| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Biological Sciences, The University of Southern Mississippi, Hattiesburg, MS 39406, USA
2 Department of Microbiology and Immunology, University of Arkansas for Medical Sciences, Little Rock, AR 72205, USA
Correspondence
Mohamed O. Elasri
mohamed.elasri{at}usm.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Expression of agr is activated in a cell-density-dependent manner as cultures enter the post-exponential growth phase. Induction of agr expression leads to RNAIII production, which results in decreased production of many surface proteins and increased production of many exoproteins (Janzon & Arvidson, 1990
; Novick et al., 1993). In addition to cell density, agr expression is also activated by sarA. The sarA locus encodes three transcripts (sarP1, 0.58 kb; sarP3, 0.84 kb; and sarP2, 1.15 kb), all of which have an identical 3' end and include the ORF (375 bp) encoding the 14.5 kDa DNA-binding protein SarA (Bayer et al., 1996). Transcripts sarP1 and sarP2 use 
A-specific promoters and are preferentially expressed from mid- to late-exponential growth phases. Transcript sarP3 uses a B-specific promoter and is expressed primarily during the post-exponential growth phase (Manna et al., 1998). There are conflicting reports regarding the impact of transcription from different sarA promoters on the production of SarA, but our studies suggest that SarA is produced in similar amounts during all growth phases (Blevins et al., 2002). SarA regulates the expression of S. aureus virulence factors via both agr-dependent (Cheung et al., 1997; Dunman et al., 2001; Heinrichs et al., 1996) and agr-independent pathways (Bayer et al., 1996; Chan & Foster, 1998a; Blevins et al., 1999). In the agr-dependent pathway, sarA activates transcription of the agr locus at the transition between the exponential and post-exponential growth phase in vitro (Chien et al., 1998; Morfeldt et al., 1996) by binding to the intergenic region between the P2 and P3 promoters (Rechtin et al., 1999). In the agr-independent pathway, sarA-mediated regulation involves a direct interaction between SarA and cis elements associated with the target genes (Blevins et al., 1999; Wolz et al., 2000; Chan & Foster, 1998a; Dunman et al., 2001; Sterba et al., 2003).
Expression of sarA is regulated by several factors. For instance, some studies suggested that B activates sarA expression (Manna et al., 1998; Bischoff et al., 2001); however, Horsburgh et al. (2002) showed no difference in the level of sarA transcript or protein between wild-type and a B mutant. The reasons for this discrepancy are not clear but it might be due to differences in the strains or growth conditions used in these studies. Another study showed that expression of sarA is repressed by SarR, a member of the SarA protein family (Manna et al., 1998). More recently, Rossi et al. (2003) identified a membrane-associated protein, MrsR, which also affects expression of sarA. Specifically, mutation of mrsA resulted in increased expression of SarA. Since MrsR does not have any DNA-binding domain, the authors suggested that its effect on sarA is mediated by other factors.
The present study was prompted by several findings that indicated the presence of accessory elements that modulate the production and/or function of sarA. For instance, while SarA activates agr in growing cells, it inhibited transcription from agr promoters in an in vitro transcription system (Chakrabarti & Misra, 2000). In addition, our studies indicate that SarA is present in equivalent amounts throughout the growth cycle despite the fact that it modulates transcription of the genes encoding S. aureus virulence factors (e.g. cna) in a temporal fashion (Blevins et al., 1999). In this report, we identify and characterize a new gene that encodes a putative membrane protein that we designate msa (modulator of sarA). Mutation of msa resulted in reduced transcription of sarA and modulated the production of several virulence factors.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
. S. aureus cells were grown at 37 °C with constant aeration (200 r.p.m. at a medium/flask volume ratio of 0.5) in tryptic soy broth (TSB) or on tryptic soy agar (TSA) supplemented with antibiotics when appropriate. Escherichia coli was also grown in TSB or on TSA. S. aureus RN4220 was used as an initial recipient for the transformation of recombinant plasmids. Phage 
11 was used as a generalized transducing phage to move plasmids from RN4220 into other S. aureus strains. Antibiotics were used in the following concentrations: for S. aureus, 10 µg erythromycin ml–1, 50 µg kanamycin ml–1, 3 µg tetracycline ml–1, 10 µg chloramphenicol ml–1; for E. coli, 100 µg ampicillin ml–1.
|
). The cna promoter in UAMS-1 was amplified by PCR using primers B1 and B3 respectively, located at positions –475 and 1596 relative to the ATG start codon (Table 2; Snodgrass et al., 1999). The PCR product was treated with Klenow and polynucleotide kinase to produce phosphorylated blunt ends. The resulting product was then cloned into the SmaI site upstream of the lux operon carried in the shuttle vector pMK4 luxABCDE. Cloning was done in E. coli strain TOP10. The resulting clones were sequenced to confirm the correct orientation of the cna promoter relative to the lux operon. Fusion of the cna promoter to the lux operon (cna-luxABCDE) resulted in the reporter plasmid pMOE59, which was introduced into strain RN4220 by electroporation. The plasmid was then introduced into RN6390 by generalized transduction using phage 11 as described previously (Blevins et al., 2002). The resulting strain was designated MOE72 and used to generate a transposon-insertion library as described below.
|
). The Tn551 insertion in MOE41 was then introduced into strains UAMS-1 and UAMS-174 via generalized transduction as described above. Mutation of msa in the recipient strains was confirmed by PCR and verified by Southern blot analysis as previously described (Gillaspy et al., 1995).
|
Complementation of msa mutant.
To complement the mutation, a 750 bp fragment encompassing the Tn551 insertion site was amplified by PCR from S. aureus RN6390 chromosomal DNA (Table 2
). Plasmid pMOE78 was constructed by cloning the PCR products into the multiple cloning site of vector pCR2.1 TOPO (Invitrogen). Cloning of msa was verified by PCR and restriction enzyme analysis. pMOE78 was ligated to the Gram-positive shuttle vector pSK265, resulting in shuttle plasmid construct pMOE83. pMOE83 was introduced into strain RN4220 by electroporation. Transformants were selected for on TSA plates containing 10 µg chloramphenicol ml–1. The plasmid was subsequently introduced into the RN6390 and UAMS-1 msa mutants via generalized transduction (Gillaspy et al., 1998).
Isolation of RNA.
For the isolation of RNA, S. aureus strains were grown without antibiotic selection and under low-aeration conditions (120 r.p.m. at a medium/flask volume ratio of 0.5) as described by Lindsay & Foster (1999). Cells were harvested at optical densities (OD560) of 0.3, 1.5 and 4.0, which correspond to the mid-exponential, late-exponential, and post-exponential growth phases respectively. Total cellular RNA was isolated from the wild-type, the msa mutants, and the complemented mutants of both RN6390 and UAMS-1. Briefly, cells were harvested by centrifugation (5000 g for 5 min at 4 °C) and resuspended in TES buffer containing 100 µg lysostaphin ml–1 (AMBI). The samples were incubated at room temperature for 10 min and were then applied to a Qiagen RNeasy Maxi column to isolate total bacterial RNA according to the manufacturer's directions. The optional on-column RNase-free DNase I (Qiagen) was used to remove contaminating DNA. After isolation of RNA, traces of contaminating DNA were further eliminated by treating RNA samples with RNase-free DNase I (DNA-free kit, Ambion) and incubating at 37 °C for 20 min. Samples were used immediately or stored at –80 °C. The quality, integrity and concentration of the RNA were determined by using an Agilent 2100 Bioanalyzer (Agilent Technologies) as described by the manufacturer. RNA preparations were tested for contaminating DNA by no-reverse-transcriptase PCR reactions.
Real-time quantitative PCR.
The primers used for real-time quantitative PCR (qPCR) were designed with Primer 3 software (Massachusetts Institute of Technology) to amplify gene fragments with an optimal size of 75–100 bp. The fragment of the gene of interest was cloned into plasmid pCR2.1 TOPO to determine PCR efficiency. Specifically, the plasmid was isolated and a series of 10-fold dilutions was prepared corresponding to 500 000 to 5 plasmid copies. For each primer set, a standard curve was generated to determine the correlation coefficient as an indicator of PCR efficiency. All qPCR reactions were run in triplicate. Melting curve analysis and agarose gel electrophoresis were done to verify primer set specificity. Sequences, correlation coefficient values and PCR efficiencies of primer sets used are listed in Table 2.
Measurements of relative levels of gene expression were done by qPCR. RNA was reverse transcribed into cDNA using iScript cDNA synthesis kit (Bio-Rad). The reverse transcriptase reactions were done at 25 °C for 5 min, 42 °C for 30 min and 85 °C for 5 min. cDNA was stored at –20 °C until needed. PCR reactions were done in 25 µl reactions by using iQ SYBR Green Supermix (Bio-Rad) as recommended by the manufacturer (Bio-Rad). The reaction mixtures contained: 5 µl cDNA; 12.5 µl iQ SYBR Green Supermix; 0.5 µl forward primer (1.5 pmol µl–1); 0.5 µl reverse primer (1.5 pmol µl–1); and 6.5 µl de-ionized H2O. PCR amplification was done using an iCycler (Bio-Rad) and the amplification parameters were as follows: 94 °C for 3 min, 1 cycle at 95 °C for 3 min, 40 cycles at 95 °C for 30 s, 50 °C for 30 s and 72 °C for 1 min. The final 80 cycles began at 55 °C and increased by 0.5 °C every 10 s. All qPCR reactions were done in triplicate and the mean CT was used for analysis of results. To verify the absence of contaminating DNA, each qPCR experiment included controls that lacked template cDNA or reverse transcriptase. The constitutively expressed gene for gyrase (gyr) was used as an endogenous control as described previously (Goerke et al., 2000). Analysis of expression of each gene was done based on at least two independent experiments. Twofold or higher changes in gene expression were considered significant.
Protease assay.
Assays for soluble proteases were done as previously described (Smeltzer et al., 1993). In brief, 300 µl of the culture supernatant was mixed with 800 µl of 3 mg azocasein ml–1 in Tris-buffered saline (pH 7.5) and incubated overnight at 37 °C. To precipitate the undegraded azocasein, 400 µl of 50 % (w/v) trichloroacetic acid was added. The precipitate was then removed by centrifugation and the amount of acid-soluble azocasein was determined by measuring A340.
Haemolysin assay.
Haemolytic activity was measured as described previously (Blevins et al., 2002). In brief, supernatants from overnight cultures were harvested and filter-sterilized with a 0.45 µm (pore-size) acetate syringe filter (CAMEO 25AS, Osmonics). After standardization, 10 µl culture supernatant was combined with 1 % rabbit blood in 10 mM Tris/HCl (pH 7.5)/0.9 % NaCl. After incubating 15 min at 37 °C, unlysed blood cells were pelleted by centrifugation. The haemolytic activity was determined by measuring the A405 of the cell-free supernatant. SDS (1 %) and TSB were used as positive and negative controls respectively.
Binding assays.
Binding of S. aureus to host proteins (fibronectin, fibrinogen or collagen) was assayed in microtitre plates coated with human fibronectin (BD Biosciences), type 1 fibrinogen from human plasma (Sigma-Aldrich,) or type 1 collagen (Sigma). Fibronectin was suspended in sterile distilled water to a starting concentration of 1 mg ml–1. Fibrinogen was suspended in phosphate-buffered saline (PBS) to a starting concentration of 15 µg ml–1. Collagen was suspended in PBS to a starting concentration of 1 mg ml–1. The host protein solutions were diluted 1 : 10 in bicarbonate buffer and 200 µl was used to coat the microtitre plate wells. A solution of BSA (2 %) was used as a negative control. After incubating overnight at 4 °C, wells were washed with PBS, and 200 µl of the test culture containing 0.5 optical density units (OD560) was added to the wells and incubated at 37 °C for 1 h. The plates were then spun at 200 g in a Jouan T20 rotor for 5 min at room temperature. The absence of cell pellets indicated binding to the host molecule bound to the walls of the wells. The presence of the cell pellet indicated lack of binding. Assays were done in triplicate and were repeated at least three times.
Western blot analysis.
Whole-cell lysates used for SarA Western blots were prepared and analysed as previously described (Blevins et al., 1999). Proteins (2 µg) for Western blot analysis were electrophoresed under denaturing conditions on precast 4–20 % gradient gels. Polyclonal anti-SarA from rabbits was used as described by Blevins et al. (1999).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
), light emission was increased relative to the wild-type parent strain when the cna-luxABCDE reporter was present in a sarA mutant (Fig. 1). RN6390 carrying the cna-luxABCDE reporter was randomly mutagenized using Tn551. Mutant libraries were screened for enhanced light production as an indication of reduced SarA activity. Once clones with enhanced bioluminescence were identified, we used PCR analysis to ensure that the insertion did not occur in sarA or agr. We also did Southern hybridization using an internal fragment of Tn551 as a probe to verify that the mutant contained a single insertion. Characterization of one such clone and comparison to the N315 genome sequence confirmed that the Tn551 insertion occurred at position 197 inside an ORF (SA1233) with no known function. SA1233 is a 402 bp ORF that encodes a protein of 133 amino acids with a predicted molecular mass of 15.6 kDa (Fig. 2a). Since this ORF and its product have not been characterized, we analysed the predicted protein sequence to gain insight into its putative function.
|
) to search for SA1233 homologues in a non-redundant database of protein sequences. We did not find any sequences that showed full-length similarity. However, a small portion of the sequence corresponding to residues 9 through 53 showed strong similarity to efflux and membrane proteins related to drug resistance. Similar results were obtained when BLAST searches were performed against the SwissProt Database (SP) and Protein Database (PDB). To confirm these results, we performed a multiple sequence alignment using the BLASTP hit sequences in the CLUSTALX environment (Thompson et al., 1997). This analysis showed that the N-terminal region (55 residues) is conserved among many eubacterial species. We also did a conserved domain search on the N-terminal region against the PFAM database (Marchler-Bauer & Bryant, 2004). This search showed the presence of an ACR_tran domain (e-value 0.91), suggesting that SA1233 belongs to the integral membrane protein family AcrB/AcrD/AcrF (pfam00873) (Fig. 2b). Further analysis of the protein sequence supported these findings by revealing the presence of hydrophobic regions and three potential transmembrane helices (Fig. 2c). These results suggest that SA1233 is a transmembrane protein that might be involved in drug resistance.
Impact of msa mutation on expression of sarA
The goal of this study was to find accessory elements that are necessary for the production or function of SarA. The reporter system (cna-luxABCDE) indicated that mutation of msa has a similar effect to mutation of sarA with respect to expression of cna. Specifically, mutation of both sarA and msa resulted in increased expression of cna (Fig. 1). To determine if the effect of msa on transcription from the cna promoter is mediated through an impact on sarA expression, we measured sarA transcription by qPCR in strains carrying the mutant msa gene. This analysis was done in the laboratory strain RN6390 and the clinical isolate UAMS-1 because previous studies showed that the regulatory functions of sarA are strain-dependent (Blevins et al., 2002; Somerville et al., 2002; Rice et al. 2004; Cassat et al., 2005). Mutation of msa resulted in a twofold and 2.85-fold decrease in sarA expression in strains RN6390 and UAMS-1 respectively (Fig. 3). Complementation studies with pMOE83 confirmed that the decrease was due to disruption of msa (Fig. 3). The decrease in sarA expression was observed in all three growth phases tested (mid-exponential, late-exponential and post-exponential) and confirmed by Western blot analysis (Fig. 4). This is consistent with the phenotype observed in the reporter strain and suggests that increased cna expression is correlated with derepression by sarA. These findings suggest that mutation of msa affects the expression of sarA at the transcriptional level in both RN6390 and UAMS-1.
|
|
). In addition, mutation of msa caused a slight reduction in RNAII expression but the reduction was not significant (less than twofold). The pattern of temporal expression of agr operons in the mutant and RN6390 was similar in that RNAIII accumulated in the post-exponential growth phase. Complementation with pMOE83 restored expression of both RNAIII and RNAII (Fig. 5).
|
). Expression of RNAII was also significantly higher in the UAMS-1 msa mutant, but only in the late-exponential growth phase. Complementation with the functional copy of msa restored expression to levels similar to wild-type, indicating that the mutation of msa is responsible for the observed changes in expression of agr operons. These results demonstrate that the effect of the msa mutation on agr is strain-dependent. The amounts of RNAIII and RNAII transcripts produced by UAMS-1 were much lower than RN6390, which is consistent with previous findings (Blevins et al., 2002). It should also be noted that cna is regulated by sarA in an agr-independent manner (Blevins et al., 1999). Additionally, induction of agr is classically associated with decreased production of surface proteins (Novick et al., 1993), but mutation of msa in UAMS-1 resulted in increased expression of agr and increased expression of the cna-luxABCDE reporter. Taken together, these results suggest that the differences we observed in expression of agr do not account for the increased expression of the cna-luxABCDE reporter in the msa mutant.
Impact of msa mutation on expression of virulence factors
We have shown that mutation of msa negatively affects transcription of sarA. We tested the effect of the msa mutation on expression of the genes encoding several virulence factors that are controlled by sarA (Tables 3 and 4). Dunman et al. (2001) developed a genome-scale transcription profile to identify genes regulated either directly or indirectly by sarA. We used this profile to select the sarA-regulated virulence genes to investigate in this study (Tables 3 and 4). Expression of virulence factors was measured in RN6390 and UAMS-1. In both strains, mutation of msa resulted in decreased expression of the genes encoding fibronectin-binding protein A (fnbA) (Tables 3 and 4). On the other hand, mutation of msa resulted in increased expression of the genes encoding aureolysin (aur) and serine protease (sspA) in both strains (Tables 3 and 4). The fact that mutation of msa had a similar effect on expression of these genes in both RN6390 and UAMS-1, together with the observation that mutation of msa had a similar effect on expression of sarA in both strains but an opposite effect on expression of agr, suggests that the altered transcription of these genes was mediated by the impact on expression of sarA rather than agr. However, the impact of the msa mutation on expression of other genes was strain-dependent. For instance, mutation of msa resulted in reduced expression of alpha toxin (hla) in RN6390 but increased expression in UAMS-1. This is also consistent with earlier reports describing the impact of mutation of sarA. Similarly, while the msa mutation affected the transcription of clumping factor (clfA) in UAMS-1, it did not have a significant effect on expression of this gene in RN6390. Conversely, transcription of protein A (spa) was reduced by mutation of msa in RN6390 but was not significantly changed in UAMS-1. Complementation studies with pMOE83 confirmed that the msa mutation was responsible for all these transcriptional changes (Tables 3 and 4). The differences in transcription profiles between the two strains highlight the importance of examining the effect of regulatory genes such as msa in clinical isolates.
|
|
). Complementation with pMOE83 reduced protease activity to wild-type levels (Fig. 6). As previously shown in UAMS-1, the sarA mutant showed an increase in proteolytic activity (approx. ninefold), the agr mutant showed no significant change in proteolytic activity, whereas the sarA/agr double mutant showed an intermediate level (approx. fivefold) of proteolytic activity in comparison to the corresponding wild-type strain (Fig. 6). Previous results from our group suggest that the serine protease encoded by sspA is the primary protease produced by UAMS-1 (Blevins et al., 2002). However, the protease assay we used will detect the activity of aureolysin and serine (V8) protease, and the fact that expression of both of these genes was increased in the msa mutant suggests that both these proteases may contribute to this phenotype. Mutation of msa had a less dramatic impact on overall proteolytic activity than mutation of sarA, suggesting that altered expression of other loci that either encode additional proteases or encode regulatory factors that modulate expression of other protease genes may also be involved (Tables 3 and 4).
|
). Complementation with pMOE83 returned haemolytic activity to near wild-type levels, indicating that the phenotype was due to mutation of msa (Fig. 7). These results are consistent with our transcriptional data (Tables 3 and 4) and with previous experiments demonstrating that mutation of sarA has a strain-dependent effect on haemolytic activity (Blevins et al., 2002). However, UAMS-1 was recently shown to carry a nonsense mutation in hla, rendering it unable to make alpha-toxin (James E. Cassat, personal communication). This suggests that the increased haemolytic activity observed in the msa mutant is a function of some other haemolytic toxin.
|
). As was observed with changes in haemolytic and proteolytic activity, complementation with pMOE83 restored the ability to bind both proteins (Table 5). Results of binding assays from the corresponding sarA and agr mutants were consistent with previous studies (Blevins et al., 2002). The decreased ability to bind fibronectin was consistent with our transcriptional analysis showing decreased fnbA transcription in the msa mutants (Tables 3 and 4). However, the ability to bind fibronectin is very sensitive to protease degradation (McGavin et al., 1997; Karlsson et al., 2001), so the increased proteolytic activity observed in msa mutants is also likely to contribute to this phenotype. Results of fibrinogen-binding assays were consistent with transcription analysis in UAMS-1. However, in RN6390, the msa mutant failed to bind fibrinogen despite the fact that there was no significant change in clfA transcription in the RN6390 msa mutant. This is consistent with the hypothesis that increased proteolytic activity may also contribute to the reduced capacity to bind fibrinogen (McAleese et al., 2001).
|
). Given the limited sensitivity of the microtitre plate assay employed in these experiments, we could not demonstrate collagen binding with UAMS-1 or UAMS-174. However, we did demonstrate that mutation of msa in both strains resulted a detectable level of collagen binding (Table 5). Once again, complementation with pMOE83 restored the wild-type phenotype (Table 5). These results are consistent with the results showing derepression of cna transcription in UAMS-1 in the post-exponential growth phase. They are also consistent with the fact that the msa mutant was isolated based on a phenotype defined using a reporter fusion with the cna promoter (cna-luxABCDE).
Temporal pattern of msa transcription
We also measured the temporal pattern of msa expression in both RN6390 and UAMS-1. The levels of transcription in all three growth phases for both strains was compared to that of strain RN6390 at mid-exponential phase. In both strains, the msa transcript level increased to a maximum in the transition from mid-exponential to late-exponential growth phase. As cells entered the post-exponential growth phase, the amount of msa transcript decreased to levels similar to mid-exponential phase. Strain RN6390 produced less msa transcript relative to UAMS-1 during mid-exponential and post-exponential growth phases but produced significantly more msa transcript during late-exponential growth phase. As expected, the msa mutants in both strains showed no transcription of msa at any growth phase (data not shown).
Impact of msa on expression of neighbouring genes
To characterize the potential effect of the mutation of msa on neighbouring genes, we analysed the expression of the two genes, lysA (SA1232) and cspA (SA1234), that flank msa. lysA is located downstream of msa and encodes diaminopimelate decarboxylase. It is the last gene of the dap operon, which is involved in lysine biosynthesis and is transcribed in the opposite direction with respect to msa (Fig. 2; Wiltshire & Foster, 2001). The gene upstream of msa, cspA (SA1234), encodes a cold-shock protein and is transcribed in the same direction as msa. In both strains, transcription of lysA and cspA was negatively affected by msa mutation. Transcription of cspA was practically abolished in the msa mutants in both RN6390 and UAMS-1 (Tables 3 and 4). Complementation with the pMOE83 resulted in re-establishment of expression to near wild-type levels, indicating the msa mutation was responsible for the phenotype. The abolition of cspA transcription brings up the possibility that some or all the phenotypic changes observed in the msa mutation are mediated by the impact of msa on cspA transcription rather than a direct effect on sarA. Further analysis is necessary to evaluate the relative contribution of cspA; however, the facts that introduction of pMOE83 complemented the defect and that cspA is upstream of msa indicate that the change in expression of the neighbouring genes is not due to a polar effect resulting from the transposon insertion in msa.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
; Bronner et al., 2004 for reviews), the agr and sarA loci clearly play central roles. We are particularly interested in the sarA locus based on studies demonstrating that it is required for biofilm formation and contributes to virulence in several animal models of staphylococcal infection (Beenken et al., 2003; Valle et al., 2003; Blevins et al., 2003; Booth et al., 1997; Cheung et al., 1994; Nilsson et al., 1996; Sifri et al., 2003). However, several lines of evidence suggest that sarA requires as yet unidentified accessory elements that modulate the transcription of sarA and/or the function SarA. For instance, mutation of sarA results in increased expression of some genes (e.g. cna) and decreased expression of others (e.g. fnbA). In addition, while mutation of sarA results in decreased expression of agr, a study using an in vitro transcription system showed that purified SarA represses transcription from agr promoters (Chakrabarti & Misra, 2000). This clearly suggests the existence of an accessory factor that is present in growing cells but was not present in the in vitro transcription system. Finally, SarA is present in roughly equivalent amounts throughout the growth cycle but represses or activates its target genes in a temporal manner (Blevins et al., 1999; Chien et al., 1999; Manna et al., 1998).
To identify genes that modulate the activity of sarA, we developed a luminescent reporter system (cna-luxABCDE) to screen for sarA-related defects. We used this reporter system to screen an RN6390 transposon-insertion library for mutants, and selected those mutants that showed a sarA phenotype in a sarA-positive background. We used the cna promoter as a reporter because regulation of the collagen-binding adhesin by sarA is well characterized and is independent of agr, thus limiting our search to genes that have a direct impact on sarA (Blevins et al., 1999). We found that mutation of msa led to increased expression from the cna promoter despite the presence of a functional sarA locus. In addition, mutation of msa had a negative impact on sarA expression in the laboratory strain RN6390 and the clinical isolate UAMS-1, suggesting that msa is important for wild-type levels of sarA transcription. We used both strains in this study because it has become evident that regulation of virulence in RN6390 is not representative of clinical isolates (Blevins et al., 2002; Somerville et al., 2002; Rice et al., 2004; Cassat et al., 2005). We used the osteomyelitis isolate UAMS-1 because it is closely related to other prominent clinical strains and, unlike RN6390, has an rsbU locus and consequently a functional sigB regulon (Cassat et al., 2005).
Since SarA activates transcription of agr, we expected that the decrease in sarA expression caused by mutation of msa would result in decreased expression of agr. However, while the onset of RNAIII expression in the RN6390 msa mutant was delayed, the amount of RNAIII in the post-exponential growth phase was comparable to that of the parent strain. In contrast, mutation of msa in UAMS-1 caused a transient increase in RNAIII expression in the mid-exponential and late-exponential growth phases. The reasons for this strain-dependent difference are unclear. However, RN6390 produces significantly more RNAIII than UAMS-1, and it is possible that the increase observed in UAMS-1 was simply masked in RN6390 given its already high level of agr expression. It is also unclear whether this difference is a function of the impact of msa on sarA expression or involves additional differences between the two strains. For instance, it is well known that all 8325-4 derivatives, including RN6390, carry a defect in rsbU, which leads to reduced production of the alternative sigma factor B (B) (Kullik et al., 1998). Additionally, Cassat et al. (2005) have demonstrated that sarT and sarU, both of which are reported to influence agr expression via a sarA-dependent mechanism (Schmidt et al., 2003), are absent in UAMS-1.
Given the impact of msa on the major global regulator sarA, we also examined the impact of mutating msa on expression of virulence factor genes known to be modulated by sarA. Expression of some of the virulence factors tested (aur, sspA and fnbA) showed a similar response to mutation of msa in both RN6390 and UAMS-1 while expression of other genes (hla, clfA and spa) was altered in a strain-dependent manner (Tables 3 and 4). As with agr, the reasons for these strain-dependent differences are not completely understood, but the results are generally consistent with previous findings showing that mutation of sarA has a disparate effect in RN6390 and clinical isolates including UAMS-1 (Blevins et al., 2002; Somerville et al., 2002; Rice et al., 2004; Cassat et al., 2005). A specific example is expression of hla, which is decreased in an RN6390 sarA mutant but increased in sarA mutants generated in UAMS-1 and other clinical isolates (Blevins et al., 2002). The fact that mutation of msa resulted in reduced expression of hla in RN6390 but increased expression of hla in UAMS-1 is therefore consistent with the hypothesis that the impact of msa on hla expression is mediated through its impact on expression of sarA. The absence of sarT and sarU may also be relevant in that regard. Specifically, Schmidt et al. (2003) showed that SarA represses sarT, which in turn represses hla. Mutation of sarA in a strain that encodes sarT (e.g. RN6390) would therefore be expected to result in decreased expression of hla due to derepression of sarT. However, other investigators have reported that SarA is capable of binding the hla promoter directly (Chien et al., 1999; Chan & Foster, 1998b) and it is possible that, in the absence of sarT, this binding results in repression of hla transcription (Cassat et al., 2005). In this scenario, mutation of sarA, or a factor like msa that is required for induction of sarA expression, would result in increased rather than decreased transcription of hla.
The impact of msa on transcription of spa was also strain-dependent. Specifically, the RN6390 msa mutant failed to express spa in all growth phases, while in UAMS-1 the msa mutant showed no significant change (less than twofold) in spa expression in the post-exponential growth phase. It is also noteworthy that, in direct contrast to agr, spa expression was significantly higher in UAMS-1 (approx. tenfold) than in RN6390 (data not shown and Blevins et al., 2002). This is perhaps to be expected in that agr represses spa transcription; however, regulation of spa transcription and protein A production is very complex. As a surface protein, it is preferentially produced in exponential growth and repressed as cultures enter the post-exponential growth phase. In addition to agr and sarA, spa expression is controlled directly or indirectly by several regulators including B, Rot, MgrA, ArlSR, SarS, SarT, SarU, SarA and RNAIII (Bronner et al., 2004). In early growth, spa is up-regulated by SarS. As the cells enter post-exponential growth, the amounts of SarA and RNAIII increase. Both SarA and RNAIII repress spa directly and via sarS, leading to a shutdown in protein A production (Arvidson & Tegmark, 2001; Cheung et al., 2004; Novick, 2003). It is important to note in that regard that all 8325-4 strains also carry a mutation in tcaR, which is an activator of sarS transcription (McCallum et al., 2004). Indeed, we have confirmed that UAMS-1 has an intact tcaR locus and that it produces significantly higher levels of sarS as well as spa (data not shown). Whether this difference accounts for the different spa phenotypes observed in the RN6390 and UAMS-1 mutants remains to be determined.
Overall, our findings show that mutation of msa affects the expression of sarA and the genes encoding several virulence factors. To the extent that mutation of sarA has been shown to result in a reduced capacity to form a biofilm (Beenken et al., 2003; Valle et al., 2003) and reduced virulence in several animal models of S. aureus infection, this suggests that msa may also make an important contribution to staphylococcal pathogenesis. Interestingly, sequence analysis of Msa indicates that it is a membrane protein, which suggests that it might interact with the external environment. We hypothesize that Msa acts as a sensor for an external signal that modulates sarA expression and thereby regulates the expression of S. aureus virulence factors. Whether Msa affects transcription of sarA directly or via an intermediate protein is not yet clear. It is also unclear what environmental signals might influence msa transcription. The drastic impact of the msa mutation on transcription of the gene encoding the cold-shock protein CspA, and the close proximity of msa and cspA genes, suggest that the changes in expression of sarA and other virulence factors might be mediated through cspA. Indeed, Katzif et al. (2003) showed that, in addition to its role in cold-shock response, CspA regulates expression of at least 14 genes. CspA homologues from other bacteria have been shown to regulate genes at the transcriptional and translational level (Yamanaka, 1999). Interestingly, neither cspA nor lysA was found to be regulated by sarA in a previous transcriptional profiling study (Dunman et al., 2001). Although these earlier experiments were done with the 8325 strain RN27 and an early generation gene chip that was limited to 
86 % of the genome of the single S. aureus strain COL, similar experiments with a UAMS-1 sarA mutant and a much more comprehensive gene chip also failed to identify cspA as part of the sarA regulon (James E. Cassat, personal communication). This suggests that the impact of msa on transcription of cspA and lysA regulation is independent of its impact on sarA.
Rossi et al. (2003) described another membrane protein, MsrR, that attenuates sarA and might be involved in sensing cell wall damage. The nature of the interaction of MsrR with sarA has not yet been determined. To our knowledge, Msa and MsrR are the only membrane-associated proteins that have been shown to regulate sarA transcription. Given the complexity of virulence regulation in S. aureus, and the fact that sarA regulates a wide variety of genes that are not directly related to virulence, we expect that membrane proteins such as MsrA and putatively Msa (whether directly or through cspA) provide the sarA regulon with external stimuli that are necessary for its functions. Other accessory elements that are located in the cytoplasm (e.g. the sarA homologues) fine-tune sarA's response to environmental conditions by activating or repressing different sets of genes.
|
ACKNOWLEDGEMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Arvidson, S. & Tegmark, K. (2001). Regulation of virulence determinants in Staphylococcus aureus. Int J Med Microbiol 291, 159–170.[CrossRef][Medline]
Bayer, M. G., Heinrichs, J. H. & Cheung, A. L. (1996). The molecular architecture of the sar locus in Staphylococcus aureus. J Bacteriol 178, 4563–4570.
Beenken, K. E., Blevins, J. S. & Smeltzer, M. S. (2003). Mutation of sarA in Staphylococcus aureus limits biofilm formation. Infect Immun 71, 4206–4211.
Bischoff, M., Entenza, J. M. & Giachino, P. (2001). Influence of a functional sigB operon on the global regulators sar and agr in Staphylococcus aureus. J Bacteriol 183, 5171–5179.
Blevins, J. S., Gillaspy, A. F., Rechtin, T. M., Hurlburt, B. K. & Smeltzer, M. S. (1999). The staphylococcal accessory regulator (sar) represses transcription of the Staphylococcus aureus collagen adhesin gene (cna) in an agr-independent manner. Mol Microbiol 33, 317–326.[CrossRef][Medline]
Blevins, J. S., Beenken, K. E., Elasri, M. O., Hurlburt, B. K. & Smeltzer, M. S. (2002). Strain-dependent differences in the regulatory roles of sarA and agr in Staphylococcus aureus. Infect Immun 70, 470–480.
Blevins, J. S., Elasri, M. O., Allmendinger, S. D., Beenken, K. E., Skinner, R. A., Thomas, J. R. & Smeltzer, M. S. (2003). Role of sarA in the pathogenesis of Staphylococcus aureus musculoskeletal infection. Infect Immun 71, 516–523.
Booth, M. C., Cheung, A. L., Hatter, K. L., Jett, B. D., Callegan, M. C. & Gilmore, M. S. (1997). Staphylococcal accessory regulator (sar) in conjunction with agr contributes to Staphylococcus aureus virulence in endophthalmitis. Infect Immun 65, 1550–1556.[Abstract]
Bronner, S., Monteil, H. & Prevost, G. (2004). Regulation of virulence determinants in Staphylococcus aureus: complexity and applications. FEMS Microbiol Rev 28, 183–200.[CrossRef][Medline]
Cassat, J. E., Dunman, P. M., McAleese, F., Murphy, E., Projan, S. J. & Smeltzer, M. S. (2005). Comparative genomics of Staphylococcus aureus musculoskeletal isolates. J Bacteriol 187, 576–592.
Chakrabarti, S. K. & Misra, T. K. (2000). SarA represses agr operon expression in a purified in vitro Staphylococcus aureus transcription system. J Bacteriol 182, 5893–5897.
Chan, P. F. & Foster, S. J. (1998a). The role of environmental factors in the regulation of virulence-determinant expression in Staphylococcus aureus 8325-4. Microbiology 144, 2469–2479.[Abstract]
Chan, P. F. & Foster, S. J. (1998b). Role of SarA in virulence determinant production and environmental signal transduction in Staphylococcus aureus. J Bacteriol 180, 6232–6241.
Cheung, A. L., Yeaman, M. R., Sullam, P. M., Witt, M. D. & Bayer, A. S. (1994). Role of the sar locus of Staphylococcus aureus in induction of endocarditis in rabbits. Infect Immun 62, 1719–1725.
Cheung, A. L., Bayer, M. G. & Heinrichs, J. H. (1997). sar genetic determinants necessary for transcription of RNAII and RNAIII in the agr locus of Staphylococcus aureus. J Bacteriol 179, 3963–3971.
Cheung, A. L., Bayer, A. S., Zhang, G., Gresham, H. & Xiong, Y. Q. (2004). Regulation of virulence determinants in vitro and in vivo in Staphylococcus aureus. FEMS Immunol Med Microbiol 40, 1–9.[CrossRef][Medline]
Chien, Y., Manna, A. C. & Cheung, A. L. (1998). SarA level is a determinant of agr activation in Staphylococcus aureus. Mol Microbiol 30, 991–1001.[CrossRef][Medline]
Chien, Y., Manna, A. C., Projan, S. J. & Cheung, A. L. (1999). SarA, a global regulator of virulence determinants in Staphylococcus aureus, binds to a conserved motif essential for sar-dependent gene regulation. J Biol Chem 274, 37169–37176.
Dunman, P. M., Murphy, E., Haney, S. & 7 other authors (2001). Transcription profiling-based identification of Staphylococcus aureus genes regulated by the agr and/or sarA loci. J Bacteriol 183, 7341–7353.
Francis, K. P., Joh, D., Bellinger-Kawahara, C., Hawkinson, M. J., Purchio, T. F. & Contag, P. R. (2000). Monitoring bioluminescent Staphylococcus aureus infections in living mice using a novel luxABCDE construct. Infect Immun 68, 3594–3600.
Gillaspy, A. F., Hickmon, S. G., Skinner, R. A., Thomas, J. R., Nelson, C. L. & Smeltzer, M. S. (1995). Role of the accessory gene regulator (agr) in pathogenesis of staphylococcal osteomyelitis. Infect Immun 63, 3373–3380.[Abstract]
Gillaspy, A. F., Lee, C. Y., Sau, S., Cheung, A. L. & Smeltzer, M. S. (1998). Factors affecting the collagen binding capacity of Staphylococcus aureus. Infect Immun 66, 3170–3178.
Goerke, C., Campana, S., Bayer, M. G., Doring, G., Botzenhart, K. & Wolz, C. (2000). Direct quantitative transcript analysis of the agr regulon of Staphylococcus aureus during human infection in comparison to the expression profile in vitro. Infect Immun 68, 1304–1311.
Heinrichs, J. H., Bayer, M. G. & Cheung, A. L. (1996). Characterization of the sar locus and its interaction with agr in Staphylococcus aureus. J Bacteriol 178, 418–423.
Hirokawa, T., Boon-Chieng, S. & Mitaku, S. (1998). SOSUI: classification and secondary structure prediction system for membrane proteins. Bioinformatics 14, 378–379.
Horsburgh, M. J., Aish, J. L., White, I. J., Shaw, L., Lithgow, J. K. & Foster, S. J. (2002). B modulates virulence determinant expression and stress resistance: characterization of a functional rsbU strain derived from Staphylococcus aureus 8325-4. J Bacteriol 184, 5457–5467.
