzymes have been identified in bacteria, in R. capsulatus (8) and in Synechocystis PCC6803. In this cyanobacterium, copper delivery to thylakoids involves the cytoplasmic and thylakoid membrane P1B-ATPases CtaA and PacS, respectively (25). The ctpA gene, encoding a metal translocating P1B-type ATPase closely related to copper pumps, is present in many bacteria. Here we identify ctpA in the purple photosynthetic bacterium R. gelatinosus, and we propose that it encodes a copper-translocating P-type ATPase needed for the activity of three copper-requiring complexes but not essential for copper tolerance.
So far, the physiological role of CtpA homologues has been studied only into two species. R. capsulatus contains only cbb3 Cox, and the role of CtpA in cbb3 assembly was clearly demonstrated (8). In B. japonicum, analysis of the Fix mutants showed that fixI was involved in cbb3 assembly but not in aa3 assembly (7). In this work we have shown that CtpA is required for both cbb3 and caa3 oxidases. Excess exogenous copper could restore, although partially, the activity of both oxidases in the ctpA− background. This is consistent with a role for CtpA in copper transport and delivery to these complexes. At this stage, the role of CtpA as an influx or efflux transporter remains to be solved. However, considering that inactivation of ctpA affects the activity of membrane proteins, it is more likely that it should transport copper from the cytoplasm to the membrane. If CtpA is an influx pump, then its inactivation should not affect the activity of the oxidases as copper is presumably present in the periplasm and available for these complexes or for the chaperons. Nonetheless, if CtpA is involved in copper import, an alternative scenario would be that in the absence of CtpA, the expression of genes encoding proteins that require copper (cbb3, caa3v and nosZ) may be repressed under copper limiting conditions. Further experiments will be undertaken to check the abundance of transcripts in the wild type and in the ctpA− mutant.
An unexpected result was the discovery of the NosZ in R. gelatinosus and the requirement of CtpA for the assembly of this enzyme. This result further supports our statement that ctpA is not specific to cbb3. NosZ biogenesis has been most thoroughly studied in P. stutzeri and requires both copper and sulfur (2, 26). Copper delivery to the periplasm involves the outer membrane channel NosA (5, 6). In many genomes, the nosA gene was absent, although nosZ is present, suggesting that NosA is not specific to NosZ. This is in agreement with data in Pseudomonas putida suggesting that other copper transporting proteins can supply NosZ with copper in this bacterium (6).
The significant sequence homology of CtpA and CopA suggests that CtpA might be also crucial for copper tolerance in R. gelatinosus and in R. capsulatus (8). However, in both strains the ctpA− mutant was not sensitive to excess copper and behaved like the wild type. This indicates that CtpA is not required for copper tolerance and suggests that other pumps involved in this process exist in these bacteria to respond to excess copper. In Pseudomonas aeruginosa, which possesses two operons encoding cbb3 and a homologue of ctpA (PA1549), the PA1549 mutant did not show any phenotype in excess copper (27). More recently, expression of ctpA (PA1549) in the copper-sensitive strain of E. coli (GG44) did not restore resistance to copper (28). Similar results were previously obtained in R. capsulatus; the ccoI− mutant was neither resistant nor hypersensitive to exogenous copper (8). These data discriminate between CtpA and CopA and assign different roles to these P-type ATPase copper transporters. CtpA may be only required to deliver copper into the membrane and the periplasm. This assumption is further supported by the phenotype of the CopA mutant in H. pylori. Despite the presence of ctpA (fixI) gene in its genome, the copA− mutant exhibits an enhanced growth sensitivity to copper. Interestingly, the micro-aerobic growth of the copA− mutant was not affected under normal copper conditions (29). Given that H. pylori sustains its micro-aerobic growth with a cbb3 oxidase, we presume that CopA is not needed for cbb3 assembly in this bacterium. In E. coli, involvement of CopA in copper delivery to membrane proteins has not been reported. A basal level of CopA is present in E. coli membranes (14). Nevertheless, no evidence involving CopA in the assembly of the heme-copper oxidase bo or the cupric NADH dehydrogenase-2 in E. coli has been found.
Two crucial questions are raised by these data. (i) Why is CtpA dispensable in copper tolerance in R. gelatinosus, R. capsulatus, H. pylori, and P. aeruginosa? (ii) How is it possible for highly similar (primary and predicted secondary structure) copper ATPases to fulfill two different roles? It is tempting to hypothesize that the expression level of CtpA may be very low regardless of the copper concentration and that excess copper may induce CopA expression but not CtpA. The only available supporting argument is provided by the copper transcriptome of P. aeruginosa. CtpA (PA1549) was not induced by excess copper, whereas CopA (PA3920) and other heavy metal translocating P-type ATPases involved in copper tolerance were induced more than 5-fold (27). Therefore, it will be interesting to constitutively express the ctpA gene in a copA− background of R. gelatinosus to test whether it can restore growth in excess copper.
The other factors to take into account are the copper chaperones or partners of CopA and CtpA. Because copper is not free either in the cytoplasm or in the periplasm but is bound to chaperones, specific interaction between the transporters and their partners may explain the difference in roles of these very similar copper transporters. In copper homeostasis, a specific interaction between CopZ and CopA or between CusF and CusB is required for copper transport. It was shown that CopZ interacts with and delivers copper to CopA (30). Similarly, in the periplasm, copper transfer occurs between CusF and CusB and necessitates a highly specific interaction between these proteins (31). Likewise, we presume that an interaction between CtpA and its cognate partners may be necessary to deliver copper to the targets. This interaction may account for the specificity of CtpA and its involvement in the assembly of cbb3, caa3, and NosZ.
In conclusion, in this paper we demonstrate that CtpA, a putative copper P1B-type transporter that may differ from CopA with respect to the “outlet” of copper, is required for the activity of three membrane cupro-enzymes. Although both proteins are highly similar, CtpA is dispensable for copper tolerance; its role may be restricted to the acquisition and delivery of copper to the copper-requiring enzymes within the membrane and the periplasm.
We acknowledge helpful discussions with A. Durand, A. S. Steunou, and M. Usdin. We are grateful to M. Argentini and D. Cornu (SICaPS Imagif) for help with MS analyses.
*This work was supported by the Agence Nationale de la Recherche (ANR BLAN06-2_147814).
The nucleotide sequence(s) reported in this paper has been submitted to the Gen-BankTM/EBI Data Bank with accession number(s) AY648960 and GQ900543.
2The abbreviations used are: