The nucleotide sequence of the sarZ gene of RN4220 was deposited in GenBank database under GenBank Accession Number AB251456. The amino acid sequence of B. subtilis OhrR (CAB13172.1) was obtained from the Genome Information Broker program from the DNA Data Bank of Japan (http://gib.genes.nig.ac.jp/). Other amino acid sequences were obtained from GenBank: S. aureus MgrA (AY266422), E. coli MarR (P27245), P. aeruginosa MexR (U23748), S. aureus SarA (U20782).
We previously reported that the cvfA gene is a virulence regulatory gene in Staphylococcus aureus. Here, we identified a novel gene named sarZ that acts as a multicopy suppressor of decreased haemolysin production in the cvfA deletion mutant. The amount of sarZ transcripts was decreased in the cvfA mutant. The sarZ-deletion mutant produced less haemolysin and attenuated virulence in a silkworm-infection model and a mouse-infection model. The amino acid sequence of the sarZ gene product had 19% identity with the transcription factor MarR in Escherichia coli, and the internal region contained a winged helix–turn–helix motif (wHTH), a known DNA binding domain. Purified recombinant SarZ protein had binding affinity for the promoter region of the hla gene that encodes α-haemolysin. SarZ mutant proteins with an amino acid substitution in the N-terminal region or in the wHTH motif had significantly decreased DNA binding. The mutated sarZ genes encoding SarZ mutant proteins with a low affinity for DNA did not complement the decreased haemolysin production or the attenuated killing ability against silkworms in the sarZ mutant. These results suggest that the DNA binding activity of the SarZ protein is required for virulence in S. aureus.
Staphylococcus aureus is a pathogenic bacterium against humans, causing various diseases such as localized skin infection, food poisoning, endocarditis and toxic shock syndrome. These diseases are caused by the regulated expression of various virulence factors. Cell surface proteins, such as fibronectin binding protein and protein A, are produced to adhere to host tissues, whereas exotoxins such as haemolysin, protease and lipase are produced to destroy host tissues (Lowy, 1998; Dinges et al., 2000; Foster, 2005). Two elements, the agr locus and the sarA gene, regulate expression of these virulence factors.
The agr locus regulates the expression of various virulence genes (Peng et al., 1988). The agr locus encodes two divergent transcripts, RNAII and RNAIII, which are transcribed from two different promoters, P2 and P3 respectively. RNAII encodes four proteins, AgrA, AgrB, AgrC and AgrD. AgrD is processed and secreted with the aid of AgrB, and functions as an autoinducing peptide (AIP). AIP is recognized by AgrC, a sensor of the two-component regulatory system, and phosphorylated AgrC leads to a second phosphorylation of the activator AgrA. Phosphorylated AgrA further facilitates the transcription of RNAII and RNAIII (Koenig et al., 2004). RNAIII regulates the expression of many virulence factors during the growth phase (Novick et al., 1993). For example, RNAIII upregulates expression of the hla gene, which encodes α-haemolysin, and downregulates expression of the spa gene, which encodes protein A, at both the transcriptional and translational levels during the post-exponential phase to stationary phase (Novick et al., 1993; Morfeldt et al., 1995; Huntzinger et al., 2005). SarA is a virulence factor regulatory protein (Cheung et al., 1992). SarA binds the promoter regions of the hla gene and the agr locus, thereby inducing transcription (Chien et al., 1998; 1999). SarA family proteins, which have amino acid sequence similarity with SarA, also contribute to virulence gene expression. Other regulatory systems, such as saeRS (Giraudo et al., 1997), srrAB (Yarwood et al., 2001), arlRS (Fournier et al., 2001) and svrA (Garvis et al., 2002), also regulate the expression of virulence genes. These genes are assumed to interact with each other and to regulate the expression of virulence genes, but the overall picture of the regulatory cascades is obscure (Novick, 2003).
We previously reported a silkworm-infection model using bacteria virulent against humans, such as S. aureus, Streptococcus pyogenes and Pseudomonas aeruginosa (Kaito et al., 2002). Therapeutic effects of antibiotics against the infection indicate that the bacterial growth in silkworm body kills the silkworm (Kaito et al., 2002; Hamamoto et al., 2004). To identify novel virulence regulatory genes, we constructed 100 S. aureus mutant strains in which conserved hypothetical genes (Kuroda et al., 2001) were deleted. We screened them based on attenuated killing ability against silkworms, and identified three novel virulence genes named cvfA, cvfB and cvfC (Kaito et al., 2005). Disrupted mutants of these genes also had attenuated virulence in mice. In the cvfA mutant, the expression of the agr locus was decreased and haemolysin, protease and nuclease production decreased. Moreover, under the genetic background of the agr-null mutation, cvfA disruption attenuated virulence in silkworms (Kaito et al., 2005). Therefore, the cvfA gene contributes to virulence via both agr-dependent and agr-independent pathways in S. aureus. The cvfA gene is conserved among many pathogenic bacteria, such as Listeria monocytogenes, S. pyogenes, Helicobactor pylori, etc. S. pyogenes cvfA-disrupted mutants produced less exotoxin and had attenuated virulence in both silkworms and mice, indicating that the cvfA gene also contributes to virulence in S. pyogenes (Kaito et al., 2005). There are three motifs in the CvfA protein; a transmembrane domain, an RNA binding domain named KH, and a metal-dependent phosphohydrolase domain named HD. Site-directed mutagenesis against the latter two motifs results in the decreased function of the cvfA gene, indicating that these motifs have an important role in CvfA function.
In present study, we screened a multicopy suppressor of decreased haemolysin production in the cvfA-deletion mutant and identified a novel gene that regulates haemolysin production. This gene encodes a novel uncharacterized protein that has sequence similarity with the transcriptional factor MarR in Escherichia coli and was named sarZ based on the sequence similarity with the sarA gene of S. aureus (Manna and Cheung, 2003; Cheung et al., 2004). Our genetic study of a sarZ-deletion mutant indicated that the sarZ gene contributes to haemolysin production and to virulence in S. aureus. Analysis of the amino acid sequence indicated that the SarZ protein contains a DNA binding motif. We purified recombinant wild-type and mutated SarZ proteins, and determined that the DNA binding activity of SarZ protein is required for haemolysin production and for virulence in S. aureus.
Multicopy supply of the sarZ gene suppresses the decreased haemolysin production in the cvfA mutant
To identify the virulence regulatory gene whose expression is activated by the cvfA gene, we screened a multicopy suppressor of the decreased haemolysin production in the cvfA mutant. The cvfA-deletion mutant was transformed with the S. aureus genome library and a colony that restored haemolysin production was isolated. Transformation of the cvfA mutant with a plasmid psup2 extracted from the colony restored haemolysin production (Fig. 1B). Sequencing of the DNA fragment inserted into psup2 revealed that there were three uncharacterized genes (SA2173, SA2174 and SA2175 annotated as hypothetical proteins in the S. aureus N315 genome database (Kuroda et al., 2001)) (Fig. 1A). To examine which gene was responsible for restoring haemolysin production in the cvfA mutant, each gene was inserted into a vector plasmid, resulting in pF1, pF2 and pF3 (Fig. 1A). We examined haemolysin production of the cvfA mutants that were transformed with these plasmids. The cvfA mutant transformed with pF2 produced a haemolytic zone on sheep blood agar plate as large as that of the parent strain and the cvfA mutant transformed with pcvfA harbouring the intact cvfA gene, whereas cvfA mutants transformed with pF1, pF3 or the vector plasmid (pSR515) did not have a haemolytic zone (Fig. 1B). Thus, the SA2174 gene in the DNA fragment inserted into pF2 is responsible for the suppression of the decreased haemolysin production in the cvfA mutant. S. aureus RN4220 strain produces two haemolysins, α-haemolysin and β-haemolysin (Peng et al., 1988). Lysis of sheep erythrocytes is a reflection of β-haemolysin. The suppression activity by the SA2174 gene was also observed using rabbit erythrocytes that are sensitive to α-haemolysin (data not shown), indicating that the SA2174 gene has the ability to suppress the decreased production of both α-haemolysin and β-haemolysin in the cvfA mutant.
The SA2174 gene was an uncharacterized gene that was named sarZ on the basis of a 14% amino acid sequence identity with SarA protein, a virulence regulator of S. aureus (Manna and Cheung, 2003). In addition, SarZ protein shared 19%, 34%, 29% and 21% amino acid sequence identity with E. coli MarR, Bacillus subtilis OhrR, S. aureus MgrA (also called NorR or Rat) and P. aeruginosa MexR respectively. The predicted secondary structure of SarZ consisted of six α-helixes and two β-sheets (data not shown). The internal α3-α4 (residues 56–78) constituted helix–turn–helix region and β1-β2 (residues 81–98) constituted winged region, resulting in the wHTH motif, a known DNA binding motif (data not shown).
Decreased amount of the sarZ transcript in the cvfA-deletion mutant
Based on the finding that the sarZ gene suppressed the decreased haemolysin production in the cvfA mutant, we hypothesized that the cvfA gene activates the sarZ gene expression at the transcriptional level. We measured the amount of the sarZ transcript in the cvfA mutant by Northern blot analysis. A 0.9 kb band was detected in the parent strain (Fig. 2). The band disappeared in the sarZ-deletion mutant, indicating that the band was sarZ transcript. The amount of sarZ transcript in the cvfA mutant was decreased to the half of the parent strain. The decreased amount of the sarZ transcript in the cvfA mutant was restored by a plasmid harbouring the intact cvfA gene. These results indicate that the cvfA gene regulates the sarZ expression at the transcriptional level.
Decreased haemolysin production and attenuated virulence by the sarZ-deletion mutant
To determine whether the sarZ gene contributes to virulence in S. aureus, we constructed a sarZ-deletion mutant. The sarZ gene was inactivated by the integration of a suicide vector into the chromosomal region of the sarZ gene by homologous recombination. The disruption of the sarZ gene was confirmed by Southern blot analysis (data not shown). The growth rate of the sarZ-deletion mutant was indistinguishable from that of the parent strain (data not shown). To examine whether the sarZ mutant produced less haemolysin, the haemolytic zone formed by the sarZ mutant on sheep blood agar plates was compared with that formed by the parent strain. The sarZ mutant formed a much smaller haemolytic zone than did the parent strain (Fig. 3). In addition, the sarZ mutant transformed with the intact sarZ gene formed a haemolytic zone as large as that of the parent strain. These results indicate that the small haemolytic zone phenotype is caused by sarZ deletion. The attenuated haemolytic activity caused by sarZ deletion was also observed using rabbit erythrocytes (Fig. 9A), indicating that the sarZ gene is required for production of both α-haemolysin and β-haemolysin.
Next, we examined the contribution of the sarZ gene to virulence in S. aureus using a silkworm-infection model and a mouse-infection model. Fifty per cent of the silkworm larvae injected with the parent strain were dead at 16 h post infection, whereas 50% of the larvae injected with the sarZ mutant were dead at 32 h post infection (Fig. 4A). Introduction of the intact sarZ gene into the sarZ mutant restored the attenuated silkworm killing ability (Fig. 4B). The number of viable cells recovered from spleen and kidney of mice infected with the sarZ mutant were 10 times lower than that from mice infected with the parent strain (Fig. 5). The decrease in the viable cell number was complemented by a plasmid harbouring the intact sarZ gene (Fig. 5). These results suggest that the sarZ gene contributes to virulence of S. aureus.
Decreased amount of the hla and hlb transcripts and RNAIII in the sarZ mutant
α-Haemolysin and β-haemolysin are encoded by the hla gene and the hlb gene respectively. We examined whether the amount of the hla and hlb transcripts were decreased in the sarZ mutant. Northern blot analysis revealed that there were less hla and hlb transcripts in the sarZ mutant than in the parent strain (Fig. 6). In addition, the amounts of the hla and hlb transcripts were increased by a plasmid harbouring the intact sarZ gene. Thus, the sarZ gene regulates the transcriptional expression of the hla and hlb genes.
Our previous study revealed that the cvfA gene upregulates RNAIII expression of the agr locus (Kaito et al., 2005). Accordingly, we examined the effect of the sarZ gene, which was a multicopy suppressor of cvfA disruption, on the expression of the agr locus. Northern blot analysis indicated that the amount of RNAIII in the sarZ mutant was lower than the parent strain (Fig. 6). The amount of RNAIII was increased by a plasmid harbouring the intact sarZ gene. This result suggests that the sarZ gene activates RNAIII expression.
DNA binding activity of SarZ protein
SarZ protein has a wHTH motif, a DNA binding domain. We hypothesized that SarZ has DNA binding activity, and examined the hypothesis using purified SarZ protein. C-terminal 6× histidine-tagged recombinant SarZ protein was overproduced in E. coli and purified using a Ni-column. A single 17 kDa band was observed by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (Fig. 7A), indicating that the purity of SarZ protein was over 95%. The DNA binding activity of the recombinant SarZ protein was examined by electrophoretic mobility shift assay using the DNA fragment of the hla promoter region. The SarZ protein and DNA complex was detected as a slower migrating band than the free probe DNA (Fig. 7B). Addition of the antibody against 6× histidine decreased migration of the complex, whereas the addition of immunoglobulin G prepared from non-immunized mouse serum did not change the migration of the complex. Therefore, we concluded that SarZ protein has DNA binding activity.
To verify the nucleotide sequence specificity of the DNA binding by SarZ protein, we performed comparative electrophoretic mobility shift assays using promoter regions of hla (two different regions), asp23 and agr (Fig. 7C). SarZ protein formed a complex with each promoter region. The dissociation constant (KD) of SarZ protein with hla promoter, hla promoter (2), asp23 promoter and agr promoter was 32 pM, 24 pM, 50 pM and 13 pM respectively. We further performed competition assays against the complex formation between SarZ and labelled hla promoter using unlabelled promoter regions of hla, agr, asp23, spa, sarA, sarU, sarV and salmon sperm DNA and poly-dI-dC (Fig. 7D, hla, salmon sperm DNA, agr, asp23; others are not shown). The complex formation of 1.8 nM SarZ protein with 6 pM labelled hla promoter was reduced to less than half saturation at 1.5 nM by all the tested unlabelled DNA. These results suggest that the specific DNA binding was not clarified, although SarZ protein has a non-specific DNA binding activity. The size of the complexes between SarZ and each DNA fragment gradually increased as the SarZ amount increased (Fig. 7C), suggesting that SarZ polymerizes along the DNA.
Construction of SarZ mutant proteins with decreased DNA binding activity
To determine the SarZ structure required for the DNA binding activity, we constructed amino acid-substituted SarZ proteins. The V43, G67, T70, V82, R92 and I96 residues of SarZ protein were predicted to be important for DNA binding on the basis of the reported results from dominant negative marR mutants (Alekshun et al., 2000) and from the crystal structures of MarR and OhrR (Alekshun et al., 2001; Hong et al., 2005). In addition to the six amino acid residues, F14 and K25 from the N-terminal region, and K103, E124, L134 and K140 from the C-terminal region were subjected to an amino acid-substituted change on the basis that these regions are involved in dimerization of MarR (Alekshun et al., 2001; Notka et al., 2002). The resulting SarZ mutant proteins were examined for DNA binding activity (Fig. 8). F14A, K25A and V43M, which are mutant proteins of the N-terminal region, and G67E, T70A, R92C and I96A, which are mutant proteins of the wHTH motif, had KD values that were eight times greater than the wild-type SarZ protein, indicating that these mutant SarZ proteins had reduced DNA binding activity (Table 1). This result suggests that the N-terminal region and the wHTH motif of SarZ protein are important for the DNA binding activity.
Table 1. Affinity of SarZ protein for DNA.
KD values for DNA binding were calculated from the results obtained in Fig. 8 using the previously described method (Riggs et al., 1970).
Haemolysin production and virulence by the SarZ mutant proteins in S. aureus cells
Next, we studied whether the expression of the SarZ mutant proteins can complement the decreased haemolysin production in the sarZ mutant. The expression of wild-type SarZ protein restored haemolysin production in the sarZ mutant. In contrast, the expression of F14A, K25A, V43M, G67E, T70A, R92C and I96A whose KD values for DNA binding were eight times greater than that of wild-type SarZ protein did not restore haemolysin production in the sarZ mutant (Fig. 9A). The expression levels of these SarZ mutant proteins were comparable to that of wild-type SarZ protein (Fig. 9B). Therefore, SarZ mutant proteins with decreased DNA binding activity are not functional for haemolysin production in S. aureus. We then examined whether the expression of these SarZ mutant proteins can complement the attenuated virulence of the sarZ mutant using a silkworm-infection model. The expression of the wild-type SarZ protein restored the silkworm killing ability of the sarZ mutant, although the expression of the F14A, K25A, G67E, T70A, R92C and I96A mutants, whose KD values for DNA binding were more than 10 times that of wild-type SarZ protein, did not restore the silkworm killing ability (Fig. 10A and B). Thus, the SarZ mutant proteins with decreased DNA binding activity are also not functional for virulence in S. aureus.
Involvement of the cvfA and sarZ genes in haemolysin production in a strain other than RN4220
RN4220 strain is a mutagenized strain by nitrosoguanidine exposure to accept foreign DNA (Peng et al., 1988; Traber and Novick, 2006). There might be many uncharacterized mutations that make the involvement of the cvfA and sarZ gene in haemolysin production be a special event in this strain. Thus, we examined the role of the cvfA and sarZ genes in a non-mutagenized strain NCTC8325-4. Disruption of the cvfA gene in NCTC8325-4 caused a decrease of haemolysin production (Fig. 11A). The decreased haemolysin production was restored by a plasmid harbouring the intact cvfA gene. This result indicates that the cvfA gene is required for haemolysin production in this strain. To know the involvement of the sarZ gene in haemolysin production in NCTC8325-4, we transformed the cvfA mutant with pF2 harbouring the sarZ gene and examined the haemolysin production. The cvfA mutant transformed with pF2 showed a larger haemolytic zone than the mutant transformed with the vector plasmid (pSR515) (Fig. 11A). Furthermore, the sarZ-deletion mutant in NCTC8325-4 decreased the haemolytic activity compared with the parental strain (Fig. 11B). The decrease of haemolysin production was restored by a plasmid harbouring the intact sarZ gene. Therefore, the sarZ gene has the ability to suppress the decreased haemolysin production in the cvfA mutant and is involved in haemolysin production in a strain other than RN4220.
In this study, we identified the sarZ gene as a multicopy suppressor of decreased haemolysin production in the cvfA mutant. The decreased amount of the sarZ transcript in the cvfA mutant suggests that the cvfA gene activates the sarZ gene at the transcriptional level. Analysis of the sarZ-deletion mutant indicated that the sarZ gene contributes to haemolysin production and virulence in S. aureus. In addition, the sarZ gene contributes to the transcriptional expression of the hla gene, the hlb gene and the agr locus. Thus, the sarZ gene seems to be activated by the cvfA gene and contributes to virulence by activating virulence gene expression. The attenuation of silkworm killing ability caused by sarZ gene disruption was much weaker than that caused by cvfA gene disruption (Fig. 4A). Therefore, we assume that the sarZ gene has a partial role in the pathway in which the cvfA gene contributes to virulence. This novel regulatory cascade might interact with the known regulatory systems, such as the sarA family genes. The interaction between the cvfA and sarZ genes and the other regulatory genes is currently under investigation.
Biochemical analysis revealed that the SarZ protein has DNA binding activity. Analysis of amino acid-substituted SarZ proteins revealed that G67, T70, R92 and I96 are important for DNA binding activity. These amino acid residues exist in the wHTH motif of SarZ protein and are assumed to have a direct role in the DNA binding. F14 and K25, which are also important for DNA binding, exist in α-helix 1 (residues 8–33) in the N-terminal region of SarZ protein. Helix α1 of MarR has an important role in dimerization of MarR protein and the dimerization is important for DNA binding (Martin and Rosner, 1995; Notka et al., 2002). F14A and K25A SarZ proteins might be defective in dimerization, resulting in defective DNA binding. Expression of the SarZ mutant proteins with reduced DNA binding activity did not complement the decreased haemolysin production and attenuated virulence of the sarZ mutant. The results suggest that the DNA binding activity of SarZ protein is required for haemolysin production and virulence in S. aureus. MarR and OhrR, which have sequence identity with SarZ protein, recognize specific nucleotide sequences of DNA and function as a transcription factor against the genes to which they specifically bind (Martin and Rosner, 1995; Fuangthong and Helmann, 2002). DNA binding of SarZ protein against various promoters was examined by comparative and competitive gel-shift assays, although the preference on the specific DNA sequence was not demonstrated. SarZ protein might require a cofactor to recognize specific DNA sequence or the promoter region used in this study might not contain the recognition sequence by SarZ. It is also possible that our assay condition to evaluate DNA binding is not optimal. Further analysis of the mechanism how SarZ protein recognizes a specific DNA sequence will help to understand the regulatory mechanism of virulence genes in S. aureus.
Bacterial strains and growth conditions
The JM109 strain of E. coli was used as a host for pMutinT3, pHY300, pND50, pCK20, pET-9a and their derivatives. The BL21(DE3)pLysS strain of E. coli was used for overexpression of SarZ protein. E. coli strains transformed with the plasmids were cultured at 37°C in Luria–Bertani (LB) broth containing 100 μg ml−1 ampicillin, 50 μg ml−1 kanamycin or 12.5 μg ml−1 chloramphenicol. S. aureus strain RN4220 (de Azavedo et al., 1985) and derivative strains were used. S. aureus strains were aerobically cultured in tryptic soy broth at 37°C, and 10 μg ml−1 erythromycin, 5 μg ml−1 tetracycline and 12.5 μg ml−1 chloramphenicol were added to the medium to maintain the plasmids. Details of bacterial strains and plasmids used in this study are shown in Table 2. Primers used in this study are listed in Table S1.
Table 2. A list of bacterial strains and plasmids used.
Strain or plasmid
Genotypes or characteristics
Source or reference
Provided by Dr Keiichi Hiramatsu (Juntendo University, Tokyo, Japan).
Provided by Dr Timothy J. Foster (Moyne Institute of Preventive Medicine, Trinity College Dublin, Dublin 2, Ireland).
Provided by Dr Naotake Ogasawara (Nara Institute of Science and Technology, Nara, Japan).
Other point-mutated plasmids were constructed in the same way with F14A plasmid.
Erm, erythromycin; Amp, ampicillin; Tet, tetracycline; Cm, chloramphenicol; Kan, kanamycin.
Transformation of E. coli, extraction of plasmid DNA from E. coli, polymerase chain reaction (PCR), and Southern blot analyses were performed as previously reported (Sambrook et al., 1989). Extraction of genomic DNA from S. aureus was performed using a QIAamp DNA Blood Kit (Qiagen Sciences, Germantown, MD). Transformation by plasmid DNA of S. aureus was performed by electroporation (Inoue et al., 2001). To construct plasmids for the complementation experiments, the S. aureus sarZ gene was amplified by PCR using oligonucleotide primers FsarZ and RsarZ and RN4220 genomic DNA as the template. The amplified DNA was inserted into pHY300 at the BamHI and HindIII sites, resulting in pC2174. To construct plasmids expressing 6× histidine-tagged SarZ for complementation experiments, PCR amplification was performed using primers FcsarZ and RcsarZ, and pC2174 as the template. RcsarZ was designed to encode a 6× histidine-residue at the C-terminus of SarZ protein. The amplified DNA fragment was cloned into pHY300 at the BamHI and HindIII sites, resulting in pCHis-Wt. To construct the plasmid for overproduction of SarZ, PCR amplification was performed using primers FopsarZ and RopsarZ, and pCHis-Wt as the template. The amplified fragment was inserted into pET-9a, resulting in pHis-Wt. These plasmids were sequenced to confirm that the sarZ gene and 6× histidine residues were intact.
Site-directed mutagenesis was performed according to the method of Li and Wilkinson (1997). The sarZ gene amplified by PCR using primers FsarZ and RsarZ was cloned into pCK20, resulting in pCK20-sarZ. Using the primer pairs for mutagenesis, such as F-F14A and R-F14A, and pCK20-sarZ as the template, mutant strands were synthesized by thermal cycling, and then template plasmid was digested with DpnI. E. coli JM109 strain was transformed with the synthesized mutated plasmids (pCK20-mutant-sarZ). The plasmids were extracted and sequenced to confirm the desired sarZ point mutation. Using primer pairs, FcsarZ and RcsarZ, and the mutated plasmid as the template, PCR amplification was performed. The amplified DNA fragments were cloned into pHY300 at the BamHI and HindIII sites, resulting in plasmids that express C-terminal 6× histidine-tagged SarZ mutant proteins, such as pCHis-F14A, for complementation experiments. These plasmids were sequenced to confirm the desired sarZ point mutation and 6× histidine residues. Using primer pairs FopsarZ and RopsarZ and these plasmids as the template, PCR amplification was performed. The amplified DNA fragments were cloned into pET-9a (Novagen, Darmstadt, Germany), resulting in plasmids for overproduction of SarZ mutant proteins, such as pHis-F14A. These plasmids were also sequenced to confirm the desired sarZ point mutation. To construct the plasmids that encode non-histidine-tagged SarZ mutant proteins, the mutated sarZ gene was excised from pCK20-mutant-sarZ and subcloned into pHY300, resulting in the plasmid for the complementation experiments, such as pC-F14A.
Screening of multicopy suppressor
To identify a multicopy suppressor, the cvfA mutant was transformed with the S. aureus genome library (Matsuo et al., 2003) and spread on 5% sheep blood agar plates. The colonies that formed a large haemolytic zone compared with the cvfA mutant were isolated. The nucleotide sequences of the extracted DNA from the colonies were determined. Three open reading frames (SA2173, SA2174 and SA2175) on the DNA fragment were amplified by PCR using primers FSA2173 and RSA2173, FSA2174 and RSA2174, or FSA2175 and RSA2175. The amplified DNA fragments were inserted into pSR515 at the SmaI site, resulting in pF1, pF2 and pF3. The cvfA mutants that were transformed with pF1, pF2 and pF3 were spotted onto sheep blood agar plates, and the haemolytic zone around the colonies was observed after overnight incubation.
Construction of the sarZ-deletion mutant and the cvfA-deletion mutant
The internal region of the sarZ gene (+10 to +211, +1 is the translation initiation site) was amplified by PCR using oligonucleotide primers FtsarZ and RtsarZ. The amplified 202 bp fragment was inserted into pCK20, resulting in pT2174. A strain resistant to chloramphenicol was obtained by transformation of the RN4220 strain with pT2174, resulting in the M2174 strain. Disruption of the sarZ gene and integration of the targeting vector to the desired chromosomal locus were confirmed by Southern blot analysis using the PCR amplified region of the sarZ gene and pCK20 as probes. To construct the cvfA-deletion mutant in NCTC8325-4 strain, phage transduction was performed as described previously (Novick, 1991). Phage 80α lysates of M1129 strain was used to infect NCTC8325-4. Similarly plasmids were transferred to recipient strains by phage transduction.
Secretion of β-haemolysin by S. aureus was examined on tryptic soy broth agar containing 5% sheep blood. An aliquot (2 μl) of overnight bacterial culture was spotted onto the plates and incubated overnight at 37°C. To examine the production of α-haemolysin by S. aureus in liquid medium, supernatants of overnight cultures (centrifuged at 15 000 rpm, 3 min) were incubated with rabbit red blood cells at 37°C for 1 h (Vandenesch et al., 1991). Reaction mixtures were centrifuged (1000 rpm, 5 min) and the increase in the OD405 of the supernatant was determined. The activity was expressed by haemolytic units corresponding to the reciprocal of the dilution of supernatant that yielded 50% lysis of erythrocytes.
Northern blot analysis
Staphylococcus aureus cells were treated with 200 μg ml−1 lysostaphin for 5 min and total RNA was extracted using an RNeasy Protect Bacteria Kit (Qiagen). RNA 2.5 μg was electrophoresed in a 1.2% agarose gel containing 6.6 M formaldehyde and transferred to a nitrocellulose membrane (GeneScreen Plus, Perkin Elmer Life Sciences, Wellesley, MA). DNA fragments of the sarZ gene, the hla gene, the hlb gene and RNAIII were amplified by PCR using primers listed in Table S1, labelled with [32P]-dCTP by random priming, and used as probes. Hybridization was performed at 42°C. To detect the sarZ transcript, total RNA extracted from the bacterial cells at the log phase (OD600 = 1) was used. To detect the hla and hlb transcripts and RNAIII, total RNA extracted from the bacterial cells at the late exponential phase (OD600 = 4) was used.
Purification of SarZ protein
Plasmid pHis-Wt was transformed into E. coli BL21(DE3)pLysS. Cultures were grown in LB broth containing 50 μg ml−1 kanamycin at 37°C to an OD600 = 0.4. SarZ overexpression was induced with isopropyl-1-thio-β-D-galactopyranoside for 4 h. After induction, bacterial cells were harvested, resuspended in lysis buffer [20 mM Tris-HCl (pH 7.9), 0.5 M NaCl, 5 mM imidazole], and lysed by freezing and thawing. Cellular debris were removed by centrifugation at 45 000 rpm for 1 h and the clear supernatant was filtrated using a 0.22 μm polyvinylidene difluoride membrane (Millipore Corporation, Tokyo, Japan). The filtrated sample was applied to a nickel-chelating resin (Invitrogen Corporation, Carlsbad, CA) according to manufacturer's instructions. The protein was eluted with elution buffer [20 mM Tris-HCl (pH 7.9), 0.5 M NaCl, 1 M imidazole] and dialysed in a buffer [20 mM Tris-HCl (pH 7.9), 50 mM NaCl, 1 mM EDTA, 5% glycerol and 1 mM dithiothreitol]. The purity of SarZ was established by sodium dodecyl sulphate-polyacrylamide gel electrophoresis and Coomassie brilliant blue staining. The SarZ mutant proteins were purified by the same method as described above using the plasmids for overproduction, such as pHis-F14.
Electrophoretic mobility shift assay
The 235 bp region of the hla promoter that contains 155 bp upstream from the translation initiation site and 80 bp downstream from the translation initiation site was amplified by PCR using primers FhlaP and RhlaP. The resulting DNA fragment was labelled with [γ32P]-dATP using T4-polynucleotide kinase. The labelled DNA fragment (6 pM) was incubated with various concentrations of SarZ protein in 25 μl binding buffer [10 mM Tris-HCl (pH 7.9), 50 mM NaCl, 5% glycerol, 2 mM dithiothreitol, 0.1 mg ml−1 BSA, 0.005% Triton-X100, 1 mM EDTA] at room temperature for 30 min. The samples were electrophoresed in a non-denaturing 5% polyacrylamide gel in 0.5× TBE buffer [45 mM Tris borate (pH 8.3), 1 mM Na2EDTA]. The gels were dried, and protein–DNA interactions were analysed by phosphoimaging using BAS-1800II (Fujifilm, Tokyo, Japan) and Image gauge v. 4.23 software (Fujifilm). The experiments were performed at least two times for wild-type SarZ protein and each SarZ mutant protein and mean values were used. The concentration of bound DNA was calculated and plotted against the concentration of SarZ, and the protein concentration that caused 50% of the DNA to be bound was determined. To determine the activity of the purified SarZ, we used the equation KD = [P]1/2−1/2[DNA]0, where [P]1/2 is the protein concentration at 50% saturation, and [DNA]0 is the total DNA concentration (Riggs et al., 1970). For the super shift assay, mouse monoclonal antibody against polyhistidine (Nacalai Tesque, Kyoto, Japan) or mouse immunoglobulin G (Dakocytomation, Glostrup, Denmark) was used. For comparative gel-shift assay, the two regions of hla (−155 to +80 and −470 to −25, +1 as the translational initiation site), agrP2P3 (281 bp), and asp23 (257 bp) were amplified by PCR using primer pairs listed in Table S1. For competition assay, the DNA fragments containing promoter regions of spa (338 bp), asp23 (257 bp), sarU (221 bp), sarV (250 bp), agrP2P3 (281 bp), sarAP1 (329 bp), sarAP2 (196 bp), sarAP3 (162 bp) was amplified by PCR using primer pairs listed Table S1 according to previous studies (Chien et al., 1999; Rechtin et al., 1999; Manna and Cheung, 2001; 2003; Manna et al., 2004).
Western blot analysis
Proteins were electrophoresed by sodium dodecyl sulphate-polyacrylamide gel electrophoresis and transferred onto a polyvinylidene difluoride membrane (Millipore Corporation) and probed with mouse anti-polyhistidine monoclonal antibody conjugated with peroxidase (Nacalai Tesque). Immunoactive bands were detected using Western Lightning (Perkin Elmer Life Sciences, Wellesley, MA).
Silkworm infection experiment
The silkworm infection experiment was performed according to the previously established method (Kaito et al., 2005). Silkworms were raised from fertilized eggs to fifth-instar larvae in our laboratory. The hatched fifth-instar larvae were fed antibiotic-free artificial food Silkmate (Katakura Industries, Japan) for 1 day. Bacterial suspensions (0.05 ml) were injected into the haemolymph of the larvae through the dorsal surface using a 27-gauge needle (Terumo, Tokyo, Japan). Overnight S. aureus cultures were diluted appropriately with saline and used for the experiment. The injected larvae were maintained without food in a safety cabinet (BHC-1303IIA; Airtech Japan, Tokyo, Japan) at 27°C with 50% humidity and survival was monitored. Statistical analyses of the survival curves were performed using one-sided rank log tests (PRISM software package, Graphpad Software, San Diego, CA). P-values of less than 0.05 were considered to be statistically significant.
Mouse infection experiment
The experiment of staphylococcal systemic infection disseminated to the spleen and kidney was performed according to the previously established method (Shi et al., 2004). Bacterial suspensions (100 μl, 4 × 108 cfu) were injected into the tail vein of 8-week-old female CD-1 mice (Charles River Laboratories, Kanagawa, Japan). At 24 h post infection, animals were killed, and spleens and kidneys were harvested and homogenized in phosphate buffered saline. After appropriate dilution, the samples were spread on tryptic soy agar plates, incubated overnight, and the number of colonies was counted. Under this assay condition, mice were not killed by bacterial injection.
We thank Drs T.J. Foster, K. Hiramatsu and N. Ogasawara for kindly providing bacterial strains and plasmids. This study was supported in part by research grants from Japan Society for the Promotion of Science (JSPS). C.K. was the recipient of a predoctoral fellowship from JSPS.