OUP user menu

Copper acquisition by the SenC protein regulates aerobic respiration in Pseudomonas aeruginosa PAO1

Emanuela Frangipani, Dieter Haas
DOI: http://dx.doi.org/10.1111/j.1574-6968.2009.01726.x 234-240 First published online: 1 September 2009

Abstract

Aerobic respiration of Pseudomonas aeruginosa involves four terminal oxidases belonging to the heme–copper family (that is, three cytochrome c oxidases and one quinol oxidase) plus one copper-independent, cyanide-insensitive quinol oxidase (CIO). The PA0114 gene encoding an SCO1/SenC-type protein, which is known to be important for copper delivery to cytochrome c in yeast, Rhodobacter spp. and Agrobacterium tumefaciens, was found to be important for copper acquisition and aerobic respiration in P. aeruginosa. A PA0114 (senC) mutant grew poorly in low-copper media and had low cytochrome cbb3-type oxidase activity, but expressed CIO at increased levels, by comparison with the wild-type PAO1. Addition of copper reversed these phenotypes, suggesting that periplasmic copper capture by the SenC protein helps P. aeruginosa to adapt to copper deprivation.

Keywords
  • Pseudomonas aeruginosa
  • copper
  • aerobic respiration
  • SCO1
  • TMPD oxidase
  • PA0114

Introduction

Some natural environments in which the metabolically versatile Gram-negative bacterium Pseudomonas aeruginosa thrives contain little bioavailable copper. This trace element is an important cofactor in numerous biological redox reactions, especially in terminal steps of respiration. Factors that can limit Cu availability are pH values >7.4 and the presence of organic ligands having high affinity for Cu. For instance, in alkaline water and soil, Cu is present as poorly soluble carbonates and hydroxides (Karthikeyan et al., 1997) and forms complexes with humic acids and organic matter (Alloway & Tills, 1984). In blood serum, Cu is firmly bound primarily to three proteins: albumin, ceruloplasmin and transcuprein, resulting in an estimated free Cu2+ concentration of about 10−13 M (Linder & Hazegh-Azam, 1996).

Pseudomonas aeruginosa has five terminal oxidases for aerobic respiration (Williams et al., 2007). Four of these enzymes – an aa3 oxidase (cox), two cbb3 oxidases (cco1 and cco2) and a quinol oxidase (cyo) – contain copper in their active sites. The fifth enzyme, a cytochrome bd-type, cyanide-insensitive quinol oxidase (CIO), does not contain copper (Comolli & Donohue, 2002; Cooper et al., 2003). The cbb3-type oxidases and CIO have high affinity for oxygen (Alvarez-Ortega & Harwood, 2007). We have shown previously that CIO is able to support aerobic growth of P. aeruginosa during severe copper deprivation imposed by a copper chelator (Frangipani et al., 2008). Thus, one strategy that enables P. aeruginosa to cope with restricted availability of copper is to use an alternative respiratory pathway that does not require copper. The two major siderophores of P. aeruginosa, pyoverdine and pyochelin, both bind Cu2+ ions (Braud et al., 2009a, b) and, in theory, might promote copper uptake. It is unlikely, however, that such a mechanism is physiologically relevant, as severe copper limitation results in a strong repression of the pyoverdine and pyochelin biosynthetic genes (Frangipani et al., 2008).

To obtain further insight into the response of P. aeruginosa to copper starvation, we mutated the PA0114 gene, which encodes a protein belonging to the SCO1/SenC family. SCO1 (synthesis of cytochrome c oxidase) proteins are known to bind copper and to be involved in the assembly of cytochrome c oxidases, both in the yeast Saccharomyces cerevisiae and in Rhodobacter spp. (Krummeck & Rodel, 1990; Nittis et al., 2001; Swem et al., 2005). Members of this family play a role in redox reactions via the reversible oxidation of an active centre disulfide bond. In S. cerevisiae, SCO1 delivers copper to the dinuclear CuA site in cytochrome c oxidase (Abajian & Rosenzweig, 2006). In Rhodobacter sphaeroides, the SCO1 homologue PrrC appears to transduce an inhibitory signal derived from the cbb3 cytochrome oxidase to the PrrBA two-component system, which is homologous to the RoxSR system of P. aeruginosa, leading to low expression of photosynthetic genes under aerobic conditions (Eraso & Kaplan, 2000). In Rhodobacter capsulatus, PrrC is termed SenC. This protein promotes optimal activity of the cytochrome cbb3 oxidase by functioning as a copper chaperone, i.e. by sequestering and delivering copper to the active site of this oxidase (Swem et al., 2005).

We found that a PA0114 (senC) mutant was severely handicapped in a low-copper medium, presumably as a consequence of malfunctioning of the heme–copper terminal oxidases. The senC mutant, like a quadruple cco1,2 cox cyo mutant, showed upregulated CIO expression in the absence of copper. The SenC phenotypes could be restored to wild-type levels by the addition of copper. These results suggest that SenC participates in copper acquisition, which most probably occurs in the periplasm of P. aeruginosa.

Materials and methods

Bacterial strains and culture conditions

Strains and plasmids used in this study are listed in Table 1. All media and solutions were prepared with deionized, double-distilled water. Bacteria were routinely grown on nutrient agar and in nutrient yeast broth (Stanisich & Holloway, 1972) at 37 °C. When required, antibiotics were added to these media at the following concentrations: 12.5 μg mL−1 tetracycline, 25 μg mL−1 kanamycin for Escherichia coli and 100 μg mL−1 tetracycline for P. aeruginosa. Growth and β-galactosidase experiments were performed in a minimal medium (OS-glucose) containing 0.5% (w/v) glucose, 0.1% (w/v) ammonium sulfate, 0.01% (w/v) Triton X-100 and salt solutions (Ornston & Stanier, 1966) from which CuSO4 was omitted unless stated otherwise. Growth in OS-glucose medium was obtained in 100-mL Erlenmeyer flasks filled with 20 mL of medium, under conditions of good aeration (shaking at 180 r.p.m.) at 37 °C.

View this table:
Table 1

Strains and plasmids used in this study

Relevant characteristicsReferences or sources
Strains
E. coli
DH5αrecA1 endA1 hsdR17 supE44 thi-1 gyrA96 relA1Δ(lacZYA-argF)U169 [φ80dlacZΔM15]F NalrSambrook et al. (1989)
HB101proA2 hsdS20 (rb mb) recA13ara-14 lacY1 galK2 rpsL20 supE44 xyl-5 mtl-1 FSambrook et al. (1989)
P. aeruginosa
PAO1 (ATCC15692)Wild typeHolloway et al. (1979)
PAO6437cioAB mutantFrangipani et al. (2008)
PAO6576PAO1 containing a 627-bp in-frame deletion in the senC (PA0114) geneThis study
PAO6593coxBA-PA0117-coIII mutantFrangipani et al. (2008))
PAO6594roxSR mutantFrangipani et al. (2008)
PAO6596PAO1 containing a deletion in the cyoABCDE locusThis study
PAO6597cyoABCDE coxBA-PA0117-coIII mutantFrangipani et al. (2008)
PAO6650PAO1 containing a cyo cco1 cco2 cox deletionFrangipani et al. (2008)
PAO6651ccoNOQP1 ccoNOQP2 mutantThis study
PAO6660PAO6437 containing a 627-bp in-frame deletion in the senC (PA0114) geneThis study
PAO6721PAO6594 containing a 627-bp in-frame deletion in the senC (PA0114) geneThis study
Plasmids
pRK2013Helper plasmid; Tra+ KmRDitta et al. (1985)
pME3087Suicide vector for allelic replacement; TcR; ColE1 repliconVoisard et al. (1994)
pME7554Plasmid carrying a translational cioA'-'lacZ fusion; TcRFrangipani et al. (2008)
pME9302Suicide construct used for deletion of the coxAB-PA0117-coIII locus; TcRFrangipani et al. (2008)
pME9303Suicide construct used for deletion of the cyoABCDE locus; TcRFrangipani et al. (2008)
pME9304Suicide construct used for deletion of the PA0114 gene; TcRThis study
pME9308Suicide construct used for deletion of the two adjacent ccoNOPQ operons; TcRFrangipani et al. (2008)

Construction of plasmids and gene replacement mutants

DNA cloning and plasmid preparations were performed according to standard methods (Sambrook et al., 1989). Large-scale preparations of plasmid DNA were obtained using JETstar 2.0 (Genomed). For inactivation of the PA0114 (senC) gene in the P. aeruginosa PAO1 chromosome, a 943-bp fragment overlapping the ATG of PA0114 and a 1024-bp fragment overlapping the TGA of PA0114 were amplified by PCR using the primer couples PA0114UPFW: 5′-CCCGGATCCATGGCCACCGTCATCGATC-3′/PA0114UPRV: 5′-CCCGAATTCCATGGGTGGGCAAGGCTC-3′ and PA0114DWFW: 5′-CGCGAATTCGCCTGATTCCCTTTCCTTCC-3′/PA0114DWRV: 5′-GGGAAGCTTCGGAACAGACCAGCACG-3′, respectively. These products were digested with BamHI–EcoRI and EcoRI–HindIII, respectively, and cloned into the corresponding sites (sites underlined in primer sequences) of the suicide vector pME3087, yielding plasmid pME9304. Plasmid pME9304, carried by E. coli DH5α, was then introduced into P. aeruginosa strains PAO1, PAO6437 and PAO6594 by triparental mating, using the helper strain E. coli HB101(pRK2013). Merodiploids were resolved as described previously (Ye et al., 1995). The resulting P. aeruginosa strains PAO6576, PAO6660 and PAO6721 each carried an in-frame PA0114 deletion. To construct strains PAO6596 (cyoABCDE) and PAO6651 (ccoNOQP1 ccoNOQP2) we used the suicide plasmids pME9303 and pME9308, respectively, as described previously (Frangipani et al., 2008). In all mutants used here, the deletions were confirmed by PCR, and PCR fragments were checked by sequencing.

β-Galactosidase assays

β-Galactosidase assays (Miller, 1972) were performed with P. aeruginosa cultures grown in OS-glucose medium, with or without 1.5 μM CuSO4. Data are mean values of three independent samples±SD.

Measurement of cytochrome c oxidase activity

Respiratory c-type cytochromes allow the artificial electron donor N,N,N′,N′-tetramethyl-p-phenylenediamine (TMPD; Fluka), to be oxidized to form a blue indophenol compound (λmax 520 nm). Cytochrome c oxidase activity was measured spectrophotometrically in whole cells by following the increase in the A520 nm at room temperature. Briefly, cells were grown in OS-glucose without added CuSO4 or supplemented with 1.5 μM CuSO4. When cells reached the late exponential growth phase, they were washed twice in 0.9% (w/v) NaCl and adjusted to an OD600≅1. Then 100 μL (c. 107 cells) was added to 1.4 mL of 33 mM potassium phosphate buffer (pH 7.0) in a cuvette. The reaction was started by the addition of 5 μL of 0.54 M TMPD. Blanks did not contain cells or TMPD. Cytochrome c oxidase activity is expressed as μmol TMPD oxidized min−1 per 107 cells, using 6.1 as the millimolar extinction coefficient for TMPD (Matsushita et al., 1982). Data are mean values of triplicates±SD.

Results

PA0114 encodes a homologue of SenC/PrrC in P. aeruginosa

The ORF PA0114 encodes a protein of the SCO1/SenC family belonging to the thioredoxin-like proteins. Considering the results obtained with SCO1/SenC in yeast and R. capsulatus (Nittis et al., 2001; Swem et al., 2005), we hypothesized that PA0114, referred to as SenCP.a. hereafter, may play a critical role in the adaptation of P. aeruginosa to low-copper environments. The SenCP.a. protein shares 33% and 37% amino acid identity with SenC of R. capsulatus and PrrC of R. sphaeroides, respectively (Fig. 1), including the conserved CxxxC domain and the conserved His-173, which form a copper-binding domain (Eraso & Kaplan, 2000; Nittis et al., 2001; Saenkham et al., 2009). SCO1, PrrC, SenC and SenCP.a. also share one transmembrane domain with the N-terminus located in the cytoplasm and the copper-binding domain located in the periplasm (Fig. 1) (Eraso & Kaplan, 2000). Contrary to the situation in Rhodobacter spp., where the prrC (senC) gene is part of a prrBCA cluster (Eraso & Kaplan, 2000), the senC. gene of P. aeruginosa PAO1 is located near the gene cluster encoding the COX terminal oxidase (PA0105–PA0108) and adjacent to a gene, PA0113, which is 59% similar to a cytochrome c oxidase folding protein annotated in Synechocystis sp.

Figure 1

Amino acid sequence alignment of PA0114 (SenCP.a.) with its homologues PrrC from Rhodobacter sphaeroides, SenC from Rhodobacter capsulatus and SCO1 from Saccharomyces cerevisiae. The conserved copper-binding domains are highlighted by dashed boxes, whereas the predicted transmembrane domains (obtained using tmpred; http://www.ch.embnet.org) are shown in boxes with solid lines. ⋆, identity;:, similarity.

Mutational loss of SenCP.a. impairs growth in a low-copper environment

To test whether the SenCP.a. protein might be involved in the assembly of terminal oxidases in P. aeruginosa, we constructed a senC in-frame deletion mutant (PAO6576). We reasoned that such a mutant should be impaired in its growth in low-copper media, similar to the behaviour of a cyo cco1 cco2 cox quadruple mutant (PAO6650). In a copper-replete environment (1.5 μM CuSO4), strain PAO6576 should grow like the wild type, whereas the growth rate of PAO6650 should remain poor. Results (Fig. 2) confirm this expectation. In OS-glucose medium without added copper, strain PAO6576 (open circles) showed a twofold decrease in growth yield after 8 h of incubation, compared with the wild-type strain PAO1 (open squares). The addition of copper to the medium restored the growth of PAO6576 (filled circles) to the rate of the wild-type PAO1 (filled squares). The same copper treatment did not have any effect on PAO6650 (filled and open triangles), whose growth resembled that of the senC mutant grown in low copper. This residual growth is essentially provided by CIO (Frangipani et al., 2008).

Figure 2

Growth of the wild-type PAO1 (squares), the quadruple cyo cco1 cco2 cox mutant PAO6650 (triangles) and the senC mutant PAO6576 (circles). Cells were grown in OS-glucose medium and growth was measured by turbidimetry. Cultures were either not amended with Cu (open symbols) or contained 1.5 μM CuSO4 (filled symbols). Each value is the average of three different cultures±SD.

To confirm that SenCP.a. plays a role in the function of copper-containing terminal oxidases, we constructed the senC cioAB and senC roxSR double mutants PAO6660 and PAO6721, respectively. Our hypothesis was that, in a low copper environment, the effect of a senC mutation would be stronger in the double mutants than in a wild-type background. Aerobic respiration would be severely compromised in the senC cioAB double mutant because all known terminal oxidases would be nonfunctional. The senC roxSR double mutant was expected to have a similar phenotype as the RoxSR two-component system is the major positive regulator of CIO (Comolli & Donohue, 2002). Both strains PAO6660 (senC cioAB; triangles in Fig. 3a) and PAO6721 (senC roxSR; circles in Fig. 3b) showed a dramatic reduction of growth during copper shortage (open symbols), compared with their parental strains PAO6437 (cioAB; squares in Fig. 3a) and PAO6594 (roxSR; diamonds in Fig. 3b). Addition of 1.5 μM CuSO4 (filled symbols in Fig. 3) fully restored the growth of the double mutants, indicating that the senC mutation affects the function of the copper-containing oxidases.

Figure 3

Effect of a senC mutation on different strains during copper shortage. (a) Growth of the cioAB mutant PAO6437 (squares) and the cioAB senC mutant PAO6660 (triangles). (b) Growth of the roxSR mutant PAO6594 (diamonds) and the roxSR senC mutant PAO6721 (circles). Cells were grown in OS-glucose medium and growth was measured by turbidimetry. Cultures were not amended with Cu (open symbols) or contained 1.5 μM CuSO4 (filled symbols). Each value is the average of three different cultures±SD.

TMPD oxidase activity is reduced in a senC mutant

The TMPD oxidase test is generally used to measure cytochrome c oxidase activity (Comolli & Donohue, 2004). To investigate which of the four heme–copper terminal oxidases were affected by the senC mutation in P. aeruginosa, we first assessed which oxidase activity could be detected by the TMPD test under the growth conditions used. To this end, we measured TMPD oxidase activity in the wild type (PAO1), a cox mutant (PAO6593), a newly constructed mutant carrying a deletion in the cyoABCDE operon (PAO6596), a cox cyo double mutant (PAO6597), a newly constructed cco1 cco2 double mutant lacking both cbb3-type oxidases (PAO6651) and a quadruple cyo cco1 cco2 cox mutant (PAO6650). The cox and cyo mutations did not affect TMPD oxidase activity, whereas a deletion inactivating the two cbb3-type terminal oxidases resulted in an almost complete loss of activity, similar to the effect seen for the quadruple mutant (Fig. 4a). The fact that the cox mutation had no detectable effect is in agreement with our finding that a coxB'-'lacZ translational fusion is very poorly expressed under the growth conditions used (data not shown). The strong impact of the cco1,2 double mutation suggests that oxidation of cytochrome c in the TMPD assay is mainly carried out by the two cbb3-type terminal oxidases, although at this stage we cannot assess the individual contribution of each of the two enzymes. The observation that both cbb3-type terminal oxidases essentially account for the TMPD oxidase activity measured is consistent with previously published data (Comolli & Donohue, 2004).

Figure 4

(a) TMPD oxidase activity of whole cells of Pseudomonas aeruginosa PAO1 (wt), a quadruple cyo cco1 cco2 cox mutant (PAO6650), a cox mutant (PAO6593), a cyo mutant (PAO6596), a cox cyo double mutant (PAO6597) and a cco1 cco2 double mutant (PAO6651). (b) Oxidase activity of whole cells of strain PAO1 (wt), the senC mutant PAO6576 and the quadruple cyo cco1 cco2 cox mutant PAO6650. Cells were grown aerobically in OS-glucose medium without added copper or in the presence of 1.5 μM CuSO4. Activity is expressed as μmol TMPD oxidized min−1 per 107 cells at pH 7.0 and 25°C. Each value is the average of three different cultures±SD.

To confirm the role of SenCP.a. in the function of terminal oxidases in P. aeruginosa, we measured oxidase activity of the wild-type PAO1, the senC mutant PAO6576 and the cyo cco1 cco2 cox quadruple mutant PAO6650, grown aerobically in OS-glucose medium without added copper or supplemented with 1.5 μM CuSO4 (Fig. 4b). In the absence of added copper, the oxidase activities of both PAO6576 and PAO6650 were very low, by comparison with the wild-type activity. The addition of 1.5 μM CuSO4 resulted in a twofold increase of the oxidase activity in the wild type and functionally complemented the SenC phenotype of strain PAO6576 to the wild-type level (Fig. 4b). As a control, the addition of copper did not affect the oxidase activity of PAO6650 (Fig. 4b). Similar results were also obtained with 20 μM CuSO4 (data not shown). These data confirm that SenCP.a. plays a role in copper acquisition in P. aeruginosa and that copper is delivered to the cbb3-type terminal oxidases and probably to COX as well, although the very low activity of the latter enzyme precludes an assessment.

SenCP.a. regulates CIO expression

We have previously shown that in P. aeruginosa, copper represses the expression of the cioAB genes in the wild type, but much less so in the quadruple cyo cco1 cco2 cox mutant (Frangipani et al., 2008). We therefore hypothesized that a senC mutant would show a similar phenotype under low-copper conditions. To test this hypothesis, we monitored the expression of a cioA'-'lacZ translational fusion, carried by plasmid pME7554, in strains PAO6576 (senC) and PAO1 during copper deprivation (Fig. 5). Under these conditions, cioA'-'lacZ expression in PAO6576 (open triangles) was higher than in the wild type (open diamonds), and this difference was most pronounced at low cell population densities (Fig. 5). When 1.5 μM CuSO4 was added to the growth medium, cioA'-'lacZ expression was repressed to low levels in both the wild type (filled diamonds) and the senC mutant (filled triangles), probably as a consequence of the restored function of the cytochrome c oxidases (Fig. 5). As a control, cioA'-'lacZ expression was measured in the roxSR mutant PAO6594 (squares) cultured under the same growth conditions; the expression was low, with and without copper added, as reported previously (Comolli & Donohue, 2002; Frangipani et al., 2008). Taken together, these results show that SenCP.a. has an important impact on the regulation of aerobic respiration in PAO1, in response to copper availability.

Figure 5

Expression of a β-galactosidase reporter plasmid containing a cioA'-'lacZ translational fusion (pME7554) in the wild-type strain PAO1 (diamonds), in a senC mutant (PAO6576; triangles) and in a roxSR mutant (PAO6594; squares). Cultures were grown in OS-glucose medium and were not amended with copper (open symbols) or contained 1.5 μM CuSO4 (filled symbols). Each value is the average of three different cultures±SD.

Discussion

The results obtained here and in a previous study (Frangipani et al., 2008) indicate that P. aeruginosa has at least two different strategies to cope with low copper availability. In the presence of the specific copper chelator bathocuproine disulfonic acid (BCS), a cioAB mutant has a strong growth handicap during severe copper deprivation under aerobic conditions (Frangipani et al., 2008), indicating that CIO allows P. aeruginosa to respire virtually without copper. This can be considered as a bypass strategy. Here, we used media containing a low copper concentration, but without addition of BCS; the Cu concentration was estimated at around 10−8 M by inductively coupled plasma MS (data not shown) and is due to contamination of the chemicals used. Under such conditions, we have found that SenCP.a. insures the delivery of copper to heme–copper terminal oxidases. A senC deletion mutant – similar to a cyo cco1 cco2 cox quadruple mutant – showed delayed growth during copper limitation, compared with the wild type (Fig. 2). This growth defect could be restored to wild-type levels by the addition of 1.5 μM CuSO4, whereas the quadruple mutant showed the same delayed growth phenotype irrespective of the copper content of the growth medium (Fig. 2). Thus, SenCP.a. insures a second strategy consisting of copper capture from low-copper media.

In the present work, we also show that the senC cioAB and senC roxSR double mutants, when grown in low-copper media, have the same handicap as that displayed by a single cioAB mutant in the presence of BCS. Thus, it appears that the senC mutation starves P. aeruginosa for copper to the same extent as does a copper chelator. SenCP.a. acts as a copper chaperone delivering copper to the cbb3-type terminal oxidases (Fig. 4). Whether SenC has the same function in COX and CYO assembly remains to be studied.

The RoxSR two-component system positively regulates CIO expression by transducing and integrating an unknown signal emanating mainly from the cbb3-1 terminal oxidase (Comolli & Donohue, 2002, 2004). This regulation appears to be particularly important during cyanogenesis under low-oxygen conditions. When endogenously produced cyanide inhibits the heme–copper oxidases, the cioAB genes are induced, mainly via the RoxSR system (Comolli & Donohue, 2002; Williams et al., 2007). Moreover, the anaerobic regulator ANR positively regulates the expression of the cbb3-2 terminal oxidase, but not that of the cbb3-1 oxidase in poorly aerated cultures (Comolli & Donohue, 2004). ANR also weakly represses the expression of the cioAB genes (Cunningham et al., 1997). Our data are compatible with a model according to which P. aeruginosa responds to copper limitation under aerobic conditions by upregulating CIO expression. This regulation indirectly involves the copper chaperone SenCP.a..

A recent study (Saenkham et al., 2009), which was published during the preparation of this manuscript, shows that the SenC homologue of Agrobacterium tumefaciens delivers Cu ions to the terminal oxidases reacting with TMPD (likely of the aa3- and cbb3-types) and that mutations in the conserved Cys71, Cys75 and His163 residues in this Cu chaperone abolish its function. The corresponding Cys84, Cys88 and His173 residues of SenCP.a. (Fig. 1) are therefore likely to form the Cu-binding site in SenCP.a. as well.

Acknowledgements

We thank Huw Williams for advice regarding the TMPD test and Cornelia Reimmann for critically reading this manuscript. This work was supported in part by the SNF project 3100A0-120365.

Footnotes

  • Editor: Dieter Jahn

References

View Abstract