INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The presence of oxygen in the environment is potentially toxic to all forms of life. This toxicity is mediated by reactive oxygen species (ROS), generated as by-products during the univalent reduction of oxygen to water, which can damage DNA, proteins, and lipids (28). These ROS include superoxide radical (O₂^·), hydrogen peroxide (H₂O₂), and hydroxyl radical (OH^·). As a defense against oxidative stress, most bacteria contain superoxide dismutases (SODs) (EC 1.15.1.1), which detoxify O₂^· to H₂O₂, which in turn is broken down to water by catalases. The SODs are metalloenzymes that are classified according to their metal cofactor. There are three main classes of SODs in bacteria, the manganese SOD (MnSOD), the iron SOD (FeSOD), and the copper zinc SOD (CuZnSOD). Recently, a new class of SOD has been described, the nickel SOD (NiSOD) (33, 34, 57). Usually, FeSODs and MnSODs exhibit strict metal cofactor specificity (6) and can be distinguished by their sensitivity to H₂O₂. However, a small group of Mn/FeSODs, named cambialistic, is active with either manganese or iron incorporated into the same active site and exhibits variable sensitivity to H₂O₂ (47, 55).

Some bacteria, such as Escherichia coli or Staphylococcus aureus, possess more than one SOD (4, 15). E. coli has three SODs, which differ in their location and temporal expression: two SODs, the FeSOD and the MnSOD, are present in the cytoplasm (32, 56), while the CuZnSOD is located in the periplasm (29). Expression of the FeSOD is thought to be constitutive, but the levels of the MnSOD fluctuate, increasing in response to O₂^· and upon changes in growth phase (17). In S. aureus, where two SODs are detected, the major enzyme, characterized as a MnSOD, is inducible by a variety of conditions (15). In contrast, the second and less-characterized SOD enzyme appears to be expressed constitutively (15).

Staphylococcus xylosus is an anaerobic facultative bacterium used as a starter culture for fermented meat products. It ensures color development by its nitrate reductase activity and protects the cured color by its catalase activity (38, 51). It also contributes to aroma, mainly by modulating the level and the nature of volatile compounds coming from lipid oxidation (5, 40, 41, 52). Antioxidant activities of S. xylosus (e.g., catalase and SOD) are thought to be involved in the development of the sensorial qualities (45). Therefore, it is crucial to characterize these enzymes and to construct mutants with the corresponding genes to understand their role. In this study, we describe the physiological and molecular characterization of the single SOD from S. xylosus. The corresponding gene was cloned and sequenced, and its regulation and its role were investigated.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains and plasmids. The bacterial strains used are listed in Table 1. The temperature-sensitive shuttle vector pBT2 (10) and the lacH promoter probe plasmid pLP1 (31) were used. The ermB cassettes from plasmid pEC5 and plasmid pEC7 were used, respectively, to interrupt the sod and the zurR genes in S. xylosus (10).

Media and culture conditions. S. xylosus was grown at 30°C in a complex medium (CM) (meat extract [10 g/liter], yeast extract [5 g/liter], NaCl [5 g/liter], Na₂HPO₄ [2 g/liter]) prepared in 0.067 M phosphate buffer, pH 6.0, or in chemically defined medium (CDM) (27). When needed, media were supplemented with erythromycin (2.5 µg/ml). To study the effect of metals, heavy metals were removed from the medium with Chelex-100 (Bio-Rad Laboratories, Hercules, Calif.) as recommended by the manufacturer, and, when needed, ultrapure MnSO₄ (Sigma) was added. Aerobic cultures were incubated on a rotary shaker at 170 rpm, and the volume of cultures did not exceed 1/10 of the total Erlenmeyer volume to ensure good aeration. Low aeration was performed by growing the cells in tubes filled up to 85% (total volume, 7 ml) with slow stirring (15 rpm) on a shaker-incubator. E. coli was grown aerobically at 37°C in Luria-Bertani medium supplemented, when needed, with ampicillin (100 µg/ml) and 0.05 mM paraquat (Sigma).

Determination of SOD and -galactosidase activities in crude extracts. For preparation of crude extracts, cells were washed once in SOD buffer containing 10 mM Tris, pH 7.0, or in -galactosidase buffer (31). The washed cells were resuspended in 500 µl of the same buffer, 1 g of glass beads (150 to 212 µm) was added, and the samples were vortexed twice for 1 min with ice cooling in between. After centrifugation (10 min, 8,000 × g, 4°C), the supernatants were collected and kept on ice. Subsequently, partially broken cells including glass beads were resuspended three times in 500 µl of buffer, vortexed, and centrifuged as described before. The combined supernatants (2 ml corresponding approximately to 2 mg of protein) were kept on ice, and aliquots were tested for SOD or -galactosidase activity. The SOD activity was measured using a RANSOD kit (Randox, Co., Antrim, United Kingdom) with 5 to 30 µg of cellular protein. The -galactosidase activity was assayed with 15 to 400 µg of cellular protein according to the method of Jankovic et al. (31). The concentration of protein was determined by the method of Bradford with bovine serum albumin as a standard (7).

Visualization of SOD activity on nondenaturing polyacrylamide gels. The SOD activity on 12.5% nondenaturing polyacrylamide gels was visualized by nitroblue tetrazolium negative staining (2). Inhibition experiments with SOD isoenzymes were done with H₂O₂ or KCN as described previously (2).

Metal depletion and reconstitution of crude extracts. Metal depletion was performed by dialyzing crude extracts against metal depletion buffer (20 mM 8-hydroxyquinoline, 2.5 M guanidium chloride, 5 mM Tris-HCl [pH 3.8], 0.1 mM EDTA) (35). Reconstitution of metal-depleted crude extracts was performed with either 0.1 mM MnCl₂ or 1 mM Fe(NH₄)₂(SO₄)₂ (35).

DNA preparation, transformation, and molecular techniques. Chromosomal DNA from S. xylosus was isolated according to the method of Marmur (39). Plasmid DNA was introduced into S. xylosus by electroporation with glycine-treated electrocompetent cells (10). DNA manipulations, Southern blotting, plasmid DNA isolation, and transformation of E. coli were performed according to standard procedures (46).

PCR conditions used for cloning the sod gene. Chromosomal DNA from S. xylosus was amplified by PCR with two degenerate primers, SOD1 and SOD2 (Table 2) (43). The PCR mix contained 100 pmol of each primer, about 300 ng of chromosomal DNA, a 0.2 mM concentration of each dNTP, 1× Taq polymerase buffer, and 1 U of Taq polymerase (Appligene). The reaction was cycled 30 times through the following temperature profile: denaturation at 95°C for 2 min, annealing of primers at 37°C for 1 min, and extension at 72°C for 1 min, followed by a final extension step of 10 min at 72°C. A 468-bp amplified fragment was obtained and sequenced. From this sequence, two specific primers, SOD4 and SOD9 (Table 2), were designed and used to screen a S. xylosus gene library (11).

RNA isolation and primer extension analysis. Preparation of RNA and primer extension reactions were done as described previously (1). The primer P3 (Table 2) complementary to the sod coding sequence was labeled at the 5' end with infrared dye IRD700. The DNA primer was elongated, and the products were analyzed on an 8% polyacrylamide-urea gel with a Li-Cor DNA sequencer to determine the transcriptional start site.

Construction of mutants by gene replacement. To inactivate the sod gene in S. xylosus, the plasmid pBSe was constructed in three steps. First, a 1-kb fragment was obtained by PCR from the sod nucleotide sequence region with the primers S1 and S2 (Table 2). The amplified fragment contained about two-thirds of the zurM gene and the complete zurR gene. It was digested with EcoRI and SstI and inserted in the temperature-sensitive shuttle vector pBT2 digested by the same enzymes. Second, the ermB cassette coming from the plasmid pEC5 was inserted between the SstI and KpnI restriction sites of the previous construct. Finally, a 1-kb fragment was obtained by PCR from the sod nucleotide sequence region with the primers S3 and S4 (Table 2). The amplified fragment contained the sod gene without its Shine Dalgarno sequence, its start codon, and its first 22 nucleotides, and a part of the stpB gene. It was digested by KpnI and SalI and inserted between the KpnI and SalI sites of the previous construct, yielding the plasmid pBSe.

Integration of promoters in front of the chromosomal -galactosidase gene lacH. The S. xylosus promoters that were analyzed with the promoter probe system are shown below (see Fig. 4). A fragment containing the complete sod promoter P1/2_sod was obtained by PCR with pS41 DNA as a template and the primers PSOD1 and PSOD2 (Table 2). The fragments containing only the promoter P1_sod or P2_sod were obtained using the primers PSOD1 and PSOD3 or PSOD2 and PSOD4 (Table 2), respectively. The BamHI-SalI fragments were cloned into the lacH promoter probe plasmid pLP1 (see Fig. 4). Successful integration of promoter fragments into pLP1 was detected on 5-bromo-4-chloro-3-indolyl--D-galactopyranoside-containing agar plates at 30°C. Promoter-containing plasmids were designated pLP21 (P1/2_sod), pLP22 (P1_sod) and pLP23 (P2_sod). Promoter sequences were verified by sequencing on both strands.

Nucleotide sequence accession number. The S. xylosus sequence determined in this study is available from the EMBL database under accession no. AJ276960.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Cloning the sod gene. To isolate the SOD-encoding gene, a PCR-based approach with degenerate primers was used. These degenerate primers were previously used to detect sod genes in different bacterial species (43). An amplified fragment was obtained and sequenced. Its deduced amino acid sequence showed a high level of similarity to SODs. Therefore, two specific primers were designed to screen an S. xylosus gene library stored as pools of plasmids (11). One pool of plasmid DNAs gave a fragment of the expected size. To identify the plasmid harboring the sod gene, E. coli QC779 (sodA sodB), which is deficient in both cytoplasmic SODs, was transformed with the plasmid mixture and plated on a rich medium with paraquat, a generator of superoxide. Without a functional cytoplasmic SOD, E. coli is unable to grow under these conditions (13). Several colonies grew, and their plasmid DNA was analyzed. All transformants harbored identical plasmids. One plasmid, designated pS41, containing an insertion of about 3.6 kb, was further studied.

View larger version (46K): [in this window] [in a new window]   FIG. 1.   Detection of SOD activity in E. coli and S. xylosus on a nondenaturing polyacrylamide gel stained for SOD activity. Lanes: 1, S. xylosus C2a; 2, E. coli QC779 (sodAsodB)/pS41 with S. xylosus sod gene; 3, E. coli QC779; 4, E. coli GC4468. Lanes contain crude extracts of aerobic cultures (15 µg of proteins were loaded for S. xylosus C2a and 30 µg were loaded for E. coli strains).

Nucleotide sequence of the sod region. The complete insert of pS41 consisting of 3,595 bp was sequenced. Four complete open reading frames (ORFs) and a truncated ORF were found on one strand (Fig. 2). The fourth ORF, encoding a polypeptide of 199 amino acids with a theoretical M_r of 22.5 and a pl of 4.67, is clearly the sod gene, since the deduced amino acid sequence of its product revealed a high level of similarity to the SOD family of proteins. The highest levels of identity were observed with the following SODs: the MnSOD (SodA) of S. aureus (accession no. AF121672) (91% identity), the SOD of Staphylococccus carnosus (accession no. AJ295150) (87% identity), the MnSODs of Bacillus subtilis (accession no. D86856), Bacillus licheniformis (accession no. AJ002279), Bacillus stearothermophilus (accession no. P00449), and Bacillus caldotenax (accession no. P28760) (68 to 72% of identical residues). The critical residues in SODs commonly used to predict the metal specificity of the enzymes (42) suggest that the S. xylosus SOD requires Mn rather than Fe as a metal cofactor. The SOD enzyme from S. xylosus (PsiPred Prediction; Protein Bioinformatics Group, Department of Biological Sciences, University of Warwick, Coventry, United Kingdom) showed a secondary structure typical of Mn/FeSODs, with two domains, the N-terminal one with two -helices and the C-terminal domain with two -helices, followed by three -strands and two -helices (data not shown).

View larger version (8K): [in this window] [in a new window]   FIG. 2.   Genetic organization surrounding the sod gene from S. xylosus. Genes are shown by arrows. Relevant restriction sites are labeled.

Determination of sod transcriptional start sites. To define the transcriptional start site of the sod gene, the 5' end of the sod transcript was mapped. A primer specific to the coding region was annealed with total RNA and extended in a primer extension assay. RNA was prepared from cells grown in CM medium and harvested at different times during growth. Two primer extension products were obtained under all conditions (data not shown). As an example, the reaction with RNA from cells in exponential growth phase is shown in Fig. 3. The sizes of the reverse transcripts placed the transcriptional start points, respectively, 25 bp and 120 bp upstream of the sod start codon. Upstream of the first start point, a putative E. coli sigma 70-like sequence was found matching four of the six bases (boldface) with the 35 (TTGACA) consensus sequence and five of the six bases with the 10 (TATAAT) consensus sequence (Fig. 3). The respective boxes of the second promoter fitted less perfectly to the consensus sequences (Fig. 3). To verify that sod possesses two functional promoters, the genomic reporter gene system described for S. xylosus (31) was used. Each presumed promoter was integrated in front of the promoterless -galactosidase gene lacH by homologous recombination (Fig. 4). The resulting strains, S. xylosus TX354 containing the first promoter P1_sod and S. xylosus TX355 containing the second promoter P2_sod, gave rise to blue colonies on 5-bromo-4-chloro-3-indolyl--D-galactopyranoside agar plates, substantiating the presence of two promoters in front of sod. A terminator-like structure immediately downstream of the sod reading frame suggests that the sod mRNA of S. xylosus is monocistronic.

View larger version (31K): [in this window] [in a new window]   FIG. 3.   Primer extension of sod transcription. Total RNA was prepared from S. xylosus C2a cells grown in nonbuffered CM medium. Thirteen micrograms of total RNA was used to extend a sod-specific primer, P3, by reverse transcriptase. One-thirtieth of the primer extension reaction was analyzed on an 8% polyacrylamide-urea gel alongside a sequencing reaction performed with the same primer. Possible sites (+1) are shown by arrows.

View larger version (28K): [in this window] [in a new window]   FIG. 4.   Genetic organization of the genomic reporter system for S. xylosus (A) and nucleotide sequences of the promoters integrated in front of the -galactosidase gene lacH (B). (A) The genetic organization of the wild-type S. xylosus lactose utilization genes lacR, lacP, and lacH is shown in the first line. The lac deletion strain TX300 harbors versions of 'lacR and 'lacH truncated at their 5'-ends and lacks lacP. The promoterless -galactosidase genes lacH and 'lacR are contained on the promoter probe plasmid pLP1. The nucleotide sequence of the region preceding the reporter gene lacH is shown below pLP1. The three restriction sites BamHI, XbaI, and SalI are available for insertion of promoter fragments. (B) The promoter regions relevant for transcription initiation and regulation are shown. Putative RNA polymerase binding sites are underlined. Transcriptional start sites are shown in boldface. Designations of the plasmids harboring the shown promoters and corresponding strains are shown in parenthesis.

Metal cofactor requirements of S. xylosus SOD. In S. xylosus, a single SOD was detected on a nondenaturing polyacrylamide gel (Fig. 5). This SOD was not inhibited by KCN (data not shown) or H₂O₂ (Fig. 5), suggesting that the enzyme is a MnSOD. To determine the exact nature of the cofactor of the S. xylosus SOD, metal depletion and reconstitution experiments were performed. The enzyme in crude extracts was depleted of metal (Fig. 5). The resulting inactive apoenzyme was dialyzed against Mn or Fe. A high activity (5.1 ± 0.2 U/mg of protein) was recovered with Mn. The Mn-reconstituted enzyme was not inhibited by H₂O₂ (Fig. 5). A very weak activity (0.4 ± 0.2 U/mg of protein) was recovered with Fe, and a faint band with the same mobility as the Mn-reconstituted enzyme was revealed after electrophoresis (data not shown).

View larger version (14K): [in this window] [in a new window]   FIG. 5.   Activity of reconstituted SOD from S. xylosus with manganese. (A) No inhibitors. (B) with H2O2. Lanes: 1, crude extract of E. coli DH5; 2, crude extract of S. xylosus; 3, apoenzyme of S. xylosus, 4, Mn-reconstituted SOD of S. xylosus. Thirty micrograms of proteins were loaded in each lane.

SOD activity and sod expression in S. xylosus. For many bacteria, the level of SOD activity fluctuates depending on the growth phase, presence of superoxide generators, and metal availability. For S. xylosus, only a slight increase of SOD activity was observed during stationary phase when cultures were grown in CM or CDM medium (Table 3). The strains S. xylosus TX353, TX354, and TX355 containing, respectively, P1/2_sod, P1_sod, and P2_sod in front of the lacH gene enabled monitoring of sod expression throughout different growth conditions by measurement of the level of -galactosidase activity. The strain S. xylosus TX302 harboring the constitutive promoter from B. subtilis, P_vegII in front of the lacH gene (31) was used to verify that there was no variation of -galactosidase activity under all growth conditions (data not shown). In agreement with SOD activity measures, expression of sod in the three strains was slightly increased in the stationary phase when cells were grown in CDM (Table 3). However, when cells were grown in CM, a 4-fold to 10-fold increase of sod expression was noticed during the stationary phase (Table 3). P2_sod and especially P1_sod were induced by stationary phase (Table 3). The low level of SOD activity measured from the stationary-phase cells grown in CM could be explained in part by the weak concentration of manganese in the medium, since in CM supplemented by 0.1 mM MnSO₄, SOD activity was higher in the stationary phase (6.5 ± 0.1 U/mg of protein) than in the nonsupplemented medium (3.2 ± 0.4 U/mg of protein).

View this table: [in this window] [in a new window]   TABLE 3.   Effect of growth phase on SOD activity and on -galactosidase expression directed by P1/2sod (TX353), P1sod (TX354), and P2sod (TX355) in S. xylosus

View this table: [in this window] [in a new window]   TABLE 4.   Effect of paraquat on SOD activity and on -galactosidase expression directed by P1/2sod (TX353), P1sod (TX354), and P2sod (TX355) in S. xylosus

View this table: [in this window] [in a new window]   TABLE 5.   Effect of metals on SOD activity and on -galactosidase expression directed by P1/2sod (TX353), P1sod (TX354), and P2sod (TX355) in S. xylosus

SOD activity in a zurR mutant. zurR encodes a putative metalloregulatory protein and belongs to the family of ferric uptake regulation proteins. In E. coli, the sodA gene encoding the MnSOD is regulated transcriptionally by Fur (53) and posttranslationally in a metal-dependent fashion (6, 44). Therefore, we measured SOD activity in a zurR mutant constructed by allelic exchange. No change of SOD activity with the zurR mutant from that of the wild type strain grown in CM or CDM media was observed (data not shown).

Physiological characterization of a sod mutant. To investigate the physiological role of the sod gene, a sod mutant was constructed by allelic exchange. The sod mutant was completely devoid of SOD activity (data not shown), proving the presence of a single expressed sod gene. The aerobic growth of the sod mutant in CM was similar to that of the wild-type strain (Fig. 6B). However, the addition of hyperbaric O₂ in the exponential growth phase resulted in a significant decrease of growth of the sod mutant compared to the wild-type strain (Fig. 6A). The same effect was noticed when 50 µM paraquat was added in the exponential growth phase (Fig. 6B), while the wild-type strain was unaffected at this concentration (Fig. 6B). In addition, growth of the sod mutant in CDM with all amino acids was at a slightly lower level than that of the wild-type strain, and when Leu, Ile and Val were lacking, this difference was dramatically increased (Fig. 6C). The sod mutant was practically unable to grow, whereas the growth of the wild-type strain was almost unaffected.

View larger version (13K): [in this window] [in a new window]   FIG. 6.   Growth of the S. xylosus wild-type strain (C2a) and an S. xylosus sod mutant (TX351) in the presence of hyperbaric oxygen or paraquat or in a chemically defined medium (CDM). Closed symbols represent the growth curves of S. xylosus C2a, and open symbols represent the growth curves of S. xylosus TX351. (A) Growth in the presence of hyperbaric oxygen (squares). Cells were grown aerobically in MC medium. In exponential growth phase, a continuous bubbling of pure oxygen in the medium was applied at a pressure of 0.2 bar (shown by O2 with an arrow). (B) Growth in the presence of paraquat. Cells were grown aerobically in MC medium. In exponential growth phase, paraquat was added (shown by an arrow) at a final concentration of 0 µM (squares) or 50 µM (triangles). (C) Growth in CDM. Cells were grown aerobically in CDM (squares) or in CDM without Leu, Ile, and Val (circles). Results are representative of two independent experiments. OD600, optical density at 600 nm.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

We determined in this study that S. xylosus contains a single SOD. In contrast, S. aureus has at least two SODs: SodA and SodX. SodX was suggested to be cell wall associated and implicated in virulence (14). This difference between the two species could be explained by the fact that S. aureus as a potential pathogen has to deal with highly microbiocidal reactive oxygen metabolites produced during the oxidative burst by phagocytes, whereas S. xylosus is rarely associated with human or animal infections.

The SOD of S. xylosus is a manganese-preferring enzyme. In crude extracts, SOD activity was completely recovered by manganese and very weakly by iron. This result questioned the cambialistic nature of S. xylosus SOD, as it has already been shown that some cambialistic SODs are less active with iron than with manganese (47, 55). However, S. xylosus SOD is not inhibited by hydrogen peroxide, and its amino acid sequence exhibits strong similarity to SodA of S. aureus which was shown to be a manganese-requiring enzyme (15). Furthermore, manganese was necessary to sustain SOD activity in the stationary phase, as was also observed by Inaoka et al. with B. subtilis (30).

In S. xylosus, upstream of the sod gene a putative zinc uptake and regulation operon (zurA, -M, and -R) was detected. This genetic organization seems to be conserved in other staphylococcal species, such as S. aureus and Staphylococcus epidermidis (25, 26; Lindsay and Foster, unpublished data). In S. xylosus, ZurR did not appear to regulate sod expression.

In E. coli, the sodA sequence contains two putative promoters, but only one has been found to be functional under normal aerobic growth conditions (50). In B. subtilis, nucleotide sequence analysis indicated that sodA possesses six putative promoters (30). According to our results, the S. xylosus sod gene possesses two functional promoters. In either a chemically defined media or a complex medium, a weak induction effect of paraquat was observed, confirming the results obtained by measuring the levels of SOD activity. For S. aureus, addition of paraquat leads to an approximately fourfold induction of sod expression (15), whereas for B. subtilis or L. monocytogenes, the addition of paraquat does not induce sod expression (30, 54). In a complex medium and not in a chemically defined medium, sod expression for S. xylosus was induced in stationary-growth phase, as was also observed for S. aureus when cells were grown in BHI medium (15), for E. coli (3), and for L. monocytogenes (54). This could reflect the need for greater protection from accumulated toxic oxidants as the cells age. However, it remains unclear why no induction was observed in a chemically defined medium. In S. xylosus, manganese did not play a role at a transcriptional level, as was shown, for instance, for E. coli sodA (49). Manganese appears to be necessary only at the posttranslational step of metal insertion at the active site.

The sod gene is not essential for aerobic growth of S. xylosus, suggesting the presence of other protective functions in S. xylosus. The sod S. xylosus mutant shares phenotypes similar to those of sodA sodB E. coli mutants, which are also sensitive to hyperbaric O₂ and to paraquat and which exhibit multiple amino acid auxotrophy (13). For E. coli, this multiple amino acid auxotrophy results from different superoxide targets. One clearly identified target is the dihydroxyacid-dehydratase, which contains a 4Fe-4S cluster and which catalyzes the penultimate step in the biosynthesis of branched-chain amino acids (8, 9, 36). Other enzymes implicated in general metabolic pathways containing 4Fe-4S clusters were found to be inactivated by superoxides and protected by SOD (18, 20, 21, 37). In S. xylosus, the lack of SOD could lead to comparable damage.

In conclusion, S. xylosus possesses one single SOD, closely related to MnSODs. The sod gene is not essential for aerobic growth but appears to be important in the protection of cell constituents against oxidative stress. The role of SOD in the inhibition of the oxidation of unsaturated free fatty acids is being studied. This will lead to an understanding of the contribution of SOD to the antioxidant properties of S. xylosus.