Available online at www.sciencedirect.com
Biochimica et Biophysica Acta 1780 (2008) 226 – 232 www.elsevier.com/locate/bbagen
Periplasmic Cu,Zn superoxide dismutase and cytoplasmic Dps concur in protecting Salmonella enterica serovar Typhimurium from extracellular reactive oxygen species Francesca Pacello a , Pierpaolo Ceci b , Serena Ammendola a , Paolo Pasquali c , Emilia Chiancone b , Andrea Battistoni a,⁎ b
a Dipartimento di Biologia, Universitá di Roma Tor Vergata, Via della Ricerca Scientifica, 00133 Roma, Italy Istituto di Biologia e Patologia Molecolari CNR, Dipartimento di Scienze Biochimiche, Università La Sapienza, P.le A. Moro, 5, 00185 Roma, Italy c Dipartimento di Sanità Alimentare e Animale, Istituto Superiore di Sanità, 00161 Rome, Italy
Received 7 August 2007; received in revised form 20 November 2007; accepted 4 December 2007 Available online 14 December 2007
Abstract Several bacteria possess periplasmic Cu,Zn superoxide dismutases which can confer protection from extracellular reactive oxygen species. Thus, deletion of the sodC1 gene reduces Salmonella enterica serovar Typhimurium ability to colonize the spleens of wild type mice, but enhances virulence in p47phox mutant mice. To look into the role of periplamic Cu,Zn superoxide dismutase and into possible additive effects of the ferritinlike Dps protein involved in hydrogen peroxide detoxification, we have analyzed bacterial survival in response to extracellular sources of superoxide and/or hydrogen peroxide. Exposure to extracellular superoxide of Salmonella Typhimurium mutant strains lacking the sodC1 and sodC2 genes and/ or the dps gene does not cause direct killing of bacteria, indicating that extracellular superoxide is poorly bactericidal. In contrast, all mutant strains display a sharp hydrogen peroxide-dependent loss of viability, the dps,sodC1,sodC2 mutant being less resistant than the dps or the sodC1,sodC2 mutants. These findings suggest that the role of Cu,Zn superoxide dismutase in bacteria is to remove rapidly superoxide from the periplasm to prevent its reaction with other reactive molecules. Moreover, the nearly additive effect of the sodC and dps mutations suggests that localization of antioxidant enzymes in different cellular compartments is required for bacterial resistance to extracytoplasmic oxidative attack. © 2007 Elsevier B.V. All rights reserved. Keywords: Cu,Zn superoxide dismutase; Dps, sodC, Salmonella enterica; Oxidative damage; NADPH oxidase
1. Introduction Several Gram-negative bacteria express periplasmic or membrane-associated Cu,Zn superoxide dismutases (Cu,ZnSODs) [1]. Although several investigations have focused on the physiological role of this antioxidant enzyme, the function of Cu,ZnSOD is not completely clarified. Cu,ZnSOD does not protect bacteria from oxygen radicals produced intracellularly as a by-product of aerobic metabolism [2,3], and is assumed to protect microorganisms against superoxide produced outside the cell [1,4] or within the
⁎ Corresponding author. Tel.: +39 0672594372; fax: +39 0672594311. E-mail address:
[email protected] (A. Battistoni). 0304-4165/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.bbagen.2007.12.001
periplasmic space [5,6]. A role in protection against exogenous superoxide could be particularly important for bacterial pathogens which are exposed to high fluxes of superoxide generated by the NADPH oxidase complex of professional phagocytes. It is supported by different investigations which have shown that inactivation of the Cu,ZnSOD gene, sodC, reduces resistance to reactive oxygen species in vitro, decreases bacterial survival within professional and nonprofessional phagocytes and results in attenuation of virulence in several animal models [3,7–19]. However, in a few cases doubts have been raised about the relevance of specific sodC genes for bacterial survival within the infected host [12,15,17]. Cu,ZnSOD was suggested to be unrelated to pathogenesis also because sodC genes are present in a number of environmental or non-pathogenic bacteria [2,6,12]. In principle, the
F. Pacello et al. / Biochimica et Biophysica Acta 1780 (2008) 226–232
presence of Cu,ZnSOD in non-pathogenic bacteria could be rationalized by the recent finding that significant amounts of superoxide are formed within the periplasm of Escherichia coli cells grown under aerobic conditions [6]. However, although significant superoxide production has been detected in stationaryphased cells starved for phosphate, maximal superoxide production in E. coli was observed in bacterial cultures actively growing on carbon sources [6]. This observation points out that there may be a separation in time between Cu,ZnSOD expression and superoxide release in the periplasmic space, because in E. coli Cu,ZnSOD is expressed exclusively in the stationary phase [5,20]. To add to the complexity, the data regarding the bactericidal activity of extracellular superoxide are contradictory [2,5,7,11–13]. Thus, while there is general agreement on the protective role of Cu, ZnSOD against killing due to reactive oxygen species generated outside the cell, some investigations have suggested that the bactericidal activity of extracellular/periplasmic superoxide is negligible [2,11,20]. The latter observations are consistent with the cytoplasmic localization of all the superoxide-sensitive biomolecules (mainly proteins containing iron–sulphur clusters) characterized so far [21] and with the inability of the superoxide anion to cross lipid membranes [22]. Moreover, although superoxide has been proposed to penetrate within the cell and attack cytosolic proteins under acidic conditions, such as those encountered in the macrophage phagolysosome [23], this possibility is still not proven and does not explain the role of Cu,ZnSOD in protecting bacteria from extracellular reactive oxygen species generated in vitro at neutral pH. To re-examine the role of Cu,ZnSOD in bacterial resistance to reactive oxygen species, we have chosen Salmonella enterica serovar Typhimurium (hereafter referred to as S. Typhimurium) as a model system. This well studied bacterium contains two distinct sodC genes which are highly expressed in stationary phased cells. One of these genes, sodC2, is the orthologue of the sodC gene of E. coli, and has a role which is likely largely unrelated to pathogenesis [14,17]. In contrast, the second gene, sodC1, is located on a lambdoid prophage [10] and greatly contributes to Salmonella virulence [7,8,13–15,17–19]. The role of Cu,ZnSOD has been analyzed in conjunction with that of the ferritin-like protein Dps (DNA-binding proteins from starved cells), a cytoplasmic protein expressed abundantly by most bacteria under a variety of stress conditions, which protects DNA against hydrogen peroxide-mediated damage through its
227
ferroxidase activity [8,24]. A recent study has shown that Dps protects Salmonella from iron-dependent killing by hydrogen peroxide, promotes Salmonella survival in murine macrophages, and enhances Salmonella virulence [25]. The results reported here indicate that periplasmic Cu,ZnSOD and cytoplasmic Dps concur in protecting bacteria from hydrogen peroxide damage. 2. Materials and methods 2.1. Growth conditions S. Typhimurium ATCC 14028 was cultured in LB broth (1% bacto tryptone w/v, 0.5% yeast extract w/v, 1% NaCl w/v) solidified by the addition of 1.75% (w/v) agar when required. Antibiotics were used at the following concentrations: chloramphenicol 30 mg/l, ampicillin 100 mg/l, kanamycin 30 mg/l.
2.2. Construction of deletion mutants All the strains used in this work are listed in Table 1. A mutation dps::cam was constructed in S. Typhimurium ATCC 14028 using the method of Datsenko and Wanner [26]. The coding sequence of this gene was replaced with the chloramphenicol resistance cassette amplified with the primers oPF1 (5′-gggacacaaacatcaagaggatatgagattatgagtaccgtgtaggctggagctgcttcg-3′) and oPF2 (5′gataaaccacaggaatttatcgaggtcgcgtgatgcggcatatgaatatcctccttagtt-3′) using pKD3 plasmid [26] as a template for PCR amplification. The construct was verified by PCR analysis. The dps::cam mutation was transferred into a clean 14028 background by P22 transduction. The new strain (dps::cam) was named PF103. The absence of the Dps protein in this mutant strain was confirmed by Western blot analysis using a rat Dps antiserum as primary antibody and a horseradish conjugate anti-rat antibody for the chemiluminescent reaction. A sodC1 deletion mutant was generated using the same oligonucleotides and the same general procedure, recently described to construct a S. Enteritidis sodC1::kan mutant [19], thus generating the strain PF099. A sodC1,sodC2 double mutant was generated by transduction of a sodC2::cam allele from strain MA7472 (provided by Lionello Bossi, Gif-sur-Yvette) into PF099, obtaining strain PF102a. To achieve the dps,sodC1,sodC2 triple mutant the two resistance cassettes were eliminated from strain PF102a by homologous recombination between the two flanking FRT sites using plasmid pCP20 [26,27], thereby obtaining strain PF111 (sodC1::scar, sodC2:: scar). Subsequently, the dps::cam mutation was introduced in the PF111 strain by P22 transduction, to obtain strain PF112 (dps::cam, sodC1::scar, sodC2::scar).
2.3. Epitope tagging of the dps gene and immunodetection The expression of the sodCI, sodCII and dps genes was studied using 3xFLAG epitope-tagged S. Typhimurium strains. The strains MA7224 and MA7225 containing epitope-tagged versions of sodC1 and sodC2, respectively, were previously described [14]. Dps epitope tagging was achieved by adding a 3xFLAG at sequence the 3′-terminus of the gene, following a described procedure [28] Briefly, a DNA fragment was amplified with primers oPF3
Table 1 Bacterial strains Name
Relevant genotype
Source
S. Typhimurium ATCC14028 PF103 PF099 MA7472 PF102a PF111 PF112 MA7223 MA7224 MA7225 PF108
dps::cam sodC1::kan sodC2::cam sodC1::kan, sodC2::cam sodC1::scar, sodC2::scar dps::cam, sodC1::scar, sodC2::scar ilvI3305::Tn10dTac-cat-43::3xFLAG-kan sodCI::3xFLAG-kan ilvI3305::Tn10dTac-cat-43::3xFLAG-kan sodCII::3xFLAG-kan ilvI3305::Tn10dTac-cat-43::3xFLAG-kan dps::3xFLAG-kan ilvI3305::Tn10dTac-cat-43::3xFLAG-kan
Lionello Bossi this work this work Lionello Bossi this work this work this work [14] [14] [14] this work
228
F. Pacello et al. / Biochimica et Biophysica Acta 1780 (2008) 226–232
(5′-ctcgataaattcctgtggtttatcgaatccaacatcgaagactacaaagaccatgacgg-3) and oPF5 (5′-caagacgtgtgcactatttaagtgcaaaacccctgtgccatatgaatatcctccttag-3′) on pSUB11. DNA was electroporated in S. Typhimurium 14028 containing plasmid pKD46 [26]. Transformants were selected and the recombination was confirmed by PCR. Subsequently, the dps::3xFLAG mutation was transferred into a clean 14028 background by P22 transduction. To have an internal standard for western blot analysis, a 3xFLAG epitope-tagged chloramphenicol acetyl transferase gene (cat) constitutively expressed under control of the tac promoter, was transduced from strain MA7223 (12), obtaining the doubly tagged strain PF108. Strains carrying the epitope-tagged genes were grown at 37 °C in 2 ml of LB medium for 20 h. 108 cells were removed immediately for the analysis of gene expression in the stationary phase by Western blotting. Aliquots of bacteria were washed in Phosphate Buffered Saline (PBS), resuspended at a concentration of 108 cells/ml in the same buffer and then treated with 250 μM H2O2 or with xanthine/xanthine oxidase. After a 1 h incubation at 37 °C the reactions were stopped with 1 U/μl catalase and 108 cells were removed and harvested by centrifugation. The pellets were boiled for 10 min in Laemmli lysis buffer and then proteins were separated by 12% SDS-PAGE and blotted onto nitrocellulose membrane (Hybond ECL, Amersham). The epitope flagged proteins were revealed by the use of anti-FLAG M2 monoclonal antibodies (Sigma-Aldrich) as primary antibody, and anti-mouse HRP-conjugated IgG (Bio-Rad) as secondary antibody. Detection was performed by enhanced chemiluminescence (ECL Advance, Amersham).
2.4. Mouse infection studies The relative ability of wild type S. Typhimurium and of the sodC1 mutant to proliferate in mice was evaluated by competition assays, carried out essentially as described by Uzzau et al. [14]. Bacteria were grown to stationary phase in LB, washed in PBS and then diluted to the required concentration. Equal number of bacteria from each strain were mixed and plated on LB agar plates; thereafter 200 colonies were picked individually and transferred onto appropriate selective media to determine the total number and percentage of bacteria from each inoculated strain. These mixtures were used to infect female wild type C57BL/ 6N or p47phox [29] isogenic mutant mice of 8 weeks (obtained from Taconic Europe). Animals were kept in a sterile environment to prevent infections. Groups of 5 animals were inoculated intraperitoneally with 0.2 ml of mixtures containing 1000 or 4000 c.f.u./mouse, respectively. When p47phox mutants showed clear symptoms of systemic disease (3–4 days after infection) animals were sacrificed and the infected spleens were removed and homogenized in PBS. Bacteria were enumerated in each infected organ by serial dilution and plating on LB agar plates. The number of bacteria recovered per spleen was 105– 106 in wild type mice, but was above 108 in p47phox mutants. Colonies were then transferred to selective plates to determine the percentage of each strain. The competitive index (CI) was calculated as (percentage of strain A recovered/ percentage of strain B recovered)/(percentage of strain A inoculated/percentage of strain B inoculated). The CI of each set of assays was analyzed statistically by using Student's t test.
2.5. Susceptibility to exogenous superoxide in vitro The superoxide anion was generated in vitro by the oxidation of xanthine catalyzed by xanthine oxidase. Overnight bacterial cultures grown on LB medium were washed in Phosphate Buffered Saline (PBS) and diluted to 106 cells/ml in PBS containing 250 μM xanthine. Aliquots of this solution were incubated at 37 °C with 0.1 U/ml xanthine oxidase (Sigma-Aldrich) and the number of viable cells after one or two hours of incubation was determined by serial dilutions plated on LB agar. To determine whether bacterial killing was due to superoxide or hydrogen peroxide, 1 U/μl Cu,ZnSOD from bovine erythrocytes (Sigma-Aldrich) or 1 U/μl catalase (Roche Diagnostics) was added to the xanthine/xanthine oxidase reaction. The percent survival following superoxide challenge was calculated for each strain by dividing the number of colony forming units (CFU) obtained upon incubation in xanthine alone by the number of CFU obtained upon incubation in xanthine oxidase. Each assay was repeated at least three times, and standard deviations were calculated. The susceptibility to exogenous superoxide was also tested by the use of KO2 (Sigma-Aldrich). The assay was carried out using overnight bacterial cultures grown in LB medium, washed in PBS and diluted to 106 cells/ml in
PBS. A 42 mM potassium superoxide solution was quickly prepared (less than 30 s) in ice-cold 50 mM NaOH, 0.5 mM diethylenetriamine pentaacetic acid, in order to reduce the spontaneous disproportionation of the superoxide anion [30], and an aliquot was immediately transferred to the bacterial suspension (0.5–1 mM KO2 final concentration). A control solution was prepared by adding 50 mM NaOH, 0.5 mM diethylenetriamine pentaacetic acid to bacterial cells. Bacteria were incubated for 1 h at 37 °C and the number of viable cells was determined by plating serial dilutions on LB agar plates. To determine whether bacterial killing was directly due to superoxide 1 U/μl catalase was added to the bacterial suspension before the addition of KO2. Bacterial survival following the challenge with superoxide anion was calculated for each strain by dividing the number of CFU obtained with the control solution by the number of CFU obtained upon incubation with KO2. Each assay was repeated at least three times, and standard deviations were calculated.
2.6. Susceptibility to hydrogen peroxide Overnight bacterial cultures grown on LB medium were washed in PBS, diluted to 106 cells/ml in PBS and incubated at 37 °C with 250 μM hydrogen peroxide. Aliquots were removed after 1 h and 2 h and the number of viable cells was determined by serial dilution and plating onto LB agar. The percent survival following hydrogen peroxide exposure was calculated for each strain by dividing the number of CFU obtained upon incubation in PBS alone by the number of CFU obtained upon incubation in hydrogen peroxide. Each assay was repeated at least three times, and standard deviations were calculated.
3. Results and discussion 3.1. Periplasmic Cu,ZnSOD protects bacteria from the phagocytic oxidative burst The role of Cu,ZnSOD in protecting bacteria from the reactive oxygen species produced by phagocytic cells is still controversial. The sodC2 gene is strictly conserved in Salmonella and several related enterobacteria. In contrast, the sodC1 gene is not present in all Salmonella strains or serovars [9]. Moreover, sodC1 is highly expressed in infected animals and, unlike sodC2, significantly contributes to Salmonella ability to replicate within the host, thus suggesting that it protects bacteria from the phagocytic oxidative burst [7,14,17,19]. However, other studies challenged this interpretation and suggested that production of large amounts of reactive oxygen species by neutrophiles rather than causing direct oxygen radical damage to microorganisms, is required to provide optimal conditions for bacterial killing by proteases [31,32]. To understand the role of Cu,ZnSOD in bacteria we have analyzed the ability of wild type and sodC1 mutant strains of S. Typhimurium to colonize the spleens of wild type and p47phox mice, which are unable to produce superoxide by the phagocytic NADPH oxidase [29]. Table 2 shows the results of competition experiments. In wild type C57BL/6N mice the sodC1 Salmonella mutant was significantly disadvantaged with respect to the wild Table 2 Competition assays in wild type and p47phox mice Mouse
Salmonella Salmonella cfu Median Number P strain A strain B inoculated CI of mice
C57BL/6N p47phox C57BL/6N p47phox
wild type wild type wild type wild type
PF099 PF099 PF099 PF099
1000 1000 4000 4000
2.425 0.42 3.022 0.383
4 5 5 5
0.0102 0.0025 0.0302 0.0001
F. Pacello et al. / Biochimica et Biophysica Acta 1780 (2008) 226–232
Fig. 1. Immunodetection of 3xFLAG-tagged proteins in S. Typhimurium ATCC14028. Strains PF108 (dps::3xFLAG cat::3xFLAG), MA7224 (sodCI::3xFLAG cat::3xFLAG) and MA7225 (sodCII::3xFLAG cat::3xFLAG) were harvested from stationary phase in LB cultures (lanes 1, 4 and 7), after 1 h of incubation in PBS (lanes 2, 5 and 8) and after 1 h of incubation in PBS and 250 μM hydrogen peroxide (lanes 3, 6 and 9), as described in materials and methods. The enzymes encoded by sodC1 and sodC2 are indicated as SodC1 and SodC2.
type ATCC14028 S. Typhimurium strain (P = 0.01 and 0.03 in the two independent experiments). Quite unexpectedly, we observed a statistically significant ability of the sodC1 mutant to outgrow the wild type strain in p47phox mice. These findings indicate that the role of the Cu,ZnSOD encoded by sodC1 is to protect Salmonella from the reactive oxygen species produced by the phagocytic NADPH oxidase. In fact, in mice lacking a functional NADPH oxidase production of the sodC1 encoded Cu,ZnSOD appears an unuseful burden which creates a disadvantage for infecting bacteria. Although our findings do not contradict the proposed role of reactive oxygen species in the activation of proteases involved in killing engulfed bacteria, they suggest that such reactive species directly cause oxidative damage to bacteria, in agreement with the classical view of the role of oxidative burst [33]. In other words, periplasmic Cu,ZnSOD appears as an important component of the antioxidant shield which protects bacteria from the phagocytic oxidative burst. 3.2. Susceptibility of S. Typhimurium to superoxide The superoxide anion produced by NADPH oxidase during the oxidative burst is rapidly converted into other more reactive oxygen species, including hydrogen peroxide. It is not yet clear whether periplasmic Cu,ZnSOD is required to prevent cellular damage caused directly by the superoxide anion or to scavenge superoxide rapidly in order to prevent its involvement in chemical reaction leading to formation of other, more toxic reactive oxygen species. Most investigations concerning the involvement of periplasmic Cu,ZnSOD in bacterial cell protection against extracellular superoxide have been carried out analyzing the effect on bacterial viability of superoxide generated in the reaction catalyzed by xanthine oxidase with xanthine as a substrate. In this assay, however, the substrate of the reaction is exhausted after a few minutes and the short-living superoxide anion is rapidly converted into the more stable hydrogen peroxide [20]. Interpretation of the results, therefore, requires a careful analysis of the oxygen species involved in cellular damage. This can be attempted by carrying out the assay in the presence of enzymes able to remove selectively superoxide (superoxide dismutase) or hydrogen peroxide (catalase). It should be noted that the assays performed in the absence of catalase invariably showed a decrease in viability of bacteria lacking sodC genes [7,13], while
229
this effect was not observed when catalase was added to the reaction mixture [2,11,20]. To better understand the role of Cu,ZnSOD in bacteria, we have analyzed the bactericidal activity of the reactive oxygen species generated by such assays on wild type and mutant Salmonellae lacking the two sodC genes and/or the dps gene. In all cases bacteria grown to the stationary phase were used, as sodC1, sodC2 and dps are maximally expressed in starved bacteria [9,14,24]. Moreover, while in E. coli the expression of dps in logarithmic phase is modulated by the exposition to sources of hydrogen peroxide (mediated by OxyR), in the stationary phase its expression is regulated by σs and is not induced by this reactive oxygen species [34]. Accordingly, Fig. 1 shows that also in S. Typhimurium the two Cu,ZnSODs and Dps are actively expressed under the conditions used in our tests and their intracellular accumulation is not significantly modified by incubation in PBS or with hydrogen peroxide. The marginal increase in accumulation of the sodC1 encoded enzyme is likely due to the lack of magnesium in PBS [35]. Similar results were obtained by incubating bacteria in PBS with xanthine/xanthine oxidase (data not shown), in line with previous observations showing that sodC1 and sodC2 are not regulated by reactive oxygen species [20]. Survival assays were carried out in the presence or absence of catalase or bovine Cu,ZnSOD in the mixture, to discriminate between the effect of superoxide or hydrogen peroxide on bacterial viability. When the assays were carried out in the presence of catalase, neither the wild type strain nor any of the mutant strains exhibited detectable loss of viability (data not shown). In contrast, when catalase was omitted, all the mutant strains were killed at significantly higher rates with respect to the wild type strain (Fig. 2). The killing rates were not affected to appreciable levels by the addition of bovine superoxide dismutase to the reaction mixture (data not shown). This result clearly indicates that the chemical agent responsible of bacterial death is hydrogen peroxide and not superoxide.
Fig. 2. Sensitivity of wild type and mutant strains of S. Typhimurium ATCC 14028 to reactive oxygen species produced in vitro by xanthine oxidase and xanthine. Results are expressed as percent survival after 1-h and 2-h incubation with xanthine and xanthine oxidase at 37 °C. The percent survival was calculated for each strain by dividing the number of CFU/ml obtained from incubation in xanthine alone by the number of CFU/ml obtained from incubation in the presence of xanthine oxidase. The results are the average of three independent experiments, each one performed in quadruplicate, and represent mean ± standard errors.
230
F. Pacello et al. / Biochimica et Biophysica Acta 1780 (2008) 226–232
Fig. 3. Susceptibility of wild type and mutant strains of S. Typhimurium ATCC 14028 to superoxide anion obtained in vitro from potassium superoxide. Results are expressed as percent survival after 1-h incubation with potassium superoxide at 37 °C. The percent survival was calculated for each strain by dividing the number of CFU/ml obtained from incubation in control solution by the number of CFU/ml obtained from incubation in the presence of potassium superoxide. The assay was also carried out in presence of 1 U/μl catalase. In this case the survival percentage was calculated for each strain by dividing the number of CFU/ml obtained from control sample by the number of CFU/ml obtained from incubation in potassium superoxide and catalase. The results are the average of three independent experiments, each one performed in quadruplicate, and represent mean ± standard errors.
Mutants lacking dps were more susceptible to hydrogen peroxide than the sodC1,sodC2 double mutant, confirming the pivotal role of Dps in bacterial resistance to this chemical agent. Interestingly, maximal susceptibility was displayed by the PF112 strain, which lacks the two sodC genes and the dps gene. Considering that Cu,ZnSODs and Dps reside in different cellular compartments, this result indicates that both periplasmic and cytoplasmic enzymes contribute to resistance to hydrogen peroxide stress in bacteria exposed to extracellular reactive oxygen species. To further verify the role of Cu,ZnSOD in resistance to superoxide and hydrogen peroxide, bacteria have been subjected to the treatment with other superoxide generators. Some authors have analyzed resistance of sodC mutant to the superoxide generator pyrogallol [36,37], although the mechanism of cell death induced by this chemical agent has not been investigated. We not only confirm that sodC mutants are more susceptible than the wild type to pyrogallol, but in addition have observed that a one hours exposure to 2 mM pyrogallol causes a significant loss in survival also of wild type bacteria incubated with catalase and superoxide dismutase (data not shown). This observation suggests that pyrogallol toxicity is not simply due to production of reactive oxygen species outside the bacterial cell and that superoxide-resistance assays based on the use of this compound should be considered with caution. We have therefore exposed wild type and mutant Salmonella strains to various amounts of potassium superoxide, ranging from 0.5 to 1 mM. Higher KO2 concentrations were not used because they caused pH changes in the solution. Although the KO2 solution was rapidly prepared under alkaline conditions to reduce the rate of spontaneous dismutation of superoxide, the decay of superoxide is a very fast second order process even under such conditions [30]. As a consequence, the actual amount of superoxide added to the bacterial suspension is much lower than the nominal concentration of KO2. To minimize sample to sample differences
due to small variations in the times of KO2 preparation and addition to bacteria, the experiment has been repeated several times with different bacterial cultures. Fig. 3 shows that wild type Salmonella is highly resistant to 1 mM KO2, while either the single dps and the double sodC1,sodC2 mutants showed comparable susceptibility to this compound. The observed qualitative differences in the relative susceptibility of strains PF103 and PF102a to KO2 and xanthine/xanthine oxidase are likely due to the different mode of exposure to superoxide. Using potassium superoxide, bacteria are exposed to a relatively high initial concentration of superoxide, while in the xanthine/ xanthine oxidase assay bacteria face a continuous flux of superoxide which interrupts upon consumption of xanthine. Interestingly, KO2 toxicity was almost completely prevented by the incubation of bacteria with catalase, confirming that hydrogen peroxide is involved in the killing of bacteria exposed to extracellular sources of superoxide. 3.3. Susceptibility of S. Typhimurium to hydrogen peroxide To understand the role of Cu,ZnSOD in conferring resistance to reactive oxygen species, the effect of hydrogen peroxide on the survival of wild type and mutant Salmonellae was analyzed (Fig. 4). In line with previous observations [25], the mutant lacking dps exhibited a very high sensitivity toward a challenge with 250 μM hydrogen peroxide. In agreement with the interpretation of the above described experiments, the PF102a strain, lacking the genes encoding for the two periplasmic Cu, ZnSODs, proved to be highly susceptible to hydrogen peroxide when compared to the wild type strain. Periplasmic Cu,ZnSOD therefore protects bacteria not only against the oxygen radical species produced by the xanthine oxidase/xanthine assay (i.e. a mixture of superoxide and hydrogen peroxide), but also against hydrogen peroxide challenge. Moreover, the PF112 strain was killed more rapidly than the mutant lacking the dps gene. The approximately additive effect of the sodC and dps mutations indicates that Cu,ZnSODs and Dps contribute to cellular defence against hydrogen peroxide independently of one another.
Fig. 4. Susceptibility of wild type and mutant strains of S. Typhimurium ATCC 14028 to hydrogen peroxide. Results are expressed as percent survival after 1-h and 2-h incubation at 37 °C. The percent survival was calculated for each strain by dividing the number of CFU/ml obtained from incubation in PBS alone by the number of CFU/ml obtained from incubation in 250 μM hydrogen peroxide. The data represent the average of three experiments, each one carried out in quadruplicate, and represent mean ± standard errors.
F. Pacello et al. / Biochimica et Biophysica Acta 1780 (2008) 226–232
A superficial reading of these results could suggest paradoxically that Cu,ZnSOD is required to protect cells from hydrogen peroxide rather than from superoxide. However, the direct participation of Cu,ZnSOD in hydrogen peroxide scavenging is extremely implausible although Cu,ZnSOD has been suggested to catalyze the reverse reaction, i.e. formation of superoxide, in the presence of high concentrations of hydrogen peroxide [38]. The recent discovery that enteric bacteria can release significant amounts of superoxide in their periplasmic space suggests that it is more reasonable to hypothesize that the inability to scavenge the superoxide anion produced endogenously has a role in the hydrogen peroxide sensitivity of the sodC mutants. E. coli and Salmonella sodC mutants show no apparent growth defects when cultivated in vitro and no specific superoxide-sensitive targets have been identified so far in the periplasm of E. coli or Salmonella. However, it is possible that superoxide generated within the periplasmic space might damage some periplasmic or membrane-component of the cell envelope somehow contributing to create a sensitizing condition towards hydrogen peroxide. An alternative possibility is that O2− might amplify the toxicity of H2O2 by promoting hydroxyl radical formation via the Haber–Weiss reaction (H2O2 + O2− → O2 + OH− + OH·). This is a quite slow reaction which can not proceed at significant rates in vivo in the absence of a redox active transition metal catalyst [39]. In bacteria, as well as in all cell types, transition metals are stably bound by other molecules which prevent their participation to reactions leading to hydroxyl radical generation. However, different studies have indicated that incubation of bacteria with hydrogen peroxide interferes with iron and copper homeostasis and enhances free iron concentration in the cytoplasm [21,40] and free copper in the periplasm [41]. A transient hydrogen peroxide-induced increase in the concentration of free periplasmic redox active metals could stimulate superoxide-mediated formation of hydroxyl radicals in the periplasm of sodC mutants and contribute to the observed phenotype. In conclusion, the present results confirm that extracellular superoxide has a negligible bactericidal activity and suggest that rapid superoxide removal by Cu,ZnSOD is required to prevent that more toxic chemical species be formed within the periplasmic space upon reaction of superoxide with hydrogen peroxide. Further, the observation that the bacterial strain lacking dps and the two sodC genes is much more susceptible to extracellular sources of reactive oxygen species than the single dps mutant indicates that full protection of bacterial cells from extracellular oxidative attack is conferred by localization of antioxidant enzymes both in the periplasm and in the cytoplasm. Acknowledgments This work was partially supported by a FIRB 2001 grant to A.B and E. C. Thanks are due to Lionello Bossi for providing us the S. Typhimurium sodC2 mutant strain used in this work. References [1] A. Battistoni, Role of prokaryotic Cu,Zn superoxide dismutase in pathogenesis, Biochem. Soc. Trans. 6 (2003) 1326–1329.
231
[2] H.M. Steinman, Function of periplasmic copper–zinc superoxide dismutase in Caulobacter crescentus, J. Bacteriol. 175 (1993) 1198–1202. [3] K.E. Wilks, K.L.R. Dunn, J.L. Farrant, K.M. Reddin, A.R. Gorringe, P.R. Langford, J.S. Kroll, periplasmic superoxide dismutase in meningococcal pathogenicity, Infect. Immun. 66 (1998) 213–217. [4] M. Lynch, H. Kuramitsu, Expression and role of superoxide dismutases (SOD) in pathogenic bacteria, Microbes Infect. 2 (2000) 1245–1255. [5] A.S. Gort, D.M. Ferber, J.A. Imlay, The regulation and role of the periplasmic copper, zinc superoxide dismutase of Escherichia coli, Mol. Microbiol. 32 (1999) 179–191. [6] S.S. Korshunov, J.A. Imlay, Detection and quantification of superoxide formed within the periplasm of Escherichia coli, J. Bacteriol. 188 (2006) 6326–6334. [7] M.A. De Groote, U.A. Ochsner, M.U. Shiloh, C. Nathan, J.M. McCord, M.C. Dinauer, S.J. Libby, A. Vazquez-Torres, Y. Xu, F.C. Fang, Periplasmic superoxide dismutase protects Salmonella from products of phagocyte NADPH-oxidase and nitric oxide synthase, Proc. Natl. Acad. Sci. U. S. A. 94 (1997) 13997–14001. [8] J.L. Farrant, A. Sansone, J.R. Canvin, M.J. Pallen, P.R. Langford, T.S. Wallis, G. Dougan, J.S. Kroll, Bacterial copper- and zinc-cofactored superoxide dismutase contributes to the pathogenesis of systemic salmonellosis, Mol. Microbiol. 25 (1997) 785–796. [9] F.C. Fang, M.A. DeGroote, J.W. Foster, A.J. Bäumler, U. Ochsner, T. Testerman, S. Bearson, J.C. Giard, Y. Xu, G. Campbell, T. Laessig, Virulent Salmonella typhimurium has two periplasmic Cu,Zn-superoxide dismutases, Proc. Natl. Acad. Sci. U. S. A. 96 (1999) 7502–7507. [10] N. Figueroa-Bossi, L. Bossi, Inducible prophages contribute to Salmonella virulence in mice, Mol. Microbiol. 33 (1999) 167–176. [11] A. Battistoni, F. Pacello, S. Folcarelli, M. Ajello, G. Donnarumma, R. Greco, M.G. Ammendolia, D. Touati, G. Rotilio, P. Valenti, Increased expression of periplasmic Cu,Zn superoxide dismutase enhances the survival of Escherichia coli invasive strains within nonphagocytic cells, Infect. Immun. 68 (2000) 30–37. [12] D.L. Piddington, F.C. Fang, T. Laessig, A.M. Cooper, I.M. Orme, N.A. Buchmeier, Cu,Zn superoxide dismutase of Mycobacterium tuberculosis contributes to survival in activated macrophages that are generating an oxidative burst, Infect. Immun. 69 (2001) 4980–4987. [13] A. Sansone, P.R. Watson, T.S. Wallis, P.R. Langford, J.S. Kroll, The role of two periplasmic copper- and zinc-cofactored superoxide dismutases in the virulence of Salmonella choleraesuis, Microbiology 148 (2002) 719–726. [14] S. Uzzau, L. Bossi, N. Figueroa-Bossi, Differential accumulation of Salmonella [Cu,Zn] superoxide dismutases SodCI and SodCII in intracellular bacteria: correlation with their relative contribution to pathogenicity, Mol. Microbiol. 46 (2002) 147–156. [15] L.M. Sly, D.G. Guiney, N.E. Reiner, Salmonella enterica serovar Typhimurium periplasmic superoxide dismutases SodCI and SodCII are required for protection against the phagocyte oxidative burst, Infect. Immun. 70 (2002) 5312–5315. [16] A. Battistoni, M. Ajello, S. Ammendola, G. Rotilio, P. Valenti, Involvement of reactive oxygen species in bacterial killing within epithelial cells, Int. J. Immunopharmacol. 17 (2004) 71–76. [17] R. Krishnakumar, M. Craig, J.A. Imlay, J.M. Slauch, Differences in enzymatic properties allow SodCI but not SodCII to contribute to virulence in Salmonella enterica serovar Typhimurium strain 14028, J. Bacteriol. 186 (2004) 5230–5238. [18] S. Ammendola, M. Ajello, P. Pasquali, J.S. Kroll, P.R. Langford, G. Rotilio, P. Valenti, A. Battistoni, Differential contribution of sodC1 and sodC2 to intracellular survival and pathogenicity of Salmonella enterica serovar Choleraesuis, Microbes Infect. 7 (2005) 698–707. [19] N. Figueroa-Bossi, S. Ammendola, L. Bossi, Differences in gene expression levels and in enzymatic qualities account for the uneven contribution of superoxide dismutases SodCI and SodCII to pathogenicity in Salmonella enterica, Microbes Infect. 8 (2006) 1569–1578. [20] K.R.C. Imlay, J.A. Imlay, Cloning and analysis of sodC, encoding the copper– zinc superoxide dismutase of Escherichia coli, J. Bacteriol. 178 (1996) 2564–2571. [21] J.A. Imlay, Iron–sulphur clusters and the problem with oxygen, Mol. Microbiol. 59 (2006) 1073–1082.
232
F. Pacello et al. / Biochimica et Biophysica Acta 1780 (2008) 226–232
[22] H.M. Hassan, I. Fridovich, Paraquat and Escherichia coli. Mechanism of production of extracellular superoxide radical, J. Biol. Chem. 254 (1979) 10846–10852. [23] S.S. Korshunov, J.A. Imlay, A potential role for periplasmic superoxide dismutase in blocking the penetration of external superoxide into the cytosol of Gram-negative bacteria, Mol. Microbiol. 43 (2002) 95–106. [24] G. Zhao, P. Ceci, A. Ilari, L. Giangiacomo, T.M. Laue, E. Chiancone, N.D. Chasteen, Iron and hydrogen peroxide detoxification properties of DNAbinding protein from starved cells. A ferritin-like DNA-binding protein of Escherichia coli, J. Biol. Chem. 277 (2002) 27689–27696. [25] T.A. Halsey, A. Vazquez-Torres, D.J. Gravdahl, F.C. Fang, S.J. Libby, The ferritin-like Dps protein is required for Salmonella enterica serovar Typhimurium oxidative stress resistance and virulence, Infect. Immun. 72 (2004) 1155–1158. [26] K.A. Datsenko, B.L. Wanner, One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products, Proc. Natl. Acad. Sci. U. S. A. 97 (2000) 6640–6645. [27] P.P. Cherepanov, W. Wackernagel, Gene disruption in Escherichia coli: TcR and KmR cassettes with the option of Flp-catalyzed excision of the antibiotic-resistance determinant, Gene 158 (1995) 9–14. [28] S. Uzzau, N. Figueroa-Bossi, S. Rubino, L. Bossi, Epitope tagging of chromosomal genes in Salmonella, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 15264–15269. [29] S.H. Jackson, J.L. Gallin, S.M. Holland, The p47 phox mouse knock-out model of chronic granulamatous disease, J. Exp. Med. 182 (1995) 751–758. [30] S. Marklund, Spectrophotometric study of spontaneous disproportionation of superoxide anion radical and sensitive direct assay for superoxide dismutase, J. Biol. Chem. 251 (1976) 7504–7507.
[31] E.P. Reeves, H. Lu, H.L. Jacobs, C.G. Messina, S. Bolsover, G. Gabella, E.O. Potma, A. Warley, J. Roes, A.W. Segal, Killing activity of neutrophils is mediated through activation of proteases by K+ flux, Nature 416 (2002) 291–297. [32] A.W. Segal, How neutrophils kill microbes, Annu Rev Immunol. 23 (2005) 197–223. [33] B.M. Babior, Phagocytes and oxidative stress, Am. J. Med. 109 (2000) 33–44. [34] S. Altuvia, M. Almiron, G. Huisman, R. Kolter, G. Storz, The dps promoter is activated by OxyR during growth and by IHF and sigma S in stationary phase, Mol. Microbiol. 13 (1994) 265–272. [35] Y.A. Golubeva, J.M. Slauch, Salmonella enterica serovar Typhimurium periplasmic superoxide dismutase SodCI is a member of the PhoPQ regulon and is induced in macrophages, J. Bacteriol. 188 (2006) 7853–7861. [36] S. Schnell, H.M. Steinman, Function and stationary-phase induction of periplasmic copper–zinc superoxide dismutase and catalase/peroxidase in Caulobacter crescentus, J. Bacteriol. 177 (1995) 5924–5929. [37] L.R. San Mateo, M.M. Hobbs, T.H. Kawula, Periplasmic copper–zinc superoxide dismutase protects Haemophilus ducreyi from exogenous superoxide, Mol. Microbiol. 27 (1998) 391–404. [38] S.I. Liochev, I. Fridovich, Reversal of the superoxide dismutase reaction revisited, Free Radic. Biol. Med. 34 (2004) 908–910. [39] B. Halliwell, J.M. Gutteridge, Role of free radicals and catalytic metal ions in human disease: an overview, Methods Enzymol. 186 (1990) 1–85. [40] S. Jang, J.A. Imlay, Micromolar intracellular hydrogen peroxide disrupts metabolism by damaging iron–sulfur enzymes, J. Biol. Chem. 282 (2007) 929–937. [41] L. Macomber, C. Rensing, J.A. Imlay, Intracellular copper does not catalyze the formation of oxidative DNA damage in Escherichia coli, J. Bacteriol. 189 (2007) 1616–1626.
