Role of ArlRS in Autolysis in Methicillin-Sensitive and Methicillin-Resistant Staphylococcus aureus Strains

Guido Memmi, Dhanalakshmi R. Nair, and Ambrose Cheung

Department of Microbiology, Dartmouth Medical School, Hanover, New Hampshire, USA

Autolysis plays an essential role in bacterial cell division and lysis with �-lactam antibiotics. Accordingly, the expression of auto-lysins is tightly regulated by several endogenous regulators, including ArlRS, a two component regulatory system that has beenshown to negatively regulate autolysis in methicillin-sensitive Staphylococcus aureus (MSSA) strains. In this study, we foundthat inactivation of arlRS does not play a role in autolysis of methicillin-resistant S. aureus (MRSA) strains, such as community-acquired (CA)-MRSA strains USA300 and MW2 or the hospital-acquired (HA)-MRSA strain COL. This contrasts with MSSAstrains, including Newman, SH1000, RN6390, and 8325-4, where autolysis is affected by ArlRS. We further demonstrated thatthe striking difference in the roles of arlRS between MSSA and MRSA strains is not due to the methicillin resistance determinantmecA. Among known autolysins and their regulators, we found that arlRS represses lytN, while no effect was seen on atl, lytM,and lytH expression in both CA- and HA-MRSA strains. Transcriptional-fusion assays showed that the agr transcripts, RNAIIand RNAIII, were significantly more downregulated in the arlRS mutant of MW2 than the MSSA strain Newman. Importantly,provision of agr RNAIII in trans to the MW2 arlRS mutant via a multicopy plasmid induced autolysis in this MRSA strain. Also,the autolytic phenotype in the arlRS mutant of MSSA strain Newman could be rescued by a mutation in either atl or lytM. To-gether, these data showed that ArlRS impacts autolysis differently in MSSA and MRSA strains.

Staphylococcus aureus, classically described as a nosocomialpathogen, has now surfaced as a common cause of community-

acquired infections, primarily due to the emergence of strainUSA300 and, to a lesser extent, USA400 (5). Recent reports sug-gest that community-acquired methicillin-resistant S. aureus(CA-MRSA) infections can occur in the hospital environment (6,27, 36, 43). The potential replacement of hospital-acquired (HA)-with CA-MRSA strains in nosocomial infections is of concernbecause CA-MRSA strains appear to be more virulent than HA-MRSA. In addition, treatment options have been hampered due toa reduction in the efficacy of antibiotics.

Autolysis, which plays an essential role in cell wall turnover,can be triggered by antibiotics or adverse physiological conditions(29, 45). S. aureus is known to carry several known or putativeautolysin (or peptidoglycan hydrolase) genes, including the majorautolysin genes atl, lytM, lytH, and lytN, none of which are essen-tial for viability. The major autolysin Atl is a 138-kDa bifunctionalprotein that undergoes proteolytic processing to generate a 62-kDa N-acetylmuramyl-L-alanine amidase and a 51-kDa endo-�-N-acetylglucosaminidase (26, 35). LytM is a 32-kDa glycyl-glycineendopeptidase that cleaves the pentaglycine linkage between pep-tidoglycan chains in a manner similar to that of lysostaphin (12,34, 38). LytH (30 kDa), LytN (46 kDa), and LytA (50 kDa) arepeptidoglycan hydrolases with N-acetylmuramyl-L-alanine ami-dase activity (17, 24, 44, 48).

For cell wall homeostasis, autolysin expression is generallytightly regulated. Negative regulators of autolysis include ArlRS,MgrA, LytSR, ClpP, SarA, and SarV, while agr and CidABC arepositive regulators (15, 16, 23, 28, 30, 32, 39, 46). In this study, wefocused on the two-component regulatory system ArlRS whichwas previously shown to be a repressor of autolysis in themethicillin-sensitive S. aureus (MSSA) strain 8325-4 (15). In thispaper, we decided to examine the role of ArlRS in clinically rele-vant MRSA strains. We found that the negative regulation of au-tolysis by ArlRS occurs in MSSA strains, including Newman, 6390,

SH1000, and 8325-4, but not in any of the MRSA strains selectedfor this study, including HA-MRSA strain COL and CA-MRSAstrains MW2 (USA400) and USA300. This difference in autolysisbetween MSSA and MRSA strains was not attributable to mecAbut can be explained in part by a difference in agr expressionbetween the two strain sets. Importantly, provision of agr in transto the arlRS mutant of MW2 via a multicopy plasmid with anexogenous promoter induced autolysis in the mutant. These stud-ies support the divergent roles of ArlRS in inducing autolysis inMRSA and MSSA strains.

MATERIALS AND METHODS

Bacterial strains and culture conditions. The E. coli strain XL-1 blue wasused in cloning experiments. The wild-type and mutant S. aureus strainsused in this study are listed in Table 1. Luria-Bertani medium (BectonDickinson) was used for culture of E. coli, while Trypticase soy broth(TSB) was used for growth of S. aureus. When appropriate, antibiotics(Sigma) were added to the media at the following concentrations: ampi-cillin at 100 �g/ml for E. coli and chloramphenicol at 10 �g/ml and eryth-romycin at 2.5 �g/ml for S. aureus. Chloramphenicol was routinely usedto maintain selection for pEPSA5- and pSK236-based plasmids.

Susceptibility testing. MICs were determined in triplicate by mi-crodilution techniques, using an inoculum of 5 � 105 CFU/ml accordingto the CLSI guideline (10). For each mutant, three independent cloneswere tested, with the MIC data reported as median values from at leastthree independent experiments for each antibiotic. Strains containing

Received 26 September 2011 Accepted 26 November 2011

Published ahead of print 2 December 2011

Address correspondence to Ambrose L. Cheung,Ambrose.Cheung@Dartmouth.edu.

Copyright © 2012, American Society for Microbiology. All Rights Reserved.

doi:10.1128/JB.06261-11

0021-9193/12/$12.00 Journal of Bacteriology p. 759–767 jb.asm.org 759

on May 28, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

pEPSA5-based plasmids were tested with and without xylose induction,but chloramphenicol was not added in these assays to avoid interferencewith �-lactam resistance evaluation.

DNA techniques. Standard molecular-cloning techniques were used,as described previously (40). Restriction enzymes and ligases were fromNew England BioLabs. I-Proof DNA polymerase (Bio-Rad) was used togenerate all DNA fragments for gene deletions, promoter fusions, andectopic expression in pEPSA5. The fidelity of all DNA sequences gener-ated by PCR was verified by fluorescently labeled dideoxynucleotide se-quencing (BigDye terminators; PE Applied Biosystems).

Construction of S. aureus mutants. All mutants were generated within-frame deletion of target genes by allelic replacement, using the

temperature-sensitive plasmid pMAD as described previously (31). Thismethod allows the disruption of genes without the insertion of an antibi-otic resistance marker. Briefly, �0.8-kb PCR products upstream anddownstream of targeted sequences were cloned into pMAD, amplified inE. coli, and transformed into S. aureus RN4220 and then into the targetstrain, followed by the process of allelic replacement as delineated previ-ously (1). All chromosomal deletions were verified by PCR and DNAsequencing. The resulting deletion strains were devoid of both arlR andarlS open reading frames (ORFs), with at least three clones each used forcharacterization. The same pMAD system was also utilized to reinsert thenative arlRS ORFs into the MW2 �arlRS for complementation. Comple-mentation was also attempted by cloning arlRS in the xylose-inducible

TABLE 1 Strains and plasmids

Strain or plasmid Comment Source

S. aureus strainsRN4220 Heavily mutagenized MSSA strain; derivative of 8325-4; agr mutant acceptor of foreign DNA 41MW2 (USA400) CA-MRSA; wild-type strain isolated in 1998 in North Dakota 3MW2 arlRS �arlRS in-frame deletion mutant of MW2 This studyMW2 mecA �mecA in-frame deletion mutant of MW2 This studyMW2 arlRS-mecA �arlRS and �mecA in-frame double-deletion mutant of MW2 This studyMW2 arlRS::arlRS �arlRS complemented with arlRS with pMAD cycling This studyMW2 arlRS with pEPSA5::arlRS �arlRS complemented with arlRS on pEPSA5 This studyUSA300 CA-MRSA; wild-type strain This studyUSA300 arlRS �arlRS in-frame deletion mutant of USA300 This studyCOL HA-MRSA, wild-type strain isolated from a human infection in the early 1960s 9, 20, 49COL arlRS �arlRS in-frame deletion mutant of COL This studyNewman MSSA; wild-type strain isolated from a human infection in 1952 11Newman arlRS �arlRS in-frame deletion mutant of Newman This studyNewman lytN �lytN in-frame deletion mutant of Newman This studyNewman lytM �lytM in-frame deletion mutant of Newman This studyNewman atl �atl in-frame deletion mutant of Newman This studyNewman arlRS-lytN �arlRS and �lytN in-frame double-deletion mutant of Newman This studyNewman arlRS-lytM �arlRS and �lytM in-frame double-deletion mutant of Newman This studyNewman arlRS-atl �arlRS and �atl in-frame double-deletion mutant of Newman This studyNewman::mecA MSSA wild-type strain expressing mecA on pEPSA5; fully methicillin resistant This studyNewman arlRS::mecA �arlRS of Newman expressing mecA on pEPSA5; fully methicillin resistant This studySH1000 MSSA; wild-type strain 8325-4 with rsbU restored 21SH1000 arlRS �arlRS in-frame deletion mutant of SH1000 This studyRN6390 MSSA; wild-type strain; agr� laboratory strain related to 8325-4; maintains hemolytic

pattern when propagated on sheep erythrocytes37

RN6390 arlRS �arlRS in-frame deletion mutant of 6390 This study8325 MSSA; wild-type strain; prophage-cured strain of NCTC8325 harboring an 11-bp deletion in

rsbU that regulates sigB activity by activating RsbV33

8325 arlRS �arlRS in-frame deletion mutant of strain 8325 This study

E. coli XL-1 blue General laboratory cloning strain 40

PlasmidspMAD Allelic replacement vector to generate S. aureus mutant strains 1pEPSA5 Ectopic expression vector (pEPSA5) for genes in S. aureus 13pMAD arlRS� pMAD containing �0.8 kb up- and downstream fragments of arlRS for deletion (DNAs

from MW2, USA300, COL, Newman, SH1000, 6390, or 8325)This study

pMAD arlRS� pMAD containing the arlRS genes and �0.8 kb up- and downstream from MW2 orNewman for complementation

This study

pMAD� MW2 mecA� pMAD containing �0.8 kb up- and downstream MW2 fragments of mecA for deletion This studypMAD Newman lytN� pMAD containing �0.8 kb up- and downstream Newman fragments of lytN for deletion This studypMAD Newman lytM� pMAD containing �0.8 kb up- and downstream Newman fragments of lytM for deletion This studypMAD Newman atl� pMAD containing �0.8 kb up- and downstream Newman fragments of atl for deletion This studypEPSA5::arlRS pEPSA5 containing 2-kb DNA fragment containing the arlRS coding regions, including their

own RBSa from MW2 or Newman at the SalI/PstI sitesThis study

pEPSA5::mecA pEPSA5 containing 2-kb DNA fragment containing the mecA coding region, including itsown RBS from MW2 at the BamHI/XbaI sites

This study

a RBS, ribosome binding site.

Memmi et al.

760 jb.asm.org Journal of Bacteriology

on May 28, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

vector pEPSA5. The sequences of DNA primers used in this study areavailable on request.

Isolation of RNA and Northern blot hybridization. Overnight cul-tures of S. aureus were diluted 1:100 in 40 ml of TSB and grown in 200-mlflasks with shaking to reach exponential phase (A650 � 0.7 in Spectronic20), using 18-mm borosilicate glass tubes. At an optical density (OD) of0.7, total RNA was extracted from 10 ml of culture by using a TRIzol-glassbead method, as described previously (30). After growth for an additional60 min, RNAs were extracted from all samples. Ten micrograms each oftotal RNA was analyzed by Northern blotting with gel-purified DNAprobes (�350 bp each) radiolabeled with [�-32P]dCTP using therandom-primed DNA-labeling kit (Roche Diagnostics GmbH) and hy-bridized under aqueous-phase conditions at 65°C. The blots were subse-quently washed, and the bands were visualized by autoradiography. Threeindependent clones were utilized to extract RNA from each strain andtested in Northern blot analysis.

Transcriptional-fusion studies of the P3 agr promoter linked to thegreen fluorescent protein (GFPuvr) reporter gene. To confirm the effectof the arlRS deletions on agr transcription, we cloned the agr P3 promoterin pALC1484, a derivative of pSK236 containing the promoterless gfpuvr

to generate transcriptional fusions, as confirmed by restriction analysisand DNA sequencing. Recombinant promoter fusion plasmids were in-troduced into S. aureus RN4220, purified, and electroporated into wild-type MW2, Newman, and their isogenic arlRS mutants. For analysis, ali-quots (200 �l) from overnight cultures of three independent clones weretransferred to microtiter wells to assay for cell density (OD at 650 nm[OD650]) and fluorescence in an FL600 fluorescence reader (BioTek In-struments). Promoter activation was plotted as the mean fluorescence/OD650 ratio, using the average values from triplicate readings from threeclones per strain.

Ectopic gene expression in S. aureus. Besides pMAD-mediated com-plementation of the mutant strains, we also utilized the plasmid pEPSA5,which can be induced for expression with xylose (1%) (11). For pEPSA5-mediated expression, PCR-amplified gene fragments were ligated intopEPSA5 and transformed first into E. coli XL-1 blue and then intoRN4220. Plasmids from positive clones of S. aureus RN4220 were thenintroduced into MW2, Newman, and their isogenic arlRS mutants asdescribed above. For reverse transcription (RT)-PCR, extracted RNA wasresuspended in diethyl pyrocarbonate (DEPC) water, treated with TurboDNase I (Ambion), and then reverse transcribed with 1 �g of total RNA,using the Transcriptor First Strand cDNA Synthesis Kit (Roche). RegularPCR was then performed on cDNAs using oligonucleotides specific foreach gene. For agr overexpression, we utilized pRN6735 (47), which is aderivative of pC194 containing the pI258 bla promoter and two-thirds ofthe structural gene (blaZ), followed by the 1,566-bp MboI fragment con-taining the RNAIII determinant lacking its own promoter. The cloned blapromoter is repressed by pI524 in the tested strain, which supplies the blarepressor in trans. Under these conditions, the bla promoter is inducibleby �-lactam compounds, such as oxacillin.

X-100-induced autolysis assays in growing cultures. The autolysisassay was performed as described previously (23). Briefly, overnight cul-tures were diluted 1:100 to an OD650 of 0.1 in TSB and grown at 37°C withshaking in the presence of Triton X-100, with the OD600 recorded hourlyfor 7 to 8 h. Each data point represents the mean and standard deviationfrom three independent experiments.

Zymogram analysis. Zymogram analysis was conducted to detect al-terations in autolysin activity as described previously, with minor altera-tions (23). Cell fractions containing autolytic enzymes from MW2, New-man, and their isogenic �arlRS mutants were extracted from the cell pelletof a 10-ml bacterial culture grown to an OD650 of 0.7 using 100 �l of 4%SDS. Equivalent amounts of extracted proteins were separated on theSDS-PAGE gel containing heat-killed RN4220 or Micrococcus lysodeikticusbacterial cells at 10 mg/ml (wet weight). The resolved proteins were al-lowed to renature overnight in water and were incubated with 0.5% meth-ylene blue to visualize clear bands representing an area of cell lysis with

RN4220 or M. lysodeikticus. The assay was repeated three times, and arepresentative experiment is shown.

RESULTSInitial characterization of arlRS mutants in autolysis. To assessthe contribution of ArlRS to autolysis, we generated in-frame de-letion mutants of both the histidine kinase and the response reg-ulator in HA-MRSA strain COL, CA-MRSA strains MW2 andUSA300, and MSSA strains Newman, SH1000, 8325-4, andRN6390 (Fig. 1C). We found that MSSA strain Newman, MRSAstrain MW2, and their corresponding arlRS mutants grew com-parably in TSB medium (Fig. 1A), similar to the growth of arlRSmutants and their respective parent strains in other MSSA andMRSA strains under study (data not shown). In the presence ofTriton X-100, however, growth of all arlRS mutants in the MSSAbackground (8325-4, RN6390, Newman, and SH1000) was se-verely affected, while all arlRS mutants of MRSA strains (COL,MW2, and USA300) remained unaffected (Fig. 1B and data notshown). Given the above findings, we elected to focus most of oursubsequent studies on MW2 and Newman, representing MRSAand MSSA strains, respectively.

To exclude chromosomal rearrangement or an ectopic pointmutation as the cause of the increased autolysis, the pMAD plas-mid was also used to introduce the native arlRS genes into the�arlRS mutants of Newman and MW2. These chromosomallycomplemented strains regained wild-type levels of growth in thepresence of Triton X-100 (Fig. 2A). Complementation was alsoobtained when arlRS were cloned into the xylose-inducible plas-mid pEPSA5 (Fig. 2A), but not when arlR or arlS was singly ex-pressed, possibly due to the high copy number of the sensor or thekinase gene provided by the plasmid or the fact that ArlR requiresArlS to function properly. We also examined the effects of arlRSmutations on genes upstream and downstream of arlRS in bothMW2 and Newman arlRS mutants by Northern blotting and didnot discern any significant polar effects as a result of the arlRSmutations (data not shown).

Given that autolysis might affect methicillin resistance, we alsoassayed the sensitivities of the arlRS mutants to oxacillin and naf-cillin in both strains MW2 and Newman. Deletion of arlRS in bothMW2 and Newman did not lead to a significant reduction in theMIC for oxacillin, nafcillin, or cefoxitin, which were the �-lactamsselected for this study.

Effect of arlRS on zymographic profiling of autolysins in ar-lRS mutants. Penicillins and �-lactams in general may enhanceautolysis by triggering murein hydrolase activity on dividing cells(19). Although the arlRS system has previously been shown to beinvolved in autolysis by negatively regulating murein hydrolaseexpression in an MSSA strain, such as 8325-4 (15), its contribu-tion to autolysis by zymogram analysis in clinically relevant MRSAstrains has not been assessed. Accordingly, we performed zymo-gram analysis in isogenic arlRS strains of MW2 and comparedthem to those of the Newman background. As shown in Fig. 1D,there was no discernible change in the lytic profiles of variousarlRS mutants compared with the parental controls, regardless ofwhether Micrococcus or S. aureus RN4220 was used as a substrate.We also deployed the Newman �arlRS strain as a cellular substratein the gel to ascertain if increased autolysis in the arlRS mutant ofNewman was due to a general increase in sensitivity to autolyticenzymes. However, these zymogram assays also showed no differ-

Role of ArlRS in Autolysis in MSSA and MRSA

February 2012 Volume 194 Number 4 jb.asm.org 761

on May 28, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

ence in the lytic band profiles between isogenic arlRS strains (datanot shown).

Effect of mecA on Triton X-100-induced autolysis in arlRSmutants. On the basis of similar autolytic defects between arlRSmutants of SigB-deficient RN6390 (also 8325-4) and strainSH1000 with SigB restored, we have excluded the role of sigma Bas the primary cause of enhanced autolysis mediated by arlRS(data not shown). Similarly, a difference in saeRS expression,which was found to be constitutive in Newman but not in 6390,SH1000, and 8325-4 (18), likely fails to explain the differences inautolysis observed between MSSA and MRSA strains. We alsotested the role of mecA by deleting the mecA gene in the arlRSmutant of MW2, as well as expressed mecA on the xylose-inducible plasmid pEPSA5 in the parental strain Newman and itsisogenic arlRS mutant. In the mecA mutant of MW2, we deletedonly the region encoding PBP2A in the staphylococcal cassettechromosome mec element (SCCmec). These data (Fig. 2B) showedthat deletion of mecA in MRSA strain MW2 did not alter growth inTriton X-100, while addition of mecA to the arlRS mutant ofMSSA strain Newman, which, as expected, increased its resistanceto oxacillin (the MIC increased from �0.5 to 64 �g/ml), also didnot restore the autolysis defect in Triton X-100. Likewise, a com-

bined mecA and arlRS deletion in MW2 did not affect growth inmedium supplemented with Triton X-100. We also evaluated theexpression of pbp2 and pbp4 and did not find any significant dif-ferences in the expression of these two genes between isogenicarlRS strains of Newman and MW2.

Effect of arlRS on expression of genes that regulate autolysisand murein hydrolase activity. To assess whether the TritonX-100-induced autolysis of the arlRS mutant in Newman, but notin MW2, was due to different impacts on autolysin gene expres-sion, Northern blots were conducted with DNA probes specific forthe genes encoding major autolysins (atl, lytM, lytN, lytH, lytA,and sle1), the two-component regulatory system (TCRS) (lytSR),and the downstream effector lrgA, which encodes a putative anti-holin that interferes with the transport of murein hydrolase acrossthe cell membrane into the cell wall (7, 8). We did not observe anydifferences in the transcription of genes encoding autolysins (atl,lytM, lytH, lytA, and sle1) between parent and the isogenic arlRSmutants. However, lytN was significantly upregulated in the arlRSmutants of both MW2 and Newman (Fig. 3A). On the other hand,the transcription of lytRS, which is a repressor of autolytic activity,and its downstream effector lrgAB was reduced in the arlRS mu-tants of both MW2 and Newman (Fig. 3A). Notably, we were able

FIG 1 Increased autolysis of the arlRS mutant of strain Newman in TSB supplemented with 0.02% Triton X-100. (A and B) Growth curves of MW2, Newman,and their isogenic arlRS mutants with and without 0.02% Triton X-100. Overnight cultures were diluted to an OD650 of 0.1 in TSB, supplemented with 0.02%Triton X-100, and grown at 37°C with shaking. The asterisk indicates statistical significance of the growth of the arlRS mutant of Newman versus the parent strainat all time points by a paired Student t test (P � 0.001). WT, wild type. (C) Map of the arlRS TCRS locus. Shown are the open reading frame designations of theMW2 and Newman genomes, secondary gene names, proposed functions, translated protein sizes (amino acids [aa]), and the region deleted in the present study.(D) Deletion of arlRS does not significantly affect murein hydrolase activity in Newman (New) and MW2. Shown are zymogram analyses of cell extracts from S.aureus MW2, Newman, and their isogenic arlRS mutants. Equivalent amounts of cell extracts were separated on an 8% SDS-polyacrylamide gel containingheat-killed S. aureus RN4220 cells (left) or Micrococcus lysodeikticus (right). The resolved gels were washed with water, incubated with lysis buffer (50 mM Tris,0.1% Triton X-100, 10 mM MgCl2, and CaCl2) for 18 to 24 h, and stained with 0.5% methylene blue. Areas of murein hydrolase activity are indicated by clearzones (the same results were obtained with heat-killed cells of Newman and MW2 arlRS mutants).

Memmi et al.

762 jb.asm.org Journal of Bacteriology

on May 28, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

to restore wild-type levels of transcription of these genes in com-plemented arlRS strains of both Newman and MW2 (data notshown). Given that these genes are affected in similar manners inboth MSSA Newman and MRSA MW2 backgrounds, it is unlikelythat differences in transcription of the genes (lytRS and lrgA) inisogenic strains of Newman and MW2 could explain the enhancedautolysis of MSSA strain Newman in Triton X-100 but not MRSAstrain MW2. To partially confirm this, we deleted lrgAB in arlRSmutants of both Newman and MW2 and found no difference ingrowth under autolysis-inducing conditions (data not shown).Cross-complementation of the arlRS mutant of Newman in transusing a pEPSA5-based construct driving lrgAB expression also did

not alleviate the Triton X-100-induced autolysis (data notshown).

Effect of arlRS deletion on expression of mgrA and agr. Inour previous studies, we described mgrA as a positive regulatorof agr and a negative regulator of autolysis by virtue of reducedagr and lrgAB expression in the mgrA mutant (22, 23). To as-certain whether the autolysis defect in MSSA strains was due toa different impact of arlRS deletion on either mgrA or agr,Northern blots were conducted with DNA probes specific formgrA, RNAII, and RNAIII (Fig. 3B). Transcription of mgrA,which constitutes two overlapping transcripts from two dis-tinct promoters, was greatly reduced in both MW2 and New-man backgrounds, although the reduction is more prominentin the Newman background. The most significant difference,however, involved transcription of both agr RNAII andRNAIII. While a modest reduction of both transcripts was ob-served with the arlRS mutant of Newman, both transcripts aremarkedly reduced in the arlRS mutant of MW2 (Fig. 3B). Tofurther confirm the effect of arlRS on agr expression, we con-ducted transcriptional fusion of the agr P3 promoter linked tothe GFPuvr reporter gene. Transcriptional-fusion studies re-vealed that agr expression is significantly downregulated in thearlRS mutant of MW2, while the effect was more modest in theNewman background (Fig. 4). Complementation of the arlRSmutants in MW2 and Newman restored expression of RNAIII,as confirmed by GFP-promoter fusion (Fig. 4) and Northernanalysis (data not shown).

Effect of RNAIII overexpression on autolysis in MW2.Northern blots and promoter fusions indicated that expressionof agr RNAIII was significantly downregulated in the arlRSmutant of MW2 (Fig. 3B and 4). As agr has been described as apositive regulator of autolysis (14), we wanted to ascertain theeffect of RNAIII overexpression in the arlRS mutant of MW2.For this experiment, pRN6735, which overexpressed RNAIIIunder a bla-inducible promoter in the presence of oxacillin(Fig. 5), was introduced into MW2 and its isogenic arlRS mu-tant. As shown in Fig. 5, overexpression of RNAIII under oxa-cillin induction was able to render the arlRS mutant of MW2prone to autolysis in the presence of 0.02% Triton X-100, in-dicating that the downregulation of agr in the arlRS mutant ofMW2 (Fig. 3B and 4) is likely a contributing factor to its resis-tance in Triton X-100-mediated autolysis compared to MSSAstrain Newman. In contrast, overexpression of RNAIII in theparental strain MW2 did not have any major effect on autolysis,even under oxacillin induction. This finding indicated that, inaddition to agr, another factor(s) regulated by arlRS may havecontributed to the propensity for autolysis in arlRS mutants.

Effect of arlRS deletion in combination with autolysins ongrowth and murein hydrolase activity. To assess if the increasedautolysis of the arlRS mutant of Newman was dependent onoverexpression of LytN (Fig. 6A) or on a mechanism involvingother known autolysins, in-frame deletion mutants were madefor lytN, atl, or lytM singularly and in combination with thearlRS mutation. All these new isogenic mutant strains grewsimilarly in TSB in the absence of Triton X-100 (data notshown). In the presence of 0.02% Triton X-100, however, mu-tation of either atl or lytM in combination with arlRS restoredthe autolytic defect while lytN did not (Fig. 6B). To determinethe effects of these mutations on cell wall murein hydrolaseactivity, zymogram analysis was performed with wild-type

FIG 2 Effects of complementation on autolysis of arlRS mutants of Newmanand MW2. (A) Complementation of arlRS on the chromosome or in trans withpEPSA5::arlRS restored wild-type levels of growth in the presence of TritonX-100. Shown are growth curves of Newman, its isogenic arlRS mutant, andcomplemented strains with and without 0.02% Triton X-100. The asteriskindicates statistical significance of the growth of the arlRS mutant versus theparent strain Newman at all time points by a paired Student t test (P � 0.001).(B) Effect of mecA on autolysis of arlRS mutants of Newman and MW2. De-letion of mecA in the arlRS mutant of MW2 did not affect growth underautolysis-inducing conditions, while cross-complementation of the arlRS mu-tant of Newman had no effect on restoration of the defect in Triton X-100-induced autolysis. Shown are growth curves of the Newman wild type and thearlRS mutant carrying pEPSA5::mecA, MW2 mecA, and mecA-arlRS mutantsin TSB-0.02% Triton X-100. The asterisk indicates statistical significance of thegrowth of the arlRS mutant versus its isogenic parent strain Newman at all timepoints by a paired Student t test (P � 0.001).

Role of ArlRS in Autolysis in MSSA and MRSA

February 2012 Volume 194 Number 4 jb.asm.org 763

on May 28, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

Newman and its isogenic mutant strains. As shown in Fig. 6A,an atl deletion with or without arlRS mutation virtually abol-ishes all murein hydrolase activity, in line with previous studies(14, 25). A lytM deletion, with or without ArlRS, leads to anoticeable increase in the �60-kDa lytic band, which may cor-respond to the 62-kDa amidase domain of Atl. Whether LytMmight play a role in processing the Atl amidase domain or

whether the effect of ArlRS on LytM or Atl may be posttran-scriptional remains to be determined.

DISCUSSION

TCRSs are widely used signal transduction devices that engage in amultitude of gene regulatory events in response to changing envi-ronmental conditions (4). In previous studies, it has been shown

FIG 3 Effects of the arlRS mutation on the expression of autolysin genes and known lytic regulators. Shown is Northern blot analysis of lytN, lrgAB, and lytRS(A) and of mgrA, RNAII, and RNAIII (B) expression. The blots were hybridized with a DNA probe specific for each gene radiolabeled with [�-32P]dCTP. Beloware shown the ethidium bromide-stained rRNAs, indicating equivalent amounts of RNA in each sample. The RNAs for each strain were extracted at OD650s of0.7, 1.1, and 1.7, corresponding to exponential, late-exponential, and early-stationary phases, respectively. The experiment was repeated at least three times withsimilar results. A representative experiment is shown.

FIG 4 Regulation of ArlRS on RNAIII. Shown is the expression of GFPuvr driven by the agr P3 promoter in overnight cultures. Promoter activity was plotted asmean fluorescence/OD650 from 3 clones in triplicate. The experiments were repeated three times, with one set shown. The asterisk indicates statistical significanceof the indicated strain compared to MW2 by a paired Student t test (P � 0.001). comp, complementation.

Memmi et al.

764 jb.asm.org Journal of Bacteriology

on May 28, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

that a mutation in the TCRS arlRS resulted in an increase in au-tolysis in a derivative of strain 8325-4 (15). However, we showhere that a deletion in TCRS arlRS leads to enhanced autolysis inTriton X-100 only in methicillin-sensitive strains (15, 28), includ-ing Newman, SH1000, 6390, and 8325-4 (Fig. 1), but not in MRSAstrains (CA-MRSA strains MW2 and USA300 and HA-MRSAstrain COL) (Fig. 1). This result is surprising, considering thatphylogenetically, staphylococcal strains Newman and NCTC8325are related to MRSA strains COL and USA300 (2). We have ex-cluded sigma B and saeRS as plausible causes to account for thedifference in autolysis between arlRS mutants of MSSA and MRSAstrains because both sigma B-positive and -negative strains(RN6390, 8325, Newman, and SH1000), and also strains withgrowth-phase-dependent expression (6390, SH100, and 8325), aswell as constitutive expression of saeRS (Newman) (18), all exhib-ited defects in autolysis in Triton X-100 in the presence of an arlRSmutation. Despite these autolytic defects, murein hydrolase activ-ities among MW2, Newman, and their arlRS isogenic mutantstrains were similar. This result, combined with our finding thatTriton X-100 induced autolysis in actively growing cells, may beattributable to a lack of sensitivity of our zymogram analysis assay.

We also could not attribute a lack of an autolysis defect inMW2 and its isogenic arlRS mutant to mecA, because the increasein autolysis was not detected in the arlRS-mecA double mutant.Furthermore, overexpression of mecA using the inducible systempEPSA5::mecA also did not rescue growth under Triton X-100 inthe �arlRS mutant of Newman, even though mecA rendered New-man and its arlRS isogenic mutant fully oxacillin resistant (Fig. 2).

S. aureus carries several known or putative peptidoglycan hy-drolases, including the major autolysin genes atl, lytM, lytH, lytN,lytA, and sle1. However, we did not find any differences in atl,lytM, lytH, lytA, and sle1 transcription between the arlRS mutantsand the isogenic parents. The observation that lytN transcription

was upregulated in similar manners between the arlRS mutantsand respective parents suggests that lytN is unlikely to explain thedivergent propensities of the arlRS mutants for autolysis betweenMW2 and Newman. In addition, a lytN mutant of Newman alsodid not exhibit a significant increase in murein hydrolase activitycompared with the parent (Fig. 6), again emphasizing a marginalrole for lytN in the different autolytic phenotypes between the twobackgrounds.

The increased rate of autolysis in arlRS mutants of MSSAstrains has been linked to downregulation of lytRS and the ensuingreduced expression of lrgAB (28), which interfere with the trans-port of murein hydrolase from the cytosol to the cell wall. As withlytN, both lytRS and lrgAB are downregulated in the arlRS mutantsof MW2 and Newman (Fig. 3). In addition, complementation oflrgAB in trans did not positively affect the growth of the arlRSmutant of strain Newman under autolytic conditions (data notshown). Together, these data suggest that lrgAB did not play amajor role in autolysis in MRSA strains, in contrast to what hasbeen found in strain 8325-4 (13).

SarA, MgrA, and AgrA have also been shown to affect autolysis,with the first two regulators being repressors of autolysis and thelast being an activator (16, 23, 46). We found that sarA transcrip-tion was not affected in the arlRS mutant. In a previous study, weshowed that arlRS transcripts were downmodulated in an mgrAmutant (23). Here, we have shown that regulation is reciprocal inthat mgrA transcription is also downmodulated in arlRS mutantsof MW2 and Newman, although the difference is more prominentin the Newman background (Fig. 3). A significant difference wasalso seen in the agr RNAII and RNAIII transcript levels in the�arlRS mutant of MW2 compared to Newman �arlRS, as assessedby Northern blot analysis and GFP-promoter fusions. Fujimotoand Bayles have also shown that an agr mutant has a lower rate ofautolysis than the wild-type RN6390 strain (16), suggesting that

FIG 5 Overexpression of RNAIII affects growth of the arlRS mutant of MW2 under autolysis-inducing conditions. (Left) Growth curves of MW2 and its isogenicarlRS mutant carrying pRN6735 with 0.02% Triton X-100. (Right) Expression of RNAIII of the arlRS mutant of MW2 with and without induction with asub-MIC concentration of oxacillin. The asterisk indicates statistical significance at OD650 for growth of the arlRS mutant of MW2 overexpressing agr at all timepoints by a paired Student t test (P � 0.001).

Role of ArlRS in Autolysis in MSSA and MRSA

February 2012 Volume 194 Number 4 jb.asm.org 765

on May 28, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

agr is likely a positive regulator of murein hydrolase activity. Toconfirm that this is also the case with MW2, we overexpressed agrRNAIII with a plasmid containing a �-lactamase-inducible pro-moter driving RNAIII. Our results showed that overexpression ofRNAIII under oxacillin induction could render the arlRS mutantof MW2 susceptible to hydrolysis in the presence of Triton X-100(Fig. 5). This effect is not due to oxacillin alone, since MW2, uponexposure to a sub-MIC level of oxacillin, did not display a dra-matic increase in autolysis compared to the control (Fig. 5). Nev-ertheless, it is clear that the dosage of RNAIII plays a crucial role inautolysis, because the augmentation in autolytic activity in the

arlRS mutant of MW2 was very mild in the absence of oxacillininduction.

Besides MW2, we also ascertained if the hyperautolytic pheno-type in the arlRS mutant of Newman could be rescued by inacti-vating the major autolysin genes atl, lytM, lytN, and lytH. Remark-ably, inactivation of Atl or LytM in the arlRS mutant of Newmanwas able to restore the growth pattern of the mutant to compara-ble to that of the parental strain in 0.02% Triton X-100, whereas alytN or lytH mutation in the arlRS mutant had no effect (Fig. 6B).Notably, an atl deletion virtually abolished murein hydrolase ac-tivity in both wild-type Newman and its isogenic arlRS mutant,while a mutation in lytM led to a significant increase in an �62-kDa band that may correspond to the Atl-derived amidase do-main in both Newman and its isogenic arlRS mutant.

Given that many of the MSSA strains (e.g., 8325-4 andSH1000) under laboratory investigation do not represent clini-cally relevant isolates, our study here on the mechanisms of autol-ysis in a CA-MRSA strain such as MW2 likely provides highlyrelevant information on autolysis under antibiotic-inducing con-ditions. In particular, our data here reveal that deletion of arlRS inMSSA strains, but not in MRSA strains, leads to a very significantincrease in Triton X-100-induced autolysis that is not attributableto alterations in the expression of sigma B, saeRS, or mecA. We alsofound that significant downregulation of RNAIII in MW2, whichis a positive regulator of autolysis, may be part of the mechanismby which a mutation of arlRS in clinically relevant MRSA strainsmay be resistant to autolysis in the presence of a small amount ofnonionic detergent or cell wall-active antibiotics. Nevertheless, itis clear that the effect of RNAIII on autolysis is dosage dependent.In this context, it will be crucial to determine the expression ofRNAIII in clinically relevant MRSA strains under antibiotic stress,which can promote autolysis. The fact that overexpression ofRNAIII in the wild-type MW2 did not induce autolysis impliesthat there is an additional factor(s) controlled by ArlRS that likelycontributes to autolysis. In a recent study, Schlag et al. describedthe role of wall teichoic acid in targeting Atl to the staphylococcalcross-wall (42). In the absence of wall teichoic acid, Atl is evenlydistributed on the cell surface, thus increasing susceptibility toautolysis. Whether arlRS affects the synthesis or distribution ofwall teichoic acid in S. aureus was not defined in our studies. Nev-ertheless, understanding the additional genetic determinants thataccount for this increased autolysis in MRSA strains could poten-tially lead to the development of novel strategies against MRSAinfections.

REFERENCES1. Arnaud M, Chastanet A, Debarbouille M. 2004. New vector for efficient

allelic replacement in naturally nontransformable, low-GC-content,gram-positive bacteria. Appl. Environ. Microbiol. 70:6887– 6891.

2. Baba T, Bae T, Schneewind O, Takeuchi F, Hiramatsu K. 2008. Genomesequence of Staphylococcus aureus strain Newman and comparative anal-ysis of staphylococcal genomes: polymorphism and evolution of two ma-jor pathogenicity islands. J. Bacteriol. 190:300 –310.

3. Baba T, et al. 2002. Genome and virulence determinants of high virulencecommunity-acquired MRSA. Lancet 359:1819 –1827.

4. Beier D, Gross R. 2006. Regulation of bacterial virulence by two-component systems. Curr. Opin. Microbiol. 9:143–152.

5. Boucher HW, Corey GR. 2008. Epidemiology of methicillin-resistantStaphylococcus aureus. Clin. Infect. Dis. 46(Suppl. 5):S344 –S349.

6. Bratu S, et al. 2005. Community-associated methicillin-resistant Staph-ylococcus aureus in hospital nursery and maternity units. Emerg. Infect.Dis. 11:808 – 813.

7. Brunskill EW, Bayles KW. 1996. Identification and molecular character-

FIG 6 Deletion of atl or lytM affects murein hydrolase activity and restoresgrowth in a Newman arlRS mutant strain under autolysis-inducing condi-tions. (A) Zymogram analysis of cell extracts from S. aureus Newman and itsisogenic arlRS, lytN, lytM, atl, and double mutants. Equivalent amounts of cellextracts were separated on an 8% SDS-polyacrylamide gel containing heat-killed S. aureus RN4220 cells. After electrophoresis, the gels were washed withwater and then buffer and finally stained with 0.5% methylene blue as de-scribed in Materials and Methods. Areas of murein hydrolase activity are in-dicated by clear zones. The arrows indicate bands of increased activity in lytMand arlRS-lytM mutants. (B) Growth curves of Newman and its isogenic arlRS,lytN, lytM, atl, and double mutants with 0.02% Triton X-100. Overnight cul-tures were diluted to an OD650 of 0.1 in TSB, supplemented with 0.02% TritonX-100, and grown at 37°C with shaking. The asterisk indicates statistical sig-nificance of growth of arlRS and arlRS-lytN mutants versus the parent strainNewman at all time points by a paired Student t test (P � 0.001).

Memmi et al.

766 jb.asm.org Journal of Bacteriology

on May 28, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

ization of a putative regulatory locus that affects autolysis in Staphylococ-cus aureus. J. Bacteriol. 178:611– 618.

8. Brunskill EW, Bayles KW. 1996. Identification of LytSR-regulated genesfrom Staphylococcus aureus. J. Bacteriol. 178:5810 –5812.

9. Chambers HF, Hartman BJ, Tomasz A. 1985. Increased amounts of anovel penicillin-binding protein in a strain of methicillin-resistant Staph-ylococcus aureus exposed to nafcillin. J. Clin. Invest. 76:325–331.

10. Clinical and Laboratory Standards Institute. 2006. Methods for dilutionantimicrobial susceptibility tests for bacteria that grow aerobically; ap-proved standard, 7th ed. Document M7-A7. Clinical and Laboratory Stan-dards Institute, Wayne, PA.

11. Duthie ES, Lorenz LL. 1952. Staphylococcal coagulase; mode of actionand antigenicity. J. Gen. Microbiol. 6:95–107.

12. Firczuk M, Mucha A, Bochtler M. 2005. Crystal structures of activeLytM. J. Mol. Biol. 354:578 –590.

13. Forsyth RA, et al. 2002. A genome-wide strategy for the identification ofessential genes in Staphylococcus aureus. Mol. Microbiol. 43:1387–1400.

14. Foster SJ. 1995. Molecular characterization and functional analysis of themajor autolysin of Staphylococcus aureus 8325/4. J. Bacteriol. 177:5723–5725.

15. Fournier B, Hooper DC. 2000. A new two-component regulatory systeminvolved in adhesion, autolysis, and extracellular proteolytic activity ofStaphylococcus aureus. J. Bacteriol. 182:3955–3964.

16. Fujimoto DF, Bayles KW. 1998. Opposing roles of the Staphylococcusaureus virulence regulators, Agr and Sar, in Triton X-100- and penicillin-induced autolysis. J. Bacteriol. 180:3724 –3726.

17. Fujimura T, Murakami K. 1997. Increase of methicillin resistance inStaphylococcus aureus caused by deletion of a gene whose product is ho-mologous to lytic enzymes. J. Bacteriol. 179:6294 – 6301.

18. Geiger T, Goerke C, Mainiero M, Kraus D, Wolz C. 2008. The virulenceregulator Sae of Staphylococcus aureus: promoter activities and responseto phagocytosis-related signals. J. Bacteriol. 190:3419 –3428.

19. Giesbrecht P, Kersten T, Maidhof H, Wecke J. 1998. Staphylococcal cellwall: morphogenesis and fatal variations in the presence of penicillin. Mi-crobiol. Mol. Biol. Rev. 62:1371–1414.

20. Gill SR, et al. 2005. Insights on evolution of virulence and resistance fromthe complete genome analysis of an early methicillin-resistant Staphylo-coccus aureus strain and a biofilm-producing methicillin-resistant Staph-ylococcus epidermidis strain. J. Bacteriol. 187:2426 –2438.

21. Horsburgh MJ, et al. 2002. sigmaB modulates virulence determinantexpression and stress resistance: characterization of a functional rsbUstrain derived from Staphylococcus aureus 8325-4. J. Bacteriol. 184:5457–5467.

22. Ingavale S, van Wamel W, Luong TT, Lee CY, Cheung AL. 2005.Rat/MgrA, a regulator of autolysis, is a regulator of virulence genes inStaphylococcus aureus. Infect. Immun. 73:1423–1431.

23. Ingavale SS, Van Wamel W, Cheung AL. 2003. Characterization of RAT,an autolysis regulator in Staphylococcus aureus. Mol. Microbiol. 48:1451–1466.

24. Jayaswal RK, Lee YI, Wilkinson BJ. 1990. Cloning and expression of aStaphylococcus aureus gene encoding a peptidoglycan hydrolase activity.J. Bacteriol. 172:5783–5788.

25. Kaneda S. 1997. Isolation and characterization of autolysin-defective mu-tants of Staphylococcus aureus that form cell packets. Curr. Microbiol.34:354 –359.

26. Komatsuzawa H, et al. 1997. Subcellular localization of the major auto-lysin, ATL and its processed proteins in Staphylococcus aureus. Microbiol.Immunol. 41:469 – 479.

27. Kourbatova EV, et al. 2005. Emergence of community-associatedmethicillin-resistant Staphylococcus aureus USA 300 clone as a cause ofhealth care-associated infections among patients with prosthetic joint in-fections. Am. J. Infect. Control 33:385–391.

28. Liang X, et al. 2005. Global regulation of gene expression by ArlRS, atwo-component signal transduction regulatory system of Staphylococcusaureus. J. Bacteriol. 187:5486 –5492.

29. Lopez R, Ronda-Lain C, Tapia A, Waks SB, Tomasz A. 1976. Suppres-sion of the lytic and bactericidal effects of cell wall inhibitory antibiotics.Antimicrob. Agents Chemother. 10:697–706.

30. Manna AC, Ingavale SS, Maloney M, van Wamel W, Cheung AL. 2004.Identification of sarV (SA2062), a new transcriptional regulator, is re-pressed by SarA and MgrA (SA0641) and involved in the regulation ofautolysis in Staphylococcus aureus. J. Bacteriol. 186:5267–5280.

31. Memmi G, Filipe SR, Pinho MG, Fu Z, Cheung A. 2008. Staphylococcusaureus PBP4 is essential for beta-lactam resistance in community-acquired methicillin resistant strains. Antimicrob. Agents Chemother. 52:3955–3966.

32. Michel A, et al. 2006. Global regulatory impact of ClpP protease ofStaphylococcus aureus on regulons involved in virulence, oxidative stressresponse, autolysis, and DNA repair. J. Bacteriol. 188:5783–5796.

33. Novick RP. 1990. The Staphylococcus as a molecular genetic system, p1– 40. In Novick RP (ed), Molecular biology of the Staphylococcus. VCHPublishers, New York, NY.

34. Odintsov SG, Sabala I, Marcyjaniak M, Bochtler M. 2004. Latent LytMat 1.3A resolution. J. Mol. Biol. 335:775–785.

35. Oshida T, et al. 1995. A Staphylococcus aureus autolysin that has anN-acetylmuramoyl-L-alanine amidase domain and an endo-beta-N-acetylglucosaminidase domain: cloning, sequence analysis, and character-ization. Proc. Natl. Acad. Sci. U. S. A. 92:285–289.

36. Popovich KJ, Weinstein RA, Hota B. 2008. Are community-associatedmethicillin-resistant Staphylococcus aureus (MRSA) strains replacing tra-ditional nosocomial MRSA strains? Clin. Infect. Dis. 46:787–794.

37. Projan SJ, Novick RP. 1997. The molecular basis of pathogenicity, p55– 81. In Crossley KB and Archer GL (ed), The staphylococci in humandisease. Churchill Livingstone, New York, NY.

38. Ramadurai L, Jayaswal RK. 1997. Molecular cloning, sequencing, andexpression of lytM, a unique autolytic gene of Staphylococcus aureus. J.Bacteriol. 179:3625–3631.

39. Rice KC, et al. 2003. The Staphylococcus aureus cidAB operon: evalua-tion of its role in regulation of murein hydrolase activity and penicillintolerance. J. Bacteriol. 185:2635–2643.

40. Sambrook J, Fritsch EF, Maniatis T. 1989. Molecular cloning: a labora-tory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold SpringHarbor, NY.

41. Schenk S, Laddaga RA. 1992. Improved method for electroporation ofStaphylococcus aureus. FEMS Microbiol. Lett. 73:133–138.

42. Schlag M, et al. 2010. Role of staphylococcal wall teichoic acid in targetingthe major autolysin Atl. Mol. Microbiol. 75:864 – 873.

43. Seybold U, et al. 2006. Emergence of community-associated methicillin-resistant Staphylococcus aureus USA300 genotype as a major cause ofhealth care-associated blood stream infections. Clin. Infect. Dis. 42:647–656.

44. Sugai M, Fujiwara T, Komatsuzawa H, Suginaka H. 1998. Identificationand molecular characterization of a gene homologous to epr (endopepti-dase resistance gene) in Staphylococcus aureus. Gene 224:67–75.

45. Tomasz A, Waks S. 1975. Mechanism of action of penicillin: triggering ofthe pneumococcal autolytic enzyme by inhibitors of cell wall synthesis.Proc. Natl. Acad. Sci. U. S. A. 72:4162– 4166.

46. Trotonda MP, Xiong YQ, Memmi G, Bayer AS, Cheung AL. 2009. Roleof mgrA and sarA in methicillin-resistant Staphylococcus aureus autolysisand resistance to cell wall-active antibiotics. J. Infect. Dis. 199:209 –218.

47. Vandenesch F, Kornblum J, Novick RP. 1991. A temporal signal, inde-pendent of agr, is required for hla but not spa transcription in Staphylo-coccus aureus. J. Bacteriol. 173:6313– 6320.

48. Wang X, Mani N, Pattee PA, Wilkinson BJ, Jayaswal RK. 1992. Analysisof a peptidoglycan hydrolase gene from Staphylococcus aureus NCTC8325. J. Bacteriol. 174:6303– 6306.

49. Yarwood JM, et al. 2002. Characterization and expression analysis ofStaphylococcus aureus pathogenicity island 3. Implications for the evolu-tion of staphylococcal pathogenicity islands. J. Biol. Chem. 277:13138 –13147.

Role of ArlRS in Autolysis in MSSA and MRSA

February 2012 Volume 194 Number 4 jb.asm.org 767

on May 28, 2020 by guest

http://jb.asm.org/

Dow

nloaded from