OUP user menu

The VraS/VraR two-component regulatory system required for oxacillin resistance in community-acquired methicillin-resistant Staphylococcus aureus

Susan Boyle-Vavra, Shaohui Yin, Robert S. Daum
DOI: http://dx.doi.org/10.1111/j.1574-6968.2006.00384.x 163-171 First published online: 1 September 2006


Methicillin/oxacillin (Oxa) resistance in Staphylococcus aureus is primarily mediated by the acquired penicillin-binding protein (PBP2a) encoded by mecA. PBP2a acts together with native PBP2 to mediate oxacillin resistance by contributing complementary transpeptidase and transglycosylase activities, respectively. The VraS/VraR two-component regulatory system is inducible by cell-wall antimicrobials (β-lactams, glycopeptides) and controls transcriptional induction of many cell-wall genes including pbp2 and itself. We investigated the role of VraS/VraR in the phenotypic expression of oxacillin resistance by inactivating vraS in community-acquired MRSA clinical isolates that lack functional genes encoding the mecA regulatory sequences mecI and mecR1. Inactivation of vraS abrogated oxacillin resistance, and complementation with the vraS operon restored the resistance phenotype. mecA transcription increased in the vraS mutants; however, PBP2a abundance was similar to that of the wild type. Although pbp2 transcription decreased in the vraS mutants, overexpression of the pbp2 operon did not restore resistance. These data demonstrate that although expressions of mecA and pbp2 are required for oxacillin resistance, they are not sufficient. Therefore, the vraS/vraR regulatory system plays a crucial role in allowing MRSA to respond to β-lactams by regulation of a gene target other than the known effectors of methicillin resistance.

  • methicillin resistance
  • Staphylococcus aureus
  • mecA
  • pbp2


Staphylococcus aureus is the most commonly isolated bacterial species from inpatients and the second most common among outpatients (Styerset alet al, 2006). Nearly all S. aureus are resistant to penicillin due to the production of β-lactamase, and methicillin-resistant S. aureus (MRSA) now accounts for c. 60% of S. aureus isolates in the United States (Styerset alet al, 2006) and 40% in the United Kingdom (Enright, 2006). The recent emergence of high-level vancomycin-resistant strains (CDC, 2004) and the rapid spread of virulent community-acquired MRSA in various continents (Okumaet alet al, 2002; Vandeneschet alet al, 2003; Ademet alet al, 2005; Enright, 2006) underscore the increasing need to understand and control resistance in this pathogen.

Penicillin is a β-lactam that acts by binding to and inactivating enzymes required for bacterial cell-wall synthesis, referred to as penicillin-binding proteins (PBPs). β-Lactam compounds that are insensitive to β-lactamase (methicillin, oxacillin, nafcillin) were introduced in the 1960s for therapy of infections caused by β-lactamase-producing penicillin-resistant S. aureus. However, MRSA strains soon evolved that had acquired the mecA gene (Matsuhashiet alet al, 1986) encoding a penicillin-binding protein (PBP2a or PBP2′) with a low affinity for β-lactams.

Staphylococcus aureus produces four native PBPs (PBP1, PBP2, PBP3 and PBP4) (Reynolds, 1988). In MRSA, incorporation of the peptidoglycan precursor into the growing peptidolgycan chain in the presence of a β-lactam is carried out by the cooperation between the transpeptidase activity of PBP2a and the transglycosylase activity supplied by the chromosomally encoded PBP2 (Pinhoet alet al, 1997, 2001). The genes encoding these proteins (mecA encoding PBBP2a and pbp2 encoding PBP2) are both induced by β-lactams (Berger-Bachi & Rohrer, 2002; Boyle-Vavraet alet al, 2002; Kurodaet alet al, 2003). Thus, β-lactams act as a signal to activate expression of both known methicillin resistance proteins, PBP2 and PBP2a. pbp2 transcription is also induced by other cell wall acting antimicrobials including the glycopeptide vancomycin (Boyle-Vavraet alet al, 2002).

The induction of mecA and pbp2 by β-lactams is mediated by distinct regulatory pathways. mecA is repressed by MecI or the homologue BlaI and de-repressed by proteolysis due to a signal-transducing protein, MecR1 or BlaR1 (Zhanget alet al, 2001). In contrast, pbp2 is transcriptionally activated by a two-component regulatory pair consisting of VraS, a membrane sensor transducing protein with a conserved histidine kinase domain, and VraR, which has a DNA-binding domain conserved among response regulators (Kurodaet alet al, 2000, 2003).

The acquired mecA gene is carried on the chromosome of MRSA strains on one of five integrated resistance islands called ‘staphylococcal chromosome cassette mec′ (SCCmec) (Itoet alet al, 2001, 2004; Maet alet al, 2002; Boyle-Vavraet alet al, 2005). SCCmec type IV is the most common SCCmec type found in epidemic community-acquired MRSA strains and SCCmec II is the most prevalent in healthcare-associated strains circulating in the United States.

The architectural differences in the mecA regulatory locus among the different SCCmec types have potential consequences for mecA regulation (Berger-Bachi & Rohrer, 2002). In SCCmec types II and III, mecR1 and mecI are cotranscribed on an operon located immediately upstream of mecA in opposite orientation relative to mecA (Itoet alet al, 2001). mecA is expressed at low levels in such strains, leading to a susceptible or a low level, heterotypic oxacillin resistance phenotype (Ryffelet alet al, 1992; Suzukiet alet al, 1992). MRSA clinical strains with SCCmec I, IV or V have deleted portions of mecI or mecR1 (Itoet alet al, 2001, 2004; Maet alet al, 2002; Boyle-Vavraet alet al, 2005). Other strains have point mutations in the mecA promoter or mecI protein coding region, resulting in decreased repressor activity (Suzukiet alet al, 1993).

As VraS/VraR regulate the induction of pbp2 transcription by oxacillin, it is conceivable that interfering with VraS/VraR-mediated signaling of pbp2 could decrease the oxacillin resistance phenotype. Moreover, whether VraS and VraR also mediate induction of mecA is not known. In this study, we sought to compare the role played by VraS and VraR in regulating oxacillin resistance in SCCmec type IV-carrying strains lacking functional mecI and mecR1 that are phenotypically oxacillin-resistant. As inactivation of vraS abolished resistance to oxacillin in both strains, the extent to which the two known mediators of methicillin resistance, PBP2 and PBP2a, were responsible for the increased susceptibility of the mutants to oxacillin was investigated.



The parent strains in which the chromosomal vraS gene was inactivated (Table 1) were chosen from MRSA isolates collected from the University of Chicago Clinical Microbiology Laboratories as part of a prospective surveillance of MRSA molecular epidemiology. All isolates were confirmed as S. aureus by Staphaurex Plus agglutination (Remel), Gram staining and mannitol salts fermentation and then stored at −70°C in skim milk (Difco). The oxacillin resistance phenotype was determined by the Vitek2 system and confirmed by the presence of the mecA gene as determined by PCR as described (Boyle-Vavraet alet al, 2005). To facilitate the use of selectable markers, parent strains were chosen that were susceptible to erythromycin (Ery) and tetracycline (Tet). Strains carrying an integrated SCCmec type IV methicillin resistance cassette were identified by PCR with genomic DNA as described (Boyle-Vavraet alet al, 2005). SCCmec type IV was defined as the presence of a ccr complex type 2 (ccrA2 and ccrB2) and mec class B complex as described (Maet alet al, 2002). vraS was inactivated in the chromosome of two community-acquired MRSA isolates (1564 and 923) that contain SCCmec type IV (ΔmecI and partial mecR1 [mecR1′]). These strains were from the ST 1 (pulsotype USA 400) and ST 8 (pulsotype USA 300) lineages, the most common community-acquired epidemic MRSA strains in the United States.

View this table:
Table 1

Strains and plasmids

DescriptionSTResistance phenotypeReference/source
Strain (plasmid)
RN4220S. aureus intermediate host strain for introduction of recombinant DNAPan susceptibleKreiswirth (1983) provided by Jean Lee
1564CA-MRSA strain from Illinois1OxaR, EryS, TetSThis report
1564-M48vraS insertion mutant of strain 15641OxaS, EryR, TetSThis report
1564-M48 (pVRASR-2)vraS insertion mutant of strain 1564 containing the vraSR operon complementation plasmid pVRASR-21OxaR, EryR, TetRThis report
1564-M48 (pVP450-vraSR)vraS insertion mutant of strain 1564 containing the vraSR complementation plasmid pVP450-vraSR1OxaR, EryR, TetRThis report
1564-M48 (pPBP2-1)1564-M48 complemented with the pbp2 operon plasmid, pPBP2-11OxaS, EryR, TetRThis report
923CA-MRSA from The University of Chicago Emergency Department8OxaR, EryS, TetSThis report
923-M2vraS insertion mutant of strain 9238OxaS, EryR, TetsThis report
923-M2 (pVRASR-2)vraS insertion mutant of strain 923 containing the vraS/vraR operon complementation plasmid, pVRASR-28OxaR, EryR, TetRThis report
pCL52.1S. aureus–E. coli shuttle vector, ori (pE194ts)TetRChia Lee
pAW8S. aureus–E. coli shuttle vector, pAMα1 gram-positive ori and ColE1 ori from E. coli.TetRAkihito Wada
pVRASR-erm-pCL52.1Allelic replacement vector containing vraS::ermCAmpR, EryR, TetRYin (2005)
pVRASR-2Entire vraS/vraR operon cloned into pAW8TetRThis report
pVP450vraS/vraR promoter region cloned into pAW8TetRYin (2005)
pVP450-vraSRContains a 1.7kb insert containing vraS and vraR ORFs cloned downstream of vraSR promoter in pVP450TetRThis report
pPBP2-1A 3.3kb fragment encompassing the entire pbp2 operon (consisting of prfA/recU and pbp2) cloned into pAW8TetRThis report
  • Ori, plasmid origin of replication; ts, temperature sensitive replicon; ST, the ST representing the genetic lineage as determined by multilocus sequence typing.

Molecular biology protocols

Standard protocols used for restriction enzyme digestion, agarose gel electrophoresis, ligation, cloning and PCR were performed as described (Ausubelet alet al, 1994). For performing PCR, template DNA was added to the Taq-Pro complete mix (Denville Scientific Inc.) in 50μL reaction volumes. Oligonucleotides were designed for each gene target from visual inspection of the DNA sequence and were synthesized by Integrated DNA Technologies. For introduction of recombinant plasmids into S. aureus, the restriction-deficient strain RN4220 was used as the intermediate host strain (Table 1).

Disruption of vraS

The vraS gene was disrupted in the chromosome of clinical MRSA strains (Table 1) by allelic exchange as previously described (Yinet alet al, 2005). Briefly, an Ery resistance cassette (ermC) was inserted into the unique EcoR1 site in the histidine kinase domain in the 3′-terminus of vraS (Fig. 1) in the plasmid, pVRASR-erm-pCL52.1. This plasmid contains a temperature-sensitive replicon from pE194 as described (Table 1) (Yinet alet al, 2005). The vraS::ermC gene was exchanged from pVRASR-erm-pCL52.1 into the chromosome by growth at 42°C and screening for loss of the Tet resistance marker of the plasmid (Yinet alet al, 2005). That the ermC cassette had inserted in the chromosomal vraS gene was determined by performing both PCR and Southern blotting for each mutant strain. vraS mutant clones were named after their respective wild-type parent strains. (For example, 1564-M48 is the vraS mutant of strain 1564.) The mutant strains were cultured in the presence of Ery to maintain the ermC insertion within vraS.

Figure 1

(a) Map depicting the vraS/vraR operon and the fragments used to clone the vraS/vraR complementation plasmids, the site of the ermC insertion in vraS within the conserved histidine kinase domain (box with diagonal lines) and the vraS Northern blotting probe used in Northern hybridizations. Details for construction of plasmids are described in Materials and methods and listed in Table 1. (b) Northern blot of strain 1548 and the vraS mutant 1548-M48 grown in the absence (−Oxa) or presence (+Oxa) of oxacillin and probed with the vraS/vraR probe shown in (a). (Lane 1) strain 1548, (Lane 2) strain 1548-M48, (Lane 3) strain 1548-M48 (pVRASR-2).

Complementation plasmids

To complement the vraS mutation, a plasmid (pVRASR-2) was constructed by inserting a vraS/vraR operon fragment (Fig. 1) into the Escherichia coliS. aureus shuttle plasmid pAW8, a low-copy-number plasmid that contains a Tet resistance marker and an origin of replication for S. aureus (pAMα1) and E. coli. A 3.3kb fragment encompassing the entire vraS/vraR operon (orf1, yvqF, vraS and vraR) (encompassing −450 to +2898 relative to the transcription start site; Yin et al, 2005) was PCR amplified using primers VraSR-F1-KpnI (5′-ATTGGTACCATGGCATTTGAGAATGC-3′) and VraSR-R-PstI (5′-GGGCTGCAGTAATTCGATACGAACTATG-3′). These primers contained 5′ terminal KpnI or PstI restriction sites (underlined bases), respectively, to facilitate cloning. The resulting fragment was digested with KpnI and PstI, ligated with KpnI- and PstI-digested pAW8 and electroporated into E. coli using Tet (5μgmL−1) as the selectable marker. pVRASR-2 was isolated from E. coli with the use of a plasmid midi isolation kit (Qiagen) and electroporated as described (Augustin & Gotz, 1990) into S. aureus strain RN4220 before introduction into vraS mutant strains. The presence of the complementation plasmid and its expression were confirmed by restriction mapping and Northern blotting, respectively.

A second complementation plasmid (pVP450-vraSR) containing a 1.7kb PCR fragment consisting solely of the vraS and vraR ORFs expressed from the vraS/vraR operon promoter was produced using a strategy similar to that described above. The insert consisting of the vraS/vraR promoter region and flanking sequences (−450 to +150) was previously inserted into the KpnI/BamHI sites of pAW8, producing pVP450 (Yinet alet al, 2005). The 1.7kb vraS/vraR fragment (+1232 to +2908 relative to the transcriptional start point Yinet alet al, 2005) was PCR amplified using primers 5′-ATTGGATCCATGAACCACTACAATAG-3′ (VRASR-F-BamH1) and 5′-GGGCTGCAGTAATTCGATACGAACTATTG-3′ (VRASR-R-PstI), digested with BamH1 and PstI, and ligated into the BamH1 and PstI sites in pVP450, creating pVP450-vraSR.

The pbp2 complementation plasmid, pPBP2-1, was constructed by PCR amplification of a 3.3kb fragment encompassing both ORFs from the pbp2 operon (prfA/recU and pbp2) using primers 5′-GCCGGTACCCACATACTTGTACTTGCCTC-3′ (PRFPBP2-F-Kpn1) and 5′-CGCGGATCCCCAATAATTTTACATAGCTAAAG-3′ (PRFPBP2-R-BamH1). This fragment was inserted into pAW8 after KpnI and BamHI digestion using a strategy similar to that described above.

Minimal inhibitory concentration (MIC) assay

The MIC of oxacillin was determined using broth dilution in tryptic soy broth (TSB) supplemented with 2% NaCl. An inoculum of 5 × 105CFUmL−1 was applied to each well of a 24-well cell culture dish as prescribed by the Clinical Laboratory Standards Institute (CLSI) (NCCLS, 2004). For the strains carrying the vraS/vraR and pbp2 complementation plasmids, Tet (5μgmL−1) was included in the medium to maintain the plasmid.

Isolation of RNA and Northern blot analysis

To evaluate the effect of the vraS mutation on transcriptional induction of various genes, overnight cultures of S. aureus were diluted 1:100 in BHI and incubated at 37°C for c. 1h (OD600nm of c. 0.2), at which time the cultures were treated with oxacillin and incubated for an additional hour or as indicated. Uninduced cultures (constitutive condition) were incubated in parallel in the absence of oxacillin. The cells were pelleted, resuspended in the appropriate volume of TE buffer containing recombinant lysostaphin (Sigma, 300μgmL−1) and incubated at room temperature for 10min. The preparation and storage of RNA and conditions for Northern blotting procedures were performed as described (Yinet alet al, 2005).

Probes for the detection of mecA, pbp2, vraS/vraR, mecR1/mecI, blaR1/blaI and blaZ were produced by PCR using the primers in Table 2. All probes were labeled with [α32P] dATP (Amersham) using the Prime-a-Gene Labeling System (Promega) as described (Yinet alet al, 2005).

View this table:
Table 2

PCR primers for amplification of Northern blotting probes

Target geneF primer sequenceR primer sequenceProduct size (bp)

Detection of PBP2a protein by Western blotting

Bacterial cultures were harvested by centrifugation and washed with 50mM Tris, 150mM NaCl and 5mM MgCl (pH 7.5). They were resuspended in the same buffer. Lysostaphin (200μgmL−1), RNase (10μgmL−1) and DNase (20UmL−1) were added to the mixture, followed by incubation at room temperature for 30min, and sonication on ice, four times for 10s each. The lysed cells were ultracentrifuged at 120000g for 40min. The resultant pellet was resuspended in 50mM sodium phosphate (pH 7.0) containing 6M urea. The membrane proteins (100μg) were separated using 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Ausubelet alet al, 1994) and electrophoretically transferred onto Immobilon-P membranes (Millipore) using a vertical slab gel apparatus (100V for 3–4h) at 4°C with a Tris-glycine transfer buffer containing 10% methanol. PBP2a was detected with a mouse anti-PBP2a monoclonal antibody (Denka Seiken Co. Ltd.) as the primary antibody (1:10000), which was detected with antimouse IgG (Promega) (1:10000) conjugated with alkaline phosphatase according to the manufacturer's instructions.


Characterization of the vraS mutants

The chromosomal vraS gene was inactivated in MRSA strains 1564 (ST 1, USA 400) and 923 (ST 8, USA 300) by allelic exchange, creating strains 1564-M48 and 923-M2, respectively (Table 1). Each mutant contained an ermC cassette within the EcoR1 site of vraS in the conserved histidine kinase domain (Fig. 1). Northern blotting confirmed that the vraS transcript of the mutant consisted of orf1, yvqF, vraS and ermC (Yinet alet al, 2005; Fig. 1b). vraR was not detected on the transcript of the vraS mutants when using a vraR-specific probe, consistent with a negative polar effect on transcription downstream of ermC (Yinet alet al, 2005). Consequently, the size of the vraS transcript in the mutants reflected the first two orfs, vraS and ermC, and was similar to that of the wild-type strains.

The vraS mutants are susceptible to oxacillin

The vraS mutation uniformly decreased the broth dilution MIC of both strains to below the oxacillin susceptibility breakpoint (≤4μgmL−1) (Table 3). The oxacillin resistance phenotype was restored in strains 1564-M48 and 923-M2 by trans-complementation using plasmid pVRASR-2 containing the entire vraS/vraR operon (orf1, yvqF, vraS and vraR) and its promoter (Table 3). The complementation plasmid (pVP450-vraSR) that contained only vraS and vraR (Fig. 1; Table 1) also complemented the mutant phenotype, albeit to a lesser extent than did the entire operon. That the vraS transcript was overexpressed in the complemented strain was confirmed by Northern blotting (Fig. 1b, lanes 3).

View this table:
Table 3

Oxacillin MICs

Strain (plasmid)Genes expressed in transMeanSD
1564-M48 (pVRASR-2)vraSR operon14.54.1
1564-M48 (pVP450-vraSR)vraS and vraR6.42.7
1564-M48 (pPBP2-1)pbp2 operon1.50.5
923-M2 (pVRASR-2)vraSR operon10.72.1
  • * Experiments were performed at least two times and at the most five times in triplicate. The concentrations of oxacillin tested were 0, 1, 2, 3, 4, 6, 8, 12, 16, 24, 32, 48, and 64μgmL−1.

Northern blotting analysis of vraS mutants

We investigated whether the oxacillin-susceptible phenotype in the vraS mutants was due to a change in expression of mecA or production of PBP2a. We performed experiments using oxacillin as mecA is inducible by this antimicrobial. No difference in growth was observed between the wild-type and mutant strains using 1μgmL−1 of oxacillin at the time points chosen for analysis. As expected, mecA transcription was induced by oxacillin in both wild-type strains (Fig. 2). Surprisingly, the mecA transcript abundance was greater in the vraS mutants than the wild-type strains in the presence of 1 or 2μgmL−1 of oxacillin, a finding that was not expected as the mutant strains were phenotypically susceptible to oxacillin. In the wild-type strains, a slight decrease in the mecA transcript abundance decreased with 2μgmL−1 compared with 1μgmL−1, a finding that could be due to growth inhibition at the higher concentration of drug.

Figure 2

Northern blot analysis for transcriptional induction of mecA, mecR1′, blaR1/blaI and blaZ in strains 923 and 1564 and their respective vraS mutants, 923-M2 and 1564-M48. Cultures were exposed to 0, 1 or 2μg of oxacillin (Oxa) mL−1 and harvested 1h later. Untreated controls were cultured in parallel with the treated samples. Six micrograms of total RNA was loaded into each lane and the transcripts were detected with the gene-specific DNA probes indicated to the left of the panel.

As strains 1564 and 923 lack mecI and contain a truncated mecR1, mecA transcription in these isolates should not be tightly repressed and mecA transcription should be controlled by BlaI/BlaR1 (Berger-Bachi & Rohrer, 2002). Like mecA, the abundance of the oxacillin-induced blaI/blaR1 transcript was slightly increased in each mutant compared with the respective isogenic parent strain. However, blaI/blaR1 did not increase to the same extent as mecA. Therefore, the increase in mecA transcript abundance did not parallel that of blaR1/blaI.

Although blaR1/blaI transcription increased slightly in the presence of oxacillin in strain 923-M2 compared with strain 923, it was not sufficient to effect an increase in the transcription of blaZ, the cognate target gene of BlaI/BlaR1. This suggests that the increase in mecA induction in the presence of oxacillin in the mutants strain was not solely mediated by BlaR1/BlaI. In strain 1564-M48, the amount of blaR1/blaI transcript abundance in the presence of oxacillin was similar to that of blaZ but both blaR1/blaI and blaZ transcript abundance was far lower than that of mecA.

Abundance of oxacillin-induced PBP2a in the vraS mutant strain 1564-M48

To assess whether the increased induction of mecA transcription affected production of the PBP2a protein, Western blotting was performed on membrane fractions isolated from strains grown in the presence or absence of oxacillin (Fig. 3). Despite the increased transcription of mecA in strain 1564-M48 relative to the wild-type parent, PBP2a abundance was similar between the wild-type and mutant strain. Moreover, a decrease in PBP2a production was not observed in 1564-M48 compared with the parent, despite the fact that the former was phenotypically susceptible to oxacillin.

Figure 3

Western blot analysis for expression of PBP2a in strains 1564 and its vraS mutant 1564-M48. Overnight cell cultures were diluted 1:100 in TSB, grown at 37°C for 1h and were treated with oxacillin (0.5μgmL−1) and harvested either 2 or 3h later. Untreated controls were cultured in parallel with the treated samples. PBP2a was detected with a mouse anti-PBP2a monoclonal antibody (Denka Seiken Co. Ltd.). The molecular weight standards (ST) were BioRad Kaleidoscope prestained markers.

Effect of the vraS mutation on pbp2 and vraS/vraR transcription

As PBP2a production was not decreased in the vraS mutant strains and as the native PBP2 transglycosylase cooperates with PBP2a to build the cell wall in the presence of oxacillin, we investigated the possibility that the oxacillin-susceptible phenotype of the vraS mutant was due to decreased pbp2 induction. Oxacillin-mediated (4μgmL−1) transcriptional induction of pbp2 was drastically decreased in the vraS mutants compared with the wild-type parent strains (Fig. 4). Vancomycin-induced pbp2 transcription was also attenuated in both vraS mutant strains (Fig. 4) as expected (Kurodaet alet al, 2003; Yinet alet al, 2005). Induction of vraS/vraR transcription by oxacillin also dramatically decreased in the vraS/vraR mutants (Fig. 4a), confirming that vraS/vraR transcriptional induction is autoregulated in MRSA strains as we found for the MSSA strain, RN4220 (Yinet alet al, 2005).

Figure 4

Northern blot analysis of induction of pbp2 and vraS/vraR operons in strains 923 and 1564, and their vraS mutants 923-M2 (M2) and 1564-M48 (M48), respectively. (a) Bacteria were exposed to either no drug (C), 2μg of vancomycinmL−1 (V) or 4μg of oxacillinmL−1 (O), were harvested 1h later and hybridized with a pbp2 or vraSR operon probe. (b) Northern blot hybridized with a pbp2-specific probe showing the restoration of pbp2 expression in the vraS mutant 1564-M48 harboring the complementation plasmid, pVRASR-2. (lane 1)Strain 1564, (lane 2)1564-M48, (lane 3) 1564-M48 (pVRASR-2) grown in the absence (−Oxa) or presence (+Oxa) of oxacillin.

Complementation of strain 1564-M48 with the vraS/vraR operon by transformation with pVRASR-2 restored the induction of pbp2 transcription (Fig. 4b, lane 3, +Oxa).

To assess directly whether the decrease in pbp2 transcription in the vraS mutants played a role in decreasing the oxacillin resistance phenotype, strain 1564-M48 was transformed with a plasmid (pPBP2-1) that contained both ORFs encoded in the pbp2 operon (recU and pbp2). The oxacillin MIC of the resulting strain, 1564-M48 (pPBP2-1), was similar to that of strain 1564-M48 (Table 3). Thus, the decrease in oxacillin induction of pbp2 expression in the vraS mutants does not explain the decrease in oxacillin resistance.


In this study, we demonstrated that the two-component regulatory system VraS/VraR plays a role in the regulation of oxacillin resistance in clinical MRSA strains that lack the mecA repressor/inducer pair, MecI/MecR1. This was demonstrated in two clinically important epidemic community-acquired MRSA strains that have been known to cause severe sepsis, necrotizing pneumonia or necrotizing fasciitis (Mongkolrattanothaiet alet al, 2003; Ademet alet al, 2005; Milleret alet al, 2005).

In MRSA isolates, the transglycosylase domain of PBP2 acts in concert with the transpeptidase of PBP2a to synthesize the cell wall despite the presence of oxacillin (Pinhoet alet al, 2001). Therefore, we first considered the possibility that the decrease in resistance was due to an unknown effect of VraS on transcription of mecA, its well-characterized regulatory genes or on production of PBP2a, the mecA gene product. Although a paradoxical increase in mecA transcription was observed, PBP2a protein abundance was similar in the wild type and mutant. Thus, although the vraS mutant strains were oxacillin susceptible, they produced wild-type amounts of PBP2a. As a decrease in pbp2 transcription was observed in the vraS mutants in this and previous studies (Kurodaet alet al, 2003; Yinet alet al, 2005), we attempted to restore the resistance phenotype by overexpressing pbp2 from a multicopy plasmid in the vraS mutants. This approach did not restore resistance although complementation with the entire vraS/vraR operon did. These findings demonstrate that VraS is required for resistance to oxacillin even when the mecA and pbp2 genes are transcribed. Thus, our data demonstrate for the first time that although expressions of both PBP2 and PBP2a are required for the oxacillin resistance phenotype, they are not sufficient. It is likely that the VraS/VraR two-component regulatory system modulates the expression of a factor or factors other than PBP2 or PBP2a that influence oxacillin resistance. Although a microarray analysis has been performed to compare the genes transcribed by a wild-type and vraS/vraR mutant strain in the presence of vancomycin (Kurodaet alet al, 2003), a similar analysis using oxacillin would suggest possible targets of vraS/vraR that are responsible for the decreased oxacillin resistance.

The idea that PBP2a production is not sufficient to mediate oxacillin resistance is consistent with previous studies (Niemeyeret alet al, 1996; Berger-Bachi & Rohrer, 2002). The vraS mutant phenotype we describe is like that of isolate BMS1 described previously, which was susceptible to oxacillin despite the fact that PBP2a was expressed (Niemeyeret alet al, 1996); the genetic basis for this paradox was not elucidated. As the role of PBP2 in oxacillin resistance was not yet appreciated, the role of pbp2 in the methicillin-susceptible phenotype of BMS1 was not studied.

In addition to mecI/mecR1 and blaR1/blaI, the regulatory genes that have been shown to affect oxacillin resistance include agr, sar (Piriz Duranet alet al, 1996) and the alternate sigma factor gene sigB (Wuet alet al, 1996). Like the vraS mutants we studied, sar, agr and sigB mutants produced wild-type amounts of PBP2a (Piriz Duranet alet al, 1996; Wuet alet al, 1996). However, in the agr and sar mutants, decreased amounts of PBP1 and PBP3 were observed.

Although PBP2a production was unaffected by the vraS mutation, the increase in mecA transcription was intriguing. The effect of the vraS mutation on mecA transcription did not appear to be mediated solely via BlaI or BlaR1 as only a slight induction of these genes and blaZ was observed compared with the dramatic induction of mecA. This supports the proposal of Cohen and Sweeny for a factor other than BlaR1 and MecR1 in the induction of blaZ (Cohen & Sweeney, 1968) and mecA (Niemeyeret alet al, 1996), respectively. VraS/VraR or its regulated genes might perform the functions of the putative BlaR2/MecR2. It remains to be determined whether VraS/VraR directly attenuates mecA induction or does so indirectly by regulation of an unidentified gene that encodes a factor that is directly involved.

Our results confirm and extend the results of others (Kurodaet alet al, 2003) who demonstrated that a vraS/vraR null mutation in strain N315 decreased the oxacillin susceptibility of that strain. However, in contrast to the strains we studied, the parent strain N315 carries SCCmec II (intact mecI and mecR1) and is phenotypically oxacillin-susceptible (Kurodaet alet al, 2003) due to efficient MecI-mediated repression (Kuwahara-Araiet alet al, 1996). We also observed a decrease in oxacillin resistance upon inactivation of vraS in two phenotypically oxacillin-resistant SCCmec II-containing strains that have the ST 5 genetic background (pulsed field type USA 100) typical of hospital-acquired MRSA strains circulating in the United States. (unpublished data). Thus, inactivation of vraS abolishes oxacillin resistance in a variety of genetic backgrounds that are phenotypically oxacillin-resistant, regardless of the presence of mecI and SCCmec types II or IV.

In summary, the data in this report provide new insight into the regulation of oxacillin resistance via the VraS/VraR two-component regulatory system and demonstrate that PBP2 and PBP2a are not sufficient for oxacillin resistance. The demonstration of this phenomenon in two of the most common virulent community-acquired clones from different lineages that contain SCCmec IV underscores the potential clinical relevance of vraS/vraR. Specifically, VraS and VraR are possible targets for combination antimicrobial therapy that could restore the usefulness of oxacillin.


S.B.-V. is a recipient of an award from The Grant Healthcare Foundation. S.Y. is a recipient of a Children's Research Foundation Award. R.S.D. is the recipient of R01 AI40481-01A1 from NIAID, RO1 CCR523379 from the CDC and support from the Grant Healthcare Foundation.


  • Authors contributionS.B.-V. and S.Y. contributed equally to the performance of the study.


View Abstract